Topic: USB1.1 24-bit audio drops?

[lordzoster]

Hallo
i experience rare and non periodic drops with echo: a first louder click and two quieter ones, at equal intervals.
I'm running this setup:
1.notebook Acer Travelmate 427 2.5GHz non-mobile P-IV, 512MB Ram, 4200Hitachi HD, USB1.1, WXP sp1 optimized (16" boot time), Battery or Line powered. No other device on USB bus; network disabled.
2.Apogee Mini Me @ 24/44.1
3.Wavelab 5; USB ASIO drivers, 11ms Input latency, correctly clock-synced.

The drops/clicks occurs even during monitoring and playback. During recording they are captured in the audio signal.
The "echo" effects is very similar to a monitoring feedback. Click loudness is directly proportional to program loudness.

Any idea?

[dunebug81]

If you suspect its because of your USB 1.1 port go to best buy (or computer store of choice) and get yourself a PCMCIA USB 2.0 card.  You can get one from best buy for about $35 and its worth every penny.

Its odd that your computer is a hella fast laptop with crap-ass USB 1.1 ports.

-Greg

[dklein]

Don't go for low latency.  You have no need unless you're trying to add live tracks to existing material.  Max out the latency period.  Also, is there any relationship between your screen turning off/on and the clicks?

[lordzoster]

thanks for the answers,
actually i don't care for low latency, i choose a "normal/relaxed" ASIO latency settings and 11ms i what i got, i'll set it higher.
no screensaver nor screen off.
i'm gonna try with a pcmcia USB2.0 i already had.

[lordzoster]

Ok i found out:
A. there's a party upon the IRQ9, i took a walk and met Mrs. Pcmcia, Mrs. Firewire, Mr. USB and Her Majesty The Video Card (ATI Radeon Mobility 7500). Anyone invited.
B. Each drop measure exactly 44samples, that is approx. 1ms since the audio is 44.1KHz.

I moved the "Hardware Acceleration" fader, in the Video Advanced Properties, totally to the left; this afternoon i'll do other tests, even with a FireWire interface, a gorgeous Terratec Mic8 FW with superb SPL Pres (somebody there at the R&D section forgot to implement quite a limiter for each mic channel... anyway it rocks.)
For the crowded IRQ9, i don't see any solution, since this notebook has a very few BIOS options (among them, "Exit and save changes" or "Exit and discard changes", useful if there were any changes possible). Does anybody knows (if and) how to change the IRQs from within WindowsXP?

tia

[dklein]

I don't think you need to worry about the IRQ9 thing.  Everything appears to be there but I'm pretty sure it's dynamically allocated.  We haven't needed to worry too much about IRQs since win2k.  There's likely some other system event that's causing the problem.

[lordzoster]

Again here to update you about my scrumblescramble with the MiniMe and my notebook.
Summary of last episodes:
glitches in the recorded/played/monitored signals: everything insisting on the IRQ9 (a.k.a. Disneyland Resort Paris) is deactivated except for firewire, usb and Her Majesty the Graphic Adapter. A deeper analysis reveals that glitches are 44samples-long drops: mathematicians are kindly invited to express their opinions about the relationships between a 44sample drop and the 44.1KHz sample rate. Respondez, s'il vous plaît.

Back to modern era.
I did a wholeful of tests [i red lot of terms in english spanglish yankeeish iddish to express this concept, i don't remember any of them: italian would'nt be well-understood by the majoirity of the spectators here, so...]
1. recording with firewire goes OK (12 hrs, 8-trks, 16/44.1)
2. using the MiniMe on a desktop system is OK
3. i'll go on testing: eg. with new USB Audio Propagamma Drivers, but they aren't so specific for the minime.
4. i'm getting bored.
Have a nice week-end, i'll try too.

[dunebug81]

If you suspect its because of your USB 1.1 port go to best buy (or computer store of choice) and get yourself a PCMCIA USB 2.0 card.

Unfortunately, the client device would still be using USB 1.1, right?  Sorta like plugging an NJB into USB 2.0.

A PCMCIA USB might use a different IRQ, etc, and that might help.

If the device is a high-speed device then it will take advantage of the faster speed.  As for the IRQ settings I dont think that has anything to do with it...it just sounds like the computer or something cant keep up with all the information being processed.

-Greg

[phatdats]

sorry if this is outdated info, but when I first got my MME (one of the first rev C's out) I was told by gary that there was a known problem with usb1.1 , the mini me and 24/96.  there were a lot of threads on this board (and the old oade forums) about usb1.1 not being able to handle 24/96 most of the time.  gary told me that they were working on a firewire upgrade (which he has since backed down from and the best I could hope for officially from my minime was 24/48....

A few people here on the board got it to work at 24/96, but the majority of us were not.

[it-goes-to-eleven]

sorry if this is outdated info, but when I first got my MME (one of the first rev C's out) I was told by gary that there was a known problem with usb1.1 , the mini me and 24/96.  there were a lot of threads on this board (and the old oade forums) about usb1.1 not being able to handle 24/96 most of the time.  gary told me that they were working on a firewire upgrade (which he has since backed down from and the best I could hope for officially from my minime was 24/48....

A few people here on the board got it to work at 24/96, but the majority of us were not.

Good to hear some have 24/96 working..  I haven't given it a serious effort on my MME.

Back to this case,  I think the original poster is bummed because he can't even do 24/44.1 reliably..

The lack of digital input on the MME makes testing difficiult.  With digi in, you can send the exact same stream over and over and simply check the data for bit accuracy.. It, at least, makes the validation much easier.

[CHURCH-AUDIO]

Hallo
i experience rare and non periodic drops with echo: a first louder click and two quieter ones, at equal intervals.
I'm running this setup:
1.notebook Acer Travelmate 427 2.5GHz non-mobile P-IV, 512MB Ram, 4200Hitachi HD, USB1.1, WXP sp1 optimized (16" boot time), Battery or Line powered. No other device on USB bus; network disabled.
2.Apogee Mini Me @ 24/44.1
3.Wavelab 5; USB ASIO drivers, 11ms Input latency, correctly clock-synced.

The drops/clicks occurs even during monitoring and playback. During recording they are captured in the audio signal.
The "echo" effects is very similar to a monitoring feedback. Click loudness is directly proportional to program loudness.

Any idea?

Acer is using a non standard Usb controller. It might also be that there is something using the usb port or making calls out to the usb port try getting a pcmcia card usb and seeing if that helps also usb 2.0 would help somewhat although for audio at 24/44.1 USB 1.1 should work ok for the $20 bucks to try another pcmcia card out you can't go wrong.