Zynewave's Forum Page

Max sent me a report with his E-MU ASIO driver diagnostics. In that report these details are of importance:

Log period: 17.05.2006 22:36:52 – 17.05.2006 23:00:34

– Current sample rate: 44100
– Input/Output latency, buffer size: 486/492, 440

ASIO driver status:
– ASIO sample position gaps: 38 (-440..47080)

The sample position is provided by the ASIO driver and it should be a continously running counter that increases with the buffer size on each buffer called by the driver. Some ASIO drivers use this samplePosition counter to notify the host if buffer underruns have occurred in the ASIO driver code. Podium checks the samplePosition only to notify in the UI that a supposedly buffer underrun has occurred, and that is why you won’t hear any clicks in the sound or see any CPU spikes in the Windows performance graphs.

In the report above it appears that the samplePosition provided by the E-MU ASIO driver occasionally is offset from -440 and up to 47080 samples. With a buffer size of 440 samples, -440 means that the samplePosition is not advanced at all. I’m guessing that the -440 is occurring at regular intervals and at some point the 47080 sample forward jump occurs to correct the accumulated sample offset error. So to me it seems that the E-MU drivers internal counter for the samplePosition is updated incorrectly.

Apparently not that many hosts check the samplePosition at all, since the E-MU drivers are not reported to cause problems in Cubase and other major hosts. I would prefer not to degrade the error check in Podium, so if someone at E-MU could look into this it would be super. Since the samplePosition jump occurs at regular intervals, it should be simple to track down what is causing this.

Technophobia, could you post your interface report here to verify yours is the same problem as Max? Also if you have contact to someone at E-MU on the productionforums site, maybe you can paste this post or post a link to this topic over there?

Thanks for looking into this Frit’s

Just a slight correction: It’s spelled Frits (without the ‘) 🙂

Even if I leave Podium as is and do not start playback/recording I get the CPU indicator flashing red for about a second every 44 seconds approx. It doesn’t actually show up as a CPU value in the indicator, i.e. it remains at 0% but flashes red briefly…?

The red flash would then indicate that there is an ASIO buffer underrun. The normal cure for this is to increase the ASIO buffer size. If you let Podium run a while to get several red flashes, then you should see some ASIO errors on the report page in the interfaces setup dialog? Post the report here or email it to me.

Once every 44 seconds you say? Are you using 44100 samplerate? If you select a different samplerate, does this affect the error interval?

P.P.S. How is the Z-reverb progressing ?

I just posted a progress update in the zReverb topic.

However, if the MIDI file in the List Browser is then dragged to a track doesn’t that create a link from the List to the Arrangement?

No. The MIDI file gets imported as a sequence of notes, all contained within the Podium project file. Sound objects are linked to their external file because the wave data are not embedded in the project file.

If that is not the case at present, could it be in a future development

It could be, but that’s a long way down the road.

The name of the sound is stored within the wav file. Podium uses this ‘name’ chunk when importing wav files. If you want it to behave like you describe, then I should modify the “export segment” command to store the user-supplied filename as the name inside the wav file, instead of the name of the sound object?