Which was is better to produce the best quality

[WeeJavaDude]

I have Adobe Priemere and CCE basic and I am trying to figure out what will produce the best quality at the fastest rate. Last night I started encoding using the CCE Adobe plug-in and it stated it would take 13 hours to encode. I have a 1Ghz 512mb machine. (Going to start looking into upgrading, but that is another thread.) I thought I had set it to only encode the work area which was 10minutes of video, but I man have set it to do the entire timeline which is 1hr and 30 minutes.

Questions I got:

1) Is is better to output from Adobe to the AVI native formate of the video and then use CCE to encode to MPEG. Does this result in PQ loss? Will this give better performance overall?

2) Is it better to use a frame grabber inbetween. Any PQ loss compared to the plug-in?


For 10 minutes of video what could I expect for encoding time on a 1Ghz machine? 13 hrs seems a bit high. Going to check tonight if I did not set the option to only do the work area.

Any opinions on what is the best way to use CCE with Adobe to produce the best PQ at the fastest rate would very much appreciated.

[flaninacupboard]

Have a look through the guides to check for frameserving from premiere - don't know for sure if it's supported, but that would be your best option.

i would guess you're encoding all 90 minutes, and probably with at least two passes. although rendering in premiere can be quite slow if you have fancy transitions etc.

Raw encoding on my athlon 650 used to be .9 real time, so 90 minutes would take about 100 minutes a pass.


If you use premiere to output to .avi (use huffyuv as it's lossless) your rendering will be done so encoding from the .avi will be much quicker (but the rendering will still be dead slow.)

[WeeJavaDude]

i would guess you're encoding all 90 minutes, and probably with at least two passes. although rendering in premiere can be quite slow if you have fancy transitions etc.

Yes it was the full 90 minutes but that should not take 13 hours even with two passes. I did a 35 minute test and it said 2 hours. That still steems a bit on the high side given that I am using a 1GHz AMD with 512MB memory.

I think the best route for me to take is framegrabbing. Someone suggested using the FrameServer for www.debugmode.com. I am going to try that and do a huffy test to see what the difference is. Thanks for the advice.