Q: MAutoVolume working

[Svama]

And alexis would help to get your vocal track without processing and with.

[Svama]

Add more details for my workflow with MAV two posts before, if interested!

[steve2KVR]

@alexis1: your comment of

achieve any minimally helpful amount of "smoothing"

sounds like you are going for compression and not just "gain riding." (Yes, I understand the basis of similarity; but that's a little like confusing a duck with an eagle because they both can fly. ) Why not put an instance of MD in front of MAV as suggested above? "Smooth" out the dynamic range first, then gain ride into the mix? Still no clue as to your signal flow, however.

My code example was mostly intended to reveal the timeline graph display, which works more effectively for me with this plugin. The settings themselves would only make sense in the context they are being used in: the input signal, sidechain signal, intended output result in the mix. I doubt they be of much use to someone else, or even for me in another project...hth.../s~

[Svama]

It's simple, mav before mdynamics, so mdynamics doesn't need too much work and the volume isn't jumping before it, it simulates a pre fx volume by hand.

[alexis1]

@alexis1: your comment of

achieve any minimally helpful amount of "smoothing"

sounds like you are going for compression and not just "gain riding." (Yes, I understand the basis of similarity; but that's a little like confusing a duck with an eagle because they both can fly. ) Why not put an instance of MD in front of MAV as suggested above? "Smooth" out the dynamic range first, then gain ride into the mix? Still no clue as to your signal flow, however.

My code example was mostly intended to reveal the timeline graph display, which works more effectively for me with this plugin. The settings themselves would only make sense in the context they are being used in: the input signal, sidechain signal, intended output result in the mix. I doubt they be of much use to someone else, or even for me in another project...hth.../s~

Thank you, steve2KVR!

To clarify, and staying away from terms like "smoothing", compression, gain riding ... I would like to have the same effect as when I do direct clip manipulation of the audio signal (by this I mean cut/cut at either end of a bit that is too loud/too soft), adjust volume up/down, move on to the next bit of audio that is too soft/too loud, repeat till done.

So what I perhaps should have written earlier is, with respect to the red/bolded/underlined bit I quoted above: I wasn't able, through manipulation of the various parameters (including the gate, though I wouldn't think that would be a big part of it), to get the leveling effect I was looking for without audible distortion. It's been a while since I used MAutoVolume, and I'm not the sharpest cookie in the drawer (see what I mean!), but this was after about an hour of trying. I learned from steve2KVR's preset (thank you ) to use the black gain line in the display to help me keep the processing from resulting in either predominant gain reduction or increasing ... my goal was (as suggested by Mr. Melda and others) to set the input gain so that some of the signal was above the line, and some below. As I mentioned, I was able to do that on this audio clip, but couldn't avoid audible distortion.

There's another MAutoVolume thread on the forum, where someone mentioned limiting the frequency of MAutoVolume's analyzer to the vocal range as a possible solution. I'm not in front of my DAW, but I only recall seeing frequency cutoffs regarding the side chain HPF/LPF ... I don't have any sidechain input though, I'm just working on an isolated vocal clip (the entire project is one track with about 2-3 seconds of vocal, and MautoVolume being the only insert on that vocal track). Is there somewhere else I can/should apply filtering so that MAutoVolume focuses on the vocal frequency? Or, maybe someone can give some further suggestions?

Thanks much!

[Svama]

Hi Alexis, sorry, have you read my workflow one side before??? Maybe this could help?

[pumafred]

There's another MAutoVolume thread on the forum, where someone mentioned limiting the frequency of MAutoVolume's analyzer to the vocal range as a possible solution. I'm not in front of my DAW, but I only recall seeing frequency cutoffs regarding the side chain HPF/LPF ... I don't have any sidechain input though, I'm just working on an isolated vocal clip (the entire project is one track with about 2-3 seconds of vocal, and MautoVolume being the only insert on that vocal track). Is there somewhere else I can/should apply filtering so that MAutoVolume focuses on the vocal frequency? Or, maybe someone can give some further suggestions?

Thanks much!

See page 16 of the Manual: https://www.meldaproduction.com/downloa ... Volume.pdf

[steve2KVR]

...at this point I'm dying to hear those 2-3 seconds of audio .../s~

[alexis1]

Dry track - MAV off: https://soundcloud.com/alexissongs/2017 ... av-off-wav




Wet track - MAV on:
https://soundcloud.com/alexissongs/2017 ... mav-on-wav




As far as I can tell it is an entirely pedestrian/unremarkable vocal snippet, but there it is ^, dry first, then processed with settings below. I just listened in my phone, it's hard to hear a difference there, but in the DAW using headphones it's identifiable. (Available to download on Soundcloud as 24 bit files).

I don't know how to do the cool "insert settings code into post", so here are the "wet" settings, where to my ears it is leveled out a fair amount (gain approximately above/below the zero line also), but with subtle distortion:

Reading from the top (Version 10.07):
1) Hold activated/blue
2) Input gain -2.4 dB
3) Output gain 3.36 dB
4) Sensitivity 100%
5) Look-ahead 100%
6) Noise Gate activated/blue
7) Range +24dB
8 ) Speed 50%
9) Noise Floor -26.8 dB
10) Speed Mode Extreme
11) Side Chain - blank/not enabled
12) Detector - not enabled

Where I am so far - The above settings level out the vocal volume across the clip well enough, but there is some subtle artifact/distortion to my ears when they are applied, pretty much across the whole clip.

So in my hands, it would be better to just cut the vocal into bits and adjust the clip gains of each segment to level it out. I don't want that ... I want to learn how to use MAutoVolume better somehow!

Thanks for reading this, and for any thoughts

[alexis1]

There's another MAutoVolume thread on the forum, where someone mentioned limiting the frequency of MAutoVolume's analyzer to the vocal range as a possible solution. I'm not in front of my DAW, but I only recall seeing frequency cutoffs regarding the side chain HPF/LPF ... I don't have any sidechain input though, I'm just working on an isolated vocal clip (the entire project is one track with about 2-3 seconds of vocal, and MautoVolume being the only insert on that vocal track). Is there somewhere else I can/should apply filtering so that MAutoVolume focuses on the vocal frequency? Or, maybe someone can give some further suggestions?

Thanks much!

See page 16 of the Manual: https://www.meldaproduction.com/downloa ... Volume.pdf

Thanks much, pumafred ... I misread the help windows on the GUI, it was all there of course!

[alexis1]

Steve2kvr isn't an allround setup with sidechain enabled. I tried for myself and it's to aggressive for me on a vocal.
Alexis it seems like you changed all parameters and ended up with your own preset.

My personal workflow is like this:

Before MAV I use a deesser (MSpectralDynamics or what you like), then MAV

1. don't use the input volume inside mav for dancing around zero, instead use the daw clip gain until you have the same effect. This helps for abing with and without MAV without volume differents!
2. play with the speed knob with fast enabled until you really hear distortion, unnatural volume changes and dial it back a bit.
3. Set up Noise floor, that it´s only catching the vocal and not all breathes and other dirt. Therefore take the quitest part of your vocal and dial the knob until it starts attaching it at noise and now turn it back a smitch, that it doesn´t affect the vocal. At this point you can also fine tune the speed, so that it doesn´t go down to fast after a vocal phrase.
4. play with sensitivity, to much sounds really unnatural to me, until 45% I'm fine.
5. Detector, highpass at 100hz doesn´t hurt!
6. Set up Output volume, so that it correlate to the input and isn´t louder or quiter!

Lookahead is still a mysterium for me. I know what it does, but don't know how to set it up in relation to the speed parameter. But cou can A/B it easily with diferrent settings!

That's it to me. And as always use your ears!

Now use your standard vocal chain after it.

And after that a second mav where you can try to use the sidechain function, feeded by the instrumental of your song.

Thanks, Svama, very helpful, I use it in a very similar way.

I find myself asking, "If it takes multiple steps and iterations to dial MautoVolume in so it sounds good without distortion, why not skip all these steps and just adjust the clip gain in the DAW once and be done with it?". In terms of your workflow - Why not stop at your step 1?

[Svama]

Then are you using waves vocal rider for vocals
But really, until now I'm preferring still the oldschool way, doing it manually, but still messing around with mav.

[steve2KVR]

...I've never actually done this "(Available to download on Soundcloud as 24 bit files)." ...a little help here ...tia.../s~

[mibby]

On SoundCloud, click on the "...More" button below the track. You will see a "download" option if it's enabled.

[steve2KVR]

...cool!...never knew that, thanx.../s~