Stereo Tool

I just uploaded a new BETA version where the AGC is improved further. "Remove remaining peaks" now causes almost no artifacts anymore.

Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... r_BETA.exe
Stand alone version: http://www.stereotool.com/download/ster ... r_BETA.exe
VST version: http://www.stereotool.com/download/vst_ ... l_BETA.dll
VST version (No SSE2): http://www.stereotool.com/download/vst_ ... 2_BETA.dll
Command line version: http://www.stereotool.com/download/ster ... d_BETA.exe


Things todo:
- Instead of lowering based on max(old method, new method) use max(min(old method, factor * lowest result of new method), new method) - this should better reduce short spikes at the corner of a processing block (loud start of a new track after silence) * NO, does not work better
- (Test if this helps) Dynamically lower the 'Remove remaining peaks' level when the AGC output level is going up for a while. * DONE, waiting for feedback
- Improve performance a bit for the new 'Remove remaining peaks' algorithm (it's heavier than before now, but it doesn't have to be. And the quality will probably get better too). * DONE
- Multiband wider bands * DONE

Hi Hans,


Bojcha contacted me about it and I had to test it to believe it. Ofcourse he was right (how can we ever doubt him )

I haven't tested it with pink noise yet but it might cause some unwanted sound colouring and overal loudness could be affected.

Hi Hans, thanks! Just for your record, the last version that worked here with Station Playlist, is the beta released on 7 May. Nothing to worry about, I'm the only one with this problem as far as I know and it might be my config.

Well, that's the thing: Noise (including pink noise) should hardly be affected. Because if 2 consecutive bands have the same amplification (which is mostly the case if you check the output, and definitely with noise), there's no colouring of any kind occurring. Still, a loud tone just between 2 bands can still come out (at least a bit) louder than it should. So solving it might improve things. AND, also important: I *think* that it might also make the output volume from Multiband more constant, which could indeed make things a bit easier for Loudness.

You can consider it like this:
- A single tone is one extreme
- Noise is just the opposite of that (so also an extreme)
- Music is somewhere in between, usually MUCH close to noise than to a single tone (just check a spectrogram and you'll see what I mean)
A single tone if affected by this, noise is not. Music maybe a bit.

Just checked it again here, but I'm not seeing any issues (even in Station Playlist). If I could reproduce it I might be able to find out what's wrong - but so far no luck...

...

LOL! It turns out that I had already 'fixed' this for all but the most extreme cases (actually I went through trouble to NOT fix the most extreme cases...). I just removed a few lines of code, and now:
- Sweep is constant (except at start and end of the range), no more jojo
- Normal music is NOT changed at all

New BETA uploaded, now with flat multiband response.

Are you shure ?

As i said on multiband is more attack now. So level must be lowered from AGC. Limit mode and Compress mode are more different.

You test it and let me know so we can bother Hans