|Subject:||Difference in latency using the MPEG-2 algorithem|
|Date:||11 Feb 2007|
I'm working on a project where I need to find out how much latency the
MPEG-2 encoder will create, and if it's constant.
I'm using a DV camera connected to an MPEG-2 hardware encoder with
composite cabel. The encoder is connected to the computer using USB.
As fare as I have heard, the algorithem should use notably longer time
encoding video with fast changing frame content, than video with small
or no change of frame content (i.e. filming a white or black screen).
The problem is that the tests that I have doen shows no difference in
latency because of the difference in video content. The main
contributor to difference in the latency seem to be the operating
system and how it decides who to what when.
Do you guys know of anyone who have done tests on this problem, or do
you have an opinion on how much might diferensiate?