Re: [xwax-devel] New xwax user, Audio interfacing experiences

[Mark H. <ma...@po...>]

On Tue, 20 Nov 2007, Rick Prokup wrote:

> Hello xwax developers!
>
> I recently discovered xwax and have been playing around with it for
> the last two weeks.  I am absolutely blown away by how well it works.
> You guys (and gals?) did a top-notch job writing xwax.  Keep up the
> good work.

Thanks!

> I have run into some issues while using these devices, however.  First
> off, version .2 from the website did not work reliably for me.  I
> downloaded the .3 version and tracked down the things that I was
> missing and got it compiled.  Using alsa devices (in my case, hw:0 and
> hw:1), xwax will work fine for hw:0 when first started up, but hw:1
> will flutter back and forth between .6 - .9 pitch while the deck is
> set to run at 33rpm solid.  Sometimes stopping and restating xwax will
> fix this, sometimes, I have to stop xwax, unplug the audio devices,
> plug them in, then give it a go again.  The most times I've had to
> repeat this cycle to get things stable is 5.  If I can leave the
> turntables running for about 5 minutes and the pitch is solid at 1.0,
> then things tend to stay stable.

This kind of flutter is usually an issue of audio being dropped somewhere. 
Have you investigated the various buffer size options? My first experiment 
would be to use the -m argument to set the buffer size to something large 
(say 1000). xwax won't be very responsive, but I'd expect the flutter to 
go away.

A second test would be to assign xwax to one audio interface and then use 
Audacity to capture audio from the second interface, and check for a 
smooth sine wave without dropouts.

> I'm curious if there is any way to skip forward and backwards through
> a track in xwax besides just spinning the record.  It would be nice to
> be able to skip 30 second blocks forward and back, since you can't
> just pick up the needle and drop it elsewhere in the track.

You should be able to do precisely that -- pick up the needle and drop it 
elsewhere! This is the 'absolute positioning' from the timecode. If you 
can't do this, it means something is wrong (when you get this working, 
you'll be _more_ impressed with xwax!)

There should be a value to the 'timecode:' on the corresponding display. 
If this is blank, you are not getting the position information 
decoded from the input audio.

Record your input audio using Audacity, and check that the waves aren't 
distorted in any way (too loud, perhaps?). xwax decodes the timecode as 
binary numbers from the waveform -- a large wave is a 1, a small wave is a 
zero.

If the wave is clear, it's possible that the phase of the first channel is 
incorrect. Note that this is a bit different to the channels being out of 
phase -- the phase of the first channel is important. This could be, for 
example, the cartridge on the deck wired incorrectly.

Look forward to heading how you get on.

Mark