Topic: Win7 Fireface UC = Blue Screen, clear steps to reproduce
Hello RME,
I am german, but I will write in english, because the program which is able to force the bluescreen with fireface UC is "Deejaysystem" VJ2 and the designer is from Spain. I first reported this bug two weeks ago. Since then, the designer of VJ2 and me, we invest a lot of time to figure out where the problem is. At first at the programm itself. Now we are sure that the only persons who can fix this is either RME or Microsoft.
I will not write every step we did. Even though it is okay to send all details and mails exchanged to the person who takes care about this issue.
It is "easy" to force the bluescreen and it will work on any computer (as far as we found out).
Here are the steps:
General conditions hardware:
? All tests are done with a RME Fireface UC. Test with a M-Audio FW 410, RME HDSP9632 runs well, no blue screen.
? The test is absolute not related to Hardware. It occur either on a Intel Atom with ICH7 notebook, but also at a Core I5 with P55 chipset. All power management functions are disable in the bios as well. Include some special CPU functions C1-state etc. For USB I also change several setting there as well (only USB 2.0 on). It also doesn?t matter which USB-Port to use. The Fireface UC was external powered, even though I tried an y-cable (to split power) as well.
General conditions OS (all 32bit):
? After figure out that this is a pretty hard to find problem I decide to setup all sets of Windows with new installations (just change haddisks). So when I talk about XP32, VISTA32 and W7/32 I mention with latest SP and Hotfixes not only for security, but also for features like stability.
? When the issue occur I also change the settings of the OS to:
All USB devices in the device manager: ?not allow to go in power save?
? Power Schema change to high-power.
Standard Sound device play/record is other than the soundcard I tested.
? Application runas ?Administrator? on VISTA/W7 and XP as Admin user
Soundcard:
? Number of buffer samples doesn?t matter, either 2048 or 64 is the same behavior.
? Sampling frequency is either 44100 or 48000.
Application:
? The problem occur up to now only with the following applications (new and old versions). I was unable to force the error with any other app under similar conditions.
- VJ2 Version 3.5.0 beta 3 (http://www.deejaysystem.com/forum/forum … sp?TID=783) . Also VJ2 Version 2.0 from 2005.
- Audio MK2 1.9.1
? The settings of the applications are standard. Even though it makes no difference if you invert all (e.g. realtime settings, adjust speed with master?)
? Following settings are needed to force the bug to appear:
- use the same soundcard (e.g. RME FFUC Analog output 1+2)but different output channel for monitor (e.g. 3+4)
- use separate channel for Deck-B (on/off) makes no difference. Also not use monitor, but use Deck B does NOT force the bug. It is only related to ?monitor? output.
- use a different soundcard (like internal Realtek etc.)for monitor does NOT force the bug as well.
? Loading of video/soundfile in Deck A. Even a wave-file is suitable.
? Monitor on/off Deck A makes no difference.
Steps to produce the bug:
1. Start to play the file of Deck A.
2. Press pause (not cue)
3. Wait for min. 10 minutes (maybe less, but more than 8)
4. Press pause again (not cue) to play
Under XP it continue to play. No bluescreen, not stuttering so far
Under VISTA it continue to play, but with stuttering, after pressing 2-3 times pause a blue screen comes up ?IRQL_NOT_LESS_OR_EQUAL? with listed driver USB?.sys.
Under W7 immediate blue screen BUGCODE_USB_DRIVER
There are now a lot of tests we did. Also a lot code changes regarding the directx api using. As I mention I can send this to the person who investigate in this topic.
The last mail from the designer regarding this issue:
"Thank you very much for your help on this, but I really believe that trying to further workaround this problem (which is NOT a Deejaysystem problem) is pointless. "
"If you want, I can give you the exact sequence of DirectSound 5 calls that are performed in Deejaysystem from program's start to the crash point. You or the driver programmer may then code a sample program that causes the error. I think that telling him "two consecutive IDirectSoundBuffer5::Stop calls followed by a pause and then twoIDirectSoundBuffer5::Play calls on different buffers could make the driver crash under some circumstances" together with the "need to wait some minutes" should give him at least a clue for start searching for problems."
Now I see a another thread here in the forum also with a Blue Screen.
I am convinced that this can be fixed! Thanks in advance for all helping here.
Sascha