More than 30dB clip volume

It would be great to be able to adjust clip volume beyond +30dB. I'm getting more 32 bit files recorded at lower levels and the need to go beyond +30dB is becoming very apparent.

Is there a technical reason why it's capped at +30db?

Because mic preamps have their own noise floor, plus mics have self-noise, I think the concept of recording in 32-bit is folly most of the time. The reason why is, because the track has been recorded at a relatively low level, you're now cranking up the signal and amplifying the mic, the preamp, and the track itself. You might protect against overload, but the problem is the low-level stuff. I don't see this as a way to avoid having an actual sound recordist on set putting the mic in the right place and keeping the levels optimized.


Having said that: I know there are plug-ins that will allow you to jack the gain up again on the input channel, so that could be a potential way to do what you want. But again... noise floor.

Philip, 32 bit files are useful for protecting against sudden loud audio, but in this case whoever recorded your audio did not take advantage of it. To need more than +30db of gain means that with the huge increase in the volume of the noise floor, the audio is likely to be unusable.
What you can try as a test is up the volume by that 30db, bounce to another track and up its volume.
Good luck.

the problem is the low-level stuff.

Weelllll....

Go to the demo around the 1:30 mark.

Incredible. Thanks for that link Jim!

Because mic preamps have their own noise floor, plus mics have self-noise, I think the concept of recording in 32-bit is folly most of the time. The reason why is, because the track has been recorded at a relatively low level, you're now cranking up the signal and amplifying the mic, the preamp, and the track itself. You might protect against overload, but the problem is the low-level stuff. I don't see this as a way to avoid having an actual sound recordist on set putting the mic in the right place and keeping the levels optimized.


Having said that: I know there are plug-ins that will allow you to jack the gain up again on the input channel, so that could be a potential way to do what you want. But again... noise floor.

32bit float recordings do not work the way you think they do.

You don't set a true gain per se, just a virtual gain. So if someone set the virtual gain too low, you are stuck with Resolve then. That does not mean you will increase the noise floor.
You can set the virtual gain on the recorder to -128db if you want and you will see just a flat line in Resolve. But you can normalize it back and you will have a perfectly nice recording. Of course only if the sound department did its job correctly in mic placement and the like.

also wondering why it is capped at 30db. why not just let the user decide?

Thanks for your responses. I think some folks have a misunderstanding of 32-bit recording and are applying their knowledge of 24/16 bit files to 32-bit.

We're all well aware of the noise floor, but this feature request wasn't about asking how to gain stage properly, it's about asking for more than +30dB of clip volume. If the only reasoning behind that limitation, as others have mentioned, is essentially "If you need more than +30 clip volume it means you should have recorded it better in the first place" then I would definitely recommend reading up some more on 32-bit recording because that thinking doesn't apply to 32-bit. Here's a great primer:
https://www.sounddevices.com/32-bit-flo ... explained/

@Charles Bennet, gain increase with 32-bit files does not increase the noise floor like you might experience with 24-bit files.

@Marc Wielage, these files were recorded by a senior recording engineer at a prominent recording studio, so we had an "actual sound recordist on set". With 32-bit, the low-level stuff is actually not a problem and gain staging becomes more like a convenience in post rather than being necessary to avoid the noise-floor / clipping.

I still think there's a valid feature request here. If you only work with 24-bit files, or don't need more than +30dB of clip gain, then definitely feel free to not use it, but for those of us who work with 32-bit files it would be a welcome change. If there are any legitimate, technical reasons why we have this limitation I would love to know!

+1


Sent from my iPhone using Tapatalk Pro

+1

@Marc Wielage, these files were recorded by a senior recording engineer at a prominent recording studio, so we had an "actual sound recordist on set". With 32-bit, the low-level stuff is actually not a problem and gain staging becomes more like a convenience in post rather than being necessary to avoid the noise-floor / clipping.

Your mistake is assuming that clipping is a big problem. It's not. Have the recordist keep an eye on levels and you'll never clip.

Trust me: the mic self-noise, preamp noise, and on-set noise will kill you if you crank it up 30+dB.

One of the problems with asking for free advice on the internet is that sometimes, the answers you get may not be what you expected or hoped for. But it doesn't make the advice any less true.

Go over to Jeff Wexler's JWSoundGroup.net and ask around from the professional location sound mixers there and ask them what they think about people who record in 32-bit. (Note that 24-bit already represents a S/N ratio of 144dB, which is enough to blow out the eardrums of any human alive. If you need more than that, something is definitely wrong.)

Your mistake is assuming that clipping is a big problem. It's not.

I must not be explaining myself properly if you think I'm arguing that clipping is a big problem. I'm pretty positive I didn't write that, rather that 32-bit recording levels are not as relevant as they are with 24-bit recordings and that increasing the gain in post does not affect the noise floor, nor is it necessary to "keep an eye on levels" to never clip. I write this to illustrate the point that since the recording levels of 32-bit files doesn't really matter, and that those levels can be set very low without consequence, then there actually is a good reason why we might need more than +30dB of clip gain in DaVinci Resolve, which is what this original post is about.

You should do some tests yourself. Plug a mic into a 32-bit recorder and set really low levels on the recorder, then raise those levels in post. No noise and a clean signal. Now imagine you can't raise your levels loud enough to normalize that recording properly because you're capped at +30dB of gain. Kinda sucks right? Hence this feature request.

If you think we shouldn't have the option of more than +30dB of gain for a specific reason feel free to mention it, but so far all your points have been about "actual sound recordists" and reducing noise in the analogue chain, which isn't what this feature request is about.

+1
This is just about supporting new tech.

+1

Trust me: the mic self-noise, preamp noise, and on-set noise will kill you if you crank it up 30+dB.

Curtis went more than +30dB in the demo I lined earlier.

The result is...pretty impressive.

I must not be explaining myself properly if you think I'm arguing that clipping is a big problem.

Quote from you:

"With 32-bit, the low-level stuff is actually not a problem and gain staging becomes more like a convenience in post rather than being necessary to avoid the noise-floor / clipping."

Try just using a sound recordist on set, have them ride the levels (which has been done in sound for more than 90 years), get the mic close to the performer, and you'll never even begin to exhaust the 144dB dynamic range of 24-bit digital audio. Gain-staging starts right at the mic; you're trying to fix it at the very end of the signal path, which goes against everything I know from half a century of sound recording.

Turning up the audio more than +30dB merely delays the decision of getting the right levels on set to trying to fix it in post. It doesn't work that way. If you need to turn something up +30dB on a mixing panel, something is very, very wrong.

Tell me what kind of mics and preamps you're using on set, and I'll give you some realistic numbers of what they measure. As far as I know, nobody even has experimental mics that will do more than 90dB, and preamps are even noisier. The Sennheiser MKH800 might be the one mic that's an exception, but I know of nobody actually using those on a film set (partly because they're somewhat fragile, partly because they're $3200).

Philip, 32 bit files are useful for protecting against sudden loud audio, but in this case whoever recorded your audio did not take advantage of it. To need more than +30db of gain means that with the huge increase in the volume of the noise floor, the audio is likely to be unusable. What you can try as a test is up the volume by that 30db, bounce to another track and up its volume. Good luck.

Listen to this man.

BTW, to Mr. Bowser: if you'd like to learn more about recording dialogue for motion pictures, I'd point you towards these books...

If you really want to understand how to edit and mix dialogue, read these:

"Dialogue Editing for Motion Pictures"
by John Purcell
https://www.amazon.com/Dialogue-Editing ... 0415828171

"Pro Tools for Film & Video"
by Dave Angell
https://www.amazon.com/Tools-Film-Video ... 0240520777

"Sound for Film & Television"
by Tomlinson Holman
https://www.amazon.com/Sound-Film-Telev ... 0240813308

"Producing Great Sound for Film & Video"
by Jay Rose
https://www.amazon.com/Producing-Great- ... 0415722071

"Audio Production & Post Production"
by Woody Woodhall
https://www.amazon.com/Audio-Production ... 0763790710

"Production Sound Mixing: The Art and Craft of Sound Recording for the Moving Image"
by John Murphy
https://www.amazon.com/Production-Sound ... 1501307088

"Audio Post Production for Television and Film"
by Hilary Wyatt & Tim Amyes
https://www.amazon.com/Audio-Post-Produ ... 0240519477

Post Sound Design: The Art and Craft of Audio Post Production for the Moving Image
by John Avarese & David Landau
https://www.amazon.com/Post-Sound-Desig ... 150132747X

(I'm quoted several times in Purcell's book, but don't let that throw you.) I've been in this game for a long, long time, and I still learned quite a bit by going through all these books and benefitting from their wisdom. I have seen some horrendous disasters in post because neophyte filmmakers thought they were going to "try something new" and try a different path for something like sound recording, and it went terribly wrong... to the extent that it cost them thousands of dollars in fixes. I could tell you the story about how all of Robert Rodriguez' sound for El Mariachi was unusable (bad sync) and had to be totally re-recorded from scratch, racking up another $100,000 to his expenses, but that's another story for another time.

Thanks for the resources Marc.

I'm sure you have a lot of great advice to give, and I don't mean to insult you, but I don't know how else to re-iterate that this feature request is not about asking for advice on how to record audio, gain stage properly, or the importance of sound recordists (which of course they are), it's about asking for the +30dB clip gain limit to be lifted.

I don't think embracing a feature request like this is going to hinder you or your work whatsoever. However it will definitely help folks like me who use 32-bit audio, which again, can have completely arbitrary recording levels because it doesn't use the traditional concept of "gain", and can be raised and lowered in post without consequence. I feel like I've explained that a lot. Simply put, 32-bit recording is a newer technology, and new tools and thinking are needed to take full advantage of it.

It's funny how we have all of these incredibly advanced tools to edit, fix, modify and completely re-shape our audio and video into these new and wildly different creations, and you're so hung up on not changing a volume slider because if that's the thing you need to fix then it means you have to start all over, and not because the data isn't perfectly intact to fix it or some other technical reason, but only because it's how its "been done in sound for more than 90 years".

I'm sure you have a lot of great advice to give, and I don't mean to insult you, but I don't know how else to re-iterate that this feature request is not about asking for advice on how to record audio, gain stage properly, or the importance of sound recordists (which of course they are), it's about asking for the +30dB clip gain limit to be lifted.

I think you're somebody who doesn't know a lot about sound who is confronted by the realization that maybe the way you recorded it was not optimum for your project. I would draw a comparison to somebody who thinks they can just point an 8K camera on a wide shot, make one exposure, and then fix all of it in post, including reframes, exposure problems, and so on. This is not a strawman argument: I come upon quite a few students who think this can be done, and they aren't aware that basically you have to get it right on set, and you can't change it all in post. Not a single one of the half-dozen books I posted says, "just record it in 32-bit and you can change it all in post." It just doesn't work that way, for the reasons stated before.

If you like, I can reach out to the sound mixers on Wexler's forum and get a topic started on whether it's a good idea to record motion picture dialogue at 32 bits. We had one about 15 years ago and it got very animated and excited very quickly, because basically somebody recording this way is typically trying to just save money by not hiring a decent sound recordist on set. (I'm not saying you are -- I'm saying that's what the upshot of the conversation became.)

Most of my post experience is with Pro Tools and not Fairlight, but I can tell you in extreme situations when we got really substandard audio recorded at very low levels (like a forensic-recovery situation), you can just insert a Trim plug-in into that channel, and that would automatically double the available range. Back then, the tracks were limited to 12dB, and the Trim plug-in I think would give you another 12dB. Note that a change of 1dB is a doubling in sound energy. What does a 30dB change do? From the Merck manuals:

...an increase of 10 decibels (dB) represents a 10-fold increase in sound intensity and a doubling of the perceived loudness. Thus, 20 dB is 100 times the intensity of 0 dB and seems 4 times as loud; 30 dB is 1,000 times the intensity of 0 dB and seems 8 times as loud.

https://www.merckmanuals.com/home/multimedia/table/measurement-of-loudness

So you're basically asking, "hey, how can I adjust Fairlight to give me 1000 times the intensity of the original recording? And that's why an engineer's jaw would drop when asked that question.

I think you're somebody who doesn't know a lot about sound who is confronted by the realization that maybe the way you recorded it was not optimum for your project.

Sorry Marc, he may have failed to mention it, but Philip has obviously been receiving files recorded in 32-bit float, which is what we work with here and you don't seem know anything about. Look it up.

when we got really substandard audio recorded at very low levels...

For context, was that audio recorded in 32-bit float?

On which device?

Sounds good Marc!

I think we’re just coming at this in very different ways. I’m trying to make a case for why we should have tools to take advantage of all of this new data we have available to us now with 32-bit files, and you’re hell bent on trying to prove that I don’t know anything about audio and how 15 year old forum posts should dictate our technological progress.

You are 100% right that I do not know a lot about audio when compared to a veteran like you. I’m sure there’s a lot you’re right about, but I can’t help thinking that I’m more right about this particular thing with 32-bit audio. The audio data is there and is perfectly clean and sounds fantastic, but in order to normalize it I need to currently boost the gain twice by bouncing out my gain changes, a process that could be made more efficient with this feature request. I haven’t introduced any noise, and there is no audible consequence to this. This is the nature of 32-bit files, because gain doesn’t really matter on the digital side (not talking about the analogue side, which you so frequently bring up).

We have the ACES color space that extends beyond human vision, 32-bit EXR files capable of inconceivably high brightness values, raw image files with a dynamic range well beyond broadcast standards. And yet we’ve created tools to juggle all of this data to fit it into our relatively narrow window of what pleases our eyes and ears. And It is my belief that we should also have a volume slider that takes advantage of all this new data in 32-bit audio files that have become more popular recently.

Philip has obviously been receiving files recorded in 32-bit float

Yes, thanks for clarity PhilB1234, these are floating point files, not 32-bit integer files. For context, I didn’t record these files, and in this particular project they are from a documentary I’m working on which was recorded in a very dynamic and fast changing environment with extremely loud and extremely quiet moments. I believe it was the sound recordists choice to not ride the levels for the quieter parts and instead favour the extremely loud when setting the (arbitrary) digital gain, knowing that you can bring up any quieter moments you need without consequence because of the 32-bit float technology. I’m now in post, as an editor, wishing I had about 4-5 more dB of latitude on some of the quieter parts to really make it work and I’m frustrated by the limit of the volume slider. Not impossible to work around but could be made more efficient, which is why I started this feature request. No amount of "do it right the first time", or "we've been doing it this way for 90 years" is going to change the fact that sometimes you need to fix problems, and new tools are needed to fix new problems.

I’ll leave it at that. I think I’ve made my case, and hopefully Blackmagic considers this feature request. If I ever end up doing any serious audio recording myself I’ll take a look at those resources you’ve linked, but until then I’m at the mercy of what I’m given and the tools available to me, which I hope continue to change with the times.

+1

I'm not an audio expert by any means I'm a YouTube video Creator and I've had situations where I needed to lift more than +30db Especially in combination with new amazing "Voice Isolation" feature - which makes the volume lower.

I don't think that having a limit ads any benefit

You are 100% right that I do not know a lot about audio when compared to a veteran like you. I’m sure there’s a lot you’re right about, but I can’t help thinking that I’m more right about this particular thing with 32-bit audio.

Here's some feedback from professional sound mixers:

https://jwsoundgroup.net/index.php?/top ... ay-or-nay/

10 replies so far in 24 hours. A couple of them pointed out that any major post facility in NY, LA, Chicago, Miami, Dallas and so on will be unable to do sessions with 32-bit audio. So you already have a workflow problem just with the 32-bit mix. And AAFs fail in this mode (at the moment) with Pro Tools. 24-bit works fine (and as I said, already has 144dB of dynamic range).

This is one of those things where if I could stop a client before they went into production, I'd say, "I beg of you, get a post supervisor and do a workflow test first, because this could be a major trainwreck for sound." I think your technique of bouncing the WAV files out and then turning them up again is kind of a sledgehammer approach to force it to work, and it's valid in a salvage operation. But that's one of those things that could be cured if you had cranked it up "somewhat" on set and given yourself -10dB of peak headroom, which is pretty standard. It's never going to clip, not even if the actor screams... but there's always the chance the mic itself will overload or the preamp will distort.

+1

I believe there is a confusion about the 'physical' volume and the capturing volume/scale. If a CG render spans a range of 0 to 53.5 or something and you linear gain down to have it sit between 0-0.7, store it, then gain back up to 53.5 you get the exact same data again. There is no increase in noise. I think we can apply that same concept to 32bit audio recording. If the actual sound was as loud as the refrigerator noise and you'd want to amp it up of course it will sound terrible but I don't think that is the case that sprouted this feature request.

We simply need tools to manage this data the same way we have tools to manage 32-bit images.
Maybe clip volume? Maybe other tools to define a base line gain level as a preprocess.

any major post facility in NY, LA, Chicago, Miami, Dallas and so on will be unable to do sessions with 32-bit audio.

Sounds like (no pun intended) they're a little behind the times and need to catch up.

To me, this would be like requiring ProRes footage because you can't use ARRIRAW.

any major post facility in NY, LA, Chicago, Miami, Dallas and so on will be unable to do sessions with 32-bit audio.

Sounds like (no pun intended) they're a little behind the times and need to catch up.

No, it's a Pro Tools problem -- I'm told a clash with AAF. The explanation is at the link I posted elsewhere.

Of course, you could to edit and mix all sound in Resolve's Fairlight, and I'd bet there's got to be some facilities using that. I don't know of any offhand, but anything is possible. Eventually, though, the signal has to come out and get distributed, and all the major studios and streaming services are requiring 24-bit for final-mix delivery, last I checked.

Eventually, though, the signal has to come out and get distributed, and all the major studios and streaming services are requiring 24-bit for final-mix delivery, last I checked.

Marc, you really should look up 32-bit float if you're going to keep commenting this feature request. No one is trying to use it for distribution. It simply offers 1528 dB of dynamic range for totally safe/worry-free recording, allowing you to do (not fix) it in post.

That last bit is where many misconceptions about 32 bit, 24/16 bit, and 32 bit *float* seem to come from. Mine were, certainly.

The staggering dB levels available *without clipping or noise* are simply mind boggling.

I thought it was nuts until someone sent me a file they had recorded for a shoot for an aircraft maintenance firm. Workers on a big jet, right up next to it, working in quiet while talking. Then starting up that engine and running it up some.

All without any adjustment to the recording levels.

The quiet part was clean, birds in background, talking clean and crisp. No noise whatsoever. And the jet engine running was about the cleanest I've ever heard.

All from one file with no recording adjustments made. You could simply mod the dB levels for playback, smoothly and efficiently, for best sound Q.

What's that line about sufficiently advanced tech being the equivalent of magic?

Sent from my SM-S908U using Tapatalk

I thought it was nuts until someone sent me a file they had recorded for a shoot for an aircraft maintenance firm. Workers on a big jet, right up next to it, working in quiet while talking. Then starting up that engine and running it up some.

We're only talking about recording dialogue on a film set with actors (or an interview). Not a jet plane engine. An aircraft engine hits about 140dB on takeoff... and that's still well-within the bounds of the 144dB dynamic range of 24-bit audio. (Of course, multiple engines could affect this.)

90% of the time you hear/see a jet engine in a film, the sound has been replaced with SFX because it sounds better. Trying to actually record one life is pretty tough. As an example: that's how the Top Gun: Maverick crew recorded the jet plane sounds around Tom Cruise. They didn't try to arrange each take to make sure they recorded "real" sound -- it was all dubbed in later, so the picture and sound could each be recorded under optimum conditions.

Marc, you really should look up 32-bit float if you're going to keep commenting this feature request. No one is trying to use it for distribution. It simply offers 1528 dB of dynamic range for totally safe/worry-free recording, allowing you to do (not fix) it in post.

I'm aware of that. Again: have you ever recorded sound? Did you consult the references I posted above? Did you check the JWSound link I posted above? How will 1528dB of dynamic range work with humans talking on a set? 130dB is already the threshold of pain, and I think by 150dB, everybody there will be deaf for days (if not forever). Explain to me why the 144dB dynamic range of 24-bit recording is insufficient. Cranking up noisy sound will increase the noise floor long before it makes the dialogue usable.

As an example let's say your mic and preamp and signal chain have a self noise of -60dbfs and you record your dialogue at -30dbfs. It makes no difference in 16,24,or 32bit float if you increase the playback level of your dialogue to -6dbfs, that differential of 30db between the noise and dialogue will remain the same. So not only is your dialogue 24db louder, so is the noise. On top of that try adding a compressor to the track and see what that does to the noise.
32 bit float is a great tool for recording in situations where sudden bursts of sound may exceed 0dbfs, but it is not a cure all for recording at too low a level in the first place. In itself it doesn't add noise to a recording, but it is not a noise reduction or removal tool for a noisy signal chain.
Going back to the original post, needing more than 30db of gain means the initial recording must be extremely low in the first place.
As to the maximum gain of +30db in Resolve that may be historical. Being a tool used only by professionals it would be expected that the audio recordings would be at good levels in the first place and not require a lot of level increase. Now the user base is much wider with more non professional users there could be an argument for increasing its ceiling.

As to the maximum gain of +30db in Resolve that may be historical. Being a tool used only by professionals it would be expected that the audio recordings would be at good levels in the first place and not require a lot of level increase. Now the user base is much wider with more non professional users there could be an argument for increasing its ceiling.

Very well-said, Charles. I appreciate your input.

Jeez, guys... No one here is claiming that 32-bit float somehow magically improves S/N ratios of whatever mics, cables or additional analog signal chains you might have going into such a recorder’s preamp. That'd be like expecting a raw format to improve the quality of your sensor or glass. Think more along the lines of bracketing in photography. Better yet, look up the difference between integer and float recording and how the preamps of a MixPre II or Zoom F6 handle gain levels. Or don't - but please stop throwing uninformed arguments into this feature request. Resolve handles the 32-float files just fine, so it should not limit their volume range and force detours through other DAWs.
The OP clearly stated that he has been receiving such files. How does that that lead to lectures on proper recording, let alone the left and right insinuations over lacking knowledge in sound?

It would be great to be able to adjust clip volume beyond +30dB. I'm getting more 32 bit files recorded at lower levels and the need to go beyond +30dB is becoming very apparent.

Is there a technical reason why it's capped at +30db?

+1

The discussion on JWSound about 32-bit dialogue recording now has 33 replies and has had 1000 views in about a week. In general, none of the sound mixers there seem to think it's a good idea, and they point out that almost none of the 32-bit A/D converters can really do 32-bits, which surprised me.

One helpful person did point out that you might be able to use the Normalize function to bring up the low sound levels, and that would average the level of the entire track and bring it up as high as it can go without clipping. But note you could have done that on the set as well.

The discussion on JWSound about 32-bit dialogue recording now has 33 replies and has had 1000 views in about a week. In general, none of the sound mixers there seem to think it's a good idea, and they point out that almost none of the 32-bit A/D converters can really do 32-bits, which surprised me.

I just read that discussion on JWSound. The problem is that you presented it with all the wrong assumptions you made here which in turn lead to the replies of the sound mixers there basing their answer on false presumptions.

I don't know why you still keep riding this horse although a lot of people here tried to correct your skewed narrative.

I just finished the post-production on a feature which will premiere next week in the cinemas all over my country. We had an audio engineer on set with the Sounddevices MixPre-10 II, boom with Schoeps CMC and several lavs. All recorded in multi-track 32-bit float. The audio guy still kept riding the gain because it was aux'ed into the camera's scratch track and for the director's monitor. Overall it was a pleasant experience and turned out great. We did a pre-production prep for the workflow to catch any surprises.

Absolutely nobody was expecting a magic pony trick out of 32-bit float recordings. You still need sound guys knowing what they are doing. When a workflow has been established it just makes their lifes a bit easier. That's it.

One helpful person did point out that you might be able to use the Normalize function to bring up the low sound levels

But thanks for this idea, Marc. Unfortunately, Resolve's Normalize function maxes out out at that same +30db of the clip's volume slider (and presumably on -30db as well).

I have bought the zoom f2, and I absolutely love 32bit float. I am an amateur, and a bad one at that, but my love is for video, not sound. So a set and forget thing like 32bit float is amazing to me. Just figure it out after recording, at my pc, not fiddleling with gain and setting in the field. I find it truly amazing, I tested with recording sound much too low in volume and much to loud, but with 32bit float you can fix the volume in post without any problems, no noise, no distortion, nothing....

Not sure this is any help, but if you raise the volume by 30 db and then make the audioclip a compound clip, you can raise it another 30 db. I am guessing that is impractical for an advanced workflow?

As a workaround until BMD fixes this (they have to, 32-bit-float will become standard at one point) I recommend the free PurestGain plugin from the fantastic Airwindows.

https://www.airwindows.com/purestgain-vst/

Also, people need to educate themselves about how 32-bit-float audio changes things before applying 16/24bit logic to it.

+1

This would be very useful for sound recordist recording sound effects. Sometimes you can't be right up on a subject and sometimes you hear stuff that you want to use in the background. I shouldn't have to use any other tool other than Resolve.

Ignore the naysayers. They do that on every post in Feature Request for some reason. It's weird.

+1, I stumbled across this thread looking for a solution to the same problem. I created an account just to reply, because I was astonished to hear some of the counterpoints.

It was frustrating for me to see Marc's stance that ignores the reality and fresh applications of 32-bit recording. I capture weddings for a living, and having a Zoom F6 with 32-bit float has saved me on many occasions.

With 32-bit, I'm able to plug into the DJ's mic out and capture speech audio perfectly, without having to watch or even set the gain (I don't always have the luxury of doing sound checks, and many of these speakers are not experienced with a handheld mic). I would've paid thousands for the peace of mind that 32-bit float provides. When I bring the audio into Fairlight and use the workarounds needed because of the arbitrary 30dB limit, it sounds just as good as (if not better than) when I used to carefully monitor gain on my old Zoom H6.

It's a needless pain that we need to do these workarounds because BMD hasn't adjusted their software to account for this technology.

For 32 bit float, this just makes sense. For those that don't understand how to use it, they needn't worry.

+1

+1 to support philipbowser

It's clear that some people outdated and stiff enough that they can't accept new information/technology, etc
The video Jim Simon posted here - is a demonstration how 32bit works.
Again, the question was about DaVinci not having more than +30db volume, not about "you don't need this feature if you know how to record, go read all of those books & hire a sound dude".

Haha, I found this thread, actually, when I was looking how to make more than +30db volume
Got Zoom F3 today
Cheers!

+1 for more than 30db clip volume

I'd like to add that perhaps simply raising the volume cap isn't the best way to implement it. It means the clip volume slider becomes more sensitive... to what max range would you place it? -100 to +100? What if that still isn't enough?

Then on the clip itself the same question comes to mind. +30 is already a decent boosting range for typical non float audio but increasing the cap would require much more precise mouse movement for such an operation or zooming in a lot vertically to achieve similar granularity.

2023-07-25 14_43_17-DaVinci Resolve Studio - Untitled Project.png (21.15 KiB) Viewed 3104 times

Even if we were to ignore this workflow optimization we still run a dangerous risk of accidentally overshooting the slider on normal level clips and blasting a fully crunched signal through our speakers.

Therefore I think the best solution would be to introduce a separate entry to the clip properties that can manage the "Base Level" for a float clip separated from the clip's volume slider. This would bring your floating point clips in line with non float and the rest of your audio workflow wouldn't change.

I'd like to add that perhaps simply raising the volume cap isn't the best way to implement it. It means the clip volume slider becomes more sensitive... to what max range would you place it? -100 to +100? What if that still isn't enough?

Then on the clip itself the same question comes to mind. +30 is already a decent boosting range for typical non float audio but increasing the cap would require much more precise mouse movement for such an operation or zooming in a lot vertically to achieve similar granularity.

2023-07-25 14_43_17-DaVinci Resolve Studio - Untitled Project.png

Even if we were to ignore this workflow optimization we still run a dangerous risk of accidentally overshooting the slider on normal level clips and blasting a fully crunched signal through our speakers.

You could leave the slider as is, anything more than +/- 30db just wouldn't move the line more, only change the dB number (btw, imo grabbing and moving the slider is not very useful anyways to make precise changes).

You could leave the slider as is, anything more than +/- 30db just wouldn't move the line more, only change the dB number (btw, imo grabbing and moving the slider is not very useful anyways to make precise changes).

+1 — just let us type any number and let the current sliders be as-is.


Sent from my iPhone using Tapatalk

You could leave the slider as is, anything more than +/- 30db just wouldn't move the line more

That's a clever solution as well.

Again, the question was about DaVinci not having more than +30db volume, not about "you don't need this feature if you know how to record, go read all of those books & hire a sound dude".

You don't need to necessarily "hire a sound dude" -- you just need to know how to record and listen. Setting levels is high on that list of skill-sets.

It's a good feature request, but until it is added to Resolve I would add an audio effect that has an output gain control. The vocal channel effect has +20db gain available. You can add extra instances for more gain.

Fairlight provides multiple options to assist in appropriately gain staging through the path including:
(1) Clip Inspector Clip Gain - maximum +30dB.
(2) Path Trim (Fairlight Mixer > Input > Path Settings) - maximum +24dB.
(3) Fairlight Mixer EQ and Dynamics modules - maximum +20dB each.
(4) Fairlight Mixer Fader (or Track Inspector > Track Level) - maximum +10dB.