Intermittent sound-bullet truncation

[Leonid Spektor]

Kevin,

	When you say "sound-bullets do not cover all the audio" do you mean the actual numeric value of bullets changes or the audio clip associated with bullets values stops playing back sooner than at other times?


If the actual bullets values change, then I would like to get a sample of your data file and MP3 file that exhibit this problem.


If playback duration changes a lot, just under a second of time or more, then again I would like to get a sample of your data file and MP3 file that exhibit this problem.


If playback duration changes a little bit, than the problem could be with computer's CPU use by other applications or components during playback interferes with audio's playback. Or user clicks with a mouse or presses a key on keyboard during playback. Or bullets were created using uncompressed media file and playback is done with compressed MP3 file or vice versa.

For accuracy we recommend using only non-compressed media during creation of bullets and during playback. The next best thing is to use the same MP3 media file for both creating bullets and playback. MP3 files are very compressed, sometimes ten times of the original non-compressed media. This means that bullet's time values will not match the same starting and ending points in compressed media as in non-compressed one. Or even using the same compressed media on two different OS systems or two widely different in performance computers or using two different versions of QuickTime. It all depends on how the MP3 media playback is implemented on given computer. MP3 playback code was designed to play long, continuos portion of media file, such as MP3 songs. It doesn't seem to handle playing back small clips with quick starts and stops from different parts of media file.


Leonid.



On Jun 1, 2012, at 09:22 , Kevin Donnelly wrote:

> Hi
> 
> We're having problems with intermittent truncation of the sound-bullets (ie 
> the sound-bullets when first assigned cover the whole span of the relevant 
> piece of audio, but on subsequent checking the same sound-bullets do not cover 
> all the audio, but a span very slightly less, meaning that the audio is 
> truncated when replayed).
> 
> We're seeing this with four transcribers, using CLAN versions as follows:
> F - 21 February 2012
> L - 12 April 2012
> E - 2 May 2012
> S - 12 July 2012
> 
> At first we thought that the mp3 conversion (using iTunes) was at fault, so we 
> asked L and F to use an mp3 converted using Audacity.  This resolved L's 
> problems, but F is still experiencing problems.  We asked S to try using the 
> original wav, but the problems remained.  S has the same version of CLAN at 
> home and at work, but only experiences the problems on the home PC where she 
> is working on our file.
> 
> The problems are intermittent, that is, on some hearings the truncation may 
> occur, and on others not.
> 
> Has anyone come across anythng like this, and if so, wht might be the cause?
> 
> I can post the mp3 if that would be useful.
> 
> -- 
> Pob hwyl / Best wishes
> 
> Kevin Donnelly
> kevindonnelly.org.uk
> 
> -- 
> You received this message because you are subscribed to the Google Groups "chibolts" group.
> To post to this group, send email to chibolts at googlegroups.com.
> To unsubscribe from this group, send email to chibolts+unsubscribe at googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/chibolts?hl=en.
> 
> 

-- 
You received this message because you are subscribed to the Google Groups "chibolts" group.
To post to this group, send email to chibolts at googlegroups.com.
To unsubscribe from this group, send email to chibolts+unsubscribe at googlegroups.com.
For more options, visit this group at http://groups.google.com/group/chibolts?hl=en.