samsc
07-07-2005, 05:32 PM
I did some h.264 test with an end users content.
All tests were done at multiple resolutions and multiple os's
Tests in Catalyst g21 v3.3 PRO on 10.3.9 and 10.4.2 with Maxtor Atlas 10k 147Gb and SATA 250Gb.
Computer is latest dual 2.7ghz G5.
Quicktime 7.0.1
Playback on single layer only.
h264 at 100% is 3 to 5 times slower than photojpeg at same image size
and the files are 3 times larger.
photojpeg 30% 1400x1050 takes around 15ms/frame
h264 100% at 1400x1050 takes 50-100ms/ frame.
DO not use h264.
Dont even think of using it....it performs really really badly- it is probably only designed for use when a single layer has to be played back- as it sucks all the cpu resources up.
----
tests did indicate that with a dual 2.7Ghz it is almost possible to say that it is capable of playing back 2 layers of photojpeg compressed content at 1400x1050
Real results will depend on content and tests performed on endusers machine
All tests were done at multiple resolutions and multiple os's
Tests in Catalyst g21 v3.3 PRO on 10.3.9 and 10.4.2 with Maxtor Atlas 10k 147Gb and SATA 250Gb.
Computer is latest dual 2.7ghz G5.
Quicktime 7.0.1
Playback on single layer only.
h264 at 100% is 3 to 5 times slower than photojpeg at same image size
and the files are 3 times larger.
photojpeg 30% 1400x1050 takes around 15ms/frame
h264 100% at 1400x1050 takes 50-100ms/ frame.
DO not use h264.
Dont even think of using it....it performs really really badly- it is probably only designed for use when a single layer has to be played back- as it sucks all the cpu resources up.
----
tests did indicate that with a dual 2.7Ghz it is almost possible to say that it is capable of playing back 2 layers of photojpeg compressed content at 1400x1050
Real results will depend on content and tests performed on endusers machine