Feature Request : Audio Ducking

Dear BMD,
Audio ducking can save hundreds of hour editing music with dialog/vocal. Sending dialog/vocal track using compressor to another track in Fairlight is not ideal way.
Hopefully next update this feature added to Resolve.
Thanks.

I don't know what would save hundreds of hours (like you said lol) when clicking on "send", then "listen" takes 5 seconds. I would understand if it's something we have to do constantly and it would add up to hours on one single project, but I can't see in which situation that could happen.

Maybe you should propose what you want to be implemented because just saying : I don't really like how it works now - doesn't help much imo

Explain what "doesn't work great", and how it would be better.

Do you mean side-chaining?

Though not something I would ever use as I mix using faders, it could possibly be an available option.

Explain how it would be better.

My thought as well, as I also find the current method quite simple.

I don't know what would save hundreds of hours (like you said lol) when clicking on "send", then "listen" takes 5 seconds. I would understand if it's something we have to do constantly and it would add up to hours on one single project, but I can't see in which situation that could happen.

Maybe you should propose what you want to be implemented because just saying : I don't really like how it works now - doesn't help much imo

Explain what "doesn't work great", and how it would be better.

LOL i don't know what to exlain, may be something that better than sending and listening compressor. Or something magic that we don't need to click anything then bummm 'it done'. LOL

I don't know what would save hundreds of hours (like you said lol) when clicking on "send", then "listen" takes 5 seconds. I would understand if it's something we have to do constantly and it would add up to hours on one single project, but I can't see in which situation that could happen.

Maybe you should propose what you want to be implemented because just saying : I don't really like how it works now - doesn't help much imo

Explain what "doesn't work great", and how it would be better.

LOL i don't know what to exlain, may be something that better than sending and listening compressor. Or something magic that we don't need to click anything then bummm 'it done'. LOL

Mmm, it would be the same, you'll have to tell which track should "send" (share or whatever the new thing would be) its content, and which track should be triggered by it.

It still would be a two very simple type of actions that need to be performed to "audio-duck" at sound track.

You need to develop your suggestions because saying, for example I would like to have better sound in fairlight - It doesn't help much.

At the opposite side, just saying for example : The audio should be better when footage is sped up, or slowed down - It looks too general, but it's not, because only developers know what goes into it and we can't propose the beginning of an idea to make it better, we can only point at the problem with examples (from samples of other DAWs, etc). They'll have to figure out about this one by themselves lol.

That's why I don't say that the actual method of doing things shouldn't change. But i don't see any compelling arguments for having something else for this very particular situation.

If you think something can be better (or more useful/different), you need to explain your suggestions a bit more, with an example of what needs to change or to be added to make it better, faster, etc.

I really think it's a better way for BMD's staff to show interest, and you'll have a better chance to have it integrated.

Thanks maxwell, may be something like this below :

Voice vs Instrument.jpg (155.66 KiB) Viewed 2675 times

There is a vocal track, and instrument track. We need a button for audio ducking, and it will generate a curve for instument track.
Maybe something like this Max,
Thanks.

I'm still not seeing any specific idea for improvement, Okie. At least nothing I would support.

Tracks can have any audio or name you wish. I would not want that to change.

Ideas that are too general run the risk of being classified as "unactionable", meaning there's nothing specific for the developers to do.

I'm still not seeing any specific idea for improvement, Okie. At least nothing I would support.

Tracks can have any audio or name you wish. I would not want that to change.

Ideas that are too general run the risk of being classified as "unactionable", meaning there's nothing specific for the developers to do.

Okay, thanks Jim,,

Please correct me if I'm wrong, but I think what the OP is asking is for a feature similar to Premiere Pro's Automatic audio ducking feature, where it analyzes a track that you specify (Dialog, for example) and it auto-keyframes the clip volume of clip(s) on another track (music track) based on the audio of that DIA track.

https://helpx.adobe.com/premiere-pro/ho ... cking.html

If that's what the OP is asking, I think that would be a nice feature to add.

Current method uses a compressor, and the way PPro does it is clip volume keyframes that are adjustable.

Please correct me if I'm wrong, but I think what the OP is asking is for a feature similar to Premiere Pro's Automatic audio ducking feature, where it analyzes a track that you specify (Dialog, for example) and it auto-keyframes the clip volume of clip(s) on another track (music track) based on the audio of that DIA track.

https://helpx.adobe.com/premiere-pro/ho ... cking.html

If that's what the OP is asking, I think that would be a nice feature to add.

Current method uses a compressor, and the way PPro does it is clip volume keyframes that are adjustable.

Thanks Serwin, exactly what i mean.

Please correct me if I'm wrong, but I think what the OP is asking is for a feature similar to Premiere Pro's Automatic audio ducking feature, where it analyzes a track that you specify (Dialog, for example) and it auto-keyframes the clip volume of clip(s) on another track (music track) based on the audio of that DIA track.

https://helpx.adobe.com/premiere-pro/ho ... cking.html

If that's what the OP is asking, I think that would be a nice feature to add.

Current method uses a compressor, and the way PPro does it is clip volume keyframes that are adjustable.

I like this feature too.

+1 for this.

Adobe's implementation of it is pretty nice. I think there's an additional control or two that could make it even better though - like an offset for how much before/after the dialog waveform to transition. Would work kinda like a matte choker.

On a related note, could something like this be used to automatically keyframe the dialog to cut out ambient? Basically a quickie dialog edit? Perhaps this already exists in Fairlight. I'm not up to speed on Fairlight yet.

Finally, people use audio ducking to avoid the music fighting the dialog. Often this is better done with notching or "scooping out" vocal frequencies from the music. But most non-sound-pros don't know this. It sure would be helpful to offer this as part of an audio ducking feature.

I think there's an additional control or two that could make it even better though - like an offset for how much before/after the dialog waveform to transition. Would work kinda like a matte choker.

With sidechain you can have a different attack and release which i guess is what you describe as before/after, you also have an additional "hold" parameter in fairlight. It looks like the adobe ducking approach is exactly what sidechain is doing. like literally. but with adjustable keyframes. The sensitivity slider is the only difference from what i can see, which looks at the time between waveform peaks?

On a related note, could something like this be used to automatically keyframe the dialog to cut out ambient? Basically a quickie dialog edit? Perhaps this already exists in Fairlight. I'm not up to speed on Fairlight yet.

that sounds exactly like a gate/expander

I actually prefer the current Resolve implementation to Adobe's.

I vote no to that idea.

My preference would be to implement sidechaining for VST plugins. This would open up many more possibilities for audio ducking, including multiband ducking which is not currently possible to do in Resolve.

I have not used Premiere, but I watched a couple videos on their auto ducking and it appears to write automation to turn down the music by a set amount during the speaking and then turns the music back up again once the speaking is finished. Does this overwrite other automation you have done manually? Do you have to run it again if you make more edits to the dialog?

If Resolve allowed for sidechaining with VSTs, then this same result could be achieved with a compressor without interfering with other automation or needing to run it again after making more edits since it runs in realtime.

FabFilter's Pro-C2 can do this (see 8:27):

Resolve's built in dynamics is missing the range control that limits how much compression can occur so that the ducked amount is held constant. It would be great to have sidechaining for VST's so we can avoid these kinds of limitations and use whatever we want.

thank Jim and takkeru, i'm using compressor. I think it's more time saving than using audio ducking for now.

thank Jim and takkeru, i'm using compressor. I think it's more time saving than using audio ducking for now.

You're welcome.

I never do voice-overs myself where the music noticeably ducks in volume while the voice is speaking, when I'm mixing voice and music I want the voice to mix into the music and duck the music just enough to allow the voice to poke through. For this task a compressor is the only way to do it and multi-band compressor/dynamics is even better (Trackspacer by Wavesfactory is brilliant). It really depends on the task you are doing, but the right compressor(s) can do any task you want and provide a lot of flexibility.

Hopefully Resolve will add sidechaining for VSTs, that will open up a lot of new possibilities for audio ducking.

My preference would be to implement sidechaining for VST plugins.

That is an idea I would vote yes to.

I actually prefer the current Resolve implementation to Adobe's.

I vote no to that idea.

I agree. I think it's better when the user learns how to side-chain a compressor, or any kind of side-chain, like a vocoder. This knowledge can be applied to any DAW, to any environment, also in the analog realm when the compressor has a side-chain input (most of them have one).

It took me about an hour to apply this knowledge to the Fairlight DAW. Every DAW is different, handles busses and channels differently. It pays off to learn the particular DAW/software you are working with.

I agree. I think it's better when the user learns how to side-chain a compressor
It took me about an hour to apply this knowledge to the Fairlight DAW.

Unfortunately Fairlight's sidechaining is extremely limited. There is no sidechaining for VST effects so you are forced to use Resolve's dynamics effect and it only provides one track to sidechain with. If BMD wants Fairlight to be a legit DAW that can replace other DAWs this needs to be addressed.

I made a request for VST sidechaining before, so for those who agree please give this thread a bump so BMD will see this as an important issue.
Audio Sidechain for VST Plugins:
viewtopic.php?f=33&t=157282

VST sidechaining would open up a lot of new possibilities for audio ducking that currently are not possible in Resolve.

VST Sidechain +1

VST Sidechain +1