Back Forum Reply New

CCE 2.66.01.0x and TMPGEnc 2.5x Plus dare comparisons...and CAVEATS

Hi,

I just finished comparing CCE SP 2.66.01.01/2.66.01.02 trial and TMPGEnc 2.5x Plus trial with a DVD compliant MPEG-2 file.

My Source: NTSC DV AVI (with transistions) and Canon GL1 3-CCD Camcorder.

WITH REGARD TO QUALITY (indentical settings for generating DVD compliant MPG file), I would definitely recommend TMPGEnc 2.5x Plus. PERIOD. No second thoughts.
The ucl for optoimal DVD settings for TMPGEnc is this:
0165394101/TMPGEnc_Template.html

TMPGEnc 2.5x Plus quality is clean, no artifacfs, no blockiness etc...

Here comes the important comparison (no audio involved):
I useed Mediaware's M2-Edit Pro trial to get all internal Info on the DVD compliant MPG file. Here it is:

lt;Bgt;Information from CCE 2.66.01.01/2.66.01.02 lt;/Bgt;
=================================================

C:\New DVD Test 2.66.mpv
------------------------

File name       : C:\New DVD Test 2.66.mpv
File size       : 14128104
Duration        : 21.187845 s
Number of frames: 635
Drop frame flag : not set
Systems         :
format          = Video-Only Stream

Video           :
format          = MPEG-2
chroma format  = 4:2:0
width          = 720
height         = 480
frame rate     = 29.97 frame/s
bit rate       = 8500000 bit/s

Audio           :
format         = Unknown
bit rate       = 0 bit/s
sampling rate  = 0 Hz
mode           =

Filter Graph     : quot;Video Rendererquot;{B87BEB7B-8D29-423F-AE4D-6582C10175AC} quot;Elecard MPEG 2 Video Decoderquot;{F50B3F13-19C4-11CF-AA9A-02608C9BABA2} quot;Elecard MPEG2 Demultiplexerquot;{731B8592-4001-46D4-B1A5-33EC792B4501} quot;C:\New DVD Test 2.66.mpv quot;{E436EBB5-524F-11CE-9F53-0020AF0BA770}-----
Frame 0   0-  634 :          3 SEQ 720 x 480 (interlaced)                 dsp asp ratio= 3/4 (720*540)                 frame rate = 29.970 frames/s                 bit rate = 8500000 b/s                 min buf size = 229376 bytes                 constrained  = false                 IQ matrix = false                 non-IQ matrix = false                 profile/level = MP@ML                 chroma format = 4:2:0

-----
Frame 0   0-   11 :         27 GOP 01:00:00:00 closed=true broken=false         0 :         35 PIC     0 I         1 :      11423 PIC     1 P         2 :      11999 PIC     2 P         3 :      12575 PIC     3 P         4 :      13151 PIC     4 I         5 :      24523 PIC     5 P         6 :      25099 PIC     6 P         7 :      25675 PIC     7 P         8 :      26251 PIC     8 I         9 :      37623 PIC     9 P        10 :      38199 PIC    10 P        11 :      38775 PIC    11 P-----
Frame 0         0 :         35 PIC     0 I              vbv_delay            = 65535              intra DC precision   = 10              picture structure    = Frame Picture              progressive frame    = 0              top field first      = 1              repeat first field   = 0              frame pred frame DCT = 0              concealment MV       = 0              quantizer scale type = 1              intra VLC format     = 1              scan format          = alternatelt;Bgt; Information from TMPGEnc 2.5x Plus lt;/Bgt;
C:\DVD Test-2.55.m2v
--------------------

File name       : C:\DVD Test-2.55.m2v
File size       : 13225988
Duration        : 21.154455 s
Number of frames: 634
Drop frame flag : not set

Systems         :
format         = MPEG-2 Program Stream
mux rate       = 4912400 b/s
Video           :
format         = MPEG-2
chroma format  = 4:2:0
width          = 720
height         = 480
frame rate     = 29.97 frame/s
bit rate       = 8500000 bit/s

Audio           :
format         = Unknown
bit rate       = 0 bit/s
sampling rate  = 0 Hz
mode           =

Filter Graph     : quot;Video Rendererquot;{B87BEB7B-8D29-423F-AE4D-6582C10175AC} quot;Elecard MPEG 2 Video Decoderquot;{F50B3F13-19C4-11CF-AA9A-02608C9BABA2} quot;Elecard MPEG2 Demultiplexerquot;{731B8592-4001-46D4-B1A5-33EC792B4501} quot;C:\DVD Test-2.55.m2v quot;{E436EBB5-524F-11CE-9F53-0020AF0BA770}-----
Frame 0   0-   -1 :       2087 SEQ 720 x 480 (interlaced)        dsp asp ratio= 3/4 (720*540)        frame rate = 29.970 frames/s        bit rate = 8500000 b/s        min buf size = 229376 bytes        constrained  = false        IQ matrix = false        non-IQ matrix = false        profile/level = MP@ML        chroma format = 4:2:0        video format = NTSC        colour primaries = ITU-R BT.470-2 System B, G        transfer = ITU-R BT.470-2 System B, G        matrix = FCC        display size = 720 x 480

-----
Frame 0   0-    8 :       2173 GOP 00:00:00:00 closed=true broken=false         0 :       2181 PIC     0 I         2 :       9776 PIC     2 P         1 :      10064 PIC     1 B         4 :      10418 PIC     4 P         3 :      10706 PIC     3 B         6 :      11024 PIC     6 P         5 :      11312 PIC     5 B         8 :      11630 PIC     8 P         7 :      11918 PIC     7 B-----
Frame 0         0 :       2181 PIC     0 I              vbv_delay            = 65535              intra DC precision   = 10              picture structure    = Frame Picture              progressive frame    = 0              top field first      = 0              repeat first field   = 0              frame pred frame DCT = 0              concealment MV       = 0              quantizer scale type = 1              intra VLC format     = 1              scan format          = alternate
If you compare the information here carefully given my M2-Edit Pro for a CCE 2.66.01.01/2.66.01.02 against TMPGEnc 2.5x Plus generated MPG file, I have some things to point here and also some questions on some entries.QUESTIONS AND CAVEATS
=====================
No. of frames shown by CCE 2.66.01.02: 635
No. of frames shown by TMPGEnc 2.5x Plus: 634

(Here the actual frames is 634 since Premiere reported 634 frames in tha source AVI). Why CCE reports 1 extra frame?

Systems format shown by shown by CCE 2.66.01.02: 635:
Video-only stream

Systems format shown by shown by TMPGEnc 2.5x Plus
MPEG-2 Program Sream
(With no audio, ideally speaking the Systems format should be Video only stream as CCE is reporting but TMPGEnc 2.5x Plus reports differently. Why?)

Extra settings shown w.r.t TMPGEnc 2.5x Plus
============================================
colour primaries  ITU-R BT.470-2 System B, G
transfer          ITU-R BT.470-2 System B, G
(Here the color primaries and and transfer information is missing in CCE. Also what does this actually mean?)
In TMPGEnc Environmental settings, I checked the Interpolate YUV...YUV 4.4.4 stuff). Is that the reason?

Extra settings shown w.r.t TMPGEnc 2.5x Plus
============================================
matrix                 FCC
display size           720 x 480
(These two are missing in CCE. Why and what does this actually mean?)The MOST SHOCKING THING I FOUND OUT IN CCE is no matter what you do, the quot;top field firstquot; is always enabled. (From the information above, it says 1 for CCE and 0 for TMPGEnc). Surprising...since DV should have the lower field set. TMPGEnc correctly shows that the upper field first is not set by showing 0 for this flag. I have seen this same setting as 1 in previous versions of CCE.Anybody who has real knowledge in the internal settings in a MPEG-2 file is badly needed here to clarify my questions and also comment on this.

Thanks.

VS

Just on a personal, non-technical note, I would argue that it is the subjective quality of the encoded file which bothers me most ( #provided, of course, that the file is accepted and multiplexed correctly by Maestro or Scenarist!!# ).

I like TMPGEnc very much, so I am not remarking on differences in subjective quality between the two encoders - I've seen great results from both of them, particularly at higher bitrates. Therefore, I now use CCE-SP where speed is important (sometimes it is important to get a disk authored quickly for someone), and I also use CCE-SP if I have to squeeze a very long piece of footage onto a DVD-R. CCE-SP allows me to very accurately tweak file size, and the advanced bitrate settings graph dialogue really is fantastically useful (well it is to ME, anyway). If I have only about an hour's worth of DV-footage to encode, where neither time or bitrate are primary issues, I'm still happy to use TMPGEnc, and I actually prefer it to CCE-SP if I am authoring half D-1 or SIF framesizes, because it's just so easy to let TMPGEnc do this, and it does a brilliant job. As I mentioned in a couple of threads recently, I am still very much undecided about ProCoder. It seems to smooth my gliding projects (where I am filming on an airfield with lots of moving grass in the frame, prior to takeoff, and during landings). This makes the grass look blurred and unrealistic - something which TMPGENc and CCE-SP don't do. I think ProCoder cheats a little here! (mastering setting used).Arky  ;o)

Arky  ;o)

Hi,

I have actually 2 mpg-2 files sitting on my local hard disk which was from CCE and TMPGEnc. I ran both at the same time with 2 different processes of Windows Media Player and couldn't believe my eyes. TMPGEnc is a class by itself (I'm talking abt subjective quality here). Also, the size of the TMPGEnc generated file is smaller than CCE generated file.

Also, just to do one more test, Arky...can you give me all settings for making DVD compliant MPG-2 file using CCE 2.66.01.02?

Also, in CCE it has a facility to input timecode on tbe generated MPG file. How do I achieve this in TMPGEnc Plus?

MY WAY OF DOING
===============
More importantly, I fed my original NTSC DV AVI to CatDV to burn in the timestamp text track, export as a NTSC DV MOV file (no compression involved) and fed it into Premiere and rendered with transistions using Etymonix MPEG-2 I-Frame codec to use the YUV 4:4:4 color space on the final rendered AVI. Etymonix suggests that their codec would preserve the exact color information if the source is RGB24 (YUV 4:1:1 NTSC DV).

Thanks.

VS
¥
Back Forum Reply New