MspectralDynamics

[MeldaProduction]

Hehe that is actually planned as part of the "mixing revolution". But jmg8 it is indeed impossible without certain measures (e.g. global receiver on the master track). Just imagine you do this:

track -> 1 -> 2 ->...

Then instance 1 starts processing, but doesn't have audio for 2 !! There is a lot of other scenarios sort of defeating causality .

[jmg8]

I KNEW IT!!! Ha ha, I guessed this was part of the revolution. Can't wait, I'll say no more about it.
Go Vojtech!

[Soarer]

You guys are awesome!
Maybe I should start with MMultiAnalyser...
But I do get the idea.
Ah priority yes! That means that the instrument with the
lowest priority is automatically frequency ducked to make
room for the higher priority instruments right?
It could be named MSmartMixer

[AudioTraveler]

So the ideal mixing situation would be to ''inter''-sidechain every instruments?

Theoretically that's a interesting approach, however if you listen to the results you will not like it, that being total spectral matching..It would improve a lot of stuff but ruin a lot also.. You want some interference between the sounds it helps masking stuff you don't want audible..

[jmg8]

So the ideal mixing situation would be to ''inter''-sidechain every instruments?

Theoretically that's a interesting approach, however if you listen to the results you will not like it, that being total spectral matching..It would improve a lot of stuff but ruin a lot also.. You want some interference between the sounds it helps masking stuff you don't want audible..

Sorry don't agree.
Firstly spectral sidechain ducking is very transparent and if used the right amount will not have any noticeable artefacts. However the benefit will be huge in comparison as it will make more room in the mix. I already mix this way sometimes, it is very cumbersome ar the moment with loads of sidechain routing. But with a dedicated plugin will be a breeze.
Secondly it makes no sense to leave unwanted frequency content in the sound, to have another sound mask it.

[AudioTraveler]

So the ideal mixing situation would be to ''inter''-sidechain every instruments?

Theoretically that's a interesting approach, however if you listen to the results you will not like it, that being total spectral matching..It would improve a lot of stuff but ruin a lot also.. You want some interference between the sounds it helps masking stuff you don't want audible..

Sorry don't agree.
Firstly spectral sidechain ducking is very transparent and if used the right amount will not have any noticeable artefacts. However the benefit will be huge in comparison as it will make more room in the mix. I already mix this way sometimes, it is very cumbersome ar the moment with loads of sidechain routing. But with a dedicated plugin will be a breeze.
Secondly it makes no sense to leave unwanted frequency content in the sound, to have another sound mask it.

You can do it, but the sounds will sound far from the original sounds and not so good sounding.. I have not only tried my self but i took it to the extreme matched every single sound to as close as +- 0.5db.

[astramistil]

You guys are awesome!
Maybe I should start with MMultiAnalyser...
But I do get the idea.
Ah priority yes! That means that the instrument with the
lowest priority is automatically frequency ducked to make
room for the higher priority instruments right?
It could be named MSmartMixer

I suggest ''complex mixing'', because it kindof act as an organism (it ''mixes'' itself by auto-reacting depending on evolving relations between elements)

Complex:
''1. A whole composed of interconnected or interwoven parts''

[astramistil]

So the ideal mixing situation would be to ''inter''-sidechain every instruments?

Theoretically that's a interesting approach, however if you listen to the results you will not like it, that being total spectral matching..It would improve a lot of stuff but ruin a lot also.. You want some interference between the sounds it helps masking stuff you don't want audible..

In an earlier post, I wrote about a ''range'' knob for each instead of -80db ducking

Otherwise you can loose ''wanted layering'', or mix could sound thin

============
One major thing I think we should consider for best results and to avoid thin mixes, is how to use M/S and L/R with this technique. I guess M/S should be 'pre' complex mixing and as jmg8 said 'static', but L/R treated independently during complex mixing

So lets say we want to keep 2-4k for G2:
-G1 70% L (dominant on L, slave on R)
-G2 60% R (dom on R)

If the overlapping G1 on R channel is unwanted, only right should be ducked, not L+R ''2-4k" range in G1

[astramistil]

So the ideal mixing situation would be to ''inter''-sidechain every instruments?

Theoretically that's a interesting approach, however if you listen to the results you will not like it, that being total spectral matching..It would improve a lot of stuff but ruin a lot also.. You want some interference between the sounds it helps masking stuff you don't want audible..

Sorry don't agree.
Firstly spectral sidechain ducking is very transparent and if used the right amount will not have any noticeable artefacts. However the benefit will be huge in comparison as it will make more room in the mix. I already mix this way sometimes, it is very cumbersome ar the moment with loads of sidechain routing. But with a dedicated plugin will be a breeze.
Secondly it makes no sense to leave unwanted frequency content in the sound, to have another sound mask it.

You can do it, but the sounds will sound far from the original sounds and not so good sounding.. I have not only tried my self but i took it to the extreme matched every single sound to as close as +- 0.5db.

Maybe you have problem with M/S and L/R balance as I wrote. It is for sure to be consider

[jmg8]

So the ideal mixing situation would be to ''inter''-sidechain every instruments?

Theoretically that's a interesting approach, however if you listen to the results you will not like it, that being total spectral matching..It would improve a lot of stuff but ruin a lot also.. You want some interference between the sounds it helps masking stuff you don't want audible..

Sorry don't agree.
Firstly spectral sidechain ducking is very transparent and if used the right amount will not have any noticeable artefacts. However the benefit will be huge in comparison as it will make more room in the mix. I already mix this way sometimes, it is very cumbersome ar the moment with loads of sidechain routing. But with a dedicated plugin will be a breeze.
Secondly it makes no sense to leave unwanted frequency content in the sound, to have another sound mask it.

You can do it, but the sounds will sound far from the original sounds and not so good sounding.. I have not only tried my self but i took it to the extreme matched every single sound to as close as +- 0.5db.

I don't actually know what you mean. I think we are talking about two different things. As what I'm talking about is very transparent and doesn't change the sound as your describing. You talk about matching sounds? We are talking about frequency ducking?

[AudioTraveler]

Ah no i was just talking in general regarding total spectral matching..

Which is how i perceived what you wrote..

astramistil wrote:
So the ideal mixing situation would be to ''inter''-sidechain every instruments?

[astramistil]

Ok guys just try:

https://dl.dropboxusercontent.com/u/172 ... mixing.RPP

this is the closet I think we can get yet.

This is not 100% what we are talking about though (because priority have to be set manually)

1-Put track on 1 and 2 (individual)
2-Pan as you want
3-Open MSpectral and keep L+R mode, or try 'L' or 'R' (I suggest 'R' on track 1 (if pannend L), and L on 2 (if panned R))

4-You can use 'Capture' so MSpectral approximate a threshold for sidechained track, then manually remove what 'hz' you want to keep (this is where ''priority'' would save alot of time)
5-Set ''D/W'' by ear, and look at MMultianalyzer 'pre+post' on [complex] tracks.

6-try the settings/suggestions in this excellent tutorial
http://soundbytesmag.net/spectraldynamicsdynamiceq/

==========

Its easy with 2 instruments, but G1+G2+K+B+Voc etc.. ouch. Though I get more used to it everytime I try, and it help alot setting all of them at the begining of a mix (making space early can help alot not to overdo tonal shaping, compression, etc.)

[jmg8]

Actually I think that the L and R should be linked/processed together. Like in most compression the channels are linked as default sometimes with an option to unlink. But unlinking can lead to an unnatural stereo balance.

[astramistil]

youre right for 1 track per channel; but actually I always use 2 tracks (crossover mixing)

G1-45%-60% L and some widening (dom)
G1-60-80% R + widen (slaved to G2), volume lower than G1 (actually, I use volume of that track to place G1 L in the stereofield)

I s-c the one on R channel, G1 L untouched (or few db, maybe in muddy area) (and sidechaining G2 L (slave))

1-But you think I should s-c G1 L+R equally? there would be a problem becauce G2 L (slave) would take dominance on certain hz from G1 L (dom) = listening conflicts = fatigue (I guess? what do you think?)

2-You think it could make the placement in stereofield clumsy, even if the two instruments (2 tracks each on each side) are in crossover? Maybe Im too close in to hear the clashes, or potential conflicts, but it sound clearer. Otherwise crossover = too much density.

[jmg8]

Yes I think both channels equally because otherwise it would sound like it was shifting from left to right. If you compress without the channels linked it will be quieter on one side and then the other. If you place an instrument in the stereo field usually you want it to stay there.