Stereo Tool

Actually in Phantom's preset I hear most improvement in the highs! So I'm not too convinced yet.....

Well, at least we're on the same page with something. I was going to suggest this a few days ago...

Can you add it back to Classic as well? Perhaps only visible with the "Extreme Tweaker" setting?

The calculation is still there (in the old MB), it's just not implemented in the new GUI. So, yes.

Help thinking, if you explain a bit how it was implemented in beta49 coupling.

I think if the coupling is between bands 1-2, 2-3, 3-4, etc.. and 1 is the reference band, all bands will all indirectly related with band 1.

One might also think that the coupling has a kind of threshold for action, based on the difference between the means of the bands. And only activate when the difference is in the set value is exceeded and if not drive it, because if the difference is greater, you can think of that required by the track and that is a spectral imbalance.
Was doing a brainstorming.

_________________
by GAP
"Less is More" (Bob Katz)

DISAGREE !! Band 1 has almost all the time the biggest amplitude at transients, so i assume the result would be pumping audio ??
Unless....this is the sound you are looking for.

Personaly i would like the audio to be as transparent as possible without causing artifacts, but this is only my personal taste, like all audio presets i make are mainly my personal taste so i can only hope for people they like em.

_________________
----------------------------------------------------------------------------
AIRCHAIN-GURU professional independant airchain consultancy.
Orban/Omnia/Vorsis/DSPX/Aphex/Inovonics
----------------------------------------------------------------------------

No link 1>2! better is 2>1, 3>2, 5>4

_________________
Urban Junglist

@phantomfm: I agree with your personal taste, mine is the same, try not to change the sound and make it as smooth as possible.
As for what the band 1, may be not saying it's the only solution.

@all:
I've been watching the coupling of Aphex 2020, for example (in the manual, do not know if someone had a chance to use it), and I think I saw something interesting and could be applied to ST.

The processor coupled to select whether 1> 2 or 2> 1, and so with the other bands (2> 3 or 3> 2, 3> 4 o 4> 3).

So I think this could be implementable in ST, in addition to adjust the percentage of each engagement.

To calculate the percentage of coupling, I thought it would be:

- If you choose the coupling 1> 2 (commands the band 1)
- So that only act in the case that the band 1 is greater than 2, otherwise no effect.
- In the coupling channel must be respected and the link between L and R. The L channel of the channel affects band 1 the band 2 L, and so on, and in turn influenced by the percentage of each link.
- Will add to the coupling part of the envelope of the band 1 on the 2 according to the percentage.
- 100% coupling is completely adding the difference of the envelope of the band 1 with 2, while 1 is higher than 2.
- To a percentage value other than 0% and less than 100%, will add the corresponding portion determined by the selected percentage.

Well, this is my thought, based on some reading on other implementations. Hope opinions.

Edit1: Something that I omitted to say is that instead of percentage, dB can be used as coupling setting (can be more precise).

It occurs to me also that the coupling, you could modify the threshold of the band with link (dynamically) to obtain the desired reduction in the coupling function.

Edit2:
In transients, the band coupling should not be respected and compressors work independently.

_________________
by GAP
"Less is More" (Bob Katz)

I think I need to go back to my original idea: Drawing a curve through the points. Or actually, through 1/point level (in absolute numbers).

But the curve start and end should probably not be allowed to move.

So something like this:


The current solution does something like this, but the outer bands are drawn too much towards the bands next to them. But if I fix that, those 'bands next to them' need to change as well...

@Hans:
act as the percentage of link if you pose?
as whether to act or not according to the adjacent band?
I do not see even this thought right, but it's just my perception.

_________________
by GAP
"Less is More" (Bob Katz)

I don't really know what you mean... Currently each band 'follows' the reciprocal of the surrounding bands, where the strength of each band depends on how far the frequencies are apart.

So for example: Band2 = 1 / (.2/Band1 + .4/Band2 + .25/Band3 + .1/Band4 + .05/Band5)
(Where .2 + .4 + .25 + .1 + .05 = 1)

The values used are the values with which the samples are multiplied; for example for -20 dB that's 0.1, for 0 dB it's 1.0 etc.