*Subtitles are available.* I want to clarify that I do not suggest using Pinnacle Dazzle, but if you decide to use it, then some models work better than other, see this video for more info: czcams.com/video/RVbqGrf4rIQ/video.html If your video is too "wavy" or is skewed on the top, you need a time base corrector (TBC). These devices are increasingly rare and expensive. Some capture cards have rudimentary TBC features, see this video for more info: czcams.com/video/bhZyk72I9Q4/video.html Also, some VCRs and camcorders have built-in TBC, like: Mitsubishi HS-MD3000 medical VCR: czcams.com/video/Xoft92VctQY/video.html or JVC GR-SX950 SVHS camcorder: czcams.com/video/qs8KF1G_mgs/video.html
@@ConsumerDV I got one, but it had bad image, so i went and recaped the Chroma and luma module, now i only get black screen 😢 send it to a Repair shop to see if the can fix it if not i will take a shot in my balls lol
Great conversion guide, with careful attention to deinterlacing details. As you point out, video pixels are not quite square, with a 10/11 width/height aspect ratio for NTSC pixels. If you apply this factor to the 720x480 frame produced by digitizing NTSC video, you get the 704x480 square-pixel frame described @7:36. However, when upscaling the frame to 720p for CZcams upload, there is an alternative scaling that facilitates cropping off the video distortion at the bottom of the NTSC frame as well. If you scale the original 720x480 frame up to 1000x736, you can simply crop off the bottom 16 lines to produce a 1000x720 frame. This scaling corrects the 10/11 pixel aspect ratio while preserving the full width of the frame, and can easily be letterboxed up to 1280x720 HD format, or left at 1000x720 for upload to CZcams. This scalling also works well with H.264 macroblock boundaries.
Thanks! 704x480 is not square-pixel, the pixels are still 10/11, but the frame is true 4:3 - 704*10/480/11 ≈ 1.3333 I see your point about 1000x736, it has AR 1.3587, which is close to 1.3636, all the while 1000 is divisible by 8. I am not sure that this size will work for H.264, quick googling returns that base macroblock is 16x16, but maybe you are right, and 8x8 macroblocks are possible. Alternatively, you can scale 720x480 to 992x728 and crop 8 pixels on the bottom instead of 16. This may not be enough though to get rid of head switching noise. Or crop 720x480 to 704x480, scale to 976x732 and crop 12 lines on the bottom.
I have recently stumbled across a JVC DVD recorder/VCR combo that has an HDMI output. I'm going to try a little Handbrake & VideoProc to see if I can clean it up more.
Very helpful video! I've been digitizing VHS and Video8 tapes through a Sony HVR-M15AE that takes analog input from my VCR/camcorder and then outputs it through a firewire cable to my pc. For software i use the ScenalyzerLive 4.0, very helpful for previewing and for joining clips together. The HVR-M15AE automatically uses the DV-codec (it's a MiniDV/HDV deck) so the colors get weird sometimes. The tapes are usually up to 3 hours long so the 50i -> 50p convertion sadly takes forever! My customers have been happy this far, but if theres ever a customer thats more needy on the quality and colors of their digitized VHS i will come back to this video surely!
Thanks! Using a DV camcorder as a converter is a foolproof method to digitize analog videos: you get resulting video encoded with an industry-standard codec with interlacing intact. The only reason against it in PAL region may be color subsampling of PAL DV camcorders, which is 4:2:0, that is, samples are taken from every other line, not from each line like in NTSC DV camcorders, which use 4:1:1 subsampling. But considering all other benefits of DV I think think is a reasonable price to pay. Regarding conversion to 50p, it seems that QTMGC is everyone's darling, but I wanted to stay within the boundaries of VirtualDub, and Yadif is not too bad. I also tried using MSU Deinterlacer, it is slightly better, but an order of magnitude slower, I would not use it if I were doing it on everyday basis.
Thank You for this process. I think it would be useful for someone to explain the import and export process for video files viewed on a computer or being uploaded to a service such as Utube versus authoring a DVD. I have home movies that will be DVD authored because I have a brother in a nursing home where it is easier to just bring a small DVD player to connect up to the TV. I also have parents to old to know how to deal with a computer so yes DVDs still are useful. When to Deinterlace or leave interlaced? When to upscale and when not to upscale. Again thank you.
DVD-video format uses MPEG-2 a.k.a. H.262 (you can read more about DVD-video on Wikipedia: en.wikipedia.org/wiki/DVD-Video). VirtualDub can read MPEG-2 but cannot write, so you need to use another tool to convert from whatever format you have into MPEG-2, then you need to author a DVD disc. For DVD, you should keep interlaced video as interlaced, because DVD does not support 50p/60p. The easiest option is to use commercial software to capture from tape directly into MPEG-2 and then author a DVD disc. Just make sure that the hardware/software you use does not deinterlace the video. If you want to capture tapes in the best possible format for preservation, then you use tools like VirtualDub or AmarecTV to capture into lossless or visually-lossless codec, and then create deliverables for different platforms: for DVD you would re-encode to MPEG-2 and author a disc, for CZcams and computer viewing you would deinterlace into 50p/60p and encode using H.264 or H.265 codec into something like MP4 or MKV.
Thank you so much for donating a person’s most valuable asset, TIME, to address my capture process and advice on capturing and DVD authoring. Yes understand the deinterlace before resize as not doing so will make deinterlacing after resize impossible due to changing line spacing in each field of the frame. Do you have a website or email address I could donate at least enough to buy a cup of coffee? Live long and prosper, thank you
I am using Sony Hi8 connected to a Sony DV camera using pass through, connect to computer using Firewire, captured using Virtualdub2. If I understand this work flow then to make a DVD I would... 1. Initially keep the frame size at 720x480i. Time marker 2:35 2. If making a DVD I do not add a deinterlace filter, keep capture interlaced as this is the NTSC standard for DVD and what the camcorder recorded. Time marker 5:25 3. I can trim the over scan and eliminate the head switch artifact with two separate crops. Time marker 8:44 4. Can I or should I resize the frame back to 704x480i? Add resize (upscaling) filter. Time marker 10:08 5. The captured file is then imported to NLE (I use Resolve), edited, exported as MOV. 6. Imported into Pinnacle Studio for DVD menu creation and burned as DVD. 7. DVD player will handle upscale and the TV deinterlace. 8. Optionally, I can upscale and convert the MOV file using NLE to 1440x1080p (progressive), export as MP4 for viewing on computer or Utube or TV if supported. Thank You
I haven't used VDub to capture DV, but if it works, great. Usually I use WinDV, but the bytes are the same. You don't need to crop anything for DVD. DVD supports both 704x480 and 720x480. It is best not to resize back to 480 lines. If you don't like the head switching noise, you can mask it by cropping, say, 4 lines at the bottom (use "Crop" filter), then by padding 4 empty lines back (use "canvas size" filter with "TC" anchor). You need to always deinterlace before resizing. Remember, that common commercial tools can capture directly to MPEG-2 and you probably can use the same tool to author a DVD. If you decide to stick with capturing into lossless (like Huffyuv) or visually-lossless (like Cineform) codec, then the steps would look like this: For DVD: 1. Capture as at 720x480i. 2. Import the captured file to Resolve (verify codec support). Make sure that Resolve treats the file as interlaced with "NTSC DV" proportions. In my case, Vegas treats Cineform files as progressive PAR 1, so I need to change it in the clip properties. Edit, export as MPEG-2 in whatever container your authoring tool understands. 3. Import into DVD authoring tool to create the menu and the file structure, burn a DVD. 4. DVD player will handle upscale and the TV deinterlace. For CZcams: 1. Capture as at 720x480i. 2. Deinterlace with doubling frame rate. Yadif is the best VirtualDub built-in deinterlacer. I use MSU Deinterlacer, you can download it separately. It is better but much slower. Other people use QTGMC with Avisynth, it is even better. 3. Crop as needed. You don't have to maintain 4:3 proportions, but it is simpler if you do. After you crop 16 pixels on the sides, for every two lines you remove three pixels on the sides. 4. Resize to at least 720p for CZcams to enable 60p. If you want to switch to square pixels, resize to 960x720 or 1440x1080. 5. Encode with H.264, specify PAR width 1, PAR height 1, save into MP4 or MOV. If you can deinterlace, crop and resize in your NLE, you can do it there. In either case, whatever you've captured will be your high-quality master.
There are playback control buttons on the bottom. On the very left there is a Stop button (with a square), then Play Input (a triangle with "I"), then Play Output (a triangle with "O"). You can use keyboard as well: Space is to play input only, Enter is to play input and output.
Thank you for great video. But I have a question. Following your guide I captured NTSC analog and created nice master AVI file with 4:2:2 color space. Unfortunately, every program I used to convert it to MP4 (Premiere, Handbrake and many others) resulted in files with 4:2:0 color space. Using VirtualDub2 internal x264 converter resulted in some kind of non-standard MP4 file which was unplayable in my Samsung TV. Files created by all other programs played perfectly. Is there a way to create x264 MP4 file while preserving 4:2:2 color space?
H.264 4:2:2 is definitely possible. Similar to your experience, when I encoded in 4:2:2, many of my desktop apps would not play it... I think VLC did, Splash did not. I haven't tried playing it on my TV. CZcams accepted it just fine. I keep Cineform 4:2:2 as a master, and render into 4:2:0 for playback. I am sorry that I don't have better and more specific advice. H.264 4:2:0 files rendered from VirtualDub using X.264 work just fine for me.
This depends. (1) You may want to digitize your analog video as close to the original as possible to become a master video, a source of all other versions. (2) Or you may want to have it in a form that is easy to edit. (3) Or you may want to have it in a form that is compatible with consumer formats like DVD or Blu-ray. (4) Or you may want to have it compatible with CZcams. Depending on a particular need, your second half of the work, AFTER you have the video captured, will be different: different codec, different scanning type, different frame size, etc. But the first half of the work - getting the video out of your camcorder or VCR to a computer - can be the same if you choose to capture it in an uncompressed or mildly-compressed format that preserves key parameters of the video: resolution, aspect ratio, scanning type, color subsampling, etc. The major difference between DVD and CZcams is that DVD video is interlaced, so you want to capture the original interlaced video from the camcorder as interlaced, and you want to keep it interlaced and author it on a DVD as interlaced, and let your TV set figure out how to show it correctly. CZcams does not handle interlaced video well, so you need to de-interlace. Before 2014 you had to deinterlace into 30p, but now you can deinterlace into 60p, converting each field into a separate frame, preserving the motion. DVD video does not support 60p, so you need to have different versions for CZcams and for DVD. I mean, you can have 30p both for DVD and CZcams, but it is the lowest common denominator. Blu-ray supported 60p for 720p only, but the latest revision of BD standard does support 60p for 1080p and 4K as well. When capturing, double-check that the codec that you are capturing into (Huffyuv, Lararith, Cineform, or whatever else) is supported by your editing tools. I am partial to Cineform because it is a high-quality codec that has been used by pros for the last fifteen years or so, but there are other good ones. It depends on how you want to watch or edit your footage later, you want to ensure compatibility and longevity. TL;DR: it is best to capture into the least destructive format and have it as a master. Then you can always make a version off it for a particular platform.
Thanks for the reply , yeah I just want to have it digitalized as similar as the original one , I sent it to someone who offers this kind of service but it looks a bit stretched and also not as fluent as the original one ( I believe he covered it to 30fps )
I've been setting up a new fresh VirtualDub2 system running Windows 7. Just did a fresh install, put on WinDV and using Dazzle to Capture the same way as in this video. When try to save the video I notice under audio compression I'm missing MPEG Layer-3, so when I hit save is gives an error (pcm_s16le: codec not currently supported in container mp4). Do you now how I can add mp3 compression to the VirtualDub2 Library?
I think I have like three different packages for MP3... I guess some comes with VDub, then I think some came with Lame? I don't remember whether Lame installed any codecs available to other programs. I found this link, it is old, but pertains to 64-bit system, so may be relevant: forum.videohelp.com/threads/271712-Lame-MP3-Encoder-on-Vista-x64#post2016711 Then again, people say that there is no benefit running 64-bit VDub, and most filters are made only for the 32-bit version. I am sorry I cannot help you better :(
I've seen so many videos uploaded to CZcams that haven't been properly deinterlaced first. They are wrong and not authentic at all, because combing only appears on the digitized footage if it hasn't been properly deinterlaced when converting to progressive scan formats. And if you get a progressive scan format that hasn't been deinterlaced, I don't think you can deinterlace it anymore, not fully at least, it will always remain wrong.
Exactly! It seems that many younger people think that this is how we used to watch TV back in the day, with combing galore. They think that combing is authentic, while in reality it was not visible on CRT television sets.
Great video although I'm curious more about some of the settings you skimmed over quickly. One question, why do you capture at 30fps when interlaced video is actually 60fps and some people say to capture at 60fps? I suppose even at 30fps you are simply getting both half set of frames changing each frame, just technically not at the correct time since one of those frames changed earlier than the other. Capturing at 60fps you would see only one half the frame changing for each frame. Both can be deinterlaced and hence you did a double frame rate selection in the deinterlaced options? So maybe doing the original capture at 60fps has no advantage??
Analog TV is 30 (well, 30×1000/1001 ≈ 29.97) for System M a.k.a. NTSC or 25 for all other systems frames per second, interlaced. Broadcast VTR recorded interlaced. Consumer grade VTR recorded interlaced. Camcorders shot and recorded interlaced. Each frame in interlaced TV comprises two fields, and normally these fields are 1/60 or 1/50 second apart, so you get 60 or 50 field per second rate, each field being a different picture made at a different time, you cannot combine them into one picture. When they wanted to transmit movies, which are normally 24 fps, they would slice movie frames into fields and arrange them into fields, creating what is known a pulldown pattern. In this cause you can sometimes weave fields into frames. It is usually much simpler for 50 Hz systems, when they just speed up a movie to 25 fps and slice each frame into two fields, then you can weave these fields back and voila! you get full resolution and characteristic motion signature. I have a short video about digitizing a 24 fps movie from an NTSC VHS cassette: czcams.com/video/DneoPwa5faY/video.html But in "normal" video all fields are different, taken 1/50 or 1/60 seconds apart. Each field contains just half the number of lines of a whole frame. So, I am capturing full information: 30 fps, interlaced. Then I separate the fields, converting them into full frames, creating 60 fps, progressive. In terms of motion 30 fps interlaced and 60 fps progressive look exactly the same. This is how you preserve full spatial and temporal resolution. Those who capture at 60 fps probably use OBS. Why? Likely because OBS has never been meant as an analog capture tool, it is a screen capture tool. Analog capture has been shoehorned in later as an afterthought, so I think that OBS gets 30 fps interlaced, then deinterlaces it into 60 fps, and then you capture 60 fps into a file. I don't use OBS for capturing, although I've heard people had good experience with it. I tried it once, it worked, but every time I tried it again, it did not work for me, so I cannot comment further :)
I thought I understood but now I'm more confused. "So, I am capturing full information: 30 fps, interlaced", but there is no setting to choose between interlaced or progressive in that menu. Are you saying the system identifies a interlaced code in the video and automatically changes from 30 pictures per second to 60 pictures per second? I am really trying to understand this but the fields are combined on the TV screen to make a full frame no? now you are telling me they are never combined into one picture, so viewing traditional TV we see only half the frame and there is black space between the lines? Must be. But then you also said "separate the fields" so that implies the fields are combined into one frame or else why do you need to separate them? The frames must be combined or you would not get combing effects. Converting fields into full frames / pictures means you are just doubling the lines so there are 2 lines now for every 1 line to get 60 fps? I'd think that would look bad. I'm obviously missing something in my understanding. I see people selecting 60 fps instead of 30 fps for the initial capture (no not OBS format but lag or huf in an avi wrapper in vdub2), then when they deinterlace they don't choose the double frame rate option like you did. I'm guessing both methods result in the same 60p video at the end of de-interlacing. Recording at 30 fps must really mean recording 60 pictures per second. Each picture having only half the lines with black space between, then these are combined together to display a frame on the PC. Do TV's display only fields as the full frame? Yes they are taken at different times so viewing on a computer monitor does not look as good when fields are combined. @@ConsumerDV
■ "there is no setting to choose between interlaced or progressive in that menu." - Are you talking about OBS or VDub? I don't know much about OBS. Analog video is either 25 or 30 fps, end of story, this is what capture devices support. These frames are nominally interlaced, but can contain progressive content, in the latter case the fields of each frame may originate from the same time. Ultimately, it is the user who makes a decision how to treat the footage: as native interlaced, as 30/25 fps progressive, or as something else, for example as a telecined movie, which originally had 24 fps and requires a different deinterlacing algorithm, inverse telecine. You choose a scene with motion and check whether there is combing. Then you try different deinterlacing algorithms and see how it looks. For standard interlaced video you should see two separate pictures in the right window for every combed picture in the left window. I suggest reading through this long page: hometheaterhifi.com/volume_7_4/dvd-benchmark-part-5-progressive-10-2000.html ■ "I am really trying to understand this but the fields are combined on the TV screen to make a full frame no?" - No. The simplest way of deinterlacing is converting each field into its own frame. All TV sets default to this mode when they don't know what to do. ■ "viewing traditional TV we see only half the frame and there is black space between the lines?" - there is no black space, hence the term "line doubler", now rarely used. Also, depending on deinterlacing algorithm it is not always half the frame. Refer to the link above. ■ "you also said "separate the fields" so that implies the fields are combined into one frame or else why do you need to separate them?" - what I meant is split one frame into fields. In real analog TV fields are pretty much independent, although there is a top and bottom field (or first and second, or even and odd). In digital video, fields are usually combined into a frame for the purposes of compression and storage, but they still can be treated very much as separate entities. ■ "Converting fields into full frames / pictures means you are just doubling the lines so there are 2 lines now for every 1 line to get 60 fps?" - Correct. Twenty years ago deinterlacers were called line doublers. ■ "I'd think that would look bad." - HD looks ok, there is enough resolution for it to look just fine. SD can show stairstepping on diagonals. ■ "I see people selecting 60 fps instead of 30 fps for the initial capture (no not OBS format but lag or huf in an avi wrapper in vdub2), then when they deinterlace they don't choose the double frame rate option like you did." - I don't know how this would work, but if the device driver allows them to select this setting, then apparently the device deinterlaces the video internally. ■ "Recording at 30 fps must really mean recording 60 pictures per second."- yes, 30 fps INTERLACED means 60 pictures per second, each picture is a field. 30 fps PROGRESSIVE means 30 pictures per second, each picture is a full frame. ■ "Each picture having only half the lines with black space between, then these are combined together to display a frame on the PC." - For interlaced video they should not be combined, see above. ■ "Do TV's display only fields as the full frame?" - this is the simplest deinterlacing method, "dumb bob". Which is why 15-20-25 years ago you would hear names like Faroudja, HQV, DVDO, the companies who made quality deinterlacers. Pioneer Kuro plasma TV set is famed for its deinterlacer, I kick myself I did not buy a 50-inch set for $1,600 on sale at Costco back then. Read the article linked above, these guys explain it very well with pictures. The article is a treasure.
The link provided does not detail how interlaced fields are displayed.... and that's where I'm confused. In the first section describing interlaced video there are 3 examples, a man, a tomato & a skier. Each one is displayed differently. The man shows one field of lines and white between them, then the second field displayed with the first field together. The tomato shows the current field with the prior field, over and over. The third shows only one field at a time with white between the lines. So which of the 3 is it? Or does it depend on the TV? perhaps old TV's display it one way new TV's a different? @@ConsumerDV
Fields are displayed in sequence with empty space between the lines of each field on a native interlaced CRT display. The next field is pained over the empty space, while the first one - and this is an important part! - is still somewhat visible, but has started to decay. So in reality you NEVER see empty space. This animation from wikipedia shows how the fields intermesh between each other: each new field is painted 1/60 s later than the prior one, but there is a time when both fields are on the screen: one is being painted and is bright, another is decaying. The careful choice of the decay time provide that the picture is not too blurry yet not too flicker-y: en.wikipedia.org/wiki/Interlaced_video#/media/File:CRT_image_creation_animation.gif There were progscan CRT displays, of course, primary computer monitors, some TV sets as well, but let us not talk about them. On a native progscan monitor, which all plasma (except Fujitsu ALiS), LCD and OLED are, fields NEVER intermesh like on the CRT. The simplest option is to convert each field into a full frame while doubling the lines, this is not optimal for film-sourced content, but at least you won't see ugly combing or ghosting. The first picture with the shot from Casablanca I believe shows Humphrey Bogart - this is a wrong picture, sadly. On a native interlaced TV you will see two fields in sequence. On a progscan TV with proper deinterlacer the video will be deinterlaced first before it hits the screen (have you watched my video about IVTC?), then the COMPLETE frame is displayed for 1/24 of a second, so you will see the full picture. The tomato shows how one field changes at the time, still preserving the prior field. They tried to simulate how the prior field is still visible, I guess this did not work for you. The skier image is another approximation, it does not show the decay, on a real CRT you will never see empty space, it will be taken by the decaying prior field. As you can see, the whole idea of interlacing was based on a very particular display technology: phosphors of CRT, when bombarded with a beam of electrons, light up and then dim relatively slow. If a TV set chooses a wrong deinterlacing method and say combines fields that do not belong to the same time, then you see combing. You will not see interlacing, as interlacing is a TECHNIQUE that works only on an interlaced monitor, you will see combing as an ARTIFACT caused by the interlaced nature of the signal and by the incorrect deinterlacing. While we are on the topic of attack and decay, plasma screens also have this feature: the tiny elements light up and and dim not immediately, but have some decay time, this is why some people prefer plasma screens for more organic look. I love my 50-inch plasma, I wish I bought the Pioneer Kuro. Early LCD panels had very low refresh rate of 60 Hz or 120 Hz, and the picture would appear and disapper almost immediately (I hate LED blinkers on cars, they have the same immediate character, compare them to good old incandescent blinkers, they light up and dim relatively slowly), so when you have motion, you would get visible stutter, or sort of "stickiness" in the picture. This has been remedied by raising LCD refresh rate to 480 Hz and then to 600 Hz, so they are flickering all the time at a very high rate, and this flickering works better for our eyes. In a way, next frame sort of blends with the prior frame instead of simply being shown after it.
Great video, thanks for posting. I used VirtualDub a week ago for the first time and it worked fine. Now, after reinstalling Windows 10 and 7, I cannot get it to work properly. I do not see "capture pin" and any codec. I tried everything, but I cannot get the codecs to show up? Any ideas? Thanks.
Thanks! All I can think of is make sure all device drivers are installed properly. As for codecs, VirtualDub comes with some bundled codecs, and picks up other ones installed on your system.
@@ConsumerDV Hey, thanks for the fast reply. I know that VirtualDub comes with codecs, which makes the whole thing even more weird, because If it comes with codecs, why can't I see them? I noticed that when I don't have a capture device plugged in, I can see the codecs. I can also see them when I capture a video, using NO codecs, and then open the video in VD and export it. Now I can pick a codec. Weird. Thanks.
What is the best way to Capture a DVI Video under windows 10 and Firewire? I want to use my pinacle Movie Box DV. Got it wokring with a Dell Firewire Card and VLC. Are there any better oprions?
As long as your software can capture DV stream, there is no difference in quality: you are getfing the data recorded on tape exactly as it is. I use WinDV, but sometimes capture with Vegas.
Hello! im having a problem. my capture pin setting doesnt let me to change anything and the crossbar option is grayed out. i bough this UCEC USB 2.0 Video Audio Capture Card Adapter. so i wonder if its just a physical limitation of this hardware or i did not have the proper drivers.
A lot of these dongles have similar appearance, but may have different hardware. I bought one like this, and just like in your case it would not allow me to save into anything but MJPEG, and the video was being deinterlaced by the hardware. Capturing neither with 3rd party tools like VirtualDub nor with the bundled software helped. I ended up returning it. With no-name products you never know. If you are in the U.S., the Diamond VC500 is around $30 (it is about £50 in the U.K., wow), and it is a decent dongle in my opinion, although I had to fight with the drivers, also I've heard that the hardware has been changed recently, but I cannot verify it, so take this recommendation with a grain of salt. Here is a quick review: czcams.com/video/j3VMQMEOWko/video.html
Thanks for this excellent tutorial. Everything is well except when I play the captured avi file, the video pauses briefly every few frames or so. The audio remains synced and is not affected. During capture I am showing no dropped frames. Do you know what my problem could be?
Hi, I am glad that it was useful! What codec and bitrate are you capturing into? Some codecs are "heavy" and need more processing. Also, higher bitrate also needs more processing. It is a balancing act. Does it stutter in VirtualDub or in an external player? What player do you use? Are you on Windows? For AVC/H.264 in TS/MP4 container I can recommend Mirillis Splash, it is very fast: mirillis.com/free-hd-video-player For codecs that are often used by VirtualDub users like Huffyuv, Lagarith, Cinefom I can recommend VLC. I like Cineform because it is not very taxing, but GoPro does not make it easy to install just a driver. VLC supports Cineform natively. Many players play faster if you make the player window smaller.
@@ConsumerDV I am on Windows 11 playing back the captured video in both VLC and VirtualDub. I am getting stuttering in both. I have tried making the window smaller with no luck there. This is all new to me so I followed your example exactly, using the Cineform codec and compressing to mp4 after cropping etc... EDIT: It seems the incoming video feed during capture is what has the stutter, so the software seems to be doing what it should. Maybe it’s my camcorder? Or it’s another problem between the camcorder and the pc? I am digitizing Hi8 tapes using a Sony Handycam CCD-TRV138 using the Pinnacle Dazzle DVC100. The camera only has a single 3.5mm A/V output to composite. Thanks for getting back so quickly!
You can play on the camcorder and check on the flip-out screen or on a TV to be sure whether the stutter is present in the original video. If not, and you see the stutter during capturing, maybe the chosen codec and bitrate are too taxing for your system, although considering that you have a Windows 11 system, I assume it is rather new and fast. During capture, check CPU and HDD load on the right side of the VDub window. If the load gets close to 100%, you may lose frames. Check the number of lost and inserted frames, the fewer you have the better.
I literally just got the DVC100 and had the same issue. I found a different driver that fixed pretty much everything! Here's the link to the website. forum.videohelp.com/threads/398965-Dazzle-DVC100-not-capturing-anymore-on-Windows-10#post2603674 If you can't find it, here's the link to the driver itself. forum.videohelp.com/attachments/56186-1607688052/Dazzle%20Drivers.zip
Hi, thank you for the helpful tutorial. In my Hi-8 footage, I get lots of horizontal white lines appearing and disappearing all the time. I think this was due to the heads being dirty or worn at the time. Can you recommend a way to remove this in VirtualDub2 or any other software?
Hi! The principle is the same, but, sadly, I cannot help neither with capturing devices nor with software, I am a Windows guy. AFAIK, there are no Mac drivers for the Dazzle.
@@ConsumerDV Thanks for the reply, I enjoy your videos a lot! I'm just starting out in HI8, its fun but I'm surprised at how tricky getting a decent capture is for me. I just found the open-source software Handbrake, and I'm hoping that with a solid initial capture, this program will help with the artifacts (and high levels of analog distortion) that I suspect is coming from either my capture device trying to deinterlace as the signal is coming in, OR it is user error due to my capture settings.
I put all configurations that you showed in the video but in the end when I save the video it appears one error message: “codec not currently supported in container mp4.” What should I do ?
Usually you capture into a lossless or visually lossless codec (I use Cineform), then you re-encode into a delivery codec, which is usually AVC/H.264 nowadays. I don't have a computer right here at the moment, but if I remember correctly, VDub2 cannot store Cineform in MP4 container, it wants AVI. Use MP4 container for your delivery when encoding into AVC.
Я так и не понял, обязательно ли надо приводить пиксели к квадратному виду? Разве не все плееры\устройства могут понимать какой пиксель в записи и выводить уже в нужном масштабе и без искажений? Как по стандарту эта штука с неквадратными пикселями работает\должна работать?
Не обязательно, но улучшает совместимость. Если вы будете загружать на ютуб, он все равно сконвертирует в квадратные пиксели. Старые контейнеры вроде AVI не имеют стандартного способа указывать пропорции пикселов, но в новых контейнерах проблем нет, и компьютерные плееры нормально проигрывают эти файлы. Главное, не забыть указать правильные пропорции при генерации файла.
Great video, thanks so much for posting this, love de-interlacing 29.97i to 59.94p, I think I'm going to use QTGMC. Are you using the DVC100 on a Windows10, 8.1 or 7 OS? I've heard a lot of Win10 systems drop frames after a certain Windows update point... I'm looking for a decent workflow and doing a lot of fact finding to get the best flow for digitizing VHS/C and Video8 footage using the DVC100.
I am using Win7 64. QTGMC usually works fine. You can have a small script with it loaded into VDub and use other filters, or you can go fully Avisynth route.
@@ConsumerDV I'm looking into using older work laptops perhaps. LENOVO LS500, they actually have built-in mini firewire 400 port, wont use the FW ports for VHS capture, but was neat the laptops had them, i can use my DVX100b and my Sony to capture Hi8 and MiniDv using the FW port. I may bulid 2 laptops with windows 8.1 and dedicate them for capture and de interlace machines. I'll read up and see stability of OS 8.1 64bit with the DVC-100. Again thanks for all the postings and yes I'm going to de interlace with Avis script, QTGMC and VirtualDub.
Two of my MiniDV camcorders and a Digital8 camcorder can convert analog to digital and output as DV: czcams.com/video/Hbytp67EYAg/video.html This may be an easier catch-all option if your camcorder can do it. I personally like DV codec, other people hate it, but it is still well supported. I prefer desktops as they are easier to upgrade and modify.
@@ConsumerDV Thanks so much again for the reply, I love video cameras, filming, editing and playing around with footage. On this note regarding MiniDv, I have Canopus 110 Firewire DV converter (DV-25 Codec), and yes I can do a pass through via my MiniDv Camera (Panasonic DVX100b, Sony TRV hi8 model) I've done a few tests and feel the footage looks nasty even when passing though via S-Video. I tried in comparison a cheap EZCAP via VirtualDub and honestly the footage looked like it had more bit depth than when captured on FCPX via the Canopus 110. I think in completing research via your youtube videos, my issue is that it's not uncompressed. I'm starting over with a new work flow and trying your Dazzle route with VirtualDub2. I agree with using a desktop, however space is an issue, I don't have a room I can dedicate at the moment for a capture suit. Again thanks for the videos, I found a Dazzle DVDRecorder HD from 2011 that's a white DVC100 Rev 1.1 for $15 CND. Got it yesterday and just today bought a red DVC100 for $15. I think it would be great to have a backup. After I find an import and archive work flow, I'll then start the QTGMC de-Interlace method for nice 59.94p footage of my old home video VHS collection.
Thanks! Try connecting your video source before starting up VDub. Connect the cables, turn the VCR on, start playing something. When the device/application senses the signal, it usually automatically switches to an appropriate mode or at least shows it as a default, then you'll be able to tweak it if necessary.
Great explanation and walkthrough - Thank you - One question when i went to MPEG Layer3 for audio the dialog box was empty despite downloading the Fraunhofer MPEG Layer-3 Audio Decoder - Thanks
Thanks! Some of the compression codecs come with VirtualDub, other are pulled from your system. I need to do a clean install to figure out which is which... Presently, for MP3 I have FFMpeg Lame MP3 with just one preset, LAME MP3 with lots of presets, MPEG Layer-3 with no presets at all, and another MPEG Layer-3 with lots of presets. I have installed FFMpeg and Lame MP3 separately at some point, so they have been pulled by VirtualDub. I think that if you install Lame MP3 package, the codec will be picked up by VirtualDub. 64-bit VirtualDub may not pick it up, but 32-bit version should.
WinDV copies DV video from a DV or Digital8 camcorder over FireWire exactly as it has been recorded on tape, and stores it in an AVI container, see more here: en.wikipedia.org/wiki/DV#File-based_media Could you tell more, how do you try playing your videos? The DV-AVI format has been supported by Windows for at least two decades. Windows Movie Maker supports it, and pretty much all video editing tools support it as well. I use Windows 7, so I cannot comment whether Windows 10 still plays this format, but I don't see why would it stop supporting it, it is still very much alive. In any case, there are zillion of other video players, for example Media Player Classic Home Cinema (MPC-HC) or VLC Media Player, maybe try them.
I was able to use this method to attain the same results! I also tried capturing the same footage using a Framemeister HD Upscaler (often used by streamers to upscale retro games in 240 or 480i) and Elgato HD60 (HDMI capture) also 60p. Here is a comparison video I made showcasing them side-by-side: czcams.com/video/WnEyspXHlfo/video.html Let me know what you think!
Nice! The VDub version is more saturated, possibly because you grab with 4:2:2 color subsampling. On the other hand, if you look at the diagonals, Elgato is no worse, and dare I say, maybe even better: i.ibb.co/g9HFYhM/video-signals-saturation-deinterlacing.jpg All in all, they are quite close if you ask me. If you want to stay with VDub, you may want to try MSU Deinterlacer filter, which in my experience is just a little bit better than Yadif, but an order of magnitude slower. Like many older filters, it works only on 32-bit VDub. Or, you may want to graduate to QTGMC deinterlacing script, you can run the script in VDub. I used QTGMC for 1970s music videos that I uploaded in 50p. I raised Elgato saturation to match VDub, the saturation "blob" on the vectorscope moved more into the red zone, so the image became a bit reddish: i.ibb.co/NLzRjJF/video-signals-adjusted-saturation2.jpg You can match it better, of course. So, all in all, Elgato is not that bad :) If you make a video about using Elgato to produce 60p videos, I'll watch it :) Or maybe you can just write a short how-to, because I thought that Elgato captures either into interlaced stream at best, or into 30p at worst.
Have you ever noticed captured 29.97i footage in VirtualDub2 have the background and foreground focus change every frame? Exaple frame 1 and 3 the foreground is in focus and frame 2 and 4 the background is in focus. Not sure if you've ever experienced this? This may cause the horizontal line jitter. As everyother frame is soft in that subject area.
I did see fields of a frame sort of flickering, when one field was slightly dimmer than another. But not frames. Do you have manual or auto focus? Although this would probably be too quick for focus change. Have you joined doom9.org or videohelp.com? Some knowledgeable people there. I mean, they are less into camcorders, but more into digitizing and converting from one format to another.
@@ConsumerDV Thanks very much, I may join those groups, I have been there reading posts and tips. Great resources, I'll maybe create an account and join. The camera has auto focus, it's a Panasonic DVX100b MiniDV, but it fluctuates every other field when separating into frames. What was your capture OS again, Windows 7 or XP using DVC-100 Capture hardware into VirtualDub2?
I am using Win7-64, but I don't think it has anything to do with this fluctuation. BTW, I was able to get rid from tons of inserted frames by deselecting "Enable audio playback" menu item.
@@ConsumerDV Good morning again, thanks for the reply. I believe you're correct about the OS, it was just a side note I wanted to know about what OS worked for you. The reason I ask is that I'm trying to keep consistent with what's working for others. I had created a fresh install of Win7 with VirtualDub2 for capture. It all worked to capture on Win7 SP1 with the DVC-100. I then wanted to make a fresh install disc with all the updates czcams.com/video/l5ADP-VZMsw/video.html So I started to update the machine and then after all the updates the Dazzle wouldn't connect. So last night I ended up reformatting the system with only SP1 on Windows 7, and low and behold the DVC-100 is now again working. So I'll not bother with creating an install version of Win7 with all the updates and rollouts. Again I really appreciate the insight from other like your self. I've been doing a few test captures of the same 1 min footage to see the best route. My source footage was shot on MiniDV, I did three tests. Captured directly to FCPX on my mac from the camera firewire (DV25), then captured in FCPX from the Canopus ADVC-110 which is analog Svideo to the DV25 codec in FCPX, again still used the MiniDV tape but used S-Video out from the camera to line in on the Canopus ADVC-110 which digitizes to DV25. Then after that I used the same original MiniDV footage from SVideo out of the Camera to the Dazzle DVC-100 and captured in VirtualDub2 lagarith codec on the Win7 machine, I did one more test just to see how bad the EZCap was, I again played from my MiniDV camera S-Video out to line in on the EzCap and captured using the lagarith codec. I'm now going over the clips comparing and that's when I notice that all methods have every other field slightly blurry, again field 1 will have the foreground in focus and field 2 will have the background in focus. It creates a slight flicker on certain objects. Not sure why this is happening. It's not the auto focus. The test shot I use, has the camera still on a table. I'll keep playing around and see what I can come up with, I didn't think digitizing would be so time consuming. I'm still going to buy a TBC. Thanks again for all your tutorials.
I forgot, maybe you have mentioned it already - do you have an older Mac with a Firewire port, or do you convert from Firewire into Thunderbolt? As I said, I noticed slight difference in color, not in focus, between two fields of my Canon Elura 100. I tried... I think a temporal smoothing filter... not sure. Not the one that is bundled with VirtualDub2. Sorry, I don't remember. It worked if the scene did not change drastically, but on scene change there was a funny effect that it tried to continue with the settings from the previous frame, looked ugly. I think I found a balance when it looked ok even with a scene change, I need to check which plugin did I use. I am not big into Avisynth, but maybe I need to dive into it for more powerful functionality. As for TBC, I think I am going to stop where I am now. So yeah, I am glad that something I put up on CZcams was useful, and for more powerful scripting stuff definitely check those forums!
Update: I finally got a Dazzle DVC100 (red color model) for $12.50, and I had to jump through a few hoops, but thank God it is fully working in action! Only thing I am having some extremely minor trouble with is that there is some ghosting during movement, mostly if the source material isn't 60p. It was worse when I had these levels filter added on Virtualdub. To be honest, it's not even that noticeable unless if you squint hard enough. Is there a fix to this? If not, I am just going to roll with it. Other than that, I have got to say that the quality is absolutely outstanding compared to my other cheap capture cards that I had in the past. I don't know why I didn't buy one of these in the first place! Thanks for the absolutely great and informal tutorial!😁🖒
If your tape is NTSC, deinterlacing to 59.94p will give you the same "look" as on TV. Similarly, if your tape is PAL/SECAM, deinterlacing to 50p will give you the TV "look". In some cases you can do better. For example, if you are capturing a Hollywood movie from an NTSC tape, then you may want to use IVTC filter instead of deinterlacing, see a short example here: czcams.com/video/DneoPwa5faY/video.html If it is a movie on a PAL/SECAM tape, you may want to treat your video as 25 fps progressive without using deinterlacing filter. If it is a European movie on an NTSC tape or vice versa, or if it is a cartoon, then more elaborate schemes of restoring the original frames may be needed, and I am not an authority on them, you may want to try forums like doom9 or videohelp. Case in point: I have a travel video that was shot in Europe, it is a mix of film and video, then it was transferred to NTSC tape. If I deinterlace to 60p, some frames still have ghosting. Another case: "Portrait Of an Album" with Frank Sinatra - it was shot on film at 24 fps, converted to video and edited on video. The effects in between the songs are made in video, so they are 30 fps interlaced. I used IVTC, so I restored the movie to 24 fps, but the effects now look ugly. I could deinterlace it to 60p, and it would look fine, but I would lose a bit of vertical resolution. Or I could process the effects scenes separately, deinterlacing them to 60p and then converting to 24 fps, throwing out extra frames. It is a black hole, once you start fiddling with it, it sucks you in :) Regaring the Dazzle 100: I like it for the overall decent quality, and that it does not care about Macrovision. But it may have hard time with timing errors, so you may see some "skewing" or "hooking" on top of the frame, especially if the tape is recorded in LP or EP mode. See comparison with another card here: czcams.com/video/bhZyk72I9Q4/video.html I did not mention in the video that the other card have some other issues like inter-field jitter, also with A/V sync. No card is perfect :(
Sorry for just now responding, been quite busy lately. Anyways, my footage is NTSC, and I'm working with tapes that have different frame rates (like your tape of Wheels: The Joy Of Cars for example). Edit: You're not going to believe this, but I figured out how to rid the ghosting. Apparently, I had to disable the noise reduction. Lesson learned: It always pays the tweak with every single setting in VirtualDub! I'll have to recapture four of my tapes again, but it will be worth it (luckily most were 30 mins anyways)! I will use the IVTC Filter for 24p only tapes, thanks for mentioning it!😁👌
I followed the same setup directions in VirtualDub2 with a dazzle dvc100 but I am not getting any picture inside of virtualdub - just a black screen even when the camera is playing the tape- any advice?
Make sure that the device is selected in the Device menu. Also, that in Video menu "video source" is correctly selected between composite and SVideo, that "capture pin" has correct video standard and frame rate, that "capture filter" also shows correct video standard, also try switching between "Overlay" and "Preview (system)", on some cards the latter works better, although I did not have issues with the DVC100. Also, try start playing video first, then start VDub and configure it - sometimes it better senses the incoming format when an actual video plays. I hope it works out for you!
@@ConsumerDV I've tried 2 different dazzle dvc100s with these settings, but I'm just getting a black screen inside VirtualDub2. I wonder if I need another driver installed? I have tried a couple already. I wonder if anyone else here has had this issue and was able to solve it...
Hey, quick question. Does the Dazzle capture card drop frames or blank out if there is either no signal or when a worn out or damaged VHS tape is being played? I have a generic Easy VHS to DVD USB card that will do that sometimes, but it will capture in interlaced MPEG1 files which gives me the opportunity to convert them to 60p in Handbrake. Quality isn't that bad imo, definitely better than a cheap Easycap USB card. Almost thinking of upgrading to A Dazzle card like you have. Thanks!
Hi, the worst I had was something like this, czcams.com/video/qs8KF1G_mgs/video.html (from 1:25 to 1:35), this dropout did not stop the capture. I haven't checked how many frames were dropped. Overall, AV sync with the Dazzle floats about 0.2-0.3 ms here and there, but usually it self-adjusts, so it never drifts away too far. To some, this much drift is a showstopper. I don't know whether the Dazzle 100 is better than an Easycap that can capture true interlaced. The good thing about the Dazzle is that it can capture from Macrovision-protected tapes, on the other hand it does not have any time base corrector functionality, so if your video exhibits skewing, the Dazzle would not correct it. The Avermedia card is better in this regard, but it has other issues. So far I haven't come across a device which is both cheap and ticks all the marks. OTOH, a used Dazzle 100/101/103/107 will cost you probably less than $15 on ebay, so maybe worth a try.
@@ConsumerDV Yes yours did better than mine would. Might give the Dazzle a try! As for macrovision, my existing usb card does pretty good, plus I have one of those Digital Video Stabilizer boxes that's powered by a 9V battery, and it helps to remove skewing in the picture if there is any. Thanks for the helpful info!👌
I was under an impression that "stabilizer" is a simply an euphemism for a macrovision stripper, but if it also helps with skewing, this is even better! I may look into getting one myself.
@@ConsumerDV Please consider it, I found mine on eBay in late 2012 for $35, and I think that's what they still go for. Might vary by tape, but to me, I think it's still a good investment for those that want to archive VHS tapes that haven't got a proper DVD release and capture it in somewhat decent quality! Replacing the 9V battery is pretty easy at least on my particular unit, just remove 4 screws, pop off the top cover, then you have access to replace the battery if desired. I replaced mine in 2020 just for good measure👌
Uncompressed takes the most space, compressed but lossless takes less space, visually lossless intermediate codecs like Cineform or DNxHD take even less space, then delivery formats like H.264 or H.265 are lossy and heavily compressed but take the least space.
No guide, sorry. Some of them come with VirtualDub, others I installed separately with their respective installers. I think X264 - an implementation of H.264 - comes with FFMPEG package. If you want MPEG-2, I believe you can get it with HCENC package.
When capturing from the Dazzle in VirtualDub2 at 29.97i 720x480, during capture it displays average frame rate. It shows live stats between 29.905 - 30.00, after capture when checking the info with file manager properties or even in VLC it notes the frame rate at 29.97. However when I bring the footage into a video editor is will display the frame rate at 29.95, even using quicktime to display the clip information it shows at 29.95. I've done more than 8 tests of capturing and each time the frame rate is slightly different, some of the rates are even 29.84, I feel this is problematic if trying to sync footage, in addition in FCPX it renders the time line as soon as it's imported because it's the frame rate is not 29.97. I just wanted to reach out as I know you use the Dazzle to capture and was wondering if you've seen your clips information? Not sure what editing programs you use. Again thanks very much for all of your posts and replies. Regards; Phil
I am not a professional video editor, so I don't know full details on how strict all the numbers should be. I edit in Vegas, and whenever I feel there is an A/V mismatch, I just correct it on the timeline either by stretching audio (usually it is spoken words, so I don't care for pitch change), or by shifting it. I do see fluctuation of frame rate when capturing with the Dazzle, so far it has not given me serious problems, usually it does not drift apart too much. I am capturing 2-hour tapes now, and they look reasonably good to me. OTOH, with another card that shows zero drift and exact 29.97 frame rate I get 3/4-second drift after one hour of capture. Frankly, I don't know how you can get the exact frame rate without genlocking the VCR, and consumer-grade VCRs cannot be genlocked. Even if an external TBC is used, it would try to match incoming rate to TBC's clock, and if the video continues to drift, it would drop or insert a frame. AFAIK, the only way to do it perfectly is to use genlock or to use a TBC that loops back to the VCR to give it a command to run faster or slower, which is basically the same as genlock, but for that you need a pro VCR. TL;DR I don't pay too much attention to that as long as the A/V is more or less in sync, and there is no visible frequent frame dropping. To me, the absence of combing and proper aspect ratio is already halfway there, but those who adjust every possible setting in their scripts for a perfect capture will not agree :) Ah, the actual question about fps on clips - I will take a look and will get back! :)
A file I captured yesterday with the Dazzle. I skipped to the very end to check A/V sync, looks good to me. MediaInfo: General Complete name : H:\My Videos\American Century\Tape 3, U778, Dazzle.avi File size : 60.4 GiB Duration : 2h 38mn Overall bit rate : 54.7 Mbps Video Format : CineForm HD Duration : 2h 38mn Bit rate : 53.2 Mbps Width : 720 pixels Height : 480 pixels Frame rate : 29.970 fps Standard : NTSC Vegas: Video: 02:38:04.141, 29.970 fps, 720x480x32, CineForm Codec Audio: 02:38:04.348, 48,000 Hz, 16 Bit, Stereo, Uncompressed
@@ConsumerDV I wanted to thank you for the reply, I read the reply last week when you posted it and I wanted to do several tests to confirm before replying back to you. I tested capture in both versions of VirtualDub one and and two with either HuffyUV or Lagarith codec, I even bought a DiamondVC500. In doing a test capture with both the DVC-100 and VC500 when opening the captured footage in a video editor like Final Cut Pro or even viewing the files in QuickTime to just use the media info, the footage frame rate was never consistent 29.97 it would always be below 29.97. I've discovered that the Windows 7 media info or VLC media clip info will show what the capture setting was set too (29.97) but it won't display what the actual frame rate is at. For myself this is a problem as when brining in footage at a different frame rate will change the length of a clip and for multicam footage, everything will shift very quickly and when dropping onto a time line 29.97 time line it will have to render on every single edit. If you're intrigued, just download the last version of QuickTime, it won't play HuffyUV or Lagarith but you can view the clip info and mine for example was 29.78. I have copied the info from Win Properties/File Info, QuickTime and VLC. Windows 10 Properties/Details Video Length 00:00:26 Frame width 720 Frame height 280 Data rate 164704kbps Total bitrate 166240kbps Frame rate 29.97 frames/second Audio Bit rate 1536kbps Channels 2 (Stereo) Audio sample rate 48.00 khz Quick Time Movie Inspector Movie FPS 29.78 Playing FPS 29.78 Data Size 523.61 MB Data Rate 166.08 mbits/sec Duration 00:00:26.44 Normal Size 720 x 480 Current Size 720 X 480 VLC Media Info Video 720 x 480 Buffer dimensions 720 x 480 Frame rate 29.970030 Orientation Top left Channels Stereo Sample rate 4800 Hz Bits per sample 16 I have done more testes and have used amarectv 3.10 capture with both the DVC-100 and the VC500 devices and they have generated steady consistent frame rate of 29.97. I also used amarectv as a passthrough to VirtualDub and recorded into VirtualDub with the HuffyUV Codec and it worked at 29.97, even during capture when looking at the average capture it would stay steady at 29.97. As mentioned I need true 29.97 footage, you may not need that. Sorry it took so long to get back to you, I was literally working on testing every single night after work, 3 hrs a night. LOL Thanks so much for you videos Regards; Phil
Hi, I appreciate the info! I've heard good things about amarecTV, but I have not used it myself. I should try it out, thanks! How do you use it as a passthrough to VirtualDub? Is it because not all codecs are available in amarecTV?
This depends on the encoder that you choose, it is all up to you. For example, I prefer Cineform because it is a high-quality intermediate codec that was very popular 10-15 years ago among pro videographers. Its bitrate depends on the quality setting and on actual video being compressed, it floats around 30-200 Mbit/s, so I get files about 16GB/hr with 35 Mbit/s for a standard def video, and in another case 58 GB/hr with 130 Mbit/s for HD video. If you choose uncompressed, the size will be even larger. Yep, these are the data rates. Choosing a space-efficient encoder like H.264 or H.265 can cause dropouts, because these encoders are more CPU-intensive. Depends on your machine. When you are capturing with VirtualDub, there is CPU utilization indication on the right side. Also, codecs like Cineform, DNxHD, ProRes are better suited for further editing, they are intermediate codecs. H.264 and H.265 are for the final encode. Anyway, after you make your final encode, you can delete the intermediate files :)
It is JVC GR-SX950 that you saw in another video. It is better as a playback device (SVHS-ET, TBC), but as a camcorder it is not that good. I think the sensor has not enough resolution.
this seems a lot of effort for some old not so hot (boring) video's. I'm amazed that anyone will try this. But good luck. I'll just trow these old videos in de dumbster. :-)
@@bigkwhit Good you found a working configuration. I use primarily Windows 7, but I tried running it on Windows 10 and it worked, but I saw quite a few frame drops. After I turned off audio playback during capture, it stopped dropping frames. I haven't tried on Windows 11, so thank you for the heads up!
There are many ways to ruin your video. Many NLEs have effects for "80s look" or "analog video look". Basically, add some jitter, sparkles, dropped frames, dull color, etc. For best effect you need to start with a video shot on a CCD-based camcorder. If all you have is a smartphone, don't shake it too much - rolling shutter artefacts are the telltale signs of a CMOS-based camera, and all Hi8 camcorders have CCD sensor, which has global shutter. Also, many effect plugins fail to reduce chroma resolution, so you don't see ugly color spills that is inherent to analog consumer-grade video.
@@ConsumerDV all I have is smartphone bro . I'm trying to upload my vlogs in my instagram with Hi8 style but I don't quite understand well what you said . Is after effect good program or not ?
Rec. 601 stipulates that 704x576 for 50 Hz ("PAL") and 704x480 for 60 Hz ("NTSC") has 4:3 proportions. It is up to you how much to crop, the goal is to ensure correct proportions of the image itself. Maybe this can help: czcams.com/video/6SqbwZdzx3g/video.html
@@ConsumerDV when we crop the video around the edges, does the quality drop ? When a video is upscaled, does the quality drop ? For maximum quality after capture, the video does not need to be modified ?
Quality drops from scaling. Cropping and scaling are different and independent operations. Scaling is inevitable when you watch on youtube or on TV. The best rule of thumb I know is not to scale like 2% or 5% or 10% So scaling something like 698x476 back to 704x480 with PAR 10/11 or into 640x480 with PAR 1 will reduce vertical detail significantly. In this case it is better to pad, that is to add black border, than to scale. Or scale to something larger like 960x720.
Yes, it can send DV video over USB at full resolution. You will need software that is capable of receiving this video, Win10/11 does not have built-in tools for it. I installed Windows Essentials 2012 on a Win10 laptop, and Windows Movie Maker recognized the camcorder, Windows automatically installed a "DVC device" driver, and I was able to capture DV video over USB. There were 80-something dropped frames at the start of the capture, no more frames were dropped during the capture, so basically I lost 3 seconds in the beginning, which I think is acceptable for capturing a 1-hour tape. See more here: czcams.com/video/LGfQpY__F3U/video.html
For some reason every time I capture video the average frame rate is around 21,22 or 23 fps depending on settings, it fluctuates quite a bit. I have it set to 25fps (UK PAL VHS tape) Over the course of an hour of capture, the sound becomes extremely out of sync when played back. (I assume this is related to the fps issue) Is it possible the dvc100 I have is just faulty? I have tried many times with various settings and keep having same problem. I have read it could be a problem with my HDD write speed also? I have set timing settings to NOT resync audio by adjusting video frames, maybe this is wrong, but I tried other way and sound was affected. I think maybe its putting in to many frames which is causing the video to become unsynced with audio? But it should still "record" the capture at 25fps right? Thanks in advance.
I don't have this problem. Do you use VirtualDub for capture? Try disabling audio playback during capture in Audio menu. Other people use AmarecTV and say that it is better in terms of A/V sync.
You need Firewire to transfer DV and HDV video. Few DV camcorders can transfer via USB. A FireWire expansion card costs about $15, easy to install. If you have a laptop, you may try converting to Thunderbolt, I haven't done it, but there are videos on CZcams that show that this conversion works. Apple-branded dongles are expensive.
@@ConsumerDV Thank you! I discovered that my camcorder only has a P2 to RCA connector, how could I record it? Using a VCR to a TV connected to a computer?
So you have a 3.5mm socket in the camcorder, and 3.5mm to RCA cable? Then you use these RCA connectors in almost any analog-to-digital converter, they all have RCA for composite video, and dual RCA for stereo audio, some of them also have SVideo input.
15 years ago this was the norm: a bunch of audio ouputs, audio input, video composite, S-Video, Firewire, USB, also a TV/FM receiver card... yeah. This is a 2008 machine. You can find a refurbished one for about $150. For my other desktop that I built myself I bought a card and installed it, it is very simple. A Firewire card can be found for as low as $15, there are no wires, just push the card into a slot on the motherboard, easier than Lego :)
@@ConsumerDV amaizing , I have bought one pci card for capture but it was poor , So , I said to buy one built in same as yours , Is there any name for pc so I can search easily
Regarding Firewire cards, some people prefer cards with TI chipset, they say these cards have better compatibility with older devices. I myself have got a card with VIA chipset, and it works flawlessly for me. I still use Windows 7, and the system was able to automatically download and install the driver. If you are asking about a complete system, then you may want to search for "refurbished Windows desktop with firewire". I built my desktop from parts.
It was a useful compromise during the early days of TV. Its usage for HDTV has been lobbied by Japanese manufacturers, who already had created equipment necessary for their analog HD system. Finally, 4K has got rid of interlaced video. Good riddance.
Interlacing is a _technique_ that meshes video fields meant to increase the perceived image rate and reduce flicker. Combing is an _artifact_ visible when interlacing is not done correctly or when interlaced video is incorrectly deinterlaced.
*Subtitles are available.* I want to clarify that I do not suggest using Pinnacle Dazzle, but if you decide to use it, then some models work better than other, see this video for more info: czcams.com/video/RVbqGrf4rIQ/video.html
If your video is too "wavy" or is skewed on the top, you need a time base corrector (TBC). These devices are increasingly rare and expensive. Some capture cards have rudimentary TBC features, see this video for more info: czcams.com/video/bhZyk72I9Q4/video.html Also, some VCRs and camcorders have built-in TBC, like: Mitsubishi HS-MD3000 medical VCR: czcams.com/video/Xoft92VctQY/video.html or JVC GR-SX950 SVHS camcorder: czcams.com/video/qs8KF1G_mgs/video.html
Panasonic NV-FS200B with TBC built-in is good right?
Sure, if it works.
@@ConsumerDV I got one, but it had bad image, so i went and recaped the Chroma and luma module, now i only get black screen 😢 send it to a Repair shop to see if the can fix it if not i will take a shot in my balls lol
Too bad. Let's hope they fix it!
Hi! Thank you a lot for this video. I wonder what capture device you recommend against the Pinnacle Dazzle for a better quality? Thanks
THIS is what I've been looking for! A capture method that preserves as much information as possible without overkill file sizes!!! Thank you!
Just use handbreak.
Excellent, excellent, excellent. 20 hours of my research would have been saved if i found this video first.!!
Great conversion guide, with careful attention to deinterlacing details. As you point out, video pixels are not quite square, with a 10/11 width/height aspect ratio for NTSC pixels. If you apply this factor to the 720x480 frame produced by digitizing NTSC video, you get the 704x480 square-pixel frame described @7:36. However, when upscaling the frame to 720p for CZcams upload, there is an alternative scaling that facilitates cropping off the video distortion at the bottom of the NTSC frame as well.
If you scale the original 720x480 frame up to 1000x736, you can simply crop off the bottom 16 lines to produce a 1000x720 frame. This scaling corrects the 10/11 pixel aspect ratio while preserving the full width of the frame, and can easily be letterboxed up to 1280x720 HD format, or left at 1000x720 for upload to CZcams. This scalling also works well with H.264 macroblock boundaries.
Thanks! 704x480 is not square-pixel, the pixels are still 10/11, but the frame is true 4:3 - 704*10/480/11 ≈ 1.3333
I see your point about 1000x736, it has AR 1.3587, which is close to 1.3636, all the while 1000 is divisible by 8. I am not sure that this size will work for H.264, quick googling returns that base macroblock is 16x16, but maybe you are right, and 8x8 macroblocks are possible.
Alternatively, you can scale 720x480 to 992x728 and crop 8 pixels on the bottom instead of 16. This may not be enough though to get rid of head switching noise.
Or crop 720x480 to 704x480, scale to 976x732 and crop 12 lines on the bottom.
У вас такой потрясающий канал! Видно, сколько сил вы вкладываете в создание роликов
Спасибо большое)
This is brilliant, I love the way you have put this together, it's got just the right amount of information and illustration.
I have recently stumbled across a JVC DVD recorder/VCR combo that has an HDMI output. I'm going to try a little Handbrake & VideoProc to see if I can clean it up more.
sensacional, eu não acredito que iria jogar no lixo a minha dazzle, com essas dicas vou melhorar em 100% muito obigado!
Your capture footage looks amazing.
An incredible video. Thank you very much.
Very helpful video! I've been digitizing VHS and Video8 tapes through a Sony HVR-M15AE that takes analog input from my VCR/camcorder and then outputs it through a firewire cable to my pc. For software i use the ScenalyzerLive 4.0, very helpful for previewing and for joining clips together. The HVR-M15AE automatically uses the DV-codec (it's a MiniDV/HDV deck) so the colors get weird sometimes. The tapes are usually up to 3 hours long so the 50i -> 50p convertion sadly takes forever! My customers have been happy this far, but if theres ever a customer thats more needy on the quality and colors of their digitized VHS i will come back to this video surely!
Thanks! Using a DV camcorder as a converter is a foolproof method to digitize analog videos: you get resulting video encoded with an industry-standard codec with interlacing intact. The only reason against it in PAL region may be color subsampling of PAL DV camcorders, which is 4:2:0, that is, samples are taken from every other line, not from each line like in NTSC DV camcorders, which use 4:1:1 subsampling. But considering all other benefits of DV I think think is a reasonable price to pay. Regarding conversion to 50p, it seems that QTMGC is everyone's darling, but I wanted to stay within the boundaries of VirtualDub, and Yadif is not too bad. I also tried using MSU Deinterlacer, it is slightly better, but an order of magnitude slower, I would not use it if I were doing it on everyday basis.
Thank You for this process. I think it would be useful for someone to explain the import and export process for video files viewed on a computer or being uploaded to a service such as Utube versus authoring a DVD. I have home movies that will be DVD authored because I have a brother in a nursing home where it is easier to just bring a small DVD player to connect up to the TV. I also have parents to old to know how to deal with a computer so yes DVDs still are useful. When to Deinterlace or leave interlaced? When to upscale and when not to upscale. Again thank you.
DVD-video format uses MPEG-2 a.k.a. H.262 (you can read more about DVD-video on Wikipedia: en.wikipedia.org/wiki/DVD-Video). VirtualDub can read MPEG-2 but cannot write, so you need to use another tool to convert from whatever format you have into MPEG-2, then you need to author a DVD disc. For DVD, you should keep interlaced video as interlaced, because DVD does not support 50p/60p.
The easiest option is to use commercial software to capture from tape directly into MPEG-2 and then author a DVD disc. Just make sure that the hardware/software you use does not deinterlace the video.
If you want to capture tapes in the best possible format for preservation, then you use tools like VirtualDub or AmarecTV to capture into lossless or visually-lossless codec, and then create deliverables for different platforms: for DVD you would re-encode to MPEG-2 and author a disc, for CZcams and computer viewing you would deinterlace into 50p/60p and encode using H.264 or H.265 codec into something like MP4 or MKV.
Thank you so much for donating a person’s most valuable asset, TIME, to address my capture process and advice on capturing and DVD authoring. Yes understand the deinterlace before resize as not doing so will make deinterlacing after resize impossible due to changing line spacing in each field of the frame.
Do you have a website or email address I could donate at least enough to buy a cup of coffee? Live long and prosper, thank you
I am using Sony Hi8 connected to a Sony DV camera using pass through, connect to computer using Firewire, captured using Virtualdub2.
If I understand this work flow then to make a DVD I would...
1. Initially keep the frame size at 720x480i. Time marker 2:35
2. If making a DVD I do not add a deinterlace filter, keep capture interlaced as this is the NTSC standard for DVD and what the camcorder recorded. Time marker 5:25
3. I can trim the over scan and eliminate the head switch artifact with two separate crops. Time marker 8:44
4. Can I or should I resize the frame back to 704x480i? Add resize (upscaling) filter. Time marker 10:08
5. The captured file is then imported to NLE (I use Resolve), edited, exported as MOV.
6. Imported into Pinnacle Studio for DVD menu creation and burned as DVD.
7. DVD player will handle upscale and the TV deinterlace.
8. Optionally, I can upscale and convert the MOV file using NLE to 1440x1080p (progressive), export as MP4 for viewing on computer or Utube or TV if supported.
Thank You
I haven't used VDub to capture DV, but if it works, great. Usually I use WinDV, but the bytes are the same.
You don't need to crop anything for DVD. DVD supports both 704x480 and 720x480. It is best not to resize back to 480 lines. If you don't like the head switching noise, you can mask it by cropping, say, 4 lines at the bottom (use "Crop" filter), then by padding 4 empty lines back (use "canvas size" filter with "TC" anchor).
You need to always deinterlace before resizing.
Remember, that common commercial tools can capture directly to MPEG-2 and you probably can use the same tool to author a DVD.
If you decide to stick with capturing into lossless (like Huffyuv) or visually-lossless (like Cineform) codec, then the steps would look like this:
For DVD:
1. Capture as at 720x480i.
2. Import the captured file to Resolve (verify codec support). Make sure that Resolve treats the file as interlaced with "NTSC DV" proportions. In my case, Vegas treats Cineform files as progressive PAR 1, so I need to change it in the clip properties. Edit, export as MPEG-2 in whatever container your authoring tool understands.
3. Import into DVD authoring tool to create the menu and the file structure, burn a DVD.
4. DVD player will handle upscale and the TV deinterlace.
For CZcams:
1. Capture as at 720x480i.
2. Deinterlace with doubling frame rate. Yadif is the best VirtualDub built-in deinterlacer. I use MSU Deinterlacer, you can download it separately. It is better but much slower. Other people use QTGMC with Avisynth, it is even better.
3. Crop as needed. You don't have to maintain 4:3 proportions, but it is simpler if you do. After you crop 16 pixels on the sides, for every two lines you remove three pixels on the sides.
4. Resize to at least 720p for CZcams to enable 60p. If you want to switch to square pixels, resize to 960x720 or 1440x1080.
5. Encode with H.264, specify PAR width 1, PAR height 1, save into MP4 or MOV.
If you can deinterlace, crop and resize in your NLE, you can do it there.
In either case, whatever you've captured will be your high-quality master.
THANK YOU! I've been looking for such an explanatory video since 2003
output/target video not playing when im playing the original video, only the input pane is playing
There are playback control buttons on the bottom. On the very left there is a Stop button (with a square), then Play Input (a triangle with "I"), then Play Output (a triangle with "O"). You can use keyboard as well: Space is to play input only, Enter is to play input and output.
Excellent information. Thanks
Thank you for great video.
But I have a question.
Following your guide I captured NTSC analog and created nice master AVI file with 4:2:2 color space.
Unfortunately, every program I used to convert it to MP4 (Premiere, Handbrake and many others) resulted in files with 4:2:0 color space.
Using VirtualDub2 internal x264 converter resulted in some kind of non-standard MP4 file which was unplayable in my Samsung TV.
Files created by all other programs played perfectly.
Is there a way to create x264 MP4 file while preserving 4:2:2 color space?
H.264 4:2:2 is definitely possible. Similar to your experience, when I encoded in 4:2:2, many of my desktop apps would not play it... I think VLC did, Splash did not. I haven't tried playing it on my TV. CZcams accepted it just fine. I keep Cineform 4:2:2 as a master, and render into 4:2:0 for playback. I am sorry that I don't have better and more specific advice. H.264 4:2:0 files rendered from VirtualDub using X.264 work just fine for me.
Hi thank you very much for thus video but what about if I don’t want to upload it to CZcams , still the same steps would work to have a nice image ?
This depends. (1) You may want to digitize your analog video as close to the original as possible to become a master video, a source of all other versions. (2) Or you may want to have it in a form that is easy to edit. (3) Or you may want to have it in a form that is compatible with consumer formats like DVD or Blu-ray. (4) Or you may want to have it compatible with CZcams.
Depending on a particular need, your second half of the work, AFTER you have the video captured, will be different: different codec, different scanning type, different frame size, etc. But the first half of the work - getting the video out of your camcorder or VCR to a computer - can be the same if you choose to capture it in an uncompressed or mildly-compressed format that preserves key parameters of the video: resolution, aspect ratio, scanning type, color subsampling, etc.
The major difference between DVD and CZcams is that DVD video is interlaced, so you want to capture the original interlaced video from the camcorder as interlaced, and you want to keep it interlaced and author it on a DVD as interlaced, and let your TV set figure out how to show it correctly. CZcams does not handle interlaced video well, so you need to de-interlace. Before 2014 you had to deinterlace into 30p, but now you can deinterlace into 60p, converting each field into a separate frame, preserving the motion. DVD video does not support 60p, so you need to have different versions for CZcams and for DVD. I mean, you can have 30p both for DVD and CZcams, but it is the lowest common denominator. Blu-ray supported 60p for 720p only, but the latest revision of BD standard does support 60p for 1080p and 4K as well.
When capturing, double-check that the codec that you are capturing into (Huffyuv, Lararith, Cineform, or whatever else) is supported by your editing tools. I am partial to Cineform because it is a high-quality codec that has been used by pros for the last fifteen years or so, but there are other good ones. It depends on how you want to watch or edit your footage later, you want to ensure compatibility and longevity.
TL;DR: it is best to capture into the least destructive format and have it as a master. Then you can always make a version off it for a particular platform.
Thanks for the reply , yeah I just want to have it digitalized as similar as the original one , I sent it to someone who offers this kind of service but it looks a bit stretched and also not as fluent as the original one ( I believe he covered it to 30fps )
I've been setting up a new fresh VirtualDub2 system running Windows 7. Just did a fresh install, put on WinDV and using Dazzle to Capture the same way as in this video. When try to save the video I notice under audio compression I'm missing MPEG Layer-3, so when I hit save is gives an error (pcm_s16le: codec not currently supported in container mp4). Do you now how I can add mp3 compression to the VirtualDub2 Library?
I think I have like three different packages for MP3... I guess some comes with VDub, then I think some came with Lame? I don't remember whether Lame installed any codecs available to other programs. I found this link, it is old, but pertains to 64-bit system, so may be relevant: forum.videohelp.com/threads/271712-Lame-MP3-Encoder-on-Vista-x64#post2016711 Then again, people say that there is no benefit running 64-bit VDub, and most filters are made only for the 32-bit version. I am sorry I cannot help you better :(
I've seen so many videos uploaded to CZcams that haven't been properly deinterlaced first. They are wrong and not authentic at all, because combing only appears on the digitized footage if it hasn't been properly deinterlaced when converting to progressive scan formats. And if you get a progressive scan format that hasn't been deinterlaced, I don't think you can deinterlace it anymore, not fully at least, it will always remain wrong.
Exactly! It seems that many younger people think that this is how we used to watch TV back in the day, with combing galore. They think that combing is authentic, while in reality it was not visible on CRT television sets.
Great video although I'm curious more about some of the settings you skimmed over quickly. One question, why do you capture at 30fps when interlaced video is actually 60fps and some people say to capture at 60fps? I suppose even at 30fps you are simply getting both half set of frames changing each frame, just technically not at the correct time since one of those frames changed earlier than the other. Capturing at 60fps you would see only one half the frame changing for each frame. Both can be deinterlaced and hence you did a double frame rate selection in the deinterlaced options? So maybe doing the original capture at 60fps has no advantage??
Analog TV is 30 (well, 30×1000/1001 ≈ 29.97) for System M a.k.a. NTSC or 25 for all other systems frames per second, interlaced. Broadcast VTR recorded interlaced. Consumer grade VTR recorded interlaced. Camcorders shot and recorded interlaced. Each frame in interlaced TV comprises two fields, and normally these fields are 1/60 or 1/50 second apart, so you get 60 or 50 field per second rate, each field being a different picture made at a different time, you cannot combine them into one picture.
When they wanted to transmit movies, which are normally 24 fps, they would slice movie frames into fields and arrange them into fields, creating what is known a pulldown pattern. In this cause you can sometimes weave fields into frames. It is usually much simpler for 50 Hz systems, when they just speed up a movie to 25 fps and slice each frame into two fields, then you can weave these fields back and voila! you get full resolution and characteristic motion signature. I have a short video about digitizing a 24 fps movie from an NTSC VHS cassette: czcams.com/video/DneoPwa5faY/video.html
But in "normal" video all fields are different, taken 1/50 or 1/60 seconds apart. Each field contains just half the number of lines of a whole frame. So, I am capturing full information: 30 fps, interlaced. Then I separate the fields, converting them into full frames, creating 60 fps, progressive. In terms of motion 30 fps interlaced and 60 fps progressive look exactly the same. This is how you preserve full spatial and temporal resolution.
Those who capture at 60 fps probably use OBS. Why? Likely because OBS has never been meant as an analog capture tool, it is a screen capture tool. Analog capture has been shoehorned in later as an afterthought, so I think that OBS gets 30 fps interlaced, then deinterlaces it into 60 fps, and then you capture 60 fps into a file. I don't use OBS for capturing, although I've heard people had good experience with it. I tried it once, it worked, but every time I tried it again, it did not work for me, so I cannot comment further :)
I thought I understood but now I'm more confused. "So, I am capturing full information: 30 fps, interlaced", but there is no setting to choose between interlaced or progressive in that menu. Are you saying the system identifies a interlaced code in the video and automatically changes from 30 pictures per second to 60 pictures per second? I am really trying to understand this but the fields are combined on the TV screen to make a full frame no? now you are telling me they are never combined into one picture, so viewing traditional TV we see only half the frame and there is black space between the lines? Must be. But then you also said "separate the fields" so that implies the fields are combined into one frame or else why do you need to separate them? The frames must be combined or you would not get combing effects. Converting fields into full frames / pictures means you are just doubling the lines so there are 2 lines now for every 1 line to get 60 fps? I'd think that would look bad. I'm obviously missing something in my understanding.
I see people selecting 60 fps instead of 30 fps for the initial capture (no not OBS format but lag or huf in an avi wrapper in vdub2), then when they deinterlace they don't choose the double frame rate option like you did. I'm guessing both methods result in the same 60p video at the end of de-interlacing. Recording at 30 fps must really mean recording 60 pictures per second. Each picture having only half the lines with black space between, then these are combined together to display a frame on the PC. Do TV's display only fields as the full frame? Yes they are taken at different times so viewing on a computer monitor does not look as good when fields are combined. @@ConsumerDV
■ "there is no setting to choose between interlaced or progressive in that menu." - Are you talking about OBS or VDub? I don't know much about OBS. Analog video is either 25 or 30 fps, end of story, this is what capture devices support. These frames are nominally interlaced, but can contain progressive content, in the latter case the fields of each frame may originate from the same time. Ultimately, it is the user who makes a decision how to treat the footage: as native interlaced, as 30/25 fps progressive, or as something else, for example as a telecined movie, which originally had 24 fps and requires a different deinterlacing algorithm, inverse telecine. You choose a scene with motion and check whether there is combing. Then you try different deinterlacing algorithms and see how it looks. For standard interlaced video you should see two separate pictures in the right window for every combed picture in the left window. I suggest reading through this long page: hometheaterhifi.com/volume_7_4/dvd-benchmark-part-5-progressive-10-2000.html
■ "I am really trying to understand this but the fields are combined on the TV screen to make a full frame no?" - No. The simplest way of deinterlacing is converting each field into its own frame. All TV sets default to this mode when they don't know what to do.
■ "viewing traditional TV we see only half the frame and there is black space between the lines?" - there is no black space, hence the term "line doubler", now rarely used. Also, depending on deinterlacing algorithm it is not always half the frame. Refer to the link above.
■ "you also said "separate the fields" so that implies the fields are combined into one frame or else why do you need to separate them?" - what I meant is split one frame into fields. In real analog TV fields are pretty much independent, although there is a top and bottom field (or first and second, or even and odd). In digital video, fields are usually combined into a frame for the purposes of compression and storage, but they still can be treated very much as separate entities.
■ "Converting fields into full frames / pictures means you are just doubling the lines so there are 2 lines now for every 1 line to get 60 fps?" - Correct. Twenty years ago deinterlacers were called line doublers.
■ "I'd think that would look bad." - HD looks ok, there is enough resolution for it to look just fine. SD can show stairstepping on diagonals.
■ "I see people selecting 60 fps instead of 30 fps for the initial capture (no not OBS format but lag or huf in an avi wrapper in vdub2), then when they deinterlace they don't choose the double frame rate option like you did." - I don't know how this would work, but if the device driver allows them to select this setting, then apparently the device deinterlaces the video internally.
■ "Recording at 30 fps must really mean recording 60 pictures per second."- yes, 30 fps INTERLACED means 60 pictures per second, each picture is a field. 30 fps PROGRESSIVE means 30 pictures per second, each picture is a full frame.
■ "Each picture having only half the lines with black space between, then these are combined together to display a frame on the PC." - For interlaced video they should not be combined, see above.
■ "Do TV's display only fields as the full frame?" - this is the simplest deinterlacing method, "dumb bob". Which is why 15-20-25 years ago you would hear names like Faroudja, HQV, DVDO, the companies who made quality deinterlacers. Pioneer Kuro plasma TV set is famed for its deinterlacer, I kick myself I did not buy a 50-inch set for $1,600 on sale at Costco back then.
Read the article linked above, these guys explain it very well with pictures. The article is a treasure.
The link provided does not detail how interlaced fields are displayed.... and that's where I'm confused. In the first section describing interlaced video there are 3 examples, a man, a tomato & a skier. Each one is displayed differently. The man shows one field of lines and white between them, then the second field displayed with the first field together. The tomato shows the current field with the prior field, over and over. The third shows only one field at a time with white between the lines. So which of the 3 is it? Or does it depend on the TV? perhaps old TV's display it one way new TV's a different? @@ConsumerDV
Fields are displayed in sequence with empty space between the lines of each field on a native interlaced CRT display. The next field is pained over the empty space, while the first one - and this is an important part! - is still somewhat visible, but has started to decay. So in reality you NEVER see empty space. This animation from wikipedia shows how the fields intermesh between each other: each new field is painted 1/60 s later than the prior one, but there is a time when both fields are on the screen: one is being painted and is bright, another is decaying. The careful choice of the decay time provide that the picture is not too blurry yet not too flicker-y: en.wikipedia.org/wiki/Interlaced_video#/media/File:CRT_image_creation_animation.gif There were progscan CRT displays, of course, primary computer monitors, some TV sets as well, but let us not talk about them.
On a native progscan monitor, which all plasma (except Fujitsu ALiS), LCD and OLED are, fields NEVER intermesh like on the CRT. The simplest option is to convert each field into a full frame while doubling the lines, this is not optimal for film-sourced content, but at least you won't see ugly combing or ghosting.
The first picture with the shot from Casablanca I believe shows Humphrey Bogart - this is a wrong picture, sadly. On a native interlaced TV you will see two fields in sequence. On a progscan TV with proper deinterlacer the video will be deinterlaced first before it hits the screen (have you watched my video about IVTC?), then the COMPLETE frame is displayed for 1/24 of a second, so you will see the full picture.
The tomato shows how one field changes at the time, still preserving the prior field. They tried to simulate how the prior field is still visible, I guess this did not work for you.
The skier image is another approximation, it does not show the decay, on a real CRT you will never see empty space, it will be taken by the decaying prior field.
As you can see, the whole idea of interlacing was based on a very particular display technology: phosphors of CRT, when bombarded with a beam of electrons, light up and then dim relatively slow.
If a TV set chooses a wrong deinterlacing method and say combines fields that do not belong to the same time, then you see combing. You will not see interlacing, as interlacing is a TECHNIQUE that works only on an interlaced monitor, you will see combing as an ARTIFACT caused by the interlaced nature of the signal and by the incorrect deinterlacing.
While we are on the topic of attack and decay, plasma screens also have this feature: the tiny elements light up and and dim not immediately, but have some decay time, this is why some people prefer plasma screens for more organic look. I love my 50-inch plasma, I wish I bought the Pioneer Kuro. Early LCD panels had very low refresh rate of 60 Hz or 120 Hz, and the picture would appear and disapper almost immediately (I hate LED blinkers on cars, they have the same immediate character, compare them to good old incandescent blinkers, they light up and dim relatively slowly), so when you have motion, you would get visible stutter, or sort of "stickiness" in the picture. This has been remedied by raising LCD refresh rate to 480 Hz and then to 600 Hz, so they are flickering all the time at a very high rate, and this flickering works better for our eyes. In a way, next frame sort of blends with the prior frame instead of simply being shown after it.
Fantastic channel
Great video, thanks for posting. I used VirtualDub a week ago for the first time and it worked fine. Now, after reinstalling Windows 10 and 7, I cannot get it to work properly. I do not see "capture pin" and any codec. I tried everything, but I cannot get the codecs to show up? Any ideas? Thanks.
Thanks! All I can think of is make sure all device drivers are installed properly. As for codecs, VirtualDub comes with some bundled codecs, and picks up other ones installed on your system.
@@ConsumerDV Hey, thanks for the fast reply. I know that VirtualDub comes with codecs, which makes the whole thing even more weird, because If it comes with codecs, why can't I see them? I noticed that when I don't have a capture device plugged in, I can see the codecs. I can also see them when I capture a video, using NO codecs, and then open the video in VD and export it. Now I can pick a codec. Weird. Thanks.
What is the best way to Capture a DVI Video under windows 10 and Firewire? I want to use my pinacle Movie Box DV. Got it wokring with a Dell Firewire Card and VLC. Are there any better oprions?
As long as your software can capture DV stream, there is no difference in quality: you are getfing the data recorded on tape exactly as it is. I use WinDV, but sometimes capture with Vegas.
Hello! im having a problem. my capture pin setting doesnt let me to change anything and the crossbar option is grayed out. i bough this UCEC USB 2.0 Video Audio Capture Card Adapter. so i wonder if its just a physical limitation of this hardware or i did not have the proper drivers.
A lot of these dongles have similar appearance, but may have different hardware. I bought one like this, and just like in your case it would not allow me to save into anything but MJPEG, and the video was being deinterlaced by the hardware. Capturing neither with 3rd party tools like VirtualDub nor with the bundled software helped. I ended up returning it. With no-name products you never know.
If you are in the U.S., the Diamond VC500 is around $30 (it is about £50 in the U.K., wow), and it is a decent dongle in my opinion, although I had to fight with the drivers, also I've heard that the hardware has been changed recently, but I cannot verify it, so take this recommendation with a grain of salt. Here is a quick review: czcams.com/video/j3VMQMEOWko/video.html
Woa thanks for this video!
Thanks for this excellent tutorial. Everything is well except when I play the captured avi file, the video pauses briefly every few frames or so. The audio remains synced and is not affected. During capture I am showing no dropped frames. Do you know what my problem could be?
Hi, I am glad that it was useful! What codec and bitrate are you capturing into? Some codecs are "heavy" and need more processing. Also, higher bitrate also needs more processing. It is a balancing act. Does it stutter in VirtualDub or in an external player? What player do you use? Are you on Windows? For AVC/H.264 in TS/MP4 container I can recommend Mirillis Splash, it is very fast: mirillis.com/free-hd-video-player For codecs that are often used by VirtualDub users like Huffyuv, Lagarith, Cinefom I can recommend VLC. I like Cineform because it is not very taxing, but GoPro does not make it easy to install just a driver. VLC supports Cineform natively. Many players play faster if you make the player window smaller.
@@ConsumerDV I am on Windows 11 playing back the captured video in both VLC and VirtualDub. I am getting stuttering in both. I have tried making the window smaller with no luck there. This is all new to me so I followed your example exactly, using the Cineform codec and compressing to mp4 after cropping etc... EDIT: It seems the incoming video feed during capture is what has the stutter, so the software seems to be doing what it should. Maybe it’s my camcorder? Or it’s another problem between the camcorder and the pc? I am digitizing Hi8 tapes using a Sony Handycam CCD-TRV138 using the Pinnacle Dazzle DVC100. The camera only has a single 3.5mm A/V output to composite. Thanks for getting back so quickly!
You can play on the camcorder and check on the flip-out screen or on a TV to be sure whether the stutter is present in the original video. If not, and you see the stutter during capturing, maybe the chosen codec and bitrate are too taxing for your system, although considering that you have a Windows 11 system, I assume it is rather new and fast. During capture, check CPU and HDD load on the right side of the VDub window. If the load gets close to 100%, you may lose frames. Check the number of lost and inserted frames, the fewer you have the better.
I literally just got the DVC100 and had the same issue. I found a different driver that fixed pretty much everything! Here's the link to the website.
forum.videohelp.com/threads/398965-Dazzle-DVC100-not-capturing-anymore-on-Windows-10#post2603674
If you can't find it, here's the link to the driver itself.
forum.videohelp.com/attachments/56186-1607688052/Dazzle%20Drivers.zip
Hi, thank you for the helpful tutorial. In my Hi-8 footage, I get lots of horizontal white lines appearing and disappearing all the time. I think this was due to the heads being dirty or worn at the time. Can you recommend a way to remove this in VirtualDub2 or any other software?
Hi, I am sorry, I am not big on restoration. There are knowledgeable people on forums like doom9 and videohelp, I usually ask there myself.
Any extra tips for working with Mac would be great, my captures look WILD 🤣
Hi! The principle is the same, but, sadly, I cannot help neither with capturing devices nor with software, I am a Windows guy. AFAIK, there are no Mac drivers for the Dazzle.
@@ConsumerDV Thanks for the reply, I enjoy your videos a lot! I'm just starting out in HI8, its fun but I'm surprised at how tricky getting a decent capture is for me. I just found the open-source software Handbrake, and I'm hoping that with a solid initial capture, this program will help with the artifacts (and high levels of analog distortion) that I suspect is coming from either my capture device trying to deinterlace as the signal is coming in, OR it is user error due to my capture settings.
I put all configurations that you showed in the video but in the end when I save the video it appears one error message: “codec not currently supported in container mp4.” What should I do ?
Usually you capture into a lossless or visually lossless codec (I use Cineform), then you re-encode into a delivery codec, which is usually AVC/H.264 nowadays. I don't have a computer right here at the moment, but if I remember correctly, VDub2 cannot store Cineform in MP4 container, it wants AVI. Use MP4 container for your delivery when encoding into AVC.
Я так и не понял, обязательно ли надо приводить пиксели к квадратному виду?
Разве не все плееры\устройства могут понимать какой пиксель в записи и выводить уже в нужном масштабе и без искажений?
Как по стандарту эта штука с неквадратными пикселями работает\должна работать?
Не обязательно, но улучшает совместимость. Если вы будете загружать на ютуб, он все равно сконвертирует в квадратные пиксели. Старые контейнеры вроде AVI не имеют стандартного способа указывать пропорции пикселов, но в новых контейнерах проблем нет, и компьютерные плееры нормально проигрывают эти файлы. Главное, не забыть указать правильные пропорции при генерации файла.
@@ConsumerDV спасибо за разъяснения! И за ролики тоже спасибо, весьма полезные.
Great video.
Great video, thanks so much for posting this, love de-interlacing 29.97i to 59.94p, I think I'm going to use QTGMC. Are you using the DVC100 on a Windows10, 8.1 or 7 OS? I've heard a lot of Win10 systems drop frames after a certain Windows update point... I'm looking for a decent workflow and doing a lot of fact finding to get the best flow for digitizing VHS/C and Video8 footage using the DVC100.
I am using Win7 64. QTGMC usually works fine. You can have a small script with it loaded into VDub and use other filters, or you can go fully Avisynth route.
@@ConsumerDV I'm looking into using older work laptops perhaps. LENOVO LS500, they actually have built-in mini firewire 400 port, wont use the FW ports for VHS capture, but was neat the laptops had them, i can use my DVX100b and my Sony to capture Hi8 and MiniDv using the FW port. I may bulid 2 laptops with windows 8.1 and dedicate them for capture and de interlace machines. I'll read up and see stability of OS 8.1 64bit with the DVC-100.
Again thanks for all the postings and yes I'm going to de interlace with Avis script, QTGMC and VirtualDub.
Two of my MiniDV camcorders and a Digital8 camcorder can convert analog to digital and output as DV: czcams.com/video/Hbytp67EYAg/video.html This may be an easier catch-all option if your camcorder can do it. I personally like DV codec, other people hate it, but it is still well supported. I prefer desktops as they are easier to upgrade and modify.
@@ConsumerDV Thanks so much again for the reply, I love video cameras, filming, editing and playing around with footage. On this note regarding MiniDv, I have Canopus 110 Firewire DV converter (DV-25 Codec), and yes I can do a pass through via my MiniDv Camera (Panasonic DVX100b, Sony TRV hi8 model) I've done a few tests and feel the footage looks nasty even when passing though via S-Video. I tried in comparison a cheap EZCAP via VirtualDub and honestly the footage looked like it had more bit depth than when captured on FCPX via the Canopus 110. I think in completing research via your youtube videos, my issue is that it's not uncompressed.
I'm starting over with a new work flow and trying your Dazzle route with VirtualDub2. I agree with using a desktop, however space is an issue, I don't have a room I can dedicate at the moment for a capture suit. Again thanks for the videos, I found a Dazzle DVDRecorder HD from 2011 that's a white DVC100 Rev 1.1 for $15 CND. Got it yesterday and just today bought a red DVC100 for $15. I think it would be great to have a backup.
After I find an import and archive work flow, I'll then start the QTGMC de-Interlace method for nice 59.94p footage of my old home video VHS collection.
Fantastic. How would you deal with a PAL/SECAM source if VirtualDub2 only offers 720x480 ?
Thanks! Try connecting your video source before starting up VDub. Connect the cables, turn the VCR on, start playing something. When the device/application senses the signal, it usually automatically switches to an appropriate mode or at least shows it as a default, then you'll be able to tweak it if necessary.
@@ConsumerDV OK
огромное спасибо !
Great explanation and walkthrough - Thank you - One question when i went to MPEG Layer3 for audio the dialog box was empty despite downloading the Fraunhofer MPEG Layer-3 Audio Decoder - Thanks
Thanks! Some of the compression codecs come with VirtualDub, other are pulled from your system. I need to do a clean install to figure out which is which... Presently, for MP3 I have FFMpeg Lame MP3 with just one preset, LAME MP3 with lots of presets, MPEG Layer-3 with no presets at all, and another MPEG Layer-3 with lots of presets. I have installed FFMpeg and Lame MP3 separately at some point, so they have been pulled by VirtualDub. I think that if you install Lame MP3 package, the codec will be picked up by VirtualDub. 64-bit VirtualDub may not pick it up, but 32-bit version should.
@@ConsumerDV Hi! how do you install those?
@@jgalianos I don't remember the details. I think I downloaded one of the binary bundles and installed it.
I am looking for a basic tutorial of WinDV. After converting from my camcorder, my converted video simply doesn't play.
WinDV copies DV video from a DV or Digital8 camcorder over FireWire exactly as it has been recorded on tape, and stores it in an AVI container, see more here: en.wikipedia.org/wiki/DV#File-based_media Could you tell more, how do you try playing your videos?
The DV-AVI format has been supported by Windows for at least two decades. Windows Movie Maker supports it, and pretty much all video editing tools support it as well. I use Windows 7, so I cannot comment whether Windows 10 still plays this format, but I don't see why would it stop supporting it, it is still very much alive. In any case, there are zillion of other video players, for example Media Player Classic Home Cinema (MPC-HC) or VLC Media Player, maybe try them.
I was able to use this method to attain the same results! I also tried capturing the same footage using a Framemeister HD Upscaler (often used by streamers to upscale retro games in 240 or 480i) and Elgato HD60 (HDMI capture) also 60p. Here is a comparison video I made showcasing them side-by-side: czcams.com/video/WnEyspXHlfo/video.html Let me know what you think!
I left a reply on your channel!
@@ConsumerDV Thanks! I uploaded a new version that fixes the framerate issue: czcams.com/video/WnEyspXHlfo/video.html
Nice! The VDub version is more saturated, possibly because you grab with 4:2:2 color subsampling. On the other hand, if you look at the diagonals, Elgato is no worse, and dare I say, maybe even better: i.ibb.co/g9HFYhM/video-signals-saturation-deinterlacing.jpg All in all, they are quite close if you ask me. If you want to stay with VDub, you may want to try MSU Deinterlacer filter, which in my experience is just a little bit better than Yadif, but an order of magnitude slower. Like many older filters, it works only on 32-bit VDub. Or, you may want to graduate to QTGMC deinterlacing script, you can run the script in VDub. I used QTGMC for 1970s music videos that I uploaded in 50p.
I raised Elgato saturation to match VDub, the saturation "blob" on the vectorscope moved more into the red zone, so the image became a bit reddish: i.ibb.co/NLzRjJF/video-signals-adjusted-saturation2.jpg You can match it better, of course.
So, all in all, Elgato is not that bad :) If you make a video about using Elgato to produce 60p videos, I'll watch it :) Or maybe you can just write a short how-to, because I thought that Elgato captures either into interlaced stream at best, or into 30p at worst.
Have you ever noticed captured 29.97i footage in VirtualDub2 have the background and foreground focus change every frame? Exaple frame 1 and 3 the foreground is in focus and frame 2 and 4 the background is in focus. Not sure if you've ever experienced this? This may cause the horizontal line jitter. As everyother frame is soft in that subject area.
I did see fields of a frame sort of flickering, when one field was slightly dimmer than another. But not frames. Do you have manual or auto focus? Although this would probably be too quick for focus change. Have you joined doom9.org or videohelp.com? Some knowledgeable people there. I mean, they are less into camcorders, but more into digitizing and converting from one format to another.
@@ConsumerDV Thanks very much, I may join those groups, I have been there reading posts and tips. Great resources, I'll maybe create an account and join.
The camera has auto focus, it's a Panasonic DVX100b MiniDV, but it fluctuates every other field when separating into frames.
What was your capture OS again, Windows 7 or XP using DVC-100 Capture hardware into VirtualDub2?
I am using Win7-64, but I don't think it has anything to do with this fluctuation. BTW, I was able to get rid from tons of inserted frames by deselecting "Enable audio playback" menu item.
@@ConsumerDV Good morning again, thanks for the reply. I believe you're correct about the OS, it was just a side note I wanted to know about what OS worked for you. The reason I ask is that I'm trying to keep consistent with what's working for others. I had created a fresh install of Win7 with VirtualDub2 for capture. It all worked to capture on Win7 SP1 with the DVC-100. I then wanted to make a fresh install disc with all the updates czcams.com/video/l5ADP-VZMsw/video.html So I started to update the machine and then after all the updates the Dazzle wouldn't connect. So last night I ended up reformatting the system with only SP1 on Windows 7, and low and behold the DVC-100 is now again working. So I'll not bother with creating an install version of Win7 with all the updates and rollouts.
Again I really appreciate the insight from other like your self.
I've been doing a few test captures of the same 1 min footage to see the best route. My source footage was shot on MiniDV, I did three tests. Captured directly to FCPX on my mac from the camera firewire (DV25), then captured in FCPX from the Canopus ADVC-110 which is analog Svideo to the DV25 codec in FCPX, again still used the MiniDV tape but used S-Video out from the camera to line in on the Canopus ADVC-110 which digitizes to DV25. Then after that I used the same original MiniDV footage from SVideo out of the Camera to the Dazzle DVC-100 and captured in VirtualDub2 lagarith codec on the Win7 machine, I did one more test just to see how bad the EZCap was, I again played from my MiniDV camera S-Video out to line in on the EzCap and captured using the lagarith codec.
I'm now going over the clips comparing and that's when I notice that all methods have every other field slightly blurry, again field 1 will have the foreground in focus and field 2 will have the background in focus. It creates a slight flicker on certain objects. Not sure why this is happening. It's not the auto focus. The test shot I use, has the camera still on a table.
I'll keep playing around and see what I can come up with, I didn't think digitizing would be so time consuming. I'm still going to buy a TBC.
Thanks again for all your tutorials.
I forgot, maybe you have mentioned it already - do you have an older Mac with a Firewire port, or do you convert from Firewire into Thunderbolt? As I said, I noticed slight difference in color, not in focus, between two fields of my Canon Elura 100. I tried... I think a temporal smoothing filter... not sure. Not the one that is bundled with VirtualDub2. Sorry, I don't remember. It worked if the scene did not change drastically, but on scene change there was a funny effect that it tried to continue with the settings from the previous frame, looked ugly. I think I found a balance when it looked ok even with a scene change, I need to check which plugin did I use. I am not big into Avisynth, but maybe I need to dive into it for more powerful functionality. As for TBC, I think I am going to stop where I am now. So yeah, I am glad that something I put up on CZcams was useful, and for more powerful scripting stuff definitely check those forums!
Update: I finally got a Dazzle DVC100 (red color model) for $12.50, and I had to jump through a few hoops, but thank God it is fully working in action! Only thing I am having some extremely minor trouble with is that there is some ghosting during movement, mostly if the source material isn't 60p. It was worse when I had these levels filter added on Virtualdub. To be honest, it's not even that noticeable unless if you squint hard enough. Is there a fix to this? If not, I am just going to roll with it. Other than that, I have got to say that the quality is absolutely outstanding compared to my other cheap capture cards that I had in the past. I don't know why I didn't buy one of these in the first place! Thanks for the absolutely great and informal tutorial!😁🖒
If your tape is NTSC, deinterlacing to 59.94p will give you the same "look" as on TV. Similarly, if your tape is PAL/SECAM, deinterlacing to 50p will give you the TV "look". In some cases you can do better. For example, if you are capturing a Hollywood movie from an NTSC tape, then you may want to use IVTC filter instead of deinterlacing, see a short example here: czcams.com/video/DneoPwa5faY/video.html If it is a movie on a PAL/SECAM tape, you may want to treat your video as 25 fps progressive without using deinterlacing filter.
If it is a European movie on an NTSC tape or vice versa, or if it is a cartoon, then more elaborate schemes of restoring the original frames may be needed, and I am not an authority on them, you may want to try forums like doom9 or videohelp. Case in point: I have a travel video that was shot in Europe, it is a mix of film and video, then it was transferred to NTSC tape. If I deinterlace to 60p, some frames still have ghosting. Another case: "Portrait Of an Album" with Frank Sinatra - it was shot on film at 24 fps, converted to video and edited on video. The effects in between the songs are made in video, so they are 30 fps interlaced. I used IVTC, so I restored the movie to 24 fps, but the effects now look ugly. I could deinterlace it to 60p, and it would look fine, but I would lose a bit of vertical resolution. Or I could process the effects scenes separately, deinterlacing them to 60p and then converting to 24 fps, throwing out extra frames. It is a black hole, once you start fiddling with it, it sucks you in :)
Regaring the Dazzle 100: I like it for the overall decent quality, and that it does not care about Macrovision. But it may have hard time with timing errors, so you may see some "skewing" or "hooking" on top of the frame, especially if the tape is recorded in LP or EP mode. See comparison with another card here: czcams.com/video/bhZyk72I9Q4/video.html I did not mention in the video that the other card have some other issues like inter-field jitter, also with A/V sync. No card is perfect :(
Sorry for just now responding, been quite busy lately. Anyways, my footage is NTSC, and I'm working with tapes that have different frame rates (like your tape of Wheels: The Joy Of Cars for example).
Edit: You're not going to believe this, but I figured out how to rid the ghosting. Apparently, I had to disable the noise reduction. Lesson learned: It always pays the tweak with every single setting in VirtualDub! I'll have to recapture four of my tapes again, but it will be worth it (luckily most were 30 mins anyways)! I will use the IVTC Filter for 24p only tapes, thanks for mentioning it!😁👌
I followed the same setup directions in VirtualDub2 with a dazzle dvc100 but I am not getting any picture inside of virtualdub - just a black screen even when the camera is playing the tape- any advice?
Make sure that the device is selected in the Device menu. Also, that in Video menu "video source" is correctly selected between composite and SVideo, that "capture pin" has correct video standard and frame rate, that "capture filter" also shows correct video standard, also try switching between "Overlay" and "Preview (system)", on some cards the latter works better, although I did not have issues with the DVC100. Also, try start playing video first, then start VDub and configure it - sometimes it better senses the incoming format when an actual video plays. I hope it works out for you!
@@ConsumerDV I've tried 2 different dazzle dvc100s with these settings, but I'm just getting a black screen inside VirtualDub2. I wonder if I need another driver installed? I have tried a couple already. I wonder if anyone else here has had this issue and was able to solve it...
Have you ever tried the iuVCR capture program ?
No, I have not.
Hey, quick question. Does the Dazzle capture card drop frames or blank out if there is either no signal or when a worn out or damaged VHS tape is being played? I have a generic Easy VHS to DVD USB card that will do that sometimes, but it will capture in interlaced MPEG1 files which gives me the opportunity to convert them to 60p in Handbrake. Quality isn't that bad imo, definitely better than a cheap Easycap USB card. Almost thinking of upgrading to A Dazzle card like you have. Thanks!
Hi, the worst I had was something like this, czcams.com/video/qs8KF1G_mgs/video.html (from 1:25 to 1:35), this dropout did not stop the capture. I haven't checked how many frames were dropped. Overall, AV sync with the Dazzle floats about 0.2-0.3 ms here and there, but usually it self-adjusts, so it never drifts away too far. To some, this much drift is a showstopper. I don't know whether the Dazzle 100 is better than an Easycap that can capture true interlaced. The good thing about the Dazzle is that it can capture from Macrovision-protected tapes, on the other hand it does not have any time base corrector functionality, so if your video exhibits skewing, the Dazzle would not correct it. The Avermedia card is better in this regard, but it has other issues. So far I haven't come across a device which is both cheap and ticks all the marks. OTOH, a used Dazzle 100/101/103/107 will cost you probably less than $15 on ebay, so maybe worth a try.
@@ConsumerDV Yes yours did better than mine would. Might give the Dazzle a try! As for macrovision, my existing usb card does pretty good, plus I have one of those Digital Video Stabilizer boxes that's powered by a 9V battery, and it helps to remove skewing in the picture if there is any. Thanks for the helpful info!👌
I was under an impression that "stabilizer" is a simply an euphemism for a macrovision stripper, but if it also helps with skewing, this is even better! I may look into getting one myself.
@@ConsumerDV Please consider it, I found mine on eBay in late 2012 for $35, and I think that's what they still go for. Might vary by tape, but to me, I think it's still a good investment for those that want to archive VHS tapes that haven't got a proper DVD release and capture it in somewhat decent quality! Replacing the 9V battery is pretty easy at least on my particular unit, just remove 4 screws, pop off the top cover, then you have access to replace the battery if desired. I replaced mine in 2020 just for good measure👌
How can i have a good file but less storage eater?
Uncompressed takes the most space, compressed but lossless takes less space, visually lossless intermediate codecs like Cineform or DNxHD take even less space, then delivery formats like H.264 or H.265 are lossy and heavily compressed but take the least space.
@@ConsumerDVThank you for answering. Do you have any guide how to put codecs?
No guide, sorry. Some of them come with VirtualDub, others I installed separately with their respective installers. I think X264 - an implementation of H.264 - comes with FFMPEG package. If you want MPEG-2, I believe you can get it with HCENC package.
@@ConsumerDV Ι didn't;t understood how to install it but thank you very much hope a video for that in the future thank you again
When capturing from the Dazzle in VirtualDub2 at 29.97i 720x480, during capture it displays average frame rate. It shows live stats between 29.905 - 30.00, after capture when checking the info with file manager properties or even in VLC it notes the frame rate at 29.97. However when I bring the footage into a video editor is will display the frame rate at 29.95, even using quicktime to display the clip information it shows at 29.95.
I've done more than 8 tests of capturing and each time the frame rate is slightly different, some of the rates are even 29.84, I feel this is problematic if trying to sync footage, in addition in FCPX it renders the time line as soon as it's imported because it's the frame rate is not 29.97.
I just wanted to reach out as I know you use the Dazzle to capture and was wondering if you've seen your clips information? Not sure what editing programs you use.
Again thanks very much for all of your posts and replies.
Regards;
Phil
I am not a professional video editor, so I don't know full details on how strict all the numbers should be. I edit in Vegas, and whenever I feel there is an A/V mismatch, I just correct it on the timeline either by stretching audio (usually it is spoken words, so I don't care for pitch change), or by shifting it. I do see fluctuation of frame rate when capturing with the Dazzle, so far it has not given me serious problems, usually it does not drift apart too much. I am capturing 2-hour tapes now, and they look reasonably good to me. OTOH, with another card that shows zero drift and exact 29.97 frame rate I get 3/4-second drift after one hour of capture. Frankly, I don't know how you can get the exact frame rate without genlocking the VCR, and consumer-grade VCRs cannot be genlocked. Even if an external TBC is used, it would try to match incoming rate to TBC's clock, and if the video continues to drift, it would drop or insert a frame. AFAIK, the only way to do it perfectly is to use genlock or to use a TBC that loops back to the VCR to give it a command to run faster or slower, which is basically the same as genlock, but for that you need a pro VCR.
TL;DR I don't pay too much attention to that as long as the A/V is more or less in sync, and there is no visible frequent frame dropping. To me, the absence of combing and proper aspect ratio is already halfway there, but those who adjust every possible setting in their scripts for a perfect capture will not agree :)
Ah, the actual question about fps on clips - I will take a look and will get back! :)
A file I captured yesterday with the Dazzle. I skipped to the very end to check A/V sync, looks good to me.
MediaInfo:
General
Complete name : H:\My Videos\American Century\Tape 3, U778, Dazzle.avi
File size : 60.4 GiB
Duration : 2h 38mn
Overall bit rate : 54.7 Mbps
Video
Format : CineForm HD
Duration : 2h 38mn
Bit rate : 53.2 Mbps
Width : 720 pixels
Height : 480 pixels
Frame rate : 29.970 fps
Standard : NTSC
Vegas:
Video: 02:38:04.141, 29.970 fps, 720x480x32, CineForm Codec
Audio: 02:38:04.348, 48,000 Hz, 16 Bit, Stereo, Uncompressed
@@ConsumerDV I wanted to thank you for the reply, I read the reply last week when you posted it and I wanted to do several tests to confirm before replying back to you. I tested capture in both versions of VirtualDub one and and two with either HuffyUV or Lagarith codec, I even bought a DiamondVC500. In doing a test capture with both the DVC-100 and VC500 when opening the captured footage in a video editor like Final Cut Pro or even viewing the files in QuickTime to just use the media info, the footage frame rate was never consistent 29.97 it would always be below 29.97.
I've discovered that the Windows 7 media info or VLC media clip info will show what the capture setting was set too (29.97) but it won't display what the actual frame rate is at. For myself this is a problem as when brining in footage at a different frame rate will change the length of a clip and for multicam footage, everything will shift very quickly and when dropping onto a time line 29.97 time line it will have to render on every single edit.
If you're intrigued, just download the last version of QuickTime, it won't play HuffyUV or Lagarith but you can view the clip info and mine for example was 29.78. I have copied the info from Win Properties/File Info, QuickTime and VLC.
Windows 10 Properties/Details
Video
Length 00:00:26
Frame width 720
Frame height 280
Data rate 164704kbps
Total bitrate 166240kbps
Frame rate 29.97 frames/second
Audio
Bit rate 1536kbps
Channels 2 (Stereo)
Audio sample rate 48.00 khz
Quick Time Movie Inspector
Movie FPS 29.78
Playing FPS 29.78
Data Size 523.61 MB
Data Rate 166.08 mbits/sec
Duration 00:00:26.44
Normal Size 720 x 480
Current Size 720 X 480
VLC Media Info
Video 720 x 480
Buffer dimensions 720 x 480
Frame rate 29.970030
Orientation Top left
Channels Stereo
Sample rate 4800 Hz
Bits per sample 16
I have done more testes and have used amarectv 3.10 capture with both the DVC-100 and the VC500 devices and they have generated steady consistent frame rate of 29.97. I also used amarectv as a passthrough to VirtualDub and recorded into VirtualDub with the HuffyUV Codec and it worked at 29.97, even during capture when looking at the average capture it would stay steady at 29.97. As mentioned I need true 29.97 footage, you may not need that.
Sorry it took so long to get back to you, I was literally working on testing every single night after work, 3 hrs a night. LOL
Thanks so much for you videos
Regards;
Phil
Hi, I appreciate the info! I've heard good things about amarecTV, but I have not used it myself. I should try it out, thanks! How do you use it as a passthrough to VirtualDub? Is it because not all codecs are available in amarecTV?
I record 10 mins with this App , the out put video is 12 giga , how I can reduce this huge amount of data
This depends on the encoder that you choose, it is all up to you. For example, I prefer Cineform because it is a high-quality intermediate codec that was very popular 10-15 years ago among pro videographers. Its bitrate depends on the quality setting and on actual video being compressed, it floats around 30-200 Mbit/s, so I get files about 16GB/hr with 35 Mbit/s for a standard def video, and in another case 58 GB/hr with 130 Mbit/s for HD video. If you choose uncompressed, the size will be even larger. Yep, these are the data rates. Choosing a space-efficient encoder like H.264 or H.265 can cause dropouts, because these encoders are more CPU-intensive. Depends on your machine. When you are capturing with VirtualDub, there is CPU utilization indication on the right side. Also, codecs like Cineform, DNxHD, ProRes are better suited for further editing, they are intermediate codecs. H.264 and H.265 are for the final encode. Anyway, after you make your final encode, you can delete the intermediate files :)
Does this work on Win 10?
Yes, it does.
8:40 what is the model of this vhs camera you used for this shot ?
It is JVC GR-SX950 that you saw in another video. It is better as a playback device (SVHS-ET, TBC), but as a camcorder it is not that good. I think the sensor has not enough resolution.
this seems a lot of effort for some old not so hot (boring) video's. I'm amazed that anyone will try this. But good luck. I'll just trow these old videos in de dumbster. :-)
The videos used are just samples. Yes, they are boring :)
where did you get that virtualdub2 version? mine is different
How different? I believe I downloaded it from the SourceForge page: sourceforge.net/projects/vdfiltermod/
I'm getting 2fps and crunchy audio with this setup. What's am I doing wrong??
No idea. Try turning off "Audio -> Enable Audio Playback", sometimes it helps with crackle and with dropped frames.
Thanks for the reply! I switched to my laptop running wondows 10 and it was better instantly
@@bigkwhit Good you found a working configuration. I use primarily Windows 7, but I tried running it on Windows 10 and it worked, but I saw quite a few frame drops. After I turned off audio playback during capture, it stopped dropping frames. I haven't tried on Windows 11, so thank you for the heads up!
Hey I want opposite . What should I do to make my videos to look like hi8 style?
There are many ways to ruin your video. Many NLEs have effects for "80s look" or "analog video look". Basically, add some jitter, sparkles, dropped frames, dull color, etc. For best effect you need to start with a video shot on a CCD-based camcorder. If all you have is a smartphone, don't shake it too much - rolling shutter artefacts are the telltale signs of a CMOS-based camera, and all Hi8 camcorders have CCD sensor, which has global shutter. Also, many effect plugins fail to reduce chroma resolution, so you don't see ugly color spills that is inherent to analog consumer-grade video.
@@ConsumerDV all I have is smartphone bro . I'm trying to upload my vlogs in my instagram with Hi8 style but I don't quite understand well what you said . Is after effect good program or not ?
I don't use After Effects, but I suppose it has some "old video" preset. Don't forget to crop your widescreen video to 4:3 proportions ;)
13:12 I looked carefully, it seems to me you cut too much. Part of the image is missing around the edges.
Rec. 601 stipulates that 704x576 for 50 Hz ("PAL") and 704x480 for 60 Hz ("NTSC") has 4:3 proportions. It is up to you how much to crop, the goal is to ensure correct proportions of the image itself. Maybe this can help: czcams.com/video/6SqbwZdzx3g/video.html
@@ConsumerDV when we crop the video around the edges, does the quality drop ? When a video is upscaled, does the quality drop ? For maximum quality after capture, the video does not need to be modified ?
Quality drops from scaling. Cropping and scaling are different and independent operations. Scaling is inevitable when you watch on youtube or on TV. The best rule of thumb I know is not to scale like 2% or 5% or 10% So scaling something like 698x476 back to 704x480 with PAR 10/11 or into 640x480 with PAR 1 will reduce vertical detail significantly. In this case it is better to pad, that is to add black border, than to scale. Or scale to something larger like 960x720.
0:48 what is the camera model ?
Panasonic GS500. PV-GS500 is 60 Hz "NTSC", NV-GS500 is 50 Hz "PAL".
@@ConsumerDV usb ?
Yes, it can send DV video over USB at full resolution. You will need software that is capable of receiving this video, Win10/11 does not have built-in tools for it. I installed Windows Essentials 2012 on a Win10 laptop, and Windows Movie Maker recognized the camcorder, Windows automatically installed a "DVC device" driver, and I was able to capture DV video over USB. There were 80-something dropped frames at the start of the capture, no more frames were dropped during the capture, so basically I lost 3 seconds in the beginning, which I think is acceptable for capturing a 1-hour tape. See more here: czcams.com/video/LGfQpY__F3U/video.html
For some reason every time I capture video the average frame rate is around 21,22 or 23 fps depending on settings, it fluctuates quite a bit.
I have it set to 25fps (UK PAL VHS tape) Over the course of an hour of capture, the sound becomes extremely out of sync when played back. (I assume this is related to the fps issue)
Is it possible the dvc100 I have is just faulty? I have tried many times with various settings and keep having same problem.
I have read it could be a problem with my HDD write speed also?
I have set timing settings to NOT resync audio by adjusting video frames, maybe this is wrong, but I tried other way and sound was affected.
I think maybe its putting in to many frames which is causing the video to become unsynced with audio?
But it should still "record" the capture at 25fps right?
Thanks in advance.
I don't have this problem. Do you use VirtualDub for capture? Try disabling audio playback during capture in Audio menu. Other people use AmarecTV and say that it is better in terms of A/V sync.
@@ConsumerDV I have audio and video disabled when capturing yes.
@@ConsumerDV using virtualdub2 with correct drivers installed
@@ConsumerDV Thanks for your reply.
I am using windows 10 64 bit, with Vdub2 using 64 bit win 7 driver I think.
It could be a faulty dazzle unit.
Try AmarecTV, other people are swear by it. I don't use it, because VDub works for me.
What if you dont have a firewire connection?
You need Firewire to transfer DV and HDV video. Few DV camcorders can transfer via USB. A FireWire expansion card costs about $15, easy to install. If you have a laptop, you may try converting to Thunderbolt, I haven't done it, but there are videos on CZcams that show that this conversion works. Apple-branded dongles are expensive.
@@ConsumerDV Thank you! I discovered that my camcorder only has a P2 to RCA connector, how could I record it? Using a VCR to a TV connected to a computer?
So you have a 3.5mm socket in the camcorder, and 3.5mm to RCA cable? Then you use these RCA connectors in almost any analog-to-digital converter, they all have RCA for composite video, and dual RCA for stereo audio, some of them also have SVideo input.
@@ConsumerDV Ok, it's a litlle easier, I see. Well, I think that's it. Thanks. I'll get a converter and be happy.
From where i can buy this pc ? It has firwwire an AC
15 years ago this was the norm: a bunch of audio ouputs, audio input, video composite, S-Video, Firewire, USB, also a TV/FM receiver card... yeah. This is a 2008 machine. You can find a refurbished one for about $150. For my other desktop that I built myself I bought a card and installed it, it is very simple. A Firewire card can be found for as low as $15, there are no wires, just push the card into a slot on the motherboard, easier than Lego :)
@@ConsumerDV amaizing , I have bought one pci card for capture but it was poor , So , I said to buy one built in same as yours ,
Is there any name for pc so I can search easily
Regarding Firewire cards, some people prefer cards with TI chipset, they say these cards have better compatibility with older devices. I myself have got a card with VIA chipset, and it works flawlessly for me. I still use Windows 7, and the system was able to automatically download and install the driver. If you are asking about a complete system, then you may want to search for "refurbished Windows desktop with firewire". I built my desktop from parts.
@@ConsumerDV thank u 🙏 so so much for ur information that definitely will help me .
I really hate interlaced video. Such a nightmare for me back in 2011.
It was a useful compromise during the early days of TV. Its usage for HDTV has been lobbied by Japanese manufacturers, who already had created equipment necessary for their analog HD system. Finally, 4K has got rid of interlaced video. Good riddance.
@@ConsumerDV Not in my area. TV here is still interlaced.
-- mgimo finished? -- ask! ))
czcams.com/video/GPAcdiZQ1zo/video.html
This is called interlacing. Not combi
Interlacing is a _technique_ that meshes video fields meant to increase the perceived image rate and reduce flicker. Combing is an _artifact_ visible when interlacing is not done correctly or when interlaced video is incorrectly deinterlaced.