Stereo Tool

P-II - 350MHz (dell)
------------------
ST Standalone v5 - All OK
ST DSP - Cannot be installed. - (message: you need atLEAST P-III)

That's odd - I think it shouldn't work on a P2... P2 and P3 behavior might be different by the way, so I'll do some tests on a real P3 to be sure.

I can't believe people are still running Stereo Tool on a P3! I actually think thats pretty cool to keep using a CPU like that for something good.

Question to Bojcha:

I've compared the previous version (BETA) against the latest (BETA2), with THE SAME settings (BETA2 2nd AGC off).

I could hear almost no difference between the two, so I've placed one on each channel and measured the differences.

Based on this measurement, BETA2 suffers _slightly_ less from the loud voice volume drop issue. Probably only because softer input sounds are played back a bit softer than in BETA. But except for that I couldn't find any differences (at least not in the Celine Dion track).

Am I missing something, or could this be correct? Tests were done based on default settings (hit RESET), Pre Amp tested at 24 and at 10. AGC's "RMS peak level smoothing" turned off, Remove Remaining Peaks set at the maximum position (to turn it off).


I'm asking this just to rule out that there's something changed in the AGC code itself that might cause a deterioration.
Changes are:
- No more bass added to band 2 (so it can now more easily separate from the lock to the band 1 level, Beyonce track is much better;
- level smoothing for input level near target output level (tracks around this input level with much dynamics came out far too soft before, now that has been greatly reduced).


My own impression is that using 2 AGC's is still a bit better than using just one, even for the 'wherever' at 0:48, but I've learned to trust your ears over my own )

i am not at home now, i'll respond with more details later...

The first AGC is Same as in BETA .. that's ok .. and i am happy with AGC1. beta/beta2.
It's true that with AGC2 things are bit better, But that is not it.. problem is in recovering down speed emidiatly.. not up-speed when the tone is already down ....

more details .. with some recording comes later ......

Hans, did you try ST5_EL34_2 PRESET ?

Ok, clear - great! (They DO behave differently in some cases!)
????
Please clarify... Recovering down speed??
Yes, but not yet in detail (actually I only checked the AGC - does indeed work a lot better than the default settings).

Yes, AGC1 in beta2 is bit different then in beta and i did't go deep into what is differences, that is why i said to load preset in beta. It did not act same as in beta 2 with same AGC1 only.

That combination agc1/2 is better for filling some gaps and even without RMSing tracks are more matched in rms level.
But i don't like when sometimes 2nd agc is going up too much and make some quiet parts in track too much loud.

Maybe it still need some super-precise setting but i only tryed what you suggest as default for beta2 agc.
Intereseting is that with this two AGCs there is still no pumping .. just leveling.

huh .. should be immediately lol

some Voice parts are lowered (leveled) by agc that's ok.. but in some moments it is lowered more .. and interesting is that same "more lowered parts" in original source are Higher. So, with this AGCs faster up speed is not solution because and agc1 and agc2 is lowering this parts in same manner. You can add 5 like this AGCs in line and all will be again same.
Maybe solution is multiband ? i already said that something is not good in multiband. Many ideas are on my mind...
Less bands? Joining some bands? beter choice of freqs.? Better filters? Add separate compress vs limit for each band ? Cross over output mix for bands? or whatever...i already get (i think) better results with some multiband settings.

i already said .. i would like to know how this can be done, but i don't.
what is read and saw, is that AGC in processors have compressors, wideband compressor, side-chain compressor, Parametric EQs even Limiter.. all that in AGC.
..and what is WideBand exactly.. definition..(?)
I am not ashamed to ask things that i don't know .. if someone can answer then there is guaranteed solution not only for AGC but for all processing. But that is something like TOP Secret.

in few minutes i'll upluad 2 tracks, One is Stereo tool and one is other processor

edit: Here it is:
1. Stereo Tool Web Stereo Preset, 2xAGCs (Hans's sugessted default settings), lowered Final Limiter to match other processor
2. other processor
http://www50.zippyshare.com/v/56016512/file.html

There are a few things that you can try:
1. Lower the Up Speed in the first AGC. That way the volume (in total) doesn't go up that fast.
2. You could try to set the target lever for the 2nd AGC a bit higher (!). That way it responds less strongly on sounds.

Basically, the idea is that the 1st AGC responds very slowly, both going up and going down. So in most tracks the volume stays roughly identical. This does of course mean that a lot of (too) loud peaks, especially at the start of a new track, remain. For that, the 2nd AGC is used.

i really understand idea .. and all logic says that it should work ... but obviously not.
i'll play with setting in AGCs more....

"Other processor" does a lot more than just AGC, because the whole sound image has changed completely. There are far more highs, and much of the dynamics is gone (it sounds compressed).

If, in Stereo Tool, you also enable Multiband, that should (if configured correctly) restore some of the lost sounds as well, which also improves the sound that comes out. (Make sure that the input level for Multiband is high enough, by increasing the 'target output level' of the AGC(s) if needed.)