Stereo Tool

Ofcourse, because 10.0 for Up speed is too much 5.0 max (for this beta)! But again nothing happens, loud tracks are still loud and dynaimc track are still dynamic. If only AGC and loudness is ON, i have to raise "Target Outrput" level by almost 3dB to get some track in same level.

Last night i have some reading about ITU BS.1770 measurement, and found something that i notice too, It racts on bass too much, but i somehow i trained myself to overlook that behavior.

- Sidechain EQ for AGC.

I think Hans' intent was different here ... correct me if I am wrong(either Hans or Bojcha):
Previously with optimal settings of AGC at 0.8/10.0 targeting say -6dB, if I played a very loud track and then a very slow track, it takes a good 20~25 secs to come at the desired output level..so softer parts following loud parts (FOR EXTENDED PERIOD) sounded annoyingly soft.
In this beta, the situation is better as it gets to the targeted level in like 1.5~2 secs.
However once you're in the middle of the soundtrack, it's pretty much the same output level in both cases. May be in this beta level is a tad softer than previous case...because of what Hans describes as 'brake'..he can elaborate as to what it implies and how it can impact. But then it explains why loud sounds remain loud and dynamic tracks retain dynamic once you're in the middle of the track.
The only thing that bothers me in this beta is pumping/breathing when just PreAmp+AGC+MB are enabled. It's less perceptible once loudness is enabled but that's obvious!

_________________
visit website

I think I am beginning to like the new AGC...unsure of what others have to think. Sounds which were either getting masked by loud parts or were sparsely audible are more distinct with this new AGC. Surely sounds like doing a good job...
The fact the future course is uncertain whether or not this AGC action would be retained or discarded, I guess I've to keep a copy of this beta for ever

_________________
visit website

No, the code is commented out. I'll enable it again.

ITU-1770 re-enabled. No further changes.

Note: There are 2 checkboxes:
ITU-1770 bass: The bass part of ITU-1770 processing. This means that low bass sounds are counted less strongly.
ITU-1770 head correction: The highs part of ITU-1770 processing. This means that highs (> 400 Hz) are counted stronger.

My advice:
- Keep BASS off unless you use a single band. (If you do turn it on for multiple bands, you'll need to reconfigure the 'Raise band 2 output level above band 1...' slider).
- Use Head to get a more constant perceived output level. May lead to increased volume (pumping) effects for loud voices.

Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... 20-003.exe
Stand alone version: http://www.stereotool.com/download/ster ... 20-003.exe
VST version: http://www.stereotool.com/download/vst_ ... 20-003.dll
VST version (No SSE2): http://www.stereotool.com/download/vst_ ... 20-003.dll
Command line version: http://www.stereotool.com/download/ster ... 20-003.exe
Linux command line version: http://www.stereotool.com/download/ster ... ETA620-003 NOT AVAILABLE
Linux GUI version: http://www.stereotool.com/download/ster ... ETA620-003 NOT AVAILABLE

Edit: Pumping effect for loud female voices with ITU-1770 indeed seems much less present than before. And: Even ITU-1770 bass sounds good now. (I've even tried the Video Phone track with 1 band, and even that sounds fine with these settings!)

Wow ST sounds better all the time! I been trialling latest uploads by hvz and it just gets better and better!

Friggin nice work Hans!

I am confused, and i was wrong.. it's not only about band freq. reaction/response for AGC. something other is in question.
First ITU (bass) on or off, for using AGC in 1-band mode, is, about 1dB
With AGC 2-band mode 1st band does not follow 2nd. (if i want to) when ITU (bass) is on. But again there is no purpose for that option.

other Test: Only AGC ON 1-band
I put 90Hz sweep at -30dB at input. Output PEEK level (depens on postamp, not important only for comparation) is -7dB
Then i send 0dB 90Hz and PEEK level is -4dB,

Same Test with 3kHz sweep:
- Input -30dB - ST Output Peek -8.5dB -- with input 0dB, output is -4dB

In 2nd test ITU (head) will just make other numbers but same difference!

So, i know that AGC is not 'linear' but currently i think it will be better to be.

Interesting is that BA (for comparation) works opposite then this ST beta.
In ST AGC goes UP faster, and more faster, in BA it slows down when it need to go UP.. and UP.
And sweep at 3kHz or 90Hz at -30dB or 0dB input, gives same dB number at output (agc only)

Hm, I though I had made it more linear, but I guess I should look into that again - 4 dB of difference is quite a lot.

One of the things that I had tried is processing 'as if' the input level is 100 times higher - that does result in a more ********** WOW

Ok... That didn't work before because of the way I was increasing the level. But it should work perfectly with the new implementation (BETA002). Wahooa!
Give me 2 minutes

Edit: Hm, more like 20 But the result seems to be MUCH better - building now, will post it tomorrow morning.

2 minutes? np, i'll wait and watch "black chicks w..."

Ah, now you know WHY it increased to 20