Stereo Tool

And can you check the 48 kHz tone.
All the tones have 100.1% MPX as expected. Only that tone has 93.6% and i checked that it was at 1.00 and that phase was at 0.

EDIT: When i try the same with composite overdrive 105% i don't get 100.1% mpx anymore with the test tones. But i get 99.0% for the test tones and 70.6% for the 48 kHz tone. Hmm... and not all the time 70.6% now. I checked again and it's at 96.3% and when i check again it's at 73.3%. So there's definitely something wrong there.

BETA085 has a new and better sounding clipper. But it also has a much higher CPU load. There's still an issue on my todo list to optimize it further.

Also a good song for tests ...especially in the beginning with very loud high`s

http://www17.zippyshare.com/v/9199655/file.html

Hm, the calibration itself IS done afterwards, but the tones are injected earlier. I guess I need to turn the composite limiter off when test tones are played.

New AGC faster dwn_spd is not good! i cannot get any good under 200%

Evanescence - My Immortal

Actually I think something is wrong with the tool that you're using to measure it. I've recorded the Stereo Tool output to a file and opened it in a wave file editor. 48 kHz is EXACTLY 1/4th of 192 kHz, due to that you get the same 8 sample values repeated over and over. So in a zoomed-out waveform it does indeed look a lot softer. But, if you zoom in and the wave editor starts to interpolate values, they suddenly reach the peak level again.

It also makes sense that you get a different behavior each time: The peak level just depends on WHICH 8 sample values you get. So if you start generating the tone again, it might start at a different spot, causing diferent sample values to be generated. However the output of the sound card is indistinguishable!

I tried making test tones with adobe audition. I tested with 53 kHz tone and 48 kHz tone. Both are set to -3dB.
In MpxTool i get exactly 100.0% with both tones, so why is there no problem when i use that? Both files played with winamp trough VAC to MpxTool.

EDIT: Looking zoomed in to my test tones both files have exactly the same audio level.

Ok I hear what you mean. I'm changing the default value to 200%, hope that suffices. (Even at 200% sudden loud spikes are still handled better than before).

For the future, I have a change planned that - if I'm right - should make the AGC much more resilient against volume drops caused by loud voices. Actually it's a pretty simple change - but I'm probably not going to implement it in version 6.20 anymore. Anyway, if that idea works I should be able to lower this value again.

For 48 kHz, try setting the start phase to 45 degrees. I've just tried it in CoolEdit (older version of Audition) and with start angle of 0 degrees I get a peak level of 0 dB, with start angle at 45 it's -3 dB. The RMS level is identical though (-3 dB).

I have some idea and it's very simple actually half is already in current AGC just some changes ... if you want to hear...