battle of the CBR/VBR/CQ modes

[therick]

wow this was quite a post! I can't honestly say that I read all of it-but I do have a question for those who have been following this post and who obviously still use TMPEGEnc for MPEG-2 (I haven't for almost a year and I have never looked back). Do the new versions of TMPEGEnc have as shitty a 2-pass VBR as the old (2.0 I guess) versions? When I first started using CCE I did some tests of the two encoders' 2-pass VBR at the same bitrates and it wasn't even close! TMPEGEnc was blocky as hell! And I remember reading on this forum that it's 2-pass sucked cause it hardly went above the average bitrate. Has this changed?

A side note-I have been fitting movies on two cds for a year-not only DVD rips, but VHS encodes...with CCE...in fact I just did an Out of Sight DVD rip with an avg. of 1646kbps...and it looks great-

[Martyn1980]

I just finished doing a 82minute 2 pass VBR on Tmpeg Plus using the new 2 pass it has, it took 4 hours on High Quality, and didn't look too great, i did aim for a 800mb filesize, but Kwags new 352x480CQ template was 830 and looked much better, so i have to agree with the CQ people if all your using is Tmpeg,then CQ is betterbut the only problem is predicting filesize, i cant overburn 30mb and its kinda pointless having 415 on each disc.

[kwag]

Martyn1980, 830MB for 84 minutes is too much for the time
You should have been able to put at least 100 minutes on that CD. What did you encode? A DivX or a Full Screen movie?

-kwag

[Martyn1980]

Both, kinda i did do a The Order on DVD with the template, but i changed some settings so i expected that.

But the last one was a full screen Divx, and a very blocky one at that, actually it was MS mpeg4, a new film just out that the kids wanted to see.

I ended up redoing it with the new 2 pass VBR in Tmpeg plus but it looked nowhere near as good, the one with your template looked as good as the original, if i can say it looked good.

I should of realised it would take more for a full screen movie, but i was in a rush, thanks for telling me.

[kwag]

Subject: A detailed analysis of TMPEG vs. CinemaCraft.


Well, because i got fed up by reading incorrect posts, I decided to do a detailed analysis and compare both encoders by using the best modes of each.
It's clear that CCE is a better MPEG-2 encoder ( that's what they say ) than TMPGEnc, and that TMPGEnc is a better MPEG-1 encoder than CCE.
That is a fact, and everyone agrees on that. But which encoder produces the best quality over the other, may it be MPEG-1 or MPEG-2? Well, this is what this is all about.
Before I start, I want to acknowledge that above 3-4Mbps, MPEG-2 will always be better than MPEG-1. No matter what encoder is used, because MPEG-2 scales, and MPEG-2 doesn't. But what really started to bother me was that some people say that CCE's X-pass is better than TMPEG's CQ modes. So here are the real life results.
I encoded "The Matrix" ( video stream only ) with the new KVCD beta 704x480 template, now using our own Quantization Matrix and GOP. The new template encodes CQ_VBR mode. The total size for the movie's video stream came out to 1,302,091KB. Then encoded the same movie, using
DVD2SVCD, with the default CCE parameters for MPEG-2, 3-pass mode ( 3 plus 1 for .vaf ), and a quality setting of 17 ( the latest recommended value ) and the noise filter "unchecked".
I had calculated the average bit rate to be 1,368Kbps, to create a file of the same size as the one created by TMPEG. It came out very close, at 1,362,127. At 60MB difference, a 4.4% advantage for CCE, but good enough for the test. A 4.4% in file size is not visible in quality. I send these parameters to a friend, who has a licensed version of CCE, so that he would encode the movie, because I don't have CCE. Not even a demo ( and I don't want it anyway ). The samples I posted are muxed with 224Kbps audio, and 30 seconds each. It's exactly the same section of both encodes. The CCE came out way smaller in size at 6,295KB, versus 7,526KB for TMPEG, which proves again my statement that CCE's X-pass mode cares about average size as specified, and doesn't care about quality. I'll explain that last statement in more detail later. Here are the two graphs made by bit rate viewer, where you can see the difference in quality, produced by TMPEG.

Here's CCE's sample graph:

Here is the clip:
http://ns1.shidima.com/kwag/cce-matrix-704x480-mpeg-2.mpg


Here's TMPEG's graph:

Here's the clip:
http://ns1.shidima.com/kwag/tmpgenc-matrix-704x480-mpeg-1.mpg


A detailed analisys of the graphs:

Take a very close look at both graphs, and note the "Peak" bit rate. In CCE, it's 1,652Kbps. In TMPEG, it's 2,254Kbs. What happened to the "INTELLIGENT" bit rate allocation of CCE. The MIN bit rate was set to 300Kbps and the MAX was set to 2,300Kbps in the encoder! It doesn't work. CCE was busy with it's muli-pass encoding, trying to balance the total bit rate assigned by the average value and not go overboard the final size. FACT: CCE's X-pass VBR mode cares about a final size. There's nothing here about quality. It will create the best possible quality "within" the constraints of the min/avg/max bit rate values given.

Another thing to look at, very closely, is the average bit rate for this clip. In TMPEG, it's 1,630Kbps. In CCE, it's 1,375Kbps. Again, this value is lower in CCE, because the encoder is worried about bit rate allocation. Even that this scene needs a lot of bit rate, because it's an action scene, CCE didn't care about increasing the bit rate.

Let's look at the quality, the Quantization values:
The average Q factor for TMPEG in this clip is 1.79, with a peak value of 3.17.
The values for CCE are 2.56, with a peak value of 3.27.
Again TMPEG's quality factor was almost twice better that CCE's value. 1.79 is way better than 2.56. It's clear in the graph.

Now if you look at both mpeg files with WinDVD, you'll see that CCE's mpeg is darker, an advantage over TMPEG for pointing out "mosquito effects" and artifacts. I believe the luminance level was set at 0-255 instead of 16-235 when it was encoded. Again, an advantage for CCE, because artifacts are damped with this color space.
But in reality, and to my surprise, CCE's mpeg file has more artifacts and "mosquito effect" than TMPEG!
If you step through the cce mpeg file, looking at the still screen shots of the clip ( by pressing "N" in WinDVD ) , you'll see the grainy (dirty?) mpeg result of CCE, specially on the fast panning scenes of the clip. I know this is not noticeable in a TV, barely on a HDTV, if viewed normally without pausing the movie. But my point is very clear.

Conclusion:

CCE is probably the best MPEG-2 software encoder.
TMPGEnc is probably the best MPEG-1 software encoder.

Which encoder produces the best quality, visually, mathematically, subjectively, every way you want to look at it, below 3Mbps?
TMPGEnc
Which mode produces the best quality?
CQ mode, at least with the new KVCD 704x480 CQ_VBR, KVCD's Q.Matrix and GOP

Please remember that CCE was used with the default Q. Matrix included with the program, and that the GOP can't be extended, as it is in TMPGEnc. That could be a factor for the large difference in quality.
As for the time to encode, TMPEG's total time was around 6 hours in a P4 @1.6Ghz using CQ-VBR mode. CCE's encoding time was almost 16 hours!, also on a P4 @1.6Ghz.

Here's a cut of DVD2SVCD log's file:

--------------------------------------------------------
- 7/11/2002 4:51:16 PM
- Free on drive C: 12224.73 mb
- Video Encoding using Cinema Craft
--------------------------------------------------------
- Editing AVS script file

Executing Frameserver.
Closing program

Executing Cinema Craft Encoder.
StreamSectors: 771186864
AudioSectors: 132753852
VideoPAPO: 9889668
ScanOffsetBytes: 235386
SeqAligningBytes: 14960088
DVDBytes: 0
VideoEndHeader: 4
SubtitleSectors: 0
EmptySectors: 238.00
PictureSectors: 1.00
PureMPEGStream: 613347866.00
Seconds: 8186.31
CalcMPEGStream: 613347866.00
Frames: 196155
CDSize: 740.00
Cut point 735.00
Variable Settings:
Frames: 196155
Anti Noise Filter: Off
Passes: 3
Image Quality: 17
VAF file creation: On
Video Encoding Mode: Multipass VBR
Min. bitrate: 300
Max. bitrate: 2300
Avg. Bitrate: 1368
Aspect Ratio: 4:3 (No borders, encoded as 4:3)

---AVS Begin---
LoadPlugin("C:\post-prod\tools\MPEG2DEC.dll")
mpeg2source("C:\THE_MATRIX_16X9LB_N_AMERICA\VIDEO_ TS\matrix-dvd2avi-176.d2v")
BilinearResize(704,352,0,0,720,480)
#TemporalSmoother(2,2)
AddBorders(0,64,0,64)
#Trim(0,196155).FadeOut(150)
#ResampleAudio(44100) # CCE 2.5 'crashfix' for Athlons
#== If you want this 'fix' permanently, edit the INI-file:
#== Under [AVSscript] set the CCEcrashfix-option to
#== CCEcrashfix=1
----AVS End----

Closing program
CCE Max Speed: 0.590
Video Encoding finished.

--------------------------------------------------------
- 7/12/2002 8:46:26 AM
- Free on drive C: 10804.96 mb


I haven't tried the new 2-pass of TMPGEnc plus against CCE's 3-pass mode. But the next run I do, will be TMPGEnc in MPEG-2 mode using our own Q. Matrix with CQ_VBR against this last run of CCE. That's my project for later today ( if I can ). So we'll see the results. As fas as quality is concerned, at 704x480, the KVCD MPEG-1 beta template at 704x480 is superior to CCE's MPEG-2 3-pass mode. Above 3Mbps, CCE WILL be superior, because MPEG-1's quality curve will not increase above that bit rate. MPEG-2 will, and using CCE will be a clear advantage.


-kwag

[Martyn1980]

Although i am a fan of your new template, it still doesn't match what im presently getting with CCE, luckily my DVD player will play very high bitrates and i usually use the settings of 300min and 4000max, with the average changing per movie, and i get amazing results using this, and would never go back to Tmpeg unless im creating VCD.

Ive got hold of a copy of Proof Of Life of a friend and i have it ripped on my hard drive, and i will test that movie using your new 704x480 template, although i will change the resolution to 576 any objections and do i need to change anything else apart from the framerate, and then with CCE using my own settings, as the final object is to get the best quality on 2 discs, not really the final filesize, although i will do a few other tests too

[Martyn1980]

I just watched those samples and i dont think theres any point in me doing tests as i already know whats going to happen, i dont like the image Tmpeg produces, never have, dont really know what it is, but theres something wrong, and the color i dont like either, I much prefer the CCE sample and although its bit noisier, ive always said Tmpeg was smoother but at the cost of detail, CCE has always produced an image similar to the original but just not as smooth, Tmpeg produces a totally different picture in my opinion, too bright, too smooth, faded colors and a kinda mist across the screen, though everyone has to remember that this is my own opinion and everyone has different tastes, the average bitrate for the Tnpeg one was reported as 1800 by VirtualDUb, i think i could get better results by using CCE 4 pass with a 200min 1800average 4500max, i do have The Matrix on DVD but it is elsewhere right now, but i will get it back soon.

I do really like your new template Kwag and was impressed with the results, and if i get better results i will switch back to Tmpeg, but this is unlikely as i dont like the Tmpeg image and my DVD doesn't support High Res Mpeg1, but supports High Res, High Bitrate Mpeg2.

[Martyn1980]

Not wanting to go on here, but i was just looking through the DVD2SVCD log you posted and it dont look like a DVD2SVCD, the AVS script looks like its from Fit CD, as it mentions the CCE, Athlon, resample audio fix, it also has the Temporal Smoother set to 2,2 which is what Fit CD defaults too, i dont think you need a Temporal Smoother on DVD sources and it adds a lot of time to the encoding time, i was wondering how it takes 16hours, i can do a 2 hours movie in roughly 8 hours on 3 pass VBR, plus Bilinear produces better quality then Bilinear and the time difference to encode isn't much.

Am i right, was that script created in Fit CD, an example being
This for the bandits trailer

---AVS Begin---
LoadPlugin("C:\PROGRA~1\DVD2SVCD\MPEG2DEC\MPEG2DEC .DLL")
mpeg2source("C:\MYDOCU~1\MYMOVI~1\BANDITS~1\ENCODE ~1\DVD2AV~1.D2V")
BicubicResize(704,432,0,1)
AddBorders(0,72,0,72)
ResampleAudio(44100)
----AVS End----

Thats DVD2SVCD

and this is what i get in Fit CD

LoadPlugin("C:\Program Files\DVD2SVCD\MPEG2Dec\MPEG2DEC.dll")
mpeg2source("C:\My Documents\My Movies\Bandits\Encoded Mpeg\DVD2AVI_PROJECT_FILE.d2v")
BicubicResize(704,446,0,0.6,19,0,682,576)
#TemporalSmoother(2,2)
AddBorders(0,65,0,65)
#Trim(0,3500).FadeOut(150)
#ResampleAudio(44100) # CCE 2.5 'crashfix' for Athlons
#== If you want this 'fix' permanently, edit the INI-file:
#== Under [AVSscript] set the CCEcrashfix-option to
#== CCEcrashfix=1

And does Virtual DUb report an incorrect average bitrate cause with the Tmpeg clip it says the Averae bitrate is 1800kbps but in that script file the average for CCE is 1346kbps, am i getting this all wrong here.

[kwag]

Ive got hold of a copy of Proof Of Life of a friend and i have it ripped on my hard drive, and i will test that movie using your new 704x480 template, although i will change the resolution to 576 any objections and do i need to change anything else apart from the framerate, and then with CCE using my own settings, as the final object is to get the best quality on 2 discs, not really the final filesize, although i will do a few other tests too

Why change anything!. The PAL templates are posted too!
http://www.kvcd.net/dvd-models.html

And go ahead and tweak them if you want, they're there as a base model. But everyone makes their own adjustments

-kwag

[kwag]

Not wanting to go on here, but i was just looking through the DVD2SVCD log you posted and it dont look like a DVD2SVCD, the AVS script looks like its from Fit CD, as it mentions the CCE, Athlon, resample audio fix, it also has the Temporal Smoother set to 2,2 which is what Fit CD defaults too, i dont think you need a Temporal Smoother on DVD sources and it adds a lot of time to the encoding time, i was wondering how it takes 16hours, i can do a 2 hours movie in roughly 8 hours on 3 pass VBR, plus Bilinear produces better quality then Bilinear and the time difference to encode isn't much.

Am i right, was that script created in Fit CD, an example being
This for the bandits trailer

---AVS Begin---
LoadPlugin("C:\PROGRA~1\DVD2SVCD\MPEG2DEC\MPEG2DEC .DLL")
mpeg2source("C:\MYDOCU~1\MYMOVI~1\BANDITS~1\ENCODE ~1\DVD2AV~1.D2V")
BicubicResize(704,432,0,1)
AddBorders(0,72,0,72)
ResampleAudio(44100)
----AVS End----

Thats DVD2SVCD

and this is what i get in Fit CD

LoadPlugin("C:\Program Files\DVD2SVCD\MPEG2Dec\MPEG2DEC.dll")
mpeg2source("C:\My Documents\My Movies\Bandits\Encoded Mpeg\DVD2AVI_PROJECT_FILE.d2v")
BicubicResize(704,446,0,0.6,19,0,682,576)
#TemporalSmoother(2,2)
AddBorders(0,65,0,65)
#Trim(0,3500).FadeOut(150)
#ResampleAudio(44100) # CCE 2.5 'crashfix' for Athlons
#== If you want this 'fix' permanently, edit the INI-file:
#== Under [AVSscript] set the CCEcrashfix-option to
#== CCEcrashfix=1

And does Virtual DUb report an incorrect average bitrate cause with the Tmpeg clip it says the Averae bitrate is 1800kbps but in that script file the average for CCE is 1346kbps, am i getting this all wrong here.

Yes, that is the script I used for the 704x480 MPEG-1 beta template. The same script I sent my friend, and that was pasted when DVD2SVCD pops-up the "Edit" AviSynth script window. We used exactly the same script for both encodes on TMPEG and CCE. But I didn't use DVD2SVCD for my 704x480 MPEG-1 encode. I read the .avs directly with TMPEG.
Temporal smoother is disabled on the script. Look at the # in the beginning of the line. Billinear was used!, What script are you reading
Look at the script I posted again!.

-kwag

[therick]

uuummmmm....how is that a valid test if you compare TMPEGEnc's CQ_VBR and CCE's X-pass VBR? I was talking about TMPEGEnc's 2-pass vs. CCE's 2-pass-and as far as CQ_VBR goes...I like to do my encodes once-and I like to fit movies on 2 CDs-that is why I use CCE's multipass on not only DVD sources, but also VHS, LD, satellite, camcorder..anything I need to encode that I want to take to DVD later.

[injunpana]

how is that a valid test if you compare TMPEGEnc's CQ_VBR and CCE's X-pass VBR? I was talking about TMPEGEnc's 2-pass vs. CCE's 2-pass

lol! Read the title and the endless pages of this thread, it's all about TmpGenc's CQ modes versus CCE's x-pass VBR, and he added his post to that. He was not necessarily responding to your post. I doubt if he was even aware of it.

[kwag]

@therick

About CCE's 3-pass vs. TMPEG Plus 2-pass?
I don't know!. I'll tell you tomorrow , as I started encoding 4 hours ago with TMPEG's 2-pass mode. It's 37% done, 6 hours 47 minutes to go.
But as the title of this thread says "battle of the CBR/VBR/CQ modes", it's a useless statement. CQ modes are clearly superior. At least with TMPEG's CQ mode over CCE's (UN)Intelligent X-pass bit rate allocation as proved above. You want to fit a movie in 2 CD's?, well, go ahead and use the 704x480 beta template. That's what it's for.
Movies already made at the quality of the sample above and fitted in 2 CD's with room to spare:
"The Matrix" 136 minutes.
"Proof of Life" 135 minutes.
More comments here: http://kvcd.net/forum/viewtopic.php?t=615&sid=b290688c2f09adbe22552b4ea829c85f


-kwag

[therick]

thanks kwag-I have some of your templates and I will try them out-

[vhelp]

late night all.


I just thought I'd throw in my own 2 cents worth, cause I finally
got CCE to work on MY pc!!

In the last week, I've drilled CCE to nowhere's end vs. TMPG.
I've kept using the same parameters for BOTH, and sing the same
avi clip(s) for both, cause I wanted to be as fare as possible.
As far as I'm concirned, and based on my knowledge of both CCE
and TMPG, I've used as equal parameters/settings as I knowingly
could.

But, to be short here, so far, I am still sticking with TMPG,
cause I still feel the quality difference vs. CCE's, using the
same parameters, are better. Yes, even on my 13" TV.
One of my theories (cce vs. tmpg) on bitrate algorithems was
expressed in better words by Kwag. Amazing how we are sometimes
in tune on some issues. ...that their is a cut-off in bitrate
to maintain space per average, in cce. Well, something like
that. This sort of explaines why, on some clips I do in both
cce vs. tmpg, that in one instance, cce shows improvment over
tmpg, and while in other test encodes, at different bitrate
combo settings, tmpg shows better. I have'nt quite nailed it
for shore, but it must have someting to do w/ that "sweet-spot"
2k or something near by. Anyways...

I'm still testing CCE (as we speak) and have a few theories of
my own, and I must say, that a lot of what Kwag had posted above
are very close, if not right on, at least w/ my theories go.

Anyways, below are my parameters for both CCE and TMPG.
Bare in mind, that, yes, I'm still a newbie when it comes to CCE,
but I'm pretty experienced w/the encoding, and my eyes tell the
trueth, well, at least to me
Ok, here's my simple settings, that led me to consistantly stay
with TMPG:

CCE:
----------
* SOURCE: Satalite Capture (same AVI clip used for test encodes)
* MPEG-2
* 3 pass (sometimes 4 passes)
* VBR: 1850/600/2520

TMPG:
----------
* SOURCE: Satalite Capture (same AVI clip used for test encodes)
* MPEG-2
* 2 pass
* VBR: 1850/600/2520

my conclusions based on above, is TMPG still shows better quality.
If I can find a good pair of sample clips to upload for you all
to view and see difference in quality, I'll post them, unless,
of course, I find CCE better than TMPG, he, he...

Again, I'm still testing with CCE, and I mean test! I don't do
quick 1 minute per day tests, I do hours, and hours - - sometimes
till 4 a.m. Well, it's almost 2 a.m. here.

My other 3rd cents worth is this. If we could just stick to some
sort of standard. ie, use the same:
* source
* movie
* scene
* MODE ie, VBR / CQ / xPASS, etc.
* MPEG-1 or MPEG-2, etc.
* etc.
for starters, we would have better judging going on
as accorately as can be. But, it seems like people keep throwing
out different movies they encoded, and hence the argument contineus
If we could just agree on a standard source, etc. i laid out above.
We could make better arguments So, what'a-ya think?

well, that's enough for now.
-vhelp

[adam]

I'm not going to get into another debate but Kwag that is the absolute most unscientific analysis I have ever heard and I don't even understand what it is trying to prove.

If you want to compare two things you need to eliminate all the variables except the ones you are trying to compare, in this case the encoder. I'm having a hard time even finding a single constant in your comparison, you have nothing but variables.

Let me get this straight you encoded the whole movie than cut out a single clip from each at a random place in the movie? If the clip only needs an avg of 1375kbits and a max of 1652 and TMPGenc used more than that then that makes TMPGenc the better encoder? Isn't it logical to think that maybe that extra bitrate that cce didn't use was allocated somewhere else where it was needed more? Isn't that the whole point of multipass vbr? If you want to test two samples then you test the whole encode and you look at the avg Q of one compared to the other, not the Q taken from a random point and only at that point. Your only reporting like %1 of your findings, what are we supposed to do with that?

Your two comparison samples arent even consistent. They have a different number of frames and a different playtime.

In one test you use mpeg1 and in the other you use mpeg2.

Your tests were run on two different computers.

CCE obviously wasn't configured correctly. A P4 @1.6Ghz does not encode at .590 with only light noise reduction filters running. It should have been running at least 3 times faster than that.

In one you modify the gop structure and in the other you don't.

In one you set all settings manually, in the other you allow a third party program to do it.

You used different avg bitrate settings in each encode. Newsflash...if you want to make a comparable file as the CQ file than you ANALYZE the bitrate of your current file, not calculate the bitrate to match the size. You obviously didnt calculate the avg correctly because cce respects min, max, and avg bitrates much moreso than TMPGenc does.

And here's the real kicker, you didn't even encode one of the files yourself. Even if your avg bitrate was calculated correctly how do you even know it was set correctly?

About the only constant that I can see in your comparison is the resolution you used, and its not even a standard resolution.

[kwag]

I'm not going to get into another debate but Kwag that is the absolute most unscientific analysis I have ever heard and I don't even understand what it is trying to prove.

You obviously can't read, or you wan't to "twist" reality, and don't want to accept facts.. I'm prooving that TMPEG is a better "overall" encoder than CCE.

If you want to compare two things you need to eliminate all the variables except the ones you are trying to compare, in this case the encoder. I'm having a hard time even finding a single constant in your comparison, you have nothing but variables.

Yes. You're right. I'm trying to eliminate CCE which you and many "dream" that it's the best encoder.

Let me get this straight you encoded the whole movie than cut out a single clip from each at a random place in the movie?

You're wrong again!. I muxed the same clips from both encodes. If there is a couple of frames in difference, it was BBmpeg when I muxed the new files.

If the clip only needs an avg of 1375kbits and a max of 1652 and TMPGenc used more than that then that makes TMPGenc the better encoder?

Of course it does!. Come on man, it's simple math. That clip is a very active clip. With lots of action. That clip NEEDS a high bit rate. But CCE just ignores it, because X-pass VBR is designed to spread the bit rate under the constrains that you specify. In this case the average bit rate. How hard is it for you to understand that Adam?

Isn't it logical to think that maybe that extra bitrate that cce didn't use was allocated somewhere else where it was needed more?

Chee Wez, I cant believe what I'm reading
This is EXACTLY a point where a high bit rate is needed!. TMPEG did use it, and CCE with it's "INTELLIGENT" bit rate allocation didnt? WAKE UP!

Isn't that the whole point of multipass vbr? If you want to test two samples then you test the whole encode and you look at the avg Q of one compared to the other, not the Q taken from a random point and only at that point. Your only reporting like %1 of your findings, what are we supposed to do with that?

BOLOGNA! The average CQ of the complete movie is lower in TMPEG. I just showed the average CQ for the clip. Here are the numbers for the complete movie, when processed with bit rate viewer.
CCE: Average bit rate 1,277. Average CQ 2.16
TMPEG: Average bit rate 1,221. Average CQ 1.81
Look at that! TMPEG has lower average bit rate, and looks better! Look at the Q, it's way lower than CCE with a lower bit rate. Who's the better encoder?
If you encode a complete movie, as I did, and you compare the same clips of both encodes, again as I did, the CQ sampled "per-part" IS valid ANYWHERE on the film compared to the SAME "per-part" of the other encode..

Your two comparison samples arent even consistent. They have a different number of frames and a different playtime.

Yes. As I said, probably because the start/end seconds is not 100% accurate when I muxed with BBmpeg. Actually the difference is like 2 or 3 frames. Do you really think that 2 or 3 frames is going to make that HUGE difference against CCE. Give me a break!

In one test you use mpeg1 and in the other you use mpeg2.
Your tests were run on two different computers.

It doesn't matter what computer you're running. If you're running the same encoder on two different CPU's, the result WILL BE THE SAME every time, if you use the same encoding parameters. I've tested this before. It never fails.

CCE obviously wasn't configured correctly. A P4 @1.6Ghz does not encode at .590 with only light noise reduction filters running. It should have been running at least 3 times faster than that.

Then you go complain to DVD2SVCD at doom9. Because DVD2SVCD was used for the encoding, but with an AviSynth script generated with FitCD. The same script was used for TMPEG and CCE.

In one you modify the gop structure and in the other you don't.

Of course. I'm making the point that the new KVCD with it's own Q. Matrix and GOP blows CCE, SVCD, (XYZ)VCD's etc., away.

In one you set all settings manually, in the other you allow a third party program to do it.

But the result is the same. Cant you see that we used the same AviSynth script?

You used different avg bitrate settings in each encode. Newsflash...if you want to make a comparable file as the CQ file than you ANALYZE the bitrate of your current file, not calculate the bitrate to match the size. You obviously didnt calculate the avg correctly because cce respects min, max, and avg bitrates much moreso than TMPGenc does.

Really!. I said that I used the average bit rate of 1,368Kbps for CCE. For TMPEG, there's no calculated average bit rate, because it's encoding in CQ mode. I used that average to try to zero in on the SIZE that TMPEG created, because it encoded in CQ mode. So the target, instead of being a CD, as it usually is calculated, was a DESTINATION SIZE. Wake up Adam, you're not analyzing the comparison very well, are you?

And here's the real kicker, you didn't even encode one of the files yourself. Even if your avg bitrate was calculated correctly how do you even know it was set correctly?

Because I was there when the encoding started. I did the set up. :P

About the only constant that I can see in your comparison is the resolution you used, and its not even a standard resolution.

It's not even a standard resolution?
704x480 IS a valid MPEG-1 "still" resolution. That's why it was chosen. Because many DVD players don't play 720x480 MPEG-1 video, but they DO play 704x480 video. And the list is growing.

I'm sorry Adam. I've provided samples and graphs, and you just don't want to accept facts.
Maybe you can provide TRUE samples, and not waka waka talk, as you always have, without any consistent proof of concepts as I have.
It's really sad, when you have something RED in front of someone, but that person INSISTS that it's BLUE. Just makes me wonder......

Have a nice night, and enjoy your lower quality video and resolutions.
The KVCD 704x480 supercedes every XVCD, XSVCD for the time being.
( And you ain't seen nothing yet 8) )

-kwag

[Martyn1980]

Kwag, sorry missed the Pal templates, also missed that that Temporal Smoother was disabled, didn't know that a #before it meant it ws disabled, i usually just remove the whole line, but that script is the exact same script Fit CD produces and if Temporal Smoother is removed then how do u explain the encoding times, because my processor runs at 1533mhz and would have done that in about 8 hours, and that with Bicubic Resizing, which is what i mean to write, not bilinear, which i wrote twice so you should of guessed i made a mistake.

Vhelp i dont know how you get Tmpeg better on its 2 pass VBR its crap on that, CQ is where i think it might come close but i dont like the 2 pass not even the new one on the Tmpeg plus version.

Am i the only one in thinking that doesn't look like a DVD2SVCD script and looks exactly like a Fit CD one.

[kwag]

Kwag, sorry missed the Pal templates, also missed that that Temporal Smoother was disabled, didn't know that a #before it meant it ws disabled, i usually just remove the whole line, but that script is the exact same script Fit CD produces and if Temporal Smoother is removed then how do u explain the encoding times, because my processor runs at 1533mhz and would have done that in about 8 hours, and that with Bicubic Resizing, which is what i mean to write, not bilinear, which i wrote twice so you should of guessed i made a mistake.

Vhelp i dont know how you get Tmpeg better on its 2 pass VBR its crap on that, CQ is where i think it might come close but i dont like the 2 pass not even the new one on the Tmpeg plus version.

Am i the only one in thinking that doesn't look like a DVD2SVCD script and looks exactly like a Fit CD one.

We processed through DVD2SVCD. Maybe that's why it's slower. Doesn't make sense, but it's in the log file.
Yes, that's a FitCD script. That's what was used. It was pasted (replacing ) over the one that DVD2SVCD creates.

-kwag

[adam]

Kwag you ran a test between two encoders without ever even having used one of the encoders yourself. I listed about 10 inconsistencies in your test and any one of them makes your results baseless.

That section of the movie needed high bitrate? Why simply because you said so? Your logic makes no sense. You obviously don't understand how vbr encoding works. The whole point is to sacrifice bitrate and even Quantization levels to achieve an overall better average quality. If you think your eyes are a better judge than any encoder's algorthms then you need to stick to manual vbr and ditch CQ and x-pass vbr altogether. You CANNOT judge the quality of an entire encode simply by analyzing a single part. TMPGEnc very well may produce a lower avg Q value than cce on this particular source if encoded at the same settings. Unfortunately since your settings are NOT the same and due to numerous other inconsistencies in your test I guess we will never know.

"But the result is the same. Cant you see that we used the same AviSynth script? "

Oh man please tell me you were kidding when you said that. You honestly think that all your settings are defined in the AVS script? Then what are all those buttons in TMPGenc and CCE for?

You used mpeg1 in one encode and mpeg2 in the other. How can you possibly draw any comparison between the inherant quality of one encoder over the other? Couldn't the quality difference be caused because of differences between mpeg1 and mpeg2? They are completely different standards you know.

704x480 is a compliant still resolution. You are not encoding stills you are encoding program streams. This is not a compliant resolution not that, that has any effect on the results of the test. I was simply making a point.

I dont think that cce is necessarily better quality than TMPGenc in all instances or even only when encoding mpeg2. Only a fool would be so close minded. Any quality differences between these two encoder can only be judged on a subjective basis because there is absolutely NO way to eliminate ALL the other variables. CCE's mpeg1 encoding quality is so bad that it is pretty much a safe bet saying that TMPGEnc is better for mpeg1. From my personal experience I also believe that TMPGenc produces better quality with DV material.

TO MY EYES cce results in better quality when encoding mpeg2 at any bitrate on most sources but not all, most notably DV.

Both encoders are very good. There are only two things about these encoders that can be objectively stated with absolute certainty.

1: CCE is much faster than TMPGenc
2: Both encoders are such high quality mpeg2 encoders and so close in quality that neither one can definitively be judged as being superior.

Trying to prove that TMPGenc is higher quality than CCE is like trying to prove that the color red is better than the color green.

If you think CQ is the highest possible encoding mode then fine use it. If you think that TMPGenc is the better encoder than fine use it. But do not try to prove either of these because it CANNOT be done.

[kwag]

Sorry Adam. But I can see that you didn't even watch the clips. That MPEG-1 clip is way better than CCE's MPEG-2.
Stop the mumbo jumbo talk and PROVE your point. You've never provided any real samples and data. Only talk.
I challenge you, to encode "The Matrix" with CCE using 3 or 4 pass mode. I don't care if you use MPEG-1 or MPEG-2. And encode the complete movie to be of a file size approximately equal to the one created by TMPEG 1,302,091KB (video stream only) in my encode. MIN=300, MAX=2,300, AVG=Your calculation. Then look at the same section clip, as the one I posted. Your result will NEVER match the sample I made with TMPEG MPEG-1. Why do I say this? Do you think that I don't know how to use CCE? Do you think that I haven't tried it before? If it was a better encoder, I would have used it in all my templates instead of TMPEG.

Until then, I rest my case.

-kwag

[therick]

alright, sorry I started up the CCE vs. TMPEGEnc thing up...but I only have one thing to say about your tests kwag...

The TMPEGEnc encoded file is bigger than the CCE encoded file...right?

And you say that the TMPEGEnc file looks better, right? Well, with more bitrate, shouldn't it?

You are not going to convince me that TMPEGEnc is a better encoder than CCE by telling me that it's MPEG-1 at a higher bitrate looks better than CCE's MPEG-2 at a lower bitrate. Nuff said.

[adam]

That MPEG-1 clip is way better than CCE's MPEG-2.

Case in point, Your comparing apples and oranges and assuming your results are somehow valid.

You've never provided any real samples and data. Only talk.

Exactly! there is no test or sample clips you can make which can prove one encoder is better than the other because the resulting quality is always subjective. There are far too many variables to account for to make a fair comparison and the actual truth is that your results will vary depending on the source. I'm not trying to PROVE anything, only to show that you have proved NOTHING.

Until then, I rest my case.

Good.

[kwag]

I'm getting out of here
This feels like a circus. It makes me laugh that you can't analyze the results correctly
Adam, you always come up with your famous "subjective" word. Is CCE or a sponsor paying you something to be completely biased?
I think something weird is going on here 8) But of course, that's also "subjective" from the point of view of everyone reading your posts.

@therick
TMPEG's sample is bigger than CCE's because it's an action scene. If you analyze the result with bit rate viewer, you'll see the same chart I posted. Because the scenes in the clip demand bit rate, TMPEG did use it.
On the other hand, CCE didn't use more bit rate, because the bit rate was spreaded throughout the complete movie. As contrary as to what Adam said, quote: "Isn't it logical to think that maybe that extra bitrate that cce didn't use was allocated somewhere else where it was needed more? ", which is total nonsense because you can see on the clip when Neo is running against the background wall that on the CCE sample his face is barely distinguishable ( by pausing on that section ), but on TMPEG's sample it's more detailed. That scene DID require a high bit rate. CCE didn't care, because it's keeping track of it's allowed bit rate to spread it evenly througout the complete picture. If you look closely, the total file size for the complete movie made with CCE was about 60MB larger than TMPEG's encode. On the cut out clip, TMPEG's size is larger than CCE. So the answer is clear right there! If the total size for the complete movie was larger for CCE, then how is it possible that in that sample TMPEG's file size is larger? Because the bit rate was completely balanced in CCE, but not in TMPEG. Again, CQ mode is optimized to sustain a fixed quality. Where did the bit rate go in CCE? Everywhere in the film, to accomodate the target file size specified.

-kwag

[Martyn1980]

Kwag if your still watching and i bet you are, i first asked you if you used Fit CD on the encode and you said you didn't, and you said, that was the Log file DVD2SVCD produced,but now you say that it was Fit CDafter all and you pasted it over DVD2SVCD log file, WHY? what was in the DVD2SVCD log file
On my Athlon XP1800+ i can encode at just under real time, with Bicubic Resize, Image Quality at 15, Noise reduction at 5, and a resolution of 704x576, and thats using DVD2SVCD only, its slower with Fit CD on its default settings but faster if i change a few things, 16 hours on a 1.4ghz Pentium 4 is very slow, or maybe its Pentiums are shitier than i thought they were, but im guessing there ar Pentium 4 users watching who will disagree and tell me their real encoding speeds.

I thought that the CCE clip looked better personally, and if the final size of the film was 60mb bigger using CCE then why did you pick that clip, was it maybe a weakness you spotted and thought you'd show everyone, if you want a fair test it has to be with a Trailer or something short, because taking a 45 second clip out of a 135min film, isn't going to produce fair results, for all i know there a bits in the CCE version that kick the crap out of Tmpeg, but you have a cheek saying adam sounds like a sponsor for CCE, he has openly admitted its far inferior at Mpeg1 and he prefers Tmpeg for DV sources, all you go on about is Tmpeg, i started on Tmpeg and moved to CCE and stayed there not because of the speed, at first i didn't know how to use it well enough to get it up to top speed, but at the quality of the video, are you going to tell me that a CQ template like yours which i personally quite like, is gonna produce better video than a 4 or 5 pass VBR with a Min 300 and a Max4000, the average is only to determine the final filesize, the bitrate is allowed to stray as high or as low as i set it, but i will still have the film on how many discs i want.

You say if CCE was better then why aren't you using it, well i dont know, why are you using an outdated video compression system like Mpeg1, its easier to use as far as i can see, i dont know if your even still watching this post and i was coming round to your templates and i do still like some of them, but i still only see them as a high quality alternative to VCD, and not a proper threat to SVCD, the quality is good but its nowhere near DVD quality whichis a phrase which gets used far too often round here without it ever being true.

sorry for the long post guys

[rendalunit]

Both encoders are very good. There are only two things about these encoders that can be objectively stated with absolute certainty.

1: CCE is much faster than TMPGenc
2: Both encoders are such high quality mpeg2 encoders and so close in quality that neither one can definitively be judged as being superior.

One thing you did not mention is the cost of each of these encoders. CCEsp is $1,950.00<- is this a fair price for an encoder? To me it seems a little steep- I'll stick with TMPGenc.

[therick]

anyone who knows anything about the internet knows that the cost of the encoder (or any software) is a non-issue...period.

[kwag]

Martyn1980, maybe I didn't explain myself correctly. Oh, BTW, I'm still subscribed to the thread 8) . I did say I used FitCD. I even posted the script. It's in the log. The same .avs script was used for both encoders.
About the passes. That's the time it took. It's stamped in the log. The number of passes in CCE was set to 3. Which is actually 4. The first creates the .vaf file. Then 3 passes, optimizing "calculated target space", as defined by min/avg/max, taking into consideration and analyzing the .vaf file.
About the CCE clip looking better, look at it in a HDTV, and you''ll see the background "noise figure" is way higher in the CCE clip than the MPEG-1 version that TMPEG created. About the clip, I muxed the audio/video with TMPEG starting exactly at 7,200 seconds to 7,230 seconds. Which is exactly 2 hours into the movie. I didn't even look for a "prefered" clip. That just happened to be the 30 second sample that starts at exactly 2 hours into the movie.
About doing a sample from a trailer, that's impossible! It's too short for an X-pass VBR encoding. You have to encode a long movie and then cut out a piece and analyze it. You can do that with CQ modes, but not with any X-pass mode. This is simple math. The longer the movie, encoded with X-pass VBR, the lower will be the quality. This of course is if you maintain the target size constant. With CQ, you're just encoding with a quality factor already pre-set to what you want. That's exactly what I did with the samples above. I encoded the MPEG-1 with TMPEG and noted the file size created. Then we encoded with CCE using the size generated by TMPEG to be a consistent and fair comparison. Contrary as to what Adam mentioned that why the CCE wasn't encoded with the same average bit rate that the TMPEG produced, that doesn't work. The test was done just like you would have done an X-pass VBR to fit a movie in X size CD. Because you can't predict the exact size or a CQ encode, that's the reason the TMPEG movie was encoded before the CCE. So that we knew what size to target for when we encoded with CCE.
Why I'm using MPEG-1 instead of MPEG-2. I'll say it again, as I've already mentioned this over 10 times. Almost every DVD player can play VCD's. But not every player can play SVCD's. About 90% of the DVD players in the market CAN'T play SVCD. Which is an MPEG-2 format. That's why I choose MPEG-1 over MPEG-2.
You mention that the 704x480 template is nowhere near DVD quality
Hey, I saw "Proof of Life" in my HDTV, and I swapped the KVCD with the original DVD many times trying to see a difference. Guess what. There's none!. That movie fits in two CD's and there is no visible difference from the original DVD. And I'm not the only one saying that. In the old 704x480, yes, it never looked like a DVD. The video was more unstable. Even though the resolution was there. And remember that from 704x480 to 720x480, there's no visual difference at all. mathematically yes. Visually, no way. It's only 16 lines, and that's nothing. I hope I made this clearer this tiime.

-kwag

[kwag]

anyone who knows anything about the internet knows that the cost of the encoder (or any software) is a non-issue...period.

And that's exactly why so many companies are going bankrupt!.

-kwag

[kwag]

And I forgot to post the result of TMPEG's MPEG-2 2-pass VBR with our Q.Matrix and GOP. The total file size of the video stream is 1,365,228.
The same parameters and .avs script were used as in CCE. Min=300, Avg=1,368 and MAX=2,300.
Here's the same sample, cut out of the mux of the .m2v and the audio file.
http://ns1.shidima.com/kwag/tmpgenc-matrix-704x480-2-pass-mpeg-2.mpg

Now you can compare "TMPEG plus" new 2-pass VBR against the same clip done with CCE in 3-pass (which is really 4 ) mode. Is CCE worth it? You decide. My choice? TMPEG. Even at MPEG-2. CCE lost it's edge.

-kwag