Panasonic GH7--Will Internal RAW Kill the GH Line?
Vložit
- čas přidán 5. 09. 2024
- A ramble about the history of LOG shooting gammas, what filmmakers notice when shooting RAW for first time, etc.
I explain why 32-bit-float has nothing to do with mic gain and clipping
maxrottersman....
You have a very interesting point of view! Then the question arises: to get the most accurate color you need to shoot at ISO 100, but will this give you the highest dynamic range? Have you heard that Nikon has raw? Does it make sense to shoot in Nraw at iso 100? And what do you think about the difference between full frame and 4/3 perspective? Field of view 50 is equivalent to 25 mm by 4/3. But the picture on video will be 4/3 flatter, provided the fields of view are the same, because at 4/3, the 25 mm lens still has the properties of a wide angle, which means it stretches the image making it flatter compared to a full frame, which with a 50 mm lens makes an image close to real while maintaining real dimensions between visible objects? Or does perspective have a greater influence in photography and the 3D effect of a lens is outweighed by its design (the fewer elements, the higher this effect)?
ISO 100 doesn't, so much, give the highest dynamic range, it gives the best tonal fidelity that the sensor (or film) is capable of. One could argue you get more "dynamic range" shooting at 800 ISO (amplifying the signal) because you get more tonal range from the shadows, but again, at the expense of tonal fidelity (more noise). That's how LOG shooting GAMMAS work essentially.
You can get the same field of view in all sensor sizes. The difference is the aperture and depth of field. Smaller sensors like MFT can't get as shallow a depth of field as APS-C or Full frame. I explain in an old view about webcams. The short answer is that you can get a lower angle of light convergence into a larger sensor (which creates blur). The angles of light convergence on smaller sensors is less. Of course, you can get a "speed booster" and get it that way but those adapters degrade the optics.
I don't believe the amount of elements in the lens means much in this stuff. Though of course, gives sharper images because each element fixes an aberration.
Hope this answers your questions.
7:45 - Yep, it's true. I think shooting VLOG has a few advantages but color fidelity isn't one of them; you can preserve some extra details in the shadows and highlights by like 2 stops overall, depending on camera model. However, let's say you're shooting on a cloudy day, outside, where there's not much dynamic range. This is where a Standard or Natural profile in 10-bit (without anything clipping) is a little more logical of a choice over VLOG, so long as you get the exposure mostly right in camera, your colors will be objectively better than someone shooting the same scene in VLOG 10-bit. I use the GH5 II. I'll eventually upgrade to the GH7 but I am not interested in RAW video. I know how to get exposure and white balance right in camera using tone mapping and the Zone System along with the waveform monitor as a guide. I'll use a SanDisk V60 card in the second slot; this allows for up to 480mbps constant write speed, more than enough for the LongGOP modes and the 4K120 LongGOP :)
Yeah, the write speed on the GH7, to handle RAW, is definitely a nice to have! Thanks for comment!
You’re completely, and I mean completely wrong about 32bit float audio. You need to look into how that actually work my friend.
Hard to listen pass that mistake, but I did.
Thanks for watching past it. If I'm wrong, please explain how I am wrong. I've spent over 100 hours studying how it actually works and I can't see that it can. You can read my conclusion here: maxrottersman.medium.com/the-three-types-of-microphone-clipping-explained-d23ad13a60f7 cheers!
I thought that a digital sensor does not have a sensitivity to light, or at least not one that you can alter through a setting.
It just has these buckets/wells that are able to catch light and iso is something that comes into play only after this light is collected.
I do find it interesting what you're saying, that iso is a measurement instead of a function. Do you have some articles where this is more throrouglhy explained?
Exactly, you can't alter a sensor's sensitivity to light! When I was young there was ISO, ASA and DIN. You can read about them on wiki, they're under "Film Speed". The problem for most photographers/filmmakers for the film era was exposing properly. How could you expose properly if you didn't know how one film stock was going to react compared to another? In early days, film (glass plates) were a cottage industry.
Today, we have those ratings so it's generally not a problem but that doesn't mean the issues have gone away. For example, photosites reach full well capacity at different levels of light depending on which filter they have: red, green or blue. That's why you'll see magenta in overexposed clouds.
Further, ISO for a Sony S sensor is not exactly the same for ISO 100 on other sensors because it's more sensitive to lower light and less to very bright light. Filmmakers get a feel for it. In other words, it's still a challenge, if you get into the weeds, of understanding on any sensor responds to different types of light.
Glad you found something interesting in my video!
dude - your shit is a level-up - refreshing to hear succinct conviction and the facts to back - subbed - hey, i have an a7siii and want to trade for a cam that shoots raw internally - i swear i can truly see the color in raw video - the filmic nature - am i nuts/placebo? - i know a7sii has external 16 bit but i love a teeny/weeny/tiny rig and atomos monitors are a big fat ass - anyway, is my desire for raw unfounded? also, 12/14/16bit - what/s the difference and if one is making the jump to raw, does it matter? what about the differences for proprietary i.e. prores, canon raw lite< n-raw, dng, etc etc. The Sigma fp beckons - do you know why? Anyway, bunch of stuffe here, but hopefully there some gist/cohesion that might inspire you to do a vid. p.s. - what about raw for AI imports - do you think AI is going to love RAW for manipulation? As I'm shooting more abstract/music video/art type stuff - i think i might qualify as someone who manipulates video intensley - lastly, cheers again for the vid ✌🏻
The A7SIII is a monster camera for film because the sensor is optimized for video. Supposedly, we can discern around 12 million colors. 8 bit can display 16 million (256x256x256). So if you can light the way you want you don't need more than 8 bit. The only reason you need more is if you can't control the light and want to make some tradeoffs.
On the other hand, most compression schemes share chroma values between pixels, reducing storage. Biology drives all this stuff. We notice blurriness, but not missing colors if they stream by quickly. I calculated it once, I believe when you watch something on Netflix you're looking at 5% of the data the camera recorded.
Anyway, RAW gives you the most control over the data because it doesn't discard anything. But in practice, you really don't want to deal with RAW because the data is huge.
If you're doing abstract/music/video type stuff I think you should try RAW only because it teaches your (or taught me) what really comes out of the camera. It allows you to do experiments that compressed video might get in the way of. Something I'm interested in too.
There's little difference between 10/12/14 bit RAW, or at least I can't see it. The main benefit is data points per pixel, not pixel blocks. If you want to experiment with RAW I'd get a cheap Canon and Magic Lantern. It gives you a lot of options. But again, I doubt you'll shoot RAW in most of your stuff. But who knows!!!
@@MaxoticsTV Cheers for such a quick and generous reponse. Could you point me to a good vid on magic lantern/canon workflow? Any thoughts on AI with Raw? So you're stating that manipulation aside, one cannot perceive the increased color-depth intuitively in an unchalleging grade?
@@thiskidkills7806 Magic Lantern is a community of weirdos ;) Ever since I got involved (and I haven't been for years, but some recently) everyone talks about documentation, tutorials, etc., but it just doesn't happen because everyone would rather work on their project. I've been playing around with Bilal Fakhouri's build. Here's a thread. www.magiclantern.fm/forum/index.php?topic=25784.0
It's one of those things just watch videos and dive in. Everyone in my experience is VERY HELPFUL and the stuff is so technical difficult (and not forgiving) that you don't have trolls. Too much work for them ;)
Anyway, join Magic Lantern's forum, look around for stuff you're interested in. It's one of these things you either fall in love with or think "life's too short" HAHA
@@MaxoticsTV duly noted 😁
@@MaxoticsTVi've tried raw video on a pixel2 phone (2017 phone). i am an amateur and have used various cameras - mft (8bit, but 10bit as well) and apsc (8bit). i was impressed with the raw a 7yr old phone can capture. did not end up actually using the raw route because of various phone-related limitations and indeed the workflow and data storage which are pretty ridiculous (at least for a hobbyist), but yeah.. raw even on a phone was impressive (for me)
I have an a7r3 and a GH6. I can tell you without a doubt the level of video quality you can get out of the Panasonic is far beyond the older generation 8 bit only video that comes out of the Sony. With the Panasonic 10 bit, you get over 1000 levels of gradation in the value; the Sony only 256. The only place is Sony might have a beat is low light performance. Take it from somebody who uses both buying the older generation Sony is not gonna beat the newer generation MFT. And I don’t think it’s a detriment to Panasonic at all to have internal raw, they need to throw everything they can to keep this platform alive. I would say their biggest problem is they need to lower the price to compete with aps-c cameras that have similar specs.
Some thoughts, it I may. The a7r3 is a high resolution camera made primarily for photography. It has a pixel pitch of 4.51um. The GH6 has 2.99um, so they're not far off. The GH6 will do sharper video IMO for a variety of reasons. If we look at the A7SIII, however, it has a pixel pitch of 8.36. Almost 3 times the GH6. It's not an issue of brands, just the physical characteristics of sensor size and pixel size. So I agree with you that a high resolution camera like the A7R series isn't going to threaten MFT, but the video optimized sensors will--including APS-C.
I don't believe the 10-bit CODECs do much, but I don't have the experience you have. Their problem is that cramming 10-bits into an 8-bit range doesn't give you colors that people can notice. You only get some improvements with banding. If you extend those 10-bits beyond the approx 6-stops of dynamic range of 8 bit you pull in noisy data because that's what at the ends of each side of the 6-stop range. I like the very gentle LOG shooting gammas, like cinelog, I believe. Trade some noise for a less contrasty image.
I like my GH6 and I'd be interested in a GH7 for the improved focus and sure, RAW would be nice here and there. I get the feeling Panasonic doesn't have a clear vision for their cameras. Thanks for comment!
@@MaxoticsTV I had an a73 as well, sold it. I still think the Panasonic beats the 24 megapixel a73… again, not in low light though. There is something else: the codecs. The Panasonic will do all intra at a much higher bit rate than the older Sony’s. I admittedly have not used the newer Sony’s, nor the a7s3, so I can’t compare the graded footage (before the degradation that happens in CZcams uploads). Suffices to say, for many reasons I am very satisfied with my MFT gear, but they do need to stay competitive, and $2200 cameras are hard to swallow when you can get FF for $1500! For my work, I actually need more depth the field and really good stabilization, so micro for third makes more sense for me right now.
@@3dtrip870 You read my mind! I wrote that the GH7 needs to be $1,500 but then figured I shouldn't go there so deleted it. I've been shooting a lot of my grandkids with the GH6 and the autofocus, though it hunts here or there, is totally fine to me. And I've always felt panny MFT footage very clean and tight. Shallow DOF is a double edged sword. Easy to overdo it. I feel the MFT gives just the right amount of separation for me. The IBIS is also just the right amount for me.
@@MaxoticsTV yep, the autofocus isn’t too bad…but I’d rather have the GH7 🤣