Stereo Tool

Glad that you are finding the issues Hans.

Sadly, I'm still getting some exceptions. I will pm you the log again.

I report no problem so far on second gen i5.

These ones are far more difficult...

I've marked them all in my code so I can look at them later (when I change the code, addresses change so I won't be able to find them again).

I see crashes in:
- ASIO
- Stokkemask (multipath clipper) - this one I understand, I'm writing outside a buffer somewhere.
- Compressor. NULL-pointers appear to pop up at several locations.

The ASIO crash may be related to a problem that Bojcha and Chris have reported, where the FM output falls silent. I've checked the code though and there's nothing weird there. And nothing that the compiler would compiler differently from before... but who knows. It could theoretically also be caused by the Stokkemask issue (I know that Bojcha is using Stokkemask, not sure about Chris?)

So for now I will fix the Stokkemask issue and wait what happens with the other crashes.

I also noticed GUI lockups when the software appeared unresponsive but continued to pass audio. Did you see anything relating to that?

I can continue to test a bit more later if that will help.

Being greedy here, but would it be possible to get control over the MB limiter attack times? Or at least could we be made aware of what they are? Based on the UI, they appear to behave much like a clipper, rather than a limiter. But that isn't a very scientific way to analyze the issue.

That is limiters in "peek" mode. It's true that they are bit tight but they are still limiters. For highs they need to be even bit tighter but for bass bit less tight. You can hear now some distprtion in few first bands. (look-ahead might help also). I've asked Hans for that.
In RMS mode they are totally different and kinda useless.

_________________
control point
control point2

True, I forgot about that one. It actually occurs inside an Intel library which will make it very hard to find, unfortunately.

Anyway, for now I'm going to fix the one crash that I do understand (the one in Stokkemask), where I see that my rewrites are causing some memory reads and writes outside the allocated memory... which can cause all kinds of weird issues throughout the entire software, so I hope (but do not expect) that this will take care of the other crashes as well.

Stokkemask issue should be solved now. Build is running.

I was looking into the compressor crashes, either the compressor is called without being initialized - which makes no sense, but I see that it's called from a separate thread, which might mean that it gets deinitialized before it's finished in the other thread. Again, that SHOULD be impossible, but there could be a bug there. If so, it would also be there in older versions though. Alternatively, it can still be caused by some memory overwrite from the Stokkemask filter...

Ok let's try it again:

Windows 32 bit:
Windows stand alone: http://www.stereotool.com/download/ster ... 51-021.exe
Winamp/DSP plugin: http://www.stereotool.com/download/dsp_ ... 51-021.exe

Changes:
- Stokkemask crash should be gone
- Stokkemask in strict mode (checkboxes for 'skip smoothing' and 'sloppy smoothing' disabled) is *slightly* less strict than before, but the sound quality is improved a lot (effect on the audio of the Stokkemask filter is reduced by more than half).
- You can set Strictness (CPU) to 3 instead of 2 as maximum. 2 should normally be sufficient though, and it's the default setting.


Reported issues that seem to have been introduced by the compiler switch that need to be fixed:
- ASIO audio drops out, might be related to ASIO crash reported by 2Sense
- Exception in: GUI Event Handler (Exception 0xc0000005 at address 0xf5f5b8, read/write=1, r/w location=0x0). Offset-info: 0xeb0c84. Reported by Fario and 2Sense
- Crashes in Compressor (Multiband, 2nd thread). NULL-pointers appear to pop up at several locations. Based on what I'm seeing in the code I'm pretty sure that this is caused by a previous crash. The occurrences that I've seen so far are always after another crash.
- In the latest build (Winamp DSP) the audio is fluttering at left and right when you enable the Pre-Limiter and using 1024 bands. Reported by Wiele. Latest DSP build is already a bit older so this might have already been fixed. I have fixed a bug in the pre-limiter.-

Edit: Just also got a loss of audio in the DSP version. Turns out to be in the FM sound card tilt correction. Will look into this tonight.

Testing Beta 021 now. So far I haven't had any issues. I will see if I can get any exceptions / restarts to occur.