In my previous posts about this interface I forgot to mention that EVERY time I started. In cases with good or bad results, I ALWAYS had two windows open according to the Windows XP task bar. The first was the small Podium splash screen, which just stayed there while the main Podium window did not open yet. The second, only seen referenced on the task bar said ASIO MME and I always had to right click on its task bar label and select CLOSE. THEN Podium would open fully. The strange thing is I thought Asio and MME were mutually exclusive methods. (?) ❓ 😯
The invisible second window is actually a dialog that is located behind the splash screen. If it is similar to the one that appears on my system, it says ‘ASIOMME.dll is already running’ with an OK button on it. This dialog is opened by the Steinberg emulated ASIO driver when Podium enumerates all available ASIO drivers on the system at startup. It typically appears when the driver (or the MME ports that it wraps) are already in use. Do you have other programs running when this happens?
The newly released 0.91 has a new report facility in the interfaces dialog. If you have the time you might try this with the MOTU again, and email/post the report to me so I can get a clue of where it goes wrong.
Regarding the constant overload bug mentioned in the 0.90 release topic:
Thanks to the helpful people at my local music-retailer, I was able to borrow a 828mkII unit. After installation on my Inspiron 8200 I had detected and fixed the bug within 5 minutes. Had I not had the device it could have taken several releases of Podium before I had figured out the cause.
And now a note to any host developers experiencing problems with MOTU ASIO drivers:
If you use the sample rate supplied by the driver in the TimeInfo structure for each ASIO buffer, don’t trust the sample rate is valid even though the flag in the structure indicates that it is valid 😕
The MOTU driver (v3.3) always sets the sample rate to 0.
I was able to set the latency to the lowest setting of 92 samples, but on my 1.7G P4, it was very sensitive at >25% cpu usage and would easily overload. At 256 samples it was ok.
I will release v0.92 on friday/saturday with this fix. Hopefully this will fix the problems reported with Creamware cards as well.