RESOLVED WONTFIX 82348
Audio Element playbackrate < .5 are rounded up
https://bugs.webkit.org/show_bug.cgi?id=82348
Summary Audio Element playbackrate < .5 are rounded up
Terry Ryan
Reported 2012-03-27 09:07:42 PDT
When using the playbackrate attribute of the audio element, rates less then .5 are rounded up to .5. The attached example shows this in action. There are several buttons that playback the content at different rates. Playback rates for values from .5-2 are accurate. The page predicts playback duration for each button press based on audio duration/rate. Under .5 always yield actual durations approximate to predicted rates Over .5 always yield actual durations approximate durations based on a .5 rate. A casual listen to the audio at a ridiculously slow playback rate (say .1) will clearly sound like only half speed.
Attachments
Html page and m4a file for testing the audio results in bug report. (339.48 KB, application/zip)
2012-03-27 09:08 PDT, Terry Ryan
no flags
Terry Ryan
Comment 1 2012-03-27 09:08:59 PDT
Created attachment 134083 [details] Html page and m4a file for testing the audio results in bug report.
Eric Carlson
Comment 2 2012-03-27 14:16:13 PDT
Not all media engines are able to play back audio at arbitrary rates. In this case the lowest rate supported is apparently 0.5, so it is being clamped to that rate.
Terry Ryan
Comment 3 2012-03-27 14:26:25 PDT
(In reply to comment #2) > Not all media engines are able to play back audio at arbitrary rates. In this case the lowest rate supported is apparently 0.5, so it is being clamped to that rate. Thanks for looking at this. Is playing at .5 the right behavior in that case? Wouldn't throwing an error be preferred?
Eric Carlson
Comment 4 2012-04-04 12:39:16 PDT
(In reply to comment #3) > (In reply to comment #2) > > Not all media engines are able to play back audio at arbitrary rates. In this case the lowest rate supported is apparently 0.5, so it is being clamped to that rate. > > Thanks for looking at this. > > Is playing at .5 the right behavior in that case? Wouldn't throwing an error be preferred? There was a discussion about this about 18 months ago on the HTML5 mailing list. The spec editor feels strongly that a UA should do whatever tricks it has to to make it appear that playback is progressing at playbackRate, for example seeking and playing a snippet, seeking again, etc. A bug was filed against the HTML spec about this, but it was closed as WONTFIX. See http://www.w3.org/Bugs/Public/show_bug.cgi?id=10837 for a discussion. WebKit previously returned the actual playback rate, ie. the rate at which the media engine was playing, but because of the spec change it was updated to always return whatever value was last set.
Terry Ryan
Comment 5 2012-04-04 12:50:23 PDT
(In reply to comment #4) > (In reply to comment #3) > > (In reply to comment #2) > > > Not all media engines are able to play back audio at arbitrary rates. In this case the lowest rate supported is apparently 0.5, so it is being clamped to that rate. > > > > Thanks for looking at this. > > > > Is playing at .5 the right behavior in that case? Wouldn't throwing an error be preferred? > > There was a discussion about this about 18 months ago on the HTML5 mailing list. The spec editor feels strongly that a UA should do whatever tricks it has to to make it appear that playback is progressing at playbackRate, for example seeking and playing a snippet, seeking again, etc. > > A bug was filed against the HTML spec about this, but it was closed as WONTFIX. See http://www.w3.org/Bugs/Public/show_bug.cgi?id=10837 for a discussion. > > WebKit previously returned the actual playback rate, ie. the rate at which the media engine was playing, but because of the spec change it was updated to always return whatever value was last set. Thanks for detailed response.
Note You need to log in before you can comment on or make changes to this bug.