how good is HC these days compared to CCE?

[dynamis]

i've done a couple movies with both, using Bitrate Redistribution, and they both look great! hc seems to have less of a halo around people's faces. maybe less mosquito noise, but i need to check again later.

the reason i'm asking is that when i do Bitrate Redistribution with a 100% sample (RB-Opt always using HC 0.21), CCE SP 2.70 takes about 3x longer (~90 minutes) to encode with 3 passes on my Q6600 than normal. HC finishes at about the same time (~30 minutes). with 20% sample, CCE finishes around 30 minutes as well. HC seems to utilize quad-cores a lot better as well because of multiple instances. with CCE i get 50-65% cpu utilization (at best!). so i'm thinking about switching to HC completely. so far HC looks great and multiple instances make it an option as fast as CCE for me. is there anything that CCE has that HC doesn't considering my setup? does it handle certain situations better or worse than HC? any help would be great. thanks!

[Fishman0919]

CCE is not optimized for quad core cpu's yet. I emailed Cinema Craft about this after I got my quad core, they responded that they are working on it.

http://www.cinemacraft.com/eng/cpu.html

[dynamis]

CCE is not optimized for quad core cpu's yet. I emailed Cinema Craft about this after I got my quad core, they responded that they are working on it.

http://www.cinemacraft.com/eng/cpu.html


thanks for the info Fisherman.

will i be missing anything if i switch to HC 0.21?

[Fishman0919]

IMO no. HC does as good a job in most case and is better in some.

[Pulp Catalyst]

HC is very good at low bit rates, but at high bitrates, i would still have to give the crown to CCE, for the moment at least, but HC is closing,

but in the case of DVD-5, HC was designed for this level of bitrate, where as CCE was designed for DVD-9,

then of course the cost has to be considered aswell, seriously, how many of us can afford the flagship CCE, CCE basic is good, but sometimes that 3rd pass is needed "We were soldiers" "Matrix" "Saving private ryan"

i would say (this isn't written in stone of course)
Aprox bitrate 2500+ CCE
any lower HC,

but overall for praticle reasons, HC takes the crown,
Cost + DVD-5 + Quad (for you anyway) HC wins

[blutach]

I see extremely good outcomes with HC at high BRs and agree that CCE at low BRs is not up to snuff.

Summing it up, I would - and do - use HC over CCE.

Regards

[kumi]

I prefer HC @ high bitrates over CCE, and like blutach says: CCE at low BRs is no good (you're far better off with ProCoder IMHO.)

[dynamis]

thanks guys for your opinions. i have one more question or so: what vbr bias do you guys use for HC? 20 (dvd-rb default) or 0 (hc default)? i've tried both, but haven't really noticed any difference.. :T thanks again.

[jjaomni]

wow!!!! trying hc now. i really thought 2 gigs of memory would be enough for while even with the quad core i just bought. but damn!! i have a commit of 2333M!! fun..

[jdobbs]

Never would have guessed it would use that much?? If using a quad core you can have 4 independent copies of HC running, though.

[dynamis]

i'm getting the same thing here. finally, an excuse to buy more ram ;)

[Boulder]

wow!!!! trying hc now. i really thought 2 gigs of memory would be enough for while even with the quad core i just bought. but damn!! i have a commit of 2333M!! fun..Try using SetMemoryMax in your script, SetMemoryMax(64) might be enough. It'll restrict the memory usage to some extent.

[trence]

Since the majority of you feel that HC is better then CCE for lower bitrates, how do you all feel on the issue of HC vs. Procoder? I am thinking about buying Procoder, if it is noticeably better than HC. I don't really care at all about speeds, I set up batch jobs for when I go to bed. Thanks for your help.

[kumi]

Personally I much prefer ProCoder over HC at higher levels of compression. Others will certainly disagree.

[blutach]

I won't disagree with you Kumi. IMHO, it is outstanding at low bitrates.

Regards

[trence]

So you would all agree that getting Procoder is worth the money over HC? And that Procoder is better than CCE for the average Full Disc encode (minus movie previews)?

[Boulder]

Apparently Procoder blurs the image so it'll make low-bitrate stuff look better. You can easily do the same for free - and having control over it - using Avisynth so I don't think that it's a valid point to purchase anything..

[kumi]

Apparently Procoder blurs the image so it'll make low-bitrate stuff look better.I've never been able to see blurring in all of my ProCoder jobs, and I've never seen anyone with comparison clips or screenshot to prove it.

I'm starting to think this is an urban legend...

[Fishman0919]

"Blur" may not be the right word... "Blends" may be better.

All ver of Procoder do this... Procoder Express, Pro 1.##, Pro 2.##, Pro 3 and Rhozet Carbon Coder. I can produce screenshot of this.

I have been doing a bunch of test with Pro 2, 3 and Carbon Coder. I had posted before that the streams of Pro 2 and Carbon are the same and that Pro 3 was different... not true. All 3 produce the same stream. The problem was that Pro 3 rounds down the bitrate more so then Pro 2 or Carbon.


Ex. I was encoding a test clip at 3661k... Pro 2 and Carbon both rounded down to 3660k, fine, but Pro 3 rounds down to 3650... very similar too how Procoder Express does. If I encoder the clip at 3650, all 3 produce the same stream bit for bit.

Edit: This larger rounding down would also be the reason Pro 3 undersizes with DVD-RB.

[Digga]

I'm starting to think this is an urban legend...I once did Casshern (very long and action laden movie) with CCE and compared it to a ProCoder 2 encode (made by a friend, we both used DVD-RB with exactly the same setting).
CCEs output was very blocky, while ProCoder softened and blurred the movie considerably.
the later was more nice to watch for both of us (subjective).

I see if I can borrow his encode and post some screen shots and/or clips.

[kumi]

All ver of Procoder do this... Procoder Express, Pro 1.##, Pro 2.##, Pro 3 and Rhozet Carbon Coder. I can produce screenshot of this.
Thank you for your offer! I'm afraid my eyes are not sensitive to the effect you are describing, and seeing screenshots of it would help me learn what to look for.

If I encoder the clip at 3650, all 3 produce the same stream bit for bit.
Hmm, interesting! Looks like I won't be upgrading...

[Fishman0919]

Don't get me wrong... what Procoder does is not a bad thing, it works well at low bitrates... but at high bitrates, other encoders (too me) seem to keep more detail.

[kumi]

Oh definetely, I'm very happy with it. I'm just curious to see the removal of detail that everyone talks about.

[Fishman0919]

Remember, MPEG 1 or 2 is all about visual perception.
Not what is really there... but what you think is there.

[AGKnotUser]

Can someone define "Low Bitrates"? Thanks.

[blutach]

It's a bit subjective, but some might consider sub 2,500 Kbps (others might say 3000 which is the DVDRB definition) as "low" and sub 2000 as very low. DVDRB Pro allows different matrices for each segment to automatically be loaded for these bitrates (and for extras). Or you can pick a matrix for the segment and edit the ECL file by hand or with Rockas Matrix Editor.

Regards

[Boulder]

Depends on the source material - so it is impossible to say what is low and what is high without testing the source ;)

For Predator, 2500kbps is very low, but for 2010: The Year We Make Contact it is quite a lot..

[AGKnotUser]

@Blutach
That's what I've been using as a guideline but I wasn't sure. Thanks for confirming it for me.

@Boulder
I'll have to look at those DVDs to get the idea. I can't understand it in the abstract. On the topic of CCE Basic for low bitrate sources should I use your Degrain.avsi?

[tom942]

"Predator" is an action movie and it happens in a forest (sleeves moving, etc), then, as a general rule, it needs more bitrate (bandwidth).

The opposite is "2010", it is very calm, black background most of the time. What I remember from, there's no hardly action scenes, so it needs less bitrate. More or less, that's it.

[AGKnotUser]

"Predator" is an action movie and it happens in a forest (sleeves moving, etc), then, as a general rule, it needs more bitrate (bandwidth).

The opposite is "2010", it is very calm, black background most of the time. What I remember from, there's no hardly action scenes, so it needs less bitrate. More or less, that's it.

Now I see, thanks for the explanation.

[djskribbles]

one thing i don't like about HC is that in soft backgrounds or sometimes smokey scenes, theres noticeable blocks (almost like macroblocking). other than that, HC is, like most people say, at least as good as CCE. but to me, the blocks make me prefer CCE as i never go below 3000kbps anyway.

[Boulder]

If you encounter such a scene, it would be useful if you could upload an unprocessed sample so that it could be analyzed further. A smoky scene is a serious challenge to any encoder.

[djskribbles]

well, i've compared scenes with CCE and HC and where it would be blocky in the HC encoded version, it would be slightly noisy in the CCE version. one particular film where i noticed it was in Gladiator. the scene after he escapes from being executed and hes racing his horse through a cloudy storm.

[Boulder]

I think it might have something to do with the flat part priority setting in CCE. Would you mind trying adjusting that setting and doing some test encodes? Then again, when flat areas look very good, there is a chance that detailed areas and areas around sharp edges degrade. It's a kind of "choose your poison" thing.

I think Hank mentioned that he would like to add a similar setting to HC as well so maybe we'll see improvements in the near future.

[Digga]

I once did Casshern (very long and action laden movie) with CCE and compared it to a ProCoder 2 encode...

...I see if I can borrow his encode and post some screen shots and/or clips.I got both versions here now, if somebody could point me to an easy way of producing samples (for Win, GUI based preferred) I gladly post some. I guess they are more meaningful than screenshots.

after some casual viewing I hardly can make out a difference between both encodes though, let alone judge which one looks better to me.
maybe the different setup is to be held responsible, or maybe my failing memory ;-)

[Clawz]

If you take the DVD of Henry V (Kenneth Branagh) and run it through CCE, HC, and Procoder (both 1.5 and 2.0) you discover an important difference. With CCE and HC, the shadows end up with noise and artifacts that are extremely annoying. Procoder does just fine. Virtually all of the movie has prevalent shadows.

[jdobbs]

My own experience is that ProCoder seems to soften images. You could do the same thing in CCE or HC using an AVISYNTH filter or a matrix.

[Fishman0919]

I've done a few HD to DVD converts with Procoder 3 and Carbon Coder... and the detail is very good... as good if not better then CCE Basic. The bitrates were kinda high (5000 +) and that may be a reason why. Procoder seems to do well at high bitrate.... it's when it gets down to lower rates (for me 3500k and down) that you can see the softer picture more noticeably.

[blutach]

I think its a real Procoder feature that it shines at low BRs (<2500). IMHO, CCE has never been real good a lower BRs, HCEnc is better.

Regards

[jdobbs]

I think its a real Procoder feature that it shines at low BRs (<2500). IMHO, CCE has never been real good a lower BRs, HCEnc is better.

RegardsI agree. The point I was trying to make is that if softening is better than blocks on lower bitrates -- you can do the same thing with a softening filter or a matrix that accomplishes the same thing. I don't think it's the encoder that makes it better -- it's the filtering.

[Fishman0919]

Here are some screenshots from the new Harry Potter movie. I removed the black bars to save space.

Original DVD

http://img219.imageshack.us/my.php?image=orgnobarsbc3.png

Mencoder, DVD-RB Pro Movie Only

http://img225.imageshack.us/my.php?image=mencodernobarsvp8.png

Carbon down convert from HD

http://img215.imageshack.us/my.php?image=hddvdcarbonconvertnobarrv9.png

Edit: Fixed link

[miggs78]

Sorry to interrupt in this discussion. I was reading earlier that anything lower than 2500+ bitrate HC is a definite winner, anything above 2500+ CCE would do it..

So how do I determine the bitrate on my DVD, well to say DVD9. Thanks..

[jdobbs]

HC is very, very good. But so is CCE... I don't think I'd agree with any comparisons that say one or the other is better at any bitrate. There are too many variables involved, and bitrate alone means nothing. The source complexity, motion, matrix, and many other factors have to be considered.

[shon3i]

I am been using CCE for all my encodes along with DVDRB-PRO and every time when new HC encoder is out i run some test, and always i get some blocks in flat areas like H264. In CCE i use minimum 4passes and 16 Complex/Flat. I get little bluring picture but without blocks in flat areas.

[gizzin]

Alot of people say that, I notice it at higher bitrates also which really shouldn't be the case. I use a bias of 30, which helps alittle I guess. Maybe try lumgain and you'll get better results. What did you encode, What was the average bitrate... a little more info would help

[tom942]

Is CCE Basic specially tweaked for DVD-5 or is it the same than CCE SP?.

I mean that this last one, works better when there is room for DVD-9, doesn´t it?.

Regards.

[smok3]

doesn't cce use adaptive matrix on scene by scene basis? (or is it even gop by gop?)

[tom942]

@smok3

Gop by gop.

Regards.

[shon3i]

Alot of people say that, I notice it at higher bitrates also which really shouldn't be the case. I use a bias of 30, which helps alittle I guess. Maybe try lumgain and you'll get better results. What did you encode, What was the average bitrate... a little more info would help
I usualy reencode, movies which need to be compressed usualy by 5-20% max, if percent higher i buy DL disk :)

[Fishman0919]

doesn't cce use adaptive matrix on scene by scene basis? (or is it even gop by gop?)


Like tom942 said but only if needed

[shon3i]

Is there some trick or matrix to force CCE SP1/SP2 encoder to keep sharp, i always use CCEAQM=1 with DVDRebuilder but CCE picture is 1-2% softer than original.

[DMF]

Geez, how do you measure that small a difference?

[jdobbs]

Is there some trick or matrix to force CCE SP1/SP2 encoder to keep sharp, i always use CCEAQM=1 with DVDRebuilder but CCE picture is 1-2% softer than original.Interesting... because the exact opposite should be true. The CCEAQM setting would lower the matrix values when necessary... making the picture sharper.

[Boulder]

I think he means that it is softer compared to the original source, which usually is true. If he wants sharper output, he should use a sharpener (LimitedSharpenFaster or SeeSaw are two good ones to try out).

[shon3i]

Yep, i mean on that. HC keep original sharpnes, but tend to be little blockly in flat areas like i said before. And CCE smooth picture, but make block-free encoding.

@Boulder, did you have some opinion how to set LimitedSharpen, to force CCE to keep original sharpnes :) Some examples?

[Boulder]

You really need to figure it out yourself, your eyes are the ones you should trust. Experiment, and don't forget to use Soothe() with LSF.

[Video Dude]

I compared HC and CCE at a low bitrate (1900 avg) using DVD-RB.

HC produced a much cleaner encode for the most part compared to CCE. I was surprised how good it looked at such a low bitrate.

But I did find HC had problems with fade from black scene changes. The entire fade was extremely blocky and was enough to distract from the movie. LUMGAIN settings did not seem to help. CCE handled the fade well and blocks were not noticeable.

[rahzel]

I am been using CCE for all my encodes along with DVDRB-PRO and every time when new HC encoder is out i run some test, and always i get some blocks in flat areas like H264. In CCE i use minimum 4passes and 16 Complex/Flat. I get little bluring picture but without blocks in flat areas.
+1... same for me man. i've been using CCE for a long time and have been happy with my results. but because many people say they like HC encoder, specifically with lower bitrates, i keep running out to try the new versions but see the exact same thing you're seeing.

to be fair, i do find HC encoder is a little less noisy and is closer to the source in most cases. but i just can't get over the blocks in flat areas. i also find that these blocks are more noticeable on certain kind of displays. and yes, i've played around with the VBR Bias. interestingly enough, i find older versions of HC produced less blocks in flat areas (0.17 and older).

[jdobbs]

Have you tried adjusting the bias?

[gizzin]

I agree with this totally, I really do think there is problem with the flat areas encoded with HC. Even at higher bitrates which shouldn't be the case. I adjusted the bias and lumgain to no avail. There still blocky. Maybe the people that don't see this using a smaller sized tv?

[rahzel]

I agree with this totally, I really do think there is problem with the flat areas encoded with HC. Even at higher bitrates which shouldn't be the case. I adjusted the bias and lumgain to no avail. There still blocky. Maybe the people that don't see this using a smaller sized tv?
same, i adjusted the vbr bias too, but it even happens with higher bitrate movies. again, older versions of HC did not do this, but ever since 0.18, i have noticed blocks. if there weren't blocks on flat areas, i would have switched to HC a long time ago.

also, as i said earlier, the blocks are more visible on certain types of TV's, not necessarily the size. oddly enough, i actually find that its more noticeable on my 20" CRT direct view TV than my 40" LCD, or 46" rear projection CRT. direct view CRT displays seem to make the blocks more visible. if anyone was wondering, direct view CRT's are those older TV's, the ones we've been using for years. rear projection CRT are those bigger TV's that come with a base attached to the TV, usually with the speakers mounted in the base.

[gizzin]

Let me say this, If you're encoding anything the bitrate will depend on the source always take this into account (ie private ryan, band of brothers etc..). In my personal opinion if you can attain a bitrate of 4000kb you shouldn't be able to notice any difference from the original (I never had blocking at this bitrate). I have done alot of encoding with HC and let me honestly say it is better than CCE. It's clearly better at lower bitrates. And at interlaced encoding. I always use HC now, I never use anything else. I also think that the bias should be set at 30. I've read people complaining that the encodes look bad, and they don't just realize that the bitrate there using for the encode is just to low. I use this as a guide for myself when encoding.

2000kb and under <- why are u backing up this dvd, buy dual layers
2000-3000kb <- u should expect blocking, noise, grain, etc.
3000-4000kb <- this is a iffy region, for the most part you shouldn't have any artifacts
4000kb and up <- golden

[rendez2k]

With HC, do people use normal or best setting? Whats the impact on time to encode vs quality gain?

[linx05]

Always 'best'. I can wait the extra time it takes. If I were impatient I would be using a transcoder ;)

[rendez2k]

Always 'best'. I can wait the extra time it takes. If I were impatient I would be using a transcoder ;)

I've come to that conclusion myself last night - I did my first encode on my new quad core and it took 33mins for the encoding part with HC in best mode! Doesn't seam worth using anything else with times like that.