Clip+:divide by zero at 3080EB08 while playing first track of particular ape+cue

[mlt]

At the end of the first track (~4:22) of one of my APE's (+cue) I hear loud noise and then I get "Divide by zero at 3080EB08" message. This behavior takes place on Sansa Clip+ with firmware versions r29093-110119 and r29086-110118 (with the same reported address). Other tracks on the same ape (even of greater length) play well. Same track plays well in WinAmp. I guess it is very specific problem. I'll appreciate any help on how I can pin point this issue.
Post Merge: January 19, 2011, 06:45:51 PMI got "Divide by zero at 3080ED78" while listening other track on the same ape.
Post Merge: January 19, 2011, 06:58:55 PMThese bugs are always reproducible for me.

[Chronon]

It seems like a good idea to post a bug report with a link to the file.

[mlt]

It seems like a good idea to post a bug report with a link to the file.

Could you suggest any lossless splitter for ape? Original file is quite huge to post.

[Chronon]

I can't suggest anything specific, but searching for "split ape" on google returns lots of promising results.

[mlt]

I can't suggest anything specific, but searching for "split ape" on google returns lots of promising results.

Unfortunately most of them imply the use of CUE. Oh well... I'll see what I can do.

[saratoga]

I can't suggest anything specific, but searching for "split ape" on google returns lots of promising results.

Unfortunately most of them imply the use of CUE. Oh well... I'll see what I can do.

Is that a problem?

[AlexP]

I can't suggest anything specific, but searching for "split ape" on google returns lots of promising results.

Unfortunately most of them imply the use of CUE. Oh well... I'll see what I can do.

You could just create a cue file that has a thirty second first track and then split that or something.

[mlt]

The use of CUE file is just a logical split of APE. I'm not aware of the way to physically losslessly cut APE. ffmpeg refused to cut ape with -acodec copy .

I'm uploading the entire file solely for the purpose of debugging. I've created cue sheet to locate buggy places.  It is quite huge download totaling 388 Mb. The password on 2-volume zip file (made with 7-zip) is "rockbox" without quotation. It unpacks well for me.

First zip volume: http://www.sendspace.com/file/87qsrz
Second zip volume: http://www.sendspace.com/file/f2plng

FlySpray bug: FS#11897

[z8080]

I checked this file.

Bug is present with build r29203.

[mlt]

Can somebody with write access test and commit a patch I attached to FlySpray? It supposedly closes FS#11897. http://www.rockbox.org/tracker/task/11897?getfile=23287 . At least demac works fine for me.

This bug is quite rare, but all players are affected.

[Chronon]

You'll get much faster attention if you go to the IRC channel and request a review of your patch there. 

EDIT: It looks like your patch was committed.  Thanks for contributing!