Stereo Tool

Stokkemask not working properly in "Leif" optimized clipper (which means all versions with CPU optimized clipper).



The CPU strictness maxed out doesn't help (here on in clipper section).

Tested with multiple presets, for some works better (not perfect, like Unveil FM) for some seems to be broken completely (Unreal Engine FM). Confirmed by Bojcha, Jesse Willem Van Der Veen and few others.

Regards,
Slawomir

Please show the settings in the other window (mainly Strictness) as well - that really affects this.

The result is the same





yesterday's settings were:



no difference

Actually I was discussing it with Bojcha once, with regard to older versions with CPU optimizations implemented - and he claimed it still didn't work. Looks like he was right.

The filter was never *perfect* in this display even without Leif's optimization. So, I need to check if it has gotten worse.

Just for clarity: Overshoots in this view don't necessarily mean that we're really overshooting, because the measurement is far more sensitive than the Stokkemask specification describes. Also, the display is always a bit 'too wide' (the top of the shape is wider than what you would get with a real analyzer), which means that you can very easily go over the red lines while on a real analyzer you would still be inside the lines.

I know, nonetheless - it's worth investigating Since it's mandatory in the NL. In Unveil it seems to work well. In other presets, like Unreal FM it doesn't work so good (In fact I don't see any difference, but it also has narrower RF sectrum in general). So please check if it stays within limits.

Actually now I tested it with both old and new - and it seems to relate to the presets more than the clipper type. Thank you for looking into that.

I have the same problem here as well but i will check it in mpx tool as well later today ...

_________________
Very proud user of Stereo Tool since 2010. Thanks Hans.

My radio station
http://amradio.ddns.net

If you check it in MpxTool you will most likely see that it is very well within limits. Part of the issue is the measurement which I cannot do more accurately without using far, FAR more CPU power.

OK

In case you wonder why MpxTool can do it: It also can't! Where my measurement is too strict, the measurement in MpxTool misses many shorter overshoots, and the spikes are drawn much less wide than they should be.

Understood MPX tool indicates less than it should (some overshoots are missed), and ST indicates "false" overshoots with regard to mask.

So different question - since there are contradicting measurements - is Stokkemask in ST working properly then? Is there any reliable way to test it?