Speed of playback

[kozikowski]

Distilling a little bit:

You are all on Windows machines. XP?

You are all using Audacity 1.3.x.

Your track specifications (on the left) are 44100, 32-bit float, Mono.

Anything with show length? All these shows are over 30 minutes long, for example?

You are all creating tracks that fail in exactly the same place every time you play them?

Does the sound pitch error ever go down? They always go up?

Are the tracks still damaged if you play them on a different computer?

Koz

[steve]

Is this only with USB interfaces?

[Gale Andrews]

So for at least randalbarnes, it's actually a recording problem, not a playback problem? And the common factor is USB external interfaces?

If it is a playback problem for some, these are not unheard of when playing back to a USB interface, especially if you are recording while playing back. That takes a lot of USB bandwidth especially at higher sample rates. Have the others tried playing the recording to the inbuilt sound device (make this change in Audacity Devices Preferences), to see if the speed up occurs?

USB interfaces do require you to match sample rates in all possible places, and if you are on Vista or 7, "Default Format" is another place you have to check. Plus, do you all have correct up-to-date drivers for the USB interface for your particular version of Windows, obtained from the manufacturer's site?

So find out what sample rates the USB device is supposed to support. Preferably look in its Manual, or try Help > Audio Device Info in Audacity. Set Default Sample Rate in Audacity Quality Preferences to a supported rate so you normally play/record at that rate (but change the project rate bottom left of Audacity if you have to). If the USB interface has a control panel, set the sample rate to match there. If you are on Vista or 7, go to Sound in the Windows Control Panel, click Recording tab, right-click over the device > Properties then hit the Advanced tab. Set the Default Format to match with that in Audacity and in any control panel the device has.


Gale

[kozikowski]

Eagerly waiting results.

I have a really rough time believing any of those settings can let you get many minutes into a show and only then screw up. Generally, if you have a bit rate mismatch or other difficulty like that, it fails right away.

I've only had one or two problems like that in the millions of years I've been doing this. I got called in to solve a communication crashing problem between a PC and a mainframe (an idea how long ago this was). Turns out, the mainframe was wired with a cable missing some of the wires, so it couldn't signal when it was filling up. Sometimes it would go an hour or more before it crashed.

These "wait a while" problems are celebrities.

Koz

[randalbarnes]

Well, I have finally found out how to duplicate the problem in my particular case. The culprit turned out to the the M-Audio Mobile Pre preamp.

Here are the details of the recording setup:

Toshiba R15-S829 convertible tablet PC
M-Audio Mobile Pre USB preamp

I can record from any of the following Mobile Pre inputs and produce the problem:

1/8 " stereo input
single 1/4 " input
two 1/4 " inputs
single XLR input
two XLR inputs

This Toshiba laptop, like many others, has its own 'power' program in addition to the Windows supplied 'power' program in the control panel.

During a recording, if I click on the Toshiba Power icon in the control panel, I'm guessing that the Mobile Pre is getting put into some strange state, because it starts to send double-rate data to Audacity, which then records it. When I am monitoring the recording, it sounds like the speaker is stuttering, which he is not doing. When I stop the recording and play it back, it is twice as fast as normal (i.e. the 'chipmunk' effect). This does not happen if I record directly into the mic input on the Toshiba (yuk!), or if I use an Audigy 2 ZS sound for the input. I really like the sound quality that I get from the Mobile Pre, so I will keep on using it but will be careful not to do anything else on the computer when I am recording (which I should have been doing anyway).

Thanks everyone, for all your help!

[Gale Andrews]

I have a really rough time believing any of those settings can let you get many minutes into a show and only then screw up. Generally, if you have a bit rate mismatch or other difficulty like that, it fails right away.

I agree, except that I've got several examples filed away where the audible results of the mismatch were only intermittently apparent (as well as those where it was immediately so). This may be machine-dependent, starting to fail when resources get low, or only be completely reproducible when generic Microsoft drivers are used (I have a real life example of that, too).

Randalbarnes, thanks for the helpful follow-up. Something like that is often the problem. Does this happen then purely by running the Toshiba program, and what does it do? Does it try to regulate the power sent to external devices somehow? And yes, even unplugging other high bandwidth USB devices like a USB modem, or turning off a memory hungry browser or photo app can make a differerence:
http://wiki.audacityteam.org/index.php? ... nd_Drivers


Gale

[randalbarnes]

Yes, I only see this behavior when the Toshiba power program is run. I'm not sure if it tries to regulate the power sent to external devices or not. It is similar to the Windows power applet in the control panel, with a few more hardware specific features.

[billw58]

"Sounds" to me like the USB interface is dropping every other sample once the laptop goes into low power mode?

-- Bill