View Full Version : Anyone getting CCE Encodes above 3.0???
Roveer
12th November 2002, 01:02
Let's cut to the chase. It's all about speed. I'm building a new encoding system and am really interested in the best CCE encode times.
Has anyone achieved CCE encode speeds above 3.0? If so, post your speed & config here. The research I've done so far on 2CPU Mobo's has left we wondering if they will do any better than the fastest P4 CPU's on a well configured system.
So, no more BS, let's just hear about the fastest darn CCE encoder there is!
Roveer
Matthew
12th November 2002, 05:17
Avisynth supports only one processor. CCE supports dual processors.
I can get absolutely kickass speeds with dual xeon (definately greater than 3) feeding an avi into CCE directly, but if I go via avisynth the speed drops right down.
Roveer
12th November 2002, 05:53
So are we saying in my example that Avisynth will be the bottleneck? I'm going from VOB's with DVD2AVI to a D2V file that I need to frameserve into CCE. Is there another process that wouldn't take hours to create avi's that can be fed into CCE to get these amazing speeds?
Thanks,
Roveer
Matthew
12th November 2002, 06:42
Yeah avisynth is the bottleneck.
You can create a huffyuv avi using DVD2AVI and then feed that into CCE but you'd need a massive HD =( [I just saved a 7 min sample and it came in at 3.8 GB]
Or you could encode 2 films using avisynth/cce simultaneously (or the one film split in two but that's an irritation).
I get 2 or above using avisynth and CCE 2.66, and that's good enough for me =)
Sp33d
12th November 2002, 14:11
What's your config Matthew. I can't get more than 1.3 out of my P4 2.0Ghz.
Matthew
13th November 2002, 00:24
2 x 2.4 ghz xeon, 1 gig RDRAM, supermicro motherboard. Encoding is done (both source and destination) on an 80 GB IBM (7200 rpm).
Actually I should correct myself, CPU usage is above 50 percent when running CCE, and when I only allow it to run on one CPU the speed drops quite a bit. So dual processors is still faster than a single processor, just not as fast as it should be. Still getting 1.75 with one processor though (chapter 1 LOTR).
asifanwar
13th November 2002, 07:39
VOB to DVD2AVI (either in FILM or not)
d2v to CCE2.66 via AVIsynth2.06 (with no resizing/filters)
Speed: 2.8x realtime
System:
Pent4 2.4GHz overclocked to 2.61Hz
512Mb PC1066 RAMBUS RAM (overclocked to 1160)
2x Maxtor 40Gb ATA133 7200rpm HD as RAID0
Gigabyte 8IHXP mobo
WinXP Pro SP1
I noticed that my speed went up notably when I switched from DDR to RAMBUS (I used to get 1.8x on my Dual Athlon1600 with 512Mb PC2100 DDR, and even a Dual Athlon2200 could only manage 2.1x with the same DDR...)
I have not tried a Huffy as I am not sure how to make one. (I would only do it for comparison purposes, cos it is too big to be practical)
Prosper
18th November 2002, 18:20
Huffy .AVI into CCE -> SVCD MPEG2 (480*480) on my Dual XP2000's is around 6x
Arky
18th November 2002, 23:50
Originally posted by Matthew
2 x 2.4 ghz xeon, 1 gig RDRAM, supermicro motherboard. Encoding is done (both source and destination) on an 80 GB IBM (7200 rpm).
Actually I should correct myself, CPU usage is above 50 percent when running CCE, and when I only allow it to run on one CPU the speed drops quite a bit. So dual processors is still faster than a single processor, just not as fast as it should be. Still getting 1.75 with one processor though (chapter 1 LOTR).
Matthew, I strongly suspect you'd see a nice speed improvement if you used seperate source and destination HDDs. Not massive, but definitely noticeable.
Very nice system, BTW! :) How much were the Xeons + MoBo, if you don't mind me asking?
Incidentally, I notice TMPGEnc 2.59 now supports Intel's hyperthreading technology, not that I imagine this would be as effective as real SMP.
Arky ;o)
Matthew
19th November 2002, 00:44
Originally posted by Arky
Matthew, I strongly suspect you'd see a nice speed improvement if you used seperate source and destination HDDs. Not massive, but definitely noticeable.
Only problem with that is that the other drives aren't particularly large (they are SCSI). In the end it's just easier to have everything done on the large IDE drive =)
Originally posted by Arky
Very nice system, BTW! :) How much were the Xeons + MoBo, if you don't mind me asking?
To be honest I don't remember (I didn't pay :D, and can't find the specs file) but it sure wasn't cheap. I'm in au so any price I gave you would be distorted anyway - our exchange rate is a shocker :( [I think the box was roughly ten grand though].
Just as a bit of trivia, for the box IBM quoted a price more than 50 percent above the chosen supplier :eek:
auenf
19th November 2002, 10:13
Originally posted by Matthew
To be honest I don't remember (I didn't pay :D, and can't find the specs file) but it sure wasn't cheap. I'm in au so any price I gave you would be distorted anyway - our exchange rate is a shocker :( [I think the box was roughly ten grand though].
which would work out to about 10 pence over your way Arky :P
Enf...
poopity poop
20th November 2002, 09:11
you get 6X on your dual 2000XP's??!?!?!?!
how?
I have dual 1900+'s with 512MB ram feeding off RAID 0 IBM deskstars:
HuffyYUV .avi -> CCE
I get no more than 2.5
Tell me how you do this, and keep it not from crashing. Right not I sometimes feed the .avi through avitynsh to stablize it. Please tell me what settings you are using with huffYUV and what version of CCE you are using. I'm still on 2.5, I'll get the next versions sometime, and using RGB(fastest) compression in huffy. How are you attining such high speeds?
auenf
20th November 2002, 12:52
hes also doing svcd, which is less frame to encode ;)
Enf...
Prosper
20th November 2002, 21:35
Originally posted by poopity poop
you get 6X on your dual 2000XP's??!?!?!?!
how?
I have dual 1900+'s with 512MB ram feeding off RAID 0 IBM deskstars:
HuffyYUV .avi -> CCE
I get no more than 2.5
Tell me how you do this, and keep it not from crashing. Right not I sometimes feed the .avi through avitynsh to stablize it. Please tell me what settings you are using with huffYUV and what version of CCE you are using. I'm still on 2.5, I'll get the next versions sometime, and using RGB(fastest) compression in huffy. How are you attining such high speeds?
AVISynth is doing it. AVS is not SMP optimized, I find that when I have to framserve to CCE, I hit around 55% utilization, otherwise it'd hit 99%. For me, TMPegenc without avs is faster than CCE with avs. Plus, yah, SVCD is a lot lower rez than DVD
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.