Which is a faster encoder TMPGE or CINEMA CRAFT ?

[energy80s]

If you want to use CCE you have to learn to use avisynth. You are lost without it. Because many users like me, are super idiots/stupids, lazy to the death, or simply too old to learn even the simpliest non GUI programing, this is a huge minus.

Couldn't agree more!!

[RabidDog]

Like the bit rate shots ... proof that 3rd pass does something..
BUT I still prefer tmpenc (and think its faster overall) simply because a) I only have to use one program to encode both audio and video
b) I dont have to learn avisynth scripting

ABove all I prize the simplicity and functionality of tmpgenc ... audio and video in one step.. not that revoutionary really! Also, it goes without saying, I find the tmpgenc output excellent.. I do mostly svcd 95%, never full dvd, sometimes vcd. I rarely if ever use any filters as they simply slow it down too much.
I am not disputing that cce encodes video only, faster, per pass, than tmpgenc.
Third edit It would have nice to see those graphs overlaid on top of each other to highlight the difference between 2 +3 (and maybe a fourth pass).

[telemike]

How about a comparison in speed between TMPGenc 1-pass CQ mode versus CCE 1-pass VBR?

[adam]

Just a suggestion for those of you who are not fans of scripting for Avisynth. Download the free program FitCD. You load your d2v and select the desired output format, ie: NTSC/PAL, DVD, SVCD etc... and it will export an avisynth script complete with some very nice optimizations. This really takes all the work out of using avisynth. If you are using a DVD source you probably don't need to run any other filters, so FitCD is all you need. Its basically a 1 click step to making the avisynth script.

Another thing to consider...DoItFast4u is an automated program for manually reauthoring DVDs. With 1 click, it will demux your set VTS sets and create avisynth scripts for each video asset.

DoItFast4u integrates with DoCCE4u, another automated program which is a frontend for CCE. If you wanted to downsample a DVD by just re-encoding the main movie in CCE, than you would just select this VTS set in DoItFast4u and hit start. After all is done, you will have your encoded file, never having had to even look at an AVS script.

For those of you who make VCDs and SVCDs, you can always just use CCE through DVD2SVCD.

Finally, there are a couple of different gui's for Avisynth which can be found on www.doom9.net. Unless you use alot of obscure filters, than this is pretty much all you need.

SatStorm: I agree with pretty much all that you said. I think CCE and TMPGenc are very similar in quality at low resolution and lower bitrates, and CCE gets better the higher you go. When you start encoding at full D1 at 4-5mbits, that's where CCE really shines.

[disturbed1]

@Adam

Sorry for the delay, thought I should actually try the settings you suggested before explaining why I use the ones I do. Of course it didn't do anything to speed up CCE.

First, 8-235 color level is the USA NTSC color standard, most PAL and Japan NTSC use a black level of 0 (0 IRE), where I live we use 7.5 IRE. Low to mid line DVD players do not convert the 0 black level correctly to be displayed on the TV. It results in overly contrasted videos . The higher priced DVD players use dedicated DSPs to broaden the color level to 0-255, which can still only be truely seen with componet outputs, not S-Video, nor composite.

The test I just did included a black level of both 0, and 8. No visual difference on my TV, with my Philips Proscreen Projector the overall image was darker with a black level of 0 compared to one of 8. The overall quantization was also higher with a black level of 0, because CCE had to give bits to the extra colors.

Zigzag is always recomended for progressive. An alternate scan technique is an advanced entropy encoding method I use with custom matrices.

zig-zag scan order:

0   1   5   6  14  15  27  28

2   4   7  13  16  26  29  42

3   8  12  17  25  30  41  43

9  11  18  24  31  40  44  53

10  19  23  32  39  45  52  54

20  22  33  38  46  51  55  60

21  34  37  47  50  56  59  61

35  36  48  49  57  58  62  63

alternate scan order:

0   4   6  20  22  36  38  52

1   5   7  21  23  37  39  53

2   8  19  24  34  40  50  54

3   9  18  25  35  41  51  55

10  17  26  30  42  46  56  60

11  16  27  31  43  47  57  61

12  15  28  32  44  48  58  62

13  14  29  33  45  49  59  63

Although altenate scanning is recomended for interlaced material, it can be used with progressive. It depends on the film your encoding, matrix coefficents, and DC precision to which one gives the better results. I use zig-zag most of the time for progressive, that time, alternate was the better choice.

CCE gave a DC precision of 10 when set to auto. I always set to 10 when the bitrate is 4500 +, 9 below 3500 and auto when using bitrates between 3500 and 4500.

Image quality priority is also dependant on the film your encoding. 27 is way too high for anything, as this will give mosqutio noise, and slight blocking is scene changes. The sweet spot is between 7-21, above that the varible bitrate is much more flatter, given too many bits to stale parts, while starving high action sections. This can be corrected using advanced mode, but if you set the image priority right in the first place, theres no need to examine every GOP. Image priority also works hand in hand with the variable rate bias.

A better analogy is:
TMPG can finish the race in two steps
CCE takes 4 to 5 steps to complete the same race.

It's common knowlege that CCE's best output is 3 passes (sometimes 4), TMPG's best ouptut is 2 pass, although great results can be achieved with one pass. When comparing you should compare best to best.

Using avisynth can achieve better output from CCE, your tailoring the script directly for CCE's output. The same can be gone with TMPG for $1950 less and an hour or two longer.

[adam]

Low to mid line DVD players do not convert the 0 black level correctly to be displayed on the TV. It results in overly contrasted videos . The higher priced DVD players use dedicated DSPs to broaden the color level to 0-255, which can still only be truely seen with componet outputs, not S-Video, nor composite.

I think you misunderstood. Camera's capture colors which tv's simply don't support, like you describe. That is why luminence ranges are compressed to the 16-235 range, so those colors don't just get clipped. Just about any source you deal with, with the exception of your own non-DV captures, is going to already have these levels compressed. By setting CCE to 0-255 you are simply preventing the further compression of your luminence scale. This does NOT create those "blacker than black" and "whiter than white" colors that you get from cameras, those are already gone.

Unless your source is in the 0-255 range than there is no reason to further compress the luma. You are decreasing the dynamic range of your colors. Regardless of what other effects this may have on quantization or compressibily, there is no good reason to further limit the range of your colors unless you have to. I would think the goal of encoding would be to best recreate your source. Why would you want to compress the colors to a smaller range than the source? Assuming the source is encoded in the 16-235 level, which will most often be the case, (DVD, captures, DV) you should just keep it that way by using a setting of 0-255.

Zigzag is always recomended for progressive. An alternate scan technique is an advanced entropy encoding method I use with custom matrices.

Ok. I was just giving the general rule, and like I said, with a modified Q Matrix the general rule goes out the window.

CCE gave a DC precision of 10 when set to auto. I always set to 10 when the bitrate is 4500 +, 9 below 3500 and auto when using bitrates between 3500 and 4500.

When you set DC to Auto is varies the DC precision according to the quantization as it encodes. It does not just select the most appropriate DC level and apply it to your movie. I believe that if you set this to auto the stream will always be tagged with a DC of 10, even though this level will only be applied during the highest of bitrate peaks. If you read the CCE manual, as well as alot of other sources, 8 should be used for anything under 4mbits and 10 should only be used for very high bitrates. 4.5mbits at a DC of 10 should actually produce worse results than 4.5mbits at DC of 9. Try analyzing some commercial DVDs in a bitrate viewer and you will be very hard pressed to find any that use a DC of 10. Most are 8 or 9.

Image quality priority is also dependant on the film your encoding. 27 is way too high for anything, as this will give mosqutio noise, and slight blocking is scene changes. The sweet spot is between 7-21, above that the varible bitrate is much more flatter, given too many bits to stale parts.

While I agree that image priority must be adjusted according to your source my tests have shown completely the opposite results as yours. When I encoded SVCDs I always fluctuated between the 7-23 range, and if you read around this is the general opinion of others as well. But with higher bitrates you do not need such a low range, and this will simply increase contour noise. 27 is most certainly not too high for anything. CCE's default is 25, and the manual recommends that this be increased for higher bitrates. Once you hit the 4-5mbit range its going to be very difficult to distinguish any differences resulting from your image quality priority settings, provided you don't use an extreme setting. However, to my eyes a setting around 25 provides better quality at the 4-5mbit range. I personally think a setting of 7 is way too low for DVD encoding.

A better analogy is:
TMPG can finish the race in two steps
CCE takes 4 to 5 steps to complete the same race.

Well, only in your opinion. As I said, and many others would agree, I prefer CCE's quality at 1 pass to TMPGenc's quality at 1 pass. Just because CCE gives the option of more than 2 passes doesn't mean you are forced to use it, and I don't see how this can be a handicap on CCE's speed attributes. CCE encodes X amount of frames in Y amount of time. TMPGenc encodes X amount of frames in roughly Y x 2 amount of time. That's really all there is to it.

It should also be noted that, even when comparing CCE's 4 passes to TMPGEnc's 2 passes, CCE still finishes about 4 hours faster. In my opinion that is no small margin.

[jaxxboss]

Absolutely amazing the effort and time devoted to all these encodes and sync prob and burning and etc.. Mind staggering!
Buy a cheap 8mb All in wonder card and play ur avi on ur computer and watch it on ur TV! Run a cheap RCA cable to the tv from the computer OR use the new wireless tech stuff.
We are all going to laugh 10 years from now when the new player will play an AVI disc, we will have like 5 -7 movies on a dvd disc playing in the player. remember : load(*),8,1 ?
LMAO!

[adam]

What do sync and burn problems and avi's have to do with this thread?

[flaninacupboard]

it's interesting the differing opinions on the cce image slider. personally, contouring or banding is the worst possible artifact you can get on a DVD. i hate it. i always have my slider at at -least- 25. i put it up to 35 when i encode anything longer than 90 minutes onto one disc. yes this means high motion scenes may suffer, but, give that high motion becomes blurred on film, and the effect of blocks from the other side of the room is blurring, this is no problem. far worse is a big wobbly band thing, cutting right from one side of the screen to the other! my UK disc of "Leon" is a single layer affair (still waiting for the special edition to come out over here....) and the opening sequence (panning across the ocean) is dominated by big wobbly bands. it's damn ugly. and this isn't even an anamorphic transfer! heaven help someone trying to squeeze something like lord of the rings on one disc, anamorphically presented, with the image slider set to 7. i couldn't watch that.

RabidDog,
yeah, i wanted to do overlaid shots, but didn't really have the time to line it all up nicely. next time i do a proper encode of a film i'll try and keep the 4 passes to show the same clip.

Jaxxboss,
this has -nothing- to do with AVI's. at least not one i could fit 6 of on a DVDR. this is about the best way to convert captures from old material (LD's, home movies, old VHS you can;t get on DVD etc) or from DV, or from DVD. if you don't have anything useful or, in fact, relevant to say, then keep quiet or go back to the newbie section.
In any case, wireless technology is hardly new. i have a television transmitter from the 1980's sat in my house, and i would never consider watching -anything- through an RCA cable, and why would i want to use my computer as a playback device when i have £300 worth of DVD player? that's just illogical.