View Full Version : Is CCE Beaten?
DVD Maniac
6th May 2006, 11:23
I have been using CCE since the early days of Rebuilder and usually with great results. Until recently, I have not tried out any of the other supported encoders on the basis that CCE has always been considered the de-facto higher quality standard. With the new release of Rebuilder I have been trying out AutoQMatEnc with some amazing results (although it is quite a lot slower).
Reading the other threads regarding quality comparisons form this and other encoders it seems that CCE is falling behind its (mostly) freeware competition.
So what's happened? I would be interested to know if the DVD producers stll use CCE or other rival products - this would obviously give an indication of any shift from their perspective - anyone in the know on this?
buzzqw
6th May 2006, 13:32
for NON interlaced encoding i use and prefer HCenc
for interlaced encoding (and want an output interlaced, like from pal dv source ) i use cce. No doubt (never tryed Procoder).
recently i have encoded 1 hr pal dv - bff shotted from my sony camera (a nice holiday in Provence, France), aiming for dvd.
I capped bitrate to min 3000 max 9000 average 8500.
Both cce and hcenc18 were feeded by this
avisource("c:\holidays\provenza.avi)
converttoyuy2(intarlaced=true) -> used for CCE
converttoyv12(interlaced=true) -> used for HCenc
i got very detailed and preserved image with cce (SP 2.70.02.06, vbr 3 pass) while with hcenc the image come out a little "washed"
just my 2 euro-cents
BHH
I find ProCoder2 to be an excellent choice for sources that otherwise would look too blocky in CCE and HC.
Procoder2_v2.04.02.mkv (http://dump.cinematik.net/kumi/Procoder2_v2.04.02_-_V01001800001002.mkv)
CCE_2.70.02.01.mkv (http://dump.cinematik.net/kumi/CCE_2.70.02.01_-_V01001800001002.mkv)
HC_0.17.mkv (http://dump.cinematik.net/kumi/HC_0.17_-_v01001800001002.mkv)
Samples encoded in DVDRB @ vbr_brate_avg=2000 to exaggerate the macroblocking.
PhillipWyllie
24th May 2006, 15:55
Maybe CCE isn't falling behind, just that the competition is catching up. I've found that CCE always produces(visually) identical output, and it's the fastest encoder I've seen. The only exception is that the colours sometimes seem a bit bland(especially on animation-bright colours etc..)
Oh a bit of advice don't cap the minimum bitrate to high, let the encoder encode as low as it can( I set my min to 0). This will allow the best distribution of bitrate.
Boulder
25th May 2006, 19:30
ProCoder looks good at low bitrates probably because it tends to smooth the image.
@PhillipWyllie: you might want to check out ColorMatrix, it'll help with the color problem.
PhillipWyllie
26th May 2006, 02:01
Colourmatrix looks complicated. I just use avisynth's tweak function( increasing the saturation to compensate).
Boulder
26th May 2006, 08:43
What's so complicated about it? You know what coefficients the encoder you use wants and then you use the correspoding mode. With hints enabled in MPEG2Source and ColorMatrix, you don't need to know if you need to use the filter or not.
PhillipWyllie
26th May 2006, 14:03
Oh colormatrix is an avisynth plugin, I thought you meant http://www.codeproject.com/vb/net/colormatrix.asp. I'll give it a try on my next project. I've just given ProCoder2 a quick shot and I'm not impressed(yet).
happycase
27th May 2006, 07:25
Yes, CCE is beaten. HC0.18 and the latest release of AutoQMatEnc and both of their just as good and often better than CCE quality proves it.
This is just my opinion, of course. CCE is still a great encoder, but HC0.18 and AutoQMatEnc give better output, but with a big increase in encoding time.
rahzel
30th May 2006, 05:16
Yes, CCE is beaten. HC0.18 and the latest release of AutoQMatEnc and both of their just as good and often better than CCE quality proves it.
This is just my opinion, of course. CCE is still a great encoder, but HC0.18 and AutoQMatEnc give better output, but with a big increase in encoding time.
i disagree that HC has clearly surpased CCE, but i agree with Phillip and that the other encoders are catching up.
i myself still prefer CCE over HC, but HC is quickly catching up. in my tests with CCE vs HC, i find HC is closer to the source, but HC is too blocky in low bitrate or smokey/cloudy scenes. CCE is not perfect either, though... i find that CCE adds edge artifacts, but they cannot be seen unless you zoom it pretty close. overall, i think CCE is still a little better, but not much at all, and of course this is just my opinion.
i have not tried AutoQMatEnc or QuEnc yet, so i cannot comment on those encoders.
happycase
31st May 2006, 08:01
I prefer HC because CCE tends to add mosquito noise, which isn't often a problem for HC. As you mentioned, HC does try to preserve as much of the source as possible, while CCE, I think, tries to have a smoother picture. When my bitrates are above 3.5 or so, I'll use HC, but if below, I'll use CCE or AutoQMatEnc.
rahzel
31st May 2006, 09:48
i think it depends what matrix you use. i myself always use AutoQ2 aka Avamat7, as i find it removes a lot of mosquito noise and blocking but keeps details. ive gone as low as 2900 for the bitrate, and CCE with AutoQ2 is still very close to the source, without smoothening the picture.
i really wonder why this matrix isnt more popular. its by far the best matrix ive tested, and it seems to work well with every bitrate. i use this with all my projects, so i pretty much use the same settings for every movie.
Boulder
31st May 2006, 09:58
It also depends on the avg bitrate quite a bit. You can also adjust the flat part priority setting (or whatever that is in the newer versions), the lower the value, the less mosquito noise there should be - at the cost of the flat areas suffering.
If you have adaptive quant matrices enabled, the matrix you set may not be used often, especially if the avg bitrate is high and/or the movie compresses well.
PhillipWyllie
1st June 2006, 06:30
If you set adaptive Q matrix to on, the quant matrix you set won't be used at all. Maybe we could have an encoder shootout,and let the Doom9 community decide what looks the best.
Boulder
1st June 2006, 08:09
The quant matrix you set will be used - unless the bitrate (or compressibility) is so high that CCE decides to use matrices that are either a half or a quarter of the one you set. The matrix you set will be the most quantizing one so it's likely that it will be used in scenes with high bitrate demands.
Trahald
3rd June 2006, 15:02
CCE's big advantage was not only great quality, but speed. Tmpgenc was/is pretty good if you are willing to wait. The other encoders have really advanced quite a bit per quality. and I agree, there are some encoding conditions where i actually prefer HC over cce. CCE still wins the speed/quality bar... at least IMHO.
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.