jsComp (Free compressor with switchable brickwall limiter)

[l0calh05t]

@echopark: I assume by "Pentium" you mean "Pentium 4", right?

If that is so it would seem to be a Pentium 4 problem (or does anyone who doesn't have problems have a P4). I wonder why... all models do support SSE2...

[l0calh05t]

Ok, for those who are getting crackles, try this version:
http://www.carolander.net/joe/jsComp_noCW.dll

You must remove the other version from your VST folder first (they use the same plugin ID). In this version, I removed the only "unusual" thing I did in the plugin, but this unusual thing (according to the MSDN documentation) should do anything at all on x86 processors, but it actually does something... maybe this causes the incompatibility with P4 processors? I'm only guessing here because I have no P4 available for testing.

[shoma]

thx l0calh05t...it's getting better...at least it's only crackling on the right channel ... ...

[echopark]

@echopark: I assume by "Pentium" you mean "Pentium 4", right?

Oh Sorry, yes Pentium 4

I will try the new version in the evening.

[l0calh05t]

Huh?? I changed something that should affect both channels equally (or rather it shouldn't actually affect anything but it's the only thing I could think of...)

[Franzel]

I have a P4 3.2 GHz and I don't get any crackles on the older version (didn't use the newer one, cause I'm not having any problems with it). I just tested again using a track with just a kickdrum in Cubase 4 and different compressor-settings, no crackles!

Edit: Did the same test with Reaper 2.54, no problems either....

[l0calh05t]

the only real change to the signal processing was to switch to full 64-bit processing, so theoretically nothing should have changed (and nothing did change on my core 2 duo). meh... maybe it's a processor bug (which do occur sometimes)

[shoma]

Now I've had some time to test a little bit more...both new versions are working flawless in SX3..tested in mono channel insert and stereo group channel insert (16bit/44.1kHz)...in wavelab 5 both versions are crackling heavily; jsComp_noCW.dll only on the right channel though...

[l0calh05t]

Could you check if the crackling occurs with all switch combinations? If there are only specific switch combinations that cause trouble it might help narrow things down a little (although honestly the problems make absolutely no sense at all up to now!)

Here's a checklist for all switch combinations:

1) Mix - Chn1/2 - Off
2) Max - Chn1/2 - Off
3) Mono - Chn1/2 - Off

4) Mix - Chn1/2 - On
5) Max - Chn1/2 - On
6) Mono - Chn1/2 - On


7) Mix - Chn3/4 - Off
Max - Chn3/4 - Off
9) Mono - Chn3/4 - Off

10) Mix - Chn3/4 - On
11) Max - Chn3/4 - On
12) Mono - Chn3/4 - On

EDIT:
and another version to try with different compiler settings:
http://www.carolander.net/joe/jsComp_slowmath.dll

[echopark]

Bin in Eile darum in deutsch:

Alle Versionen haben nach wie vor das selbe Problem auf dem Pentium 4.
Sämtliche Einstellungen am Plugin ändern nichts.

Es klingt so als wenn der linke und rechte Kanal auseinander laufen.
Als wäre ein kleines Delay drauf zwischen links und rechts.

Der Klang klingt irgendwie bröselig / körnig


Auf meinem Hauptcomputer AMD X2 Dualcore läuft das Plugin ohne Probleme.
Dürfte wirlklich ein CPU Problem sein.

[shoma]

Bin in Eile darum in deutsch:

Alle Versionen haben nach wie vor das selbe Problem auf dem Pentium 4.
Sämtliche Einstellungen am Plugin ändern nichts.

Es klingt so als wenn der linke und rechte Kanal auseinander laufen.
Als wäre ein kleines Delay drauf zwischen links und rechts.

Der Klang klingt irgendwie bröselig / körnig


Auf meinem Hauptcomputer AMD X2 Dualcore läuft das Plugin ohne Probleme.
Dürfte wirlklich ein CPU Problem sein.

and now in english:

"I'm in a hurry, therefore in german:

All versions still have the same problem on the Pentium 4
All settings on the plugin change nothing.

It sounds as if the left and right channels are diverging.
Like there is a delay between left and right.

The sound sounds somehow corned / grainy


On my main computer AMD X2 Dual Core the plugin is running without problems.
Might be really a CPU problem."


Well, that's exactly what I can reproduce on my machine...all suggested settings cause no change in the result...it's crackling with a kind of "grainy" sound...like a bit of timestretching would be applied

All this applies of course only to Wavelab 5 on my machine...SX3 still is without problems...

[echopark]

Shoma, thanks for your translation !

[shoma]

no prob

@ l0calh05t

I just tried to render a file in Wavelab 5, hoping the file would be rendered with crackling so you can hear how it sounds...but WL failed to render the file; the process starts and just before it would be ready an error pops up...


"Ein schwerer Fehler wurde durch das PlugIn "jsComp"verursacht.
WaveLab kann unter diesen Umständen nicht mehr einwandfrei funktionieren. Es wird geraten, Ihre Arbeit zu sichern und WaveLab neu zu starten.
Um das Problem zu beheben, wenden Sie sich bitte an den Hersteller des PlugIns."

which in english might mean something like:

"A serious error has occured inside the plugin "jsComp".
WaveLab could control it, but is is advisable to save your work and restart WaveLab.Moreover, we advise you to come into contact with the technical support of this plugin manufacturer."


damn...

...anyway...I like your plugin in SX3...

[l0calh05t]

Danke für die Übersetzung, aber ich verstehe das auch so ganz gut
Und damit andere Forenmitglieder es auch verstehen, ab hier weiter auf Englisch.

Ok, I "fixed" a few things that weren't causing any trouble here, but might theoretically cause problems on specific systems or hosts, so please try this new version and report any improvements (or new problems...) I'm sorry I can't just solve the problems, because I have no system on which they occur...

http://www.carolander.net/joe/jsComp.dll

[l0calh05t]

UPDATE:
-CPU usage reduced (especially with no input signal)
-P4 bug fixed?