Prepping a file to an Engineer's specs
Vložit
- čas přidán 13. 09. 2024
- Sometimes a client / engineer will have a special request for file processing. In this video I walk you though and example of one such request.
My control room setup is a:
Shure SM7B: geni.us/Qn3S (Amazon)
Shure A7WS windscreen: geni.us/NXFA (Amazon)
Neewer Scissor arm Mount: geni.us/5QBL (Amazon)
See home studio setups here: kit.co/mikedelg...
Mike, your explanation of 0dB has got to be the best analogy I've ever heard - for beginners or full time professionals . I'm a full time VO guy and I've got to say BOOTH JUNKIE is the only VO related channel I actually subscribe to. I've been trained by some really great VO coaches and I've come across what I would I can only describe as some "predatory" VO "Coaches" - i.e. they eat your money. Booth Junkie is unlike any channel I've ever seen - it's not "coaching" and it's not a "gearhead" channel or that God Awful "Unboxing" (although if unboxing is your thing folks no harm no foul - just not my thing)
Your channel is extremely relevant content for the industry. Whether you are a beginner, enthusiast, or a full time working VO guy you seem to tune into what is really relevant in the biz. I first came across your channel when I put in my Whisper Room - and the Rolls headphone amp vid saved me from a ton of pain and heartache attempting to chain my two Apogee Quartets (yes I got two of them because I have a travel rig and my simple brain learned one so I have two (it's all about workflow in a volume driven business) All I have to say is THANK YOU for being the real deal. Best of luck. Keep recording something amazing!
You tube dosen't deserve this comment, wonderful.
Also, you can expand this analogy to use ocean floor as sound floor and show how it affects the area of depth operation for the sub.
So agree! So grateful as a newbie for you Booth Junkie!!
I really love all your videos. I've been a voice actor for over 35 years and have worked as an audio engineer, on a nationally syndicated radio show. In this particular video, I know how to achieve this, however, on most of your videos, I pick up some valuable information. And as they say, you're never too old to learn something new. Thank you, so much, for teaching this old dog some new tricks. You're a true pro! :)
Thanks for the kind words Bennie! I'm quite sure with that experience your depth of knowledge far eclipses mine!
I just started using Reaper Trial Version for VO and I am really looking at buying a license. Since I've had it I have just mainly been playing with it (HUNT AND FIND) , BUT.... couldn't find how to NORMALIZE. Brought up this video and within seconds I had the answer....so easy. All the videos are real informative and straight forward. Have really learned a bunch. Keep up the AWESOME work... We all appreciate you so much. Thanks Mike!
I'm pretty sure they were asking for the left channel normalized AFTER the processing. Your effects chain will almost certainly change the level.
Yep. You've got to normalize after applying the effects.
On the processed track you could do it at least two different ways.
Method 1
1) select the track
2) select from the menus Item, then Apply track/take effects to items as new take". This renders the trac kwith the fx chain to a new take and makes it the active take.
3) disable the track's FXso you aren't apply the chain to the new take that already inclyudes the FX processing
4) normalize the item
5) trim by to 3db
Method 2
1) select menu Track, then Render / Freeze Tracks, then "Render tracks to mono stem tracks (and mute originals)" Not that this will render every track not just the one selected. So don't use this method if you have a lot of tracks.
2) select the rendered right channel track and apply the item processing to normalize it
3) trim by 3db
The first method could probably be scripted to be normalize post fx items.
That was my thought too, as the compressor will greatly reduce the maximum volume level from -3dBFS, and even with makeup gain it won't match.
My thoughts exactly! Plus, you don't wanna normalize before you apply your default chain, because the noise floor would be different, all the compression settings would need to be adjusted, etc.. Other than that, a good video. It would be nice if there was a setting in normalizes that allowed you to set the max volume, but it's no big deal to just drop the volume. Good video except for the backward processing on the second track.
Thanks, I was always rendering the audio track, then putting it back into Reaper to Normalize. You've saved me !
Left channel was to be untreated, but yeah, his work flow resulted in the right side being normalized before processing. I would have duplicated the pre-normalized track first and then gone nutz.
Thanks for this!
I'm just starting out with recording for public use projects. This helps a bunch
REAPER user here too, love it. It seems the second request was asking also for -3dBfs. I would right click on the processed track and "render/freeze track" then sub menu "render tracks to mono stem tracks (and mute original)" … then normalise that track to -3dBfs so they are a better A/B comparison for the Engineer
DO NOT FORGET ABOUT YOUR PAN LAW SETTINGS. Pan laws (almost always) lower the volume of panned tracks, the default in Reaper is -3db at 100% pan.
If you’re still learning about recording, mixing or anything else like me, I suggest disabling this to better your skills at balancing tracks knowingly without aids. ESPECIALLY if doing client work and something ‘hidden’ and automatic like pan law could ruin their whole processing chain, make you look like a rookie and/or cost you the gig.
Very good point. Thanks for the tip. I didn't know REAPER did that! I'm gonna change my settings right now. Actually, I wouldn't be worried about looking like a rookie with an engineer asking for those specs anyway. Normalizing is a rookie move to begin with, and why they would want a stereo file instead of just two mono files is beyond me. I can't figure out what good that would do. They would just have to split it anyway.
Great video. Although as a mix engineer, I'd never ask or expect an artist to supply in such a fashion. Plus, once any dynamic processing has been applied, such as normalisation, to any level, mix decisions have already been impacted. The best way to supply a vocal file to a mix engineer, is to record completely dry and safely use up the dynamic range of whatever bit depth you're recording at. All mix decisions should really be left to the mix and engineer. If it's a budget production where the vox artist is expected to basically supply a finished sound, say for a talking book etc. Then fair enough, but even in that scenario you'd never do what was asked of by that particular engineer. Again, great video.
It's 5AM and Ive been watching your videos for hours. Thanks to EposVox for referring me here! Love your content!
You, Sir, are a legend! And such a beautiful beard too!
Another great tutorial. Hugh thanks, C xx
It's funny when you think about it, you'll charge hundreds of dollars to vo a 30sec spot, but sit down for a half hour showing thousands of folks ( practically for free) how to do stuff !! Thanks... from a fellow lefty named Mike
I watched old videos of yours before the beard. The bead suits you perfectly man. Lookin good!
Your videos helped me get much better audio on my screencasts (different user/channel)
Took me a decade of audio production to learn what FS meant! Thanks!
I was banging my head against the wall looking for a way to change the default setting for normalization from 0 db to -3 db only to find out, that's not how it works. This video kind of confirmed that for me. Thanks.
GREAT JOB! I learned so much from this video...thank u so much...keep em coming!!
NOOB QUESTION: Wouldn't compression affect the levels, post normalization? By which I mean, since the engineer requested processed track normalized to -3dB, wouldn't adding compression AFTER normalization cause the processed file to be lower than the -3dB request? Or is it an industry assumption that the engineer would expect normalization prior to compression?
If the compressor uses make-up gain (most do) then no, it would automatically adjust the output to make up for the gain reduction. Usually you have to specifically tell the compresor to not use make up gain.
Yeah, I’d have started with duplicate un-normalised files and I’d normalise the second to -3dB after processing.
Hi!
Sounds, like this video somewhere was mixed down to mono, 'couse there is no difference between r and l chanels: not hearing that panning.
Thanks Mike! YOU are amazing!!!!! So kind, so helpful, your closing line is so encouraging! Please keep on doing what you do, your are absolutely GREAT at it and it's awesome to know that your around... somewhere in the world! I'm preparing to start my career in voice acting and I am from Mauritius! (It's an island in the Indian Ocean ;)!) Your treasured advice will precious for my work to become "amazing" :D wish me the best!
Go ahead with your videos. Simple easy and educative!
I'd love to see a video done on using reaper to meet ACX requirements
Fantastic video
I loved the submarine example
Dude you look so much like a friend I used to go fishing with, he is a DJ
Maybe that is him. 👀
You are great at teaching! Glad I found this channel
Why would it want -3db ?
Love your videos
Ridiculously useful to me a new Reaper user. Am used to using Audacity and it has a macro that preps for peak and RMS. Is there such a macro in Reaper or is it just another chain?
Ps yes I know about the newer Reaper online course. It's awesome, helped me set up my Reaper to VO-friendly preferences so quickly and shortened my learning curve hugely. Haven't gotten through all of it, will do soon!
Thank you for these wonderful tutorials! They are wildly helpful and encouraging for me in regards to my work. You rock Mike!
One question, when i normalize, my volume gets very, very loud. Is this .. well, normal?
Normalization raises the level to the highest possible level before clipping, so yes it is normal.
When you record, you keep recording levels low because digital clipping is hard clipping and unforgiving, so it must be avoided at all costs. With 24-bit recording you have more bits than you have analog noise floor (self-noise a.k.a. Johnson noise in all electronic parts puts your analog noise floor at about 20 bits in the digital domain unless you cryogenically cool them) so you have enough S/N to give yourself 12-20 dB headroom to eliminate digital clipping. Once you're done you can normalize in preparation for _truncating least significant bits._ If you don't normalize, it's not going to make the file more subject to noise like analog tape, so for the most part it's an unnecessary step, as it normalizing for -3 dB instead of 0 dBFS. But if that's what gets you paid...
I'm a dialogue editor/mixer for a living and cannot understand why any engineer wouldn't want just raw audio from you guys. There's noting you can do to it that we can't do way better.
That "engineer" will probably parallel process already existing fx on the raw file to save time and work, and then call it a day.
Have you ever taken part in anime dubbing? If yes, which one, i will watch it next
Great stuff, as always!
New to the channel. Great stuff. Thank you. Btw, what drum kit you got back there?
Just a video tip. Your video gets darker at times so set the exposure on your camera to manual. This will fix the issue
Love your videos. Would love to see a video explaining the processing you typically use, such as the REaGate, De-easer, ReaComp, ReaEQ and Master Limiter that are showing in this video. Would love to see a brief explanation of what each are, what our settings are for them, and why you do them in that order. I don't use that particular program (I use Adobe Audition currently) but could easily setup the same kinds of filters... I just don't know what to setup to make my voice sound better. I'm not looking to do VO work but rather starting a podcast but still want to sound as professional as I can. Thanks!
Is it normal to record at a lower level and then normalize to increase volume? If so, why?
You adjust the gain on your microphone to record at the highest level without peaking, but that's not necessarily the highest level of output. By using normalize and compressor you can level out the (bring up the quiets, bring down the louds) so it's more consistent, and also bring up the overall volume so it's louder overall without sacrificing quality. Search "normalize and compression for video" for some good instructional videos about them.
I love your hair-style today! so real!!
I have no doubt that everything is setup as nicely as possibly but why is it so hard to hear you vs a traditional youtube video. I'm listening on a pair of monitor speakers as well.
Mike, you have added 2 more items to your Fx chain than you had in your presets video. Can you do a short video talking about your current VO Fx chain? Thanks
I realize your comment is two months old, and the video itself is from 2016, but I noticed that as well. Looking at the two added FX in Mike's chain in this video, I see a de-esser and a limiter. Both are present in Reaper. The limiter can probably be used as is... the default settings are to -3db. The de-esser might need some tweaking depending on your voice and where your "ssssss" fall in the frequency range. By default, it's set to 4k.
Glad you're making vids again. Is that a non stock windscreen on your sm7b?
Mine is an older mic, but I'm pretty sure it still comes with both windscreens now. The bigger one is called the A7WS windscreen. It let's you get a bit closer to the mic. It's this one: amzn.to/2cmgCCb
I absolutely love your videos, but there's a pretty big flaw in this one. You would want to normalize the channel with FX AFTER all the FX are applied, not before. Notice the huge difference in the wave form? If I have to ever normalize (which I don't usually do because I like the control of using an expander first, then alternate very subtle multiband compression with an expander, de-esser and de-breather in between. Much better quality than just normalizing, which tends to bring up the noise floor, breaths, mouth clicks, etc..Normalize last, and only if you have to. I still don't get it when a producer asks for a normalized version of a file anyway. With all the other stuff they're going to do anyway, why do they care about the 20 seconds it takes to normalize? Kind of says to me, "This guy/gal doesn't know what they're doing. Yikes!"
I'm pretty sure this was a "comparison" track request. They wanted to compare a "default" recording to what his editing skills did with it. If it is in stereo, both samples are synced up and its very easy to switch which one you listening to by adjusting which speaker is playing. Normalizing simply made sure the volumes on both tracks were the same. (Although I personally would have duplicated the tracks BEFORE normalizing the "default" track.)
Mike, You have helped me more than most & many but your Reaper vids are getting OLD(Not current) and I need refreshers at times. The old vids don't have current program info. Please UPDATE the old Reaper vids with new ones for me and the new kids on the block! We NEED them!! Please be the Reaper KING and keep us up to date. This is my second job which I LOVE but I need refresher info at times and when it's not up to date it can make my head spin. We need you Dude! Don't move on too fast. We need the Reaper King to stay current. Thank you!
Mike, it looks like your recording FX chain lowers the gain staging maybe 6db. It's hard to tell on REAPER's master mixer in the video. Didn't the engineer's request for the right stereo track ask that it be normalized again to -3dBFS? "The identical voice recording with any processing you would normally use. Normalized to -3dBFS" So he/she wants it back up. What about dBTP? That is what I am getting requests for now in light of LUFS.
Thanks
:-Don
Hey Mike,
@ 4:26 when you dropped the ceiling to -3 part of the file recorded looks like its being clipped.
Is that just a visual underlay of the recording or is it actually being clipped?
So, since Im still learning allvthe tech stuff and terms I have a simple yet possibly stupid question,....
Does this mean it can create that professional sound required for higher paying gigs?
Very helpful, thanks
hi from the UK, just wanted to say that as a relative newbie to the business, I am just starting my third audio book narration I really like your method of teaching. Do you do training and demos for VO? Cheers Tony
On the topic on normalization: Have you tried the SWS Extension for the Reaper plugins?
I believe they have a "normalize to XdB" option, which would make the process even easier. I still don't quite understand the different kinds of normalization, because it seems that Reaper and SWS have different ones, and I think it would be nice if you could compare and explain the differences.
I'm wondering why the engineer didn't just take the raw track and do it him/herself.
Mike - care to do A fun comparison... Original sm7 VS. New one. The sm7b... I have an original If you want to borrow it...
Mike I would like to know at which level usually you record your file, and if you record at 48khz 24bit. thanks!!! Great voice and great channel!!!
Thanks again for your videos. Can Reaper be used to measure and adjust the overall loudness of a file, such as the ACX requirement of -23 to -18 RMS?
Yes it this feature was added in 2021
Great Video again... btw. I would normalize the second Part AFTER adding the Effects, so there is more Punch in it...
Greetings from Germany,
Chris :-)
maybe a beginner question, is there a "best order" to make your effect chain? EX: normalization -> EQ -> de-esser or other.
In general, you normalize last. That's because normalization turns up EVERYTHING -- including the noise. So, if you know you don't want noise, hum, HVAC, breaths, etc -- why turn them up loud and THEN try to get rid of them? The order of operations will vary for each artist, but a common one is: noise reduction; noise gate; compression; EQ; normalize.
@@scott_pinzon thanks for the reply, I have made many alterations to my setup to more match what you had said over the years, solid points for sure.
The spec assumes in-the-box processing. What if you're using an analog vocal strip for some of your processing? Should you bypass everything except gain?
Thanks Mike. Question though about lowering to -3dB. It appears visually that you've clipped the peak that was normalized to 0dB. Is there an adjustment made to that peak that's just not shown?
I DO VOICE-OVERS MYSELF BUT I LOVE THIS GUY'S VOICE
is it a Yellow Submarine? in mixing do you suggest normalizing individual track before splitting phrase/transient to adjust item dynamics(volume).?
Wow man thanks heaps :-)
EXCELLENTE!
Think you could do the same tutorial but with a program like Audacity?
Jake Morris I have audacity also and was wondering the same thing 🎤
Normalize is an effect under "Effects" and you can add another track under "Tracks."
1. Track > Add Stereo Track
2. Copy and paste your edited track to one channel.
3. Copy and paste the original to the other channel.
4. Highlight the entire track and normalize it. It will let you type in -3db, etc.
You can also click on the track dropdown on the left to split the stereo track.
Your process chain appears to have reduced the level on the right channel. would you not be better off normalising last?
Pan Law?
There is a hiiiis in this mic
trying to get this set up for audiobooks. can you discuss
I don't understand why would anyone ask a stereo file. Would be easily to ask for two mono file.
When I normalize it brings back my breaths that were once blocked by the noise gate. How would I correct this?
You have the right workflow, since you're gating before normalizing. Normalizing turns up EVERYTHING, so best to ditch the sounds you don't want, before making everything louder.
Three possible solutions. One, your raw file may need to be recorded louder, so there's not such a dramatic volume difference between raw and normalized.
Two, re-check your noise gate threshold. In your DAW, use whatever allows you to check amplitude statistics (that's what it's called in Audition) to learn your RMS Average Volume. Noise gate threshold should be within 2dB (+ or -) of your RMS Average. (Long way of saying, your noise gate might not be aggressive enough; if the threshold is too high, it won't catch enough of the breath sounds.)
Three... just apply noise gate a second time after normalizing... but this is likely to also emphasize sibilance, so you might THEN need a de-esser. Best to get those levels and thresholds correct the first time! (I realize you posted 2 years ago, but this is a really common problem... hope this helps someone.)
Mike, thank you for these videos. In your own tracks, do you typically normalize to 0 dBFS or do you normalize to some other value?
I do not normalize unless the client requests it (and so far it's been very rare when they do.)
You can get inter sample clipping if you normalize to 0dB. You need to normalize to -3dB to make sure you avoid that. It is mostly a problem if you later convert the file to a compressed format like mp3.
Hey Mike, I need someone to walk me through how to go from raw voice over to a mixed master for AXC. Do you have anything like that? I need the chain to follow. Just asking. Tom
Try udemy.com. Look for Paul Jenkins course: The complete audio production and narration course. This course is designed for ACX -everything you'll ever need to know. I have done the course and recommend it. Hope this helps.
@@fmj4138 Hey, thanks for the response. I will check it out.
Yeah... but why? What the hell?
Hey man! I'm using the Behringer Xenyx X1204 USB, would the Shure SM7 B sound as good as yours? I heard that you also use a Cloud filter or something, is that boosting your voice?
Hey just to chime in here, the "sounding as good of yours" would be much more due to the space you're recording in rather than the mixer you use, reference his videos talking about room acoustics. If you're looking to do voice recording, specifically audiobooks and such, you want a very sound absorbent, "dead" room with lots of sound absorption panels, bass traps, etc.. And a badass voice. Hope this helps a bit.
Thank you man! It helped a lot!
When i normalize, my audio becomes very, very loud. Is this.. well, normal?
I would write that engineer back and ask him what he's drinking. Normalization is a total rookie move. I never normalize. I use an upward expander so I don't get the noise floor issues that you can see in the wave form in this video. Especially with that mic, which tends to be noisy anyway because you've got to crank the gain so high to get a decent level out of it. Everything about what that engineer is asking for is just bizarre. Why in the world would he want the file in stereo instead of two mono files? I've never heard of a request like that. Can someone explain in which context that would be useful?
If the engineer processes a lot of tracks and has a workflow that is based on this standard, or an automated process to separate and blend or select one or the other.
What is the upward expander you mention? Normalization that ignores things below a certain level? Or does it multiply amplitude and change the dynamic range?
i will pay the Shure Sm7b mic
wht els do i need with it to start recording ?
only the interface or need more things to get it work on my pc ?
Egy Knight
Interface and a cloudlifter for the sm7b.
Maybe I missed something, but what on Earth did you add the pitch-shift effect for? Also, as a musician, you always hear people saying to never directly copy and paste a part to make it stereo, or it can cause some comb-filtering/distortion issues. Or if you do that, offset the second track a bit to get rid of that issue. That being said, I've actually done this a fair amount of times during my music-recording, when I'm feeling lazy, and I don't think I've ever had a problem doing this haha.
As an engineer I think it’s a bit odd to be asking such things of your client.
I don't hear left or right difference, somehow your recorded. video makes it both channels become sterio.
From where I live, Québec/Canada, you cannot legally call yourself a "sound engineer" if you don't have a "real engineer degree".
Sure, you can call yourself an "acoustic wizard" and brag about your diploma from Hogwarts School of Witchcraft and Wizardry, as long as you don't use the word "engineer".
Miiiiiiiiiiike, Where are you, I need your help!
I managed to do my first short thing on ACX, 2200 words, and they accepted the file.
Now I've taken on something a bit larger and this whole procedure for normalization is just not working out for me, I'm committed to this, I have a PreSonus interface and an NT2 - A, some of my work is on my little CZcams channel.
Here's the problem.
I finish all the edits, breath removal, tighten the delivery timing and then I figure it's time to normalize.
But when I do I get peaks the make the red lights light and then I grab the overall volume and reduce it since acx wants -3 peak values, I do what you showed.
Still when the file plays back the red lights light.
I did install those sws extensions to figure out the LUFS, but it still too loud.
Please advise me, I've listened to you, Bill DeWees, VOBS and VO Buzz.
I'm old, I don't have much hair yet and I'm dragging it out by the handful!
(okay, i'm not, but I have to send in good files)
Mostly great stuff (as usual). However, never gate. Expand, sure. All day long. Gate? Never. Ever.
Isn't gating the best way to get rid of unwanted little noises like breaths and lip smacks? Why would you say "Gate, never ever" when it is a common workflow for virtually all voiceover? Educate me!
@@scott_pinzon Gating is binary ~ sound/no sound. The gate is either open or closed. What I’ve had a lot of success with is serial expansion, which turns down the noises, but never goes all the way to digital black. You don’t want black defining the pauses between sentences, you want tone on those spaces. A gate blacks out tone. Expansion leaves tone intact.
need a new lamp budy?
Not sure what monitors you're using but the sound from you 7b is absolute mud. Sounds like there is a loawpass filter on or something. Don't make ur videos sound like that if ur an Audio Engineer