Stereo Tool

Hi Hans, in last beta vers. I again have some error with configure soundcard in standalone. in all menu's i have only D/E1 ??
Previously, this error has occurred and I have not had the opportunity to select the sound card in the dropdown menu




Sorry !!!! This is my Windows, i restart and now all is OK

Odd. D/E1 means: Device / (-1 lol.... That's the default device) Error 1 - (MMSYSERR_ERROR - that's an internal error in Windows!!!).

This happens when I ask Windows information about the device (name, max sample rate etc.). As you can see I get an internal error as reply.

This _should_ never happen (it's an internal error) - anyway for future versions I'll clarify the message ("Error 1 (Device -1)" would be a lot clearer than D/E1).

I have updated the AGC in the BETA. Should work much better now. AGC window layout still needs an update/clarification. But the functionality is there now.

Indeed !
The sliders inside AGC need clarification.
I think I am not not an absolute beginner, but it is more or less trial and error principle adjusting the sliders.
It is not clear what the sliders do or what values are used.
The same thing in many other Stereo Tool modules.
If you are not an expert it is more or less impossible to adjust things (I can do it, but never sure what exactly I do).

What do these 00000-99999 sliders adjust exactly ?
Ok, 00000 is less than 99999.
So, more or less.
But the the details are hidden for me behind a black cloud.

Please, a short explanation plus a simple AGC only preset (AGC.sts) to help me and others to understand how these sliders inside the AGC GUI work in detail !?

I have uploaded a new version where the AGC window should be a lot clearer. Also, AGC gating is added. Uploaded as BETA (see link somewhere in this thread - no time to search now...)

As I say before 4.22 is the best version , and what is need is only the vlevel inside (replaced Pre-Limiter) that`s all . 4.22 takes little more CPU but there is settings "low" "mid" and "high" ..... cool and everything works . There`s nothing to say more

Hi DJ Doggy,

What do you mean by '4.22 is better than 5.00'? If you find anything that sounds better in version 4.22, please let me know - most sounds are better (cleaner) in version 5.00, and a lot of artifacts are gone.

Note: The new AGC may send audio with a different volume level into the multiband compressor, so it's difficult to compare the two versions with AGC or Pre Limiter enabled. If you turn it off and manually regulate the input volume, both versions should sound very close (also, turn Bass Boost and Phase Rotation off in version 5.00, since they didn't exist in version 4.22).

Easiest way to test:
- Load version 5.00, press RESET to turn everything off
- Then load version 4.22, and select a preset
- Turn Pre Limiter off
- Manually configure Pre Amp so Multiband gets a 'decent' input level
- Then you can compare the two versions.

Very very nice release 5.0 !

Hans,

I just measured with a sweep from 20Hz to 20KHz and noticed that the filters cause a 'jojo' effect on the output. Differences in multiband output are huge. Seems that the combined output of the filters is not flat.

So, if this can be as should, i expect:
- More attack on multiband again (because wider bands - so AGC 'max level' maybe need to go down again)
- Better fix for AGC output. Wider bands should better levels signal. Current problem with AGC's output when needs goes down)
- Improvemenent of Final Limiter and Loudness because of steadier output out of the multiband.
- Maybe solve problems with some Voices vibrations/distortion. - Here i think that some voices are on frequencies between bands, so they are louder and push the final limiter and loudness.

Ehm, Luke, are you Bojcha now? Because this afternoon he wrote exactly (and I mean exactly) the same thing...

Anyway, this behavior has been there since version 1.00, but I'll look into it - I *think* it has little effect on music (that's why I didn't try to fix it before), but I never thought about this voices vibration issue, and you might be right there (no idea, but it's possible).

The only 'simple' solution that I can use will combine the input of 3 bands (say I'm processing band 4, then I would look at band 3+4+5 and then only use the result on band 4). Any other solution will double the multiband CPU load, and require a lot of extra code.