clip.start in a playlist shows a spurious frame when starting clip - rtmp and pseudostreaming

[GoogleCodeExporter]

What steps will reproduce the problem?
1. given a playlist where start is configured for the clips

What is the expected output? What do you see instead?
Expected: clip starts at start position
Instead: there's a spurious frame displayed before the clip actually starts

Samples:

rtmp:
http://flowplayer.blacktrash.org/test/rtmp-start.html

pseudostreaming:
http://flowplayer.blacktrash.org/test/pseudo-start.html

Original issue reported on code.google.com by blacktrashproduct on 13 May 2012 at 9:38

[GoogleCodeExporter]

Interestingly enough the best look and feel for this scenario is achieved with 
the otherwise hacky php-pseudostreaming:
http://flowplayer.blacktrash.org/test/php-pseudo-start.html
Perhaps because it's so snappy, snappier than any true streaming protocol 
(until the server is overloaded of course).

Original comment by blacktrashproduct on 14 May 2012 at 12:08

[GoogleCodeExporter]

I don't see an issue with either of them apart from buffer start times. your 
php one takes a while to start with a frame with his head down, the rtmp one 
starts with his head up with banjo in hand. screenshot perhaps. 

Original comment by electrot...@gmail.com on 14 May 2012 at 2:58

[GoogleCodeExporter]

You will see the difference when you jump back and forth between the clips 
using the playlist buttons. rtmp flickers (image disappears for short moment), 
as if it were restarting; the impression with php-pseudostreaming is way 
smoother.

Set priority to low.

Original comment by blacktrashproduct on 14 May 2012 at 3:39

• Added labels: Priority-Low
• Removed labels: Priority-Medium

[GoogleCodeExporter]

ahh gotcha, I honestly don't know thats part of the video buffering, its 
grabbing a frame then buffering. 

Original comment by electrot...@gmail.com on 14 May 2012 at 4:03

[GoogleCodeExporter]

on pseudostreaming it's getting partial content, cancelling it, then starting 
again thats what you are seeing, very strange and needs debugging. 

Original comment by electrot...@gmail.com on 14 May 2012 at 4:07

[GoogleCodeExporter]

its doing the same on the php implementation. there is perhaps something that 
can be done about it though. 

Original comment by electrot...@gmail.com on 14 May 2012 at 4:09