Mpeg2Dec2 or Mpeg2Dec3?
I've been reading that Tom Barry's mpeg2dec2 is faster for Pentium 4 processors, but I see kwag's script useds Mark's Mpeg2Dec3.
Is there a quality or feature difference between these? I have a P4 so I'd like to take advantage of it. Thanks |
I guess a better question is does anyone use this option "idct=5"
example: Mpeg2Source("D:\Projects\project.d2v",idct=5) The documentation for Mpeg2Dec3 says this is faster for P4 machines. I notice Kwag doesn't use it even though he has a P4. I don't really understand this setting and am curious if there is a quality impact by using this and if anyone has actually noticed faster frameserving and overall reduced encode time with this option. |
Hi audi2honda,
I hadn't paid attention to that :!: I guess I'll have to run a clip with and without that option, to see how it impacts speed and/or quality :idea: -kwag |
Done :!:
Results are identical in both file size and speed. Did a 30 second clip, and there was no difference. -kwag |
what version of mpeg2dec3 do you have? a new version with optimized code was just released about a week ago.
|
Quote:
Edit: The date is when I downloaded it. Do you have a newer version :?: -kwag |
Give this one a try. You have to set idct=5 in your .d2v file and in your .avs script after the mpegsource call like I put in that simple example above. Let me know if you notice any performance. I'm playing with it too.
http://forum.doom9.org/showthread.ph...ght=idct+speed |
Quote:
http://nic.dnsalias.com/ :wink: |
Ok, same result with 1.08. No diff with or without "idct=5" :)
As: Mpeg2Source("F:\KPAX\VIDEO_TS\movie.d2v") Or : Mpeg2Source("F:\KPAX\VIDEO_TS\movie.d2v", idct=5 ) -kwag |
oh well worth a try :?
|
Maybe on a higher end P4 (2.4+Ghz) there's a difference. Mine is a poor 1.6Ghz CPU :roll:
-kwag |
This is a bit off-topic, but just thought I'd add that there definetly IS a noticeable speed increase if you have a P4 processor and use "idct=5" in your .d2v while using MPEG2DEC2 in Avisynth 2.08. I guess MPEG2DEC3 must need a bit more optimizing. :wink:
-d&c |
i'm going to try it now. Kwag did you put it in your .d2v file also?
|
Quote:
-kwag |
Just ran a short sampler encode with idct=4 vs idct=5 in Avinsynth 2.08 with MPEG2DEC2 to verify my previous statement :wink: :
time with idct=4 in .d2v was 3:25 time with idct=5 in .d2v was 3:05 that's with a P4 2.26Ghz, btw. It's a pity if the same speed boost doesn't ring true for MPEG2DEC3. -d&c |
Then what about MPEG2DEC2 vs. MPEG2DEC3 comparison both on 2.52 :?:
-kwag |
I'm running some tests with mpeg2dec3 and will have results in about 10 minutes. THen I'll use mpeg2dec2
|
I'm creating a new project file with DVD2AVI 1.77.3 to compare against the one I have, which was done with DVD2AVI 1.76.
-kwag |
Well, I see no diff with my CPU. I just tested with DVD2AVI 1.77.3, and verified the project file. It's using SSE-2,MMX. Here's the line:
iDCT_Algorithm=5 (1:MMX 2:SSEMMX 3:FPU 4:REF 5:SSE2MMX) Maybe my P(iss)4 CPU being 1.6Ghz, it's not good enough to see a difference :!: -kwag |
I don't see any difference either. I ran a few times and there was no difference between the plugins. I have a P4 1.8. I've got a 2.8ghz at work that I can try a few tests on tomorrow, but it seems the difference is negligable.
I'd try mpeg2dec2, but i'm not giving up AVS 2.5 |
Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.