Stereo Tool

I have tested Beta70A.

Ok, no crashes or any other of the formerly known problems with Acon chainer or inside Wavelab 6.
But, there is still something wrong with the processing.
Using the DSP version with BA Live or with Winamp 5 the sound is really good.

When I use instead the VST version (same preset as in DSP) in Acon chainer (BA Live hosted) or inside Wavelab 6 the sound is muddy (less dynamics, less highs, etc.).
This difference is not introduced by BA Live, Acon or Wavelab (always the same sound quality with the VST version).
It must be inside the Stereo Tool VST version.
I have compared both versions (VST and DSP) inside Wavelab 6 by using my DSPtoVST Wrapper (so that I can use the DSP as if it is a VST plugin).
Even this wrapped DSP is still much better (audio quality) than the native Stereo Tool VST plugin.

Please check this !
If I had more time I would show some waveform screenshots.
But at the moment I am too busy with other things.
There is definetly a difference in processing (sound quality) DSP versus VST !

Michi, thanks for testing the BETA70A! If that version works, VST GUI support should now be ok.

Which leaves the audio processing part. I have also received reports that the AGC behaves differently (faster or slower, don't remember which it was), so my first guess would be that something is wrong with the sample rate. I had hoped that it would be fixed when the GUI was fixed but now I'll have to look into it.

I use a preset without any AGC and Pre-Amp.
I use Noise Gate (heavy), Stereo Image (light), Multiband (heavy), SB comp (light) and Final Limiter (x1.00) only.

I cannot find the original KVR forums discussion where xlutop pointed to his free DSPtoVST Wrapper, but here you can find the download 'winampvst_dsp.zip'
http://www.xlutop.com/buzz/zip/
So this way you can compare VST and DSP in every VST app you want side by side.
Important note (to prevent any problems with other DSPs and make it easy to use with Stereo Tool DSP):
1) This dsp.dll always refers to a traditional (old V2) Winamp installation directory ('...\Program Files\Winamp\Plugins') and loads automatically the first DSP (in alphabetical order) inside the plugin folder.
So best thing you can do, is to have Stereo Tool as the only DSP plugin inside this plugin folder to load it automatically when you select DSP.dll inside your VST hosting application !
2) Do not close or minimize (!) the DSP.dll window !
When you minimize it, the hosted DSP GUI window (Stereo Tool) also will disappear.
It won't crash, so you can try what happens (try to load other DSPs or whatever).
I have used this under Win XP pro SP2 without any crash.
And most important:
You get the same audio quality as with the native Stereo Tool DSP version hosted in a regular Winamp DSP host.

If the sample rate is incorrect, it will affect everything except Pre Amp and Post Amp

So I'll at least fix that today. (Just a single command suffices to fix it, but it's not there now. In the Winamp implementation it is!).

Indeed beta70A is working like a charm. All known VST hosts are hosting it without any fuss. More importantly the last worked with settings are now recalled(every single parameter).
I did complain about sampling rate glitch a few posts back (actually couple of times if I remember correct). Surprisingly that seems to have been solved. Plays(or processes) at the same rate now irrespective of sampling frequency. No change of tonality now as it used to happen a few betas back.
As far as AGC(or multiband) gain rush is concerned, DSP has always been slower than Stereo Tool VST/Standalone. A reason why DSP version sounds different than Standalone/VST. Bear in mind there's no difference between Standalone and VST whatsoever. I can say that confidently because I am not using any extensive hosting chains to monitor. Currently I use ASIO to output/input stream from or to Stereo Tool with mere 64 samples latency. So thats like ~1.5 ms x 3 = 4.5 ms(at 44.1KHz) for the processed audio to reach my audio hardware. I have no clue how much is the latency between VST plugin and the host (in samples). Not talking about the Stereo Tool's processing latency(4096). But I guess it shouldn't be more than 128 samples(3ms at 44.1KHz). For there's no difference between the processed audio from the hosted VST vs audio from a player->ST standalone->SoundCard whereas I can very well make out a difference in change of latencies (to say 72 samples).

The fact that DSP's gain rush is(or probably) slow, gives more of a warm processed audio and porabably more time to react to highs than the other versions.

_________________
visit website

Come on boys !
Who has stolen this command line from the VST version ?
So, do you think that the DSP version is the black sheep that behaves odd (and should somehow be corrected) ?
Most of the time I use the DSP version and I have only done short tests of VST and standalone version (but without that detailed listening for differences).

If it is not possible to make SA/VST/DSP sound almost the same, then IMO it would be necessary to create special presets for (and limited to) the DSP version, because this difference is not small at all.


@ phoenix
And finally I have tweaked this whole J River app monster for my needs.
I start to like it (using the negative skin effect I even can put my sunglasses down).
But:
With version 14 I had always stutters when I enabled Stereo Tool DSP.
I had to update to version 15.
With Kernel Streaming (BA Pipeline 2) as output it works.
And this way when I compare the usage of the DSP inside J River with the usage of the DSP via BA Live there is indeed a noticeable difference.
J River internal usage of the DSP sounds better (no doubt).
Even in conjunction with my cheap Realtek HD onboard (no expensive ASIO card, sorry !) I can hear this difference (it sounds cleaner, richer, etc.).
It is closer to the quality I know from Wavelab or Audition (v1.5 !).
Thank you for this hint !

They should - and can - be identical, I'll try to figure out what's going on here.

I've compared the sound of the VST (in VST_Checker) vs. the DSP (in Winamp) plugin.
Using a 44.1 kHz input file, I hear no difference, and see no difference in the GUI display.
Using a 176.4 kHz input file, the VST sounds very (and way too) bright.
Alternatively, if you were to use a lower input sampling rate I expect the VST to sound dull.

I'm going to build a version which properly supports all sample rates; with this change the 176.4 kHz file sounds the same as the 44.1 kHz file.

BETA71:
- Display de-esser and voice vibration filter effects ■ DONE
- Check which 'New Loudness bass processing' sliders must be kept, and place them at correct location. ■ DISABLED THE UNNECESSARY SLIDERS, UPDATED DEFAULTS ACCORDING TO BOJCHA'S AND CHRIS'S FEEDBACK - WAITING FOR COMPLAINTS...
- VST: Check audio issues (muffled, different AGC up speed). ■ DONE, waiting for reply

Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... 510-71.exe
Stand alone version: http://www.stereotool.com/download/ster ... 510-71.exe
VST version: http://www.stereotool.com/download/vst_ ... 510-71.dll
VST version (No SSE2): http://www.stereotool.com/download/vst_ ... 510-71.dll
Command line version: http://www.stereotool.com/download/ster ... 510-71.exe
Linux command line version: http://www.stereotool.com/download/ster ... BETA510-71 (not available)
Linux GUI version: http://www.stereotool.com/download/ster ... BETA510-71 (not available)

Tomorrow I'm going to look at the presets. So if anyone has updates of presets posted here earlier (Bojcha? Phoenix/Orban8400?) please post them... I've already scanned the presets forum, and found a few interesting new ones that I'm considering to include:

Bojcha - exp27_streaming_RC1 (probably very outdated?!)
Gtxx
Gtxx_6_10
ncsflorida - WSOC
SuperH (AAC+) - update
My own Orban 8400-like preset for Phoenix.

So far I've tested these presets with only 1 song, so I'll have to check if they work well with a bunch of tracks before I add them.

I'll have to scan through this thread to see if there are more - but since a lot has changed recently, updated versions might be useful anyway.