Q: MAutoVolume working

Official support for: meldaproduction.com
RELATED
PRODUCTS
MAutoVolume$55.00Buy

Post

I'll give that a go, but I was trying out the "Using side-chain to keep a track 'on top'" method.
DarkStar, ... Interesting, if true
Inspired by ...

Post

Ok, Vojtech, understood. But since so many people suggest that, to accommodate analog-emulating plugins, -18dbfs is best (and I have become somewhat obsessive about this), why does your plugin force us to -16dbfs? I think your friends at Waves :D have a target level in their leveller, no?

Post

Getting there :)

Here are six audio clips (blue track) - the fourth is the same as the third but with a lower volume, and the sixth is the same as the fifth but with a lower volume. I put them through MAutoVolume at increasing 'Output Gain' values and recorded the results:

Image

There are a couple of anomalies (marked with X on the orange track). I have not yet worked out how to get them 'auto-volumised'.
DarkStar, ... Interesting, if true
Inspired by ...

Post

...ok...
@pumafred...I've never come across the "-18dvfs" reference as being "standard" before and am not quite sure how that means anything in the digital domain; unless the developer/s had in fact designed that into their particular plugin/s. So -16dbfs doesn't seem to make much difference one way or the other to me. You have to start somewhere. I dropped the waves plugin after finding the input/output gain controls much more flexible in MAV...did I miss something?

@DarkStar...could you post your settings? I've never seen anything like your screen shot. I do know from experience that all of the settings must be "tuned" to the particular source/sidechain material to achieve the response I want...so it can be quite "fiddly" at times...hth.../s~
mba m2 15" | 16gig.ram | 1tb ssd | macOS 26.1 Tahoe
logic 11.2.2  | reaper 7.75 | cubase 14.0.4
focusrite.2i2 | A&H CQ18t

Post

They are audio clips in Reaper, not in MAutoVolume (MAV), plus annotation by me. But here are the MAV settings; I changed very little from the defaults (still experimenting).

Code: Select all

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
Last edited by DarkStar on Thu May 11, 2017 6:16 pm, edited 1 time in total.
DarkStar, ... Interesting, if true
Inspired by ...

Post

steve2KVR wrote:...ok...
@pumafred...I've never come across the "-18dvfs" reference as being "standard" before and am not quite sure how that means anything in the digital domain; unless the developer/s had in fact designed that into their particular plugin/s.
"Most plugins began life as an emulation of a piece of analog gear, i.e. a compressor, EQ, etc. The goal with the plugin was to emulate the sonic characteristics and/or behavior of the analog gear. To achieve this the plugin had to emulate the analog gear’s sweet spot. This means plugins also have sweet spots just like their real life 3D counterparts. And the sweet spot for a modeled plugin is basically the same as the piece of gear it emulates – 0dBVU.
...
...a signal measuring 0dBVU (or 0VU) on a VU meter is roughly equivalent to (drum roll please….) -18dBFS (RMS not Peak) on your DAW meters. -18dBFS (RMS not Peak) or thereabouts is The Magic Number! Feed your DAW a level of around -18dBFS (RMS) when recording and your convertors will be happy, your plugins will be happy"
https://mixcoach.com/levels-demystified ... ed-part-5/

Post

Of course, this is only valid for "analog emulating" plugins, or if you want to send audio out and back from DAW to analog gear, I believe.

Post

...oh, ya...io to hardware makes sense, gain staging and all that (50+yrs in analog), I was coming from purely ITB and my understanding that Melda is not based on emulations (with a few exceptions).
When I input/record into the 2i2 I make sure there is "headroom" before cliping, but as high as I can get, which I target for anything between -24dbsf and -12dbfs PFM into the DAW, and then use the input gain parameter on the first plugin after that (as needed).../s~
mba m2 15" | 16gig.ram | 1tb ssd | macOS 26.1 Tahoe
logic 11.2.2  | reaper 7.75 | cubase 14.0.4
focusrite.2i2 | A&H CQ18t

Post

If you want to get rid of those spikes in the audio, increase the speed. I've done similar tests. By increasing the speed I could get vocals that looked like a block. However the sound is bad. If the speed is too high it causes artifacts similar to pumping on a compressor or sucking sounds. IMO it's better to let MAV do most of the work and then use a compressor afterwords on the few peaks left. For what I've tried so far it sounds more natural.

Post

Chandlerhimself wrote:If you want to get rid of those spikes in the audio, increase the speed. I've done similar tests. By increasing the speed I could get vocals that looked like a block. However the sound is bad. If the speed is too high it causes artifacts similar to pumping on a compressor or sucking sounds. IMO it's better to let MAV do most of the work and then use a compressor afterwords on the few peaks left. For what I've tried so far it sounds more natural.
Very helpful post. Consistent with my experience in general with MautoVolume.

Post

Indeed - that's what I ended up doing. :)

Also: instead of getting one MAV to do all the work, try putting two in sequence and let each do some of the work - that can smooth things out nicely.
DarkStar, ... Interesting, if true
Inspired by ...

Post

I seem to recall Vojtech saying NOT to use more than one instance of MAV on a track. (??) Please correct me if I'm wrong!

Post

DarkStar, ... Interesting, if true
Inspired by ...

Post

...interesting, always interesting around here...generally I "pre-process" before MAV...ie, a vocal is always done in a track-stack, as if it were a mini mixer, where the source track will have
MADEq: for highpass and remove "nasties"...done to taste
MComp: fast attack/release to tame the "spikes"...done to taste
MComp: slower attack/release to "calm" things down...done to taste
...this track is also almost always were the swipe-edit assembly is done.
There can be several effects buses etc included in the stack as well, just depends. Then MAV goes on the stack master track/fader, where the sidechain input source come from a dedicated sub mix of some sort...hth.../s~
mba m2 15" | 16gig.ram | 1tb ssd | macOS 26.1 Tahoe
logic 11.2.2  | reaper 7.75 | cubase 14.0.4
focusrite.2i2 | A&H CQ18t

Post

I'm also trying to achieve good and faster results with mav than manually editing vocals, but until now I didn't get results I like so I started doing it manually again.
Maybe there should be a starting preset for vocals, bass, guitar..... and you only need to tweak the speed to taste if it is that easy and possible.
If someone have good starting points please share. I've already spent so much time and was only disappointed at the end.
Often the midrange get very busy and overemphasised. Tried the filters in the bottom for no heavy reaction on the highs, maybe that's the problem?! Are you using the filters e.g. for vocals?

Post Reply

Return to “MeldaProduction”