Log in

View Full Version : CCE different decoders test!!!!!


KeyserSoze
2nd December 2003, 01:28
I did some tests with CCE 2.50 with 3 different Decoders(MPEG2Dec3 v1.10,MPEGDecoder_YV12 v2.03a,MPEGDecoder v2.03a) via Avisynth(2.08+2.53)on LOTR2 PAL DVD.
(Progressive frames,zigzag scanning order,3*multipassVBR,bias 20,IQP 20,Bitrate 0-3247-7500).

Testing 5000 frames with BatchCCEWS:

Settings of mpeg2dec3.avs(Avisynth2.53):

LoadPlugin("D:\DVDCCE\MPEG2Dec3\MPEG2Dec3.dll")
MPEG2Source("D:\DVDCCE\Rips\LOTR2cut\temp\LOTR2.d2v")
AddAudio()
ConvertToYUY2

Mpegdecoder_YV12(Avisynth2.53:
LoadPlugin("D:\DVDCCE\MPEGDecoder_YV12\MPEGDecoder.dll")
MPEGSource("D:\DVDCCE\Rips\LOTR2cut\temp\LOTR2.d2v",5000)
AddAudio()
ConvertToYUY2

Mpegdecoder(Avisynth2.08):
LoadPlugin("D:\DVDCCE\MPEGDecoder\MPEGDecoder.dll")
MPEGSource("D:\DVDCCE\Rips\LOTR2cut\temp\LOTR2.d2v",5000)
AddAudio()

Results:

Mpeg2dec3.mpv 79228 Kb(CCE speed 2.225)
MpegDecYV12.mpv 79229 kb(CCE speed 2.495)@ Athlon XP2400+
MpegDec.mpv 79229 kb(CCE speed 2.425)

In BitRate Viewer(full) i get this results of the previous .mpv(checked 3 times , each time i restarted my PC) :

==============Peak Bitrate+Q====== Avg Bitrate+Q

Mpeg2dec3.mpv--- 5662+Q6.26====== 2596+Q2.64
MpegDecYV12.mpv- 5678+Q6.23====== 2596+Q2.61
MpegDec.mpv----- 5678+Q6.23====== 2596+Q2.61

Also i got from Windvd 5.1+Progressive Deinterlace Option, the following stills:

http://users.otenet.gr/~pamiga2/Mpeg2dec3.png
http://users.otenet.gr/~pamiga2/mpegdecoderUV12.png
http://users.otenet.gr/~pamiga2/mpegdecoderold.png
http://users.otenet.gr/~pamiga2/original.png

There seems to be a difference between Mpeg2dec3 and Mpegdecoder in quality and Bitrate Viewer !!!!!!!!
Any comments for the best one between Mpeg2dec3 and Mpegdecoder ?

(No difference between the two Mpegdecoder 2.03 version in .png stills)

By the way i got the following info from Bitrate Viewer for the original Vobs of LOTR2 PAL:

Nom. bitrate: 7500000 Bit/Sec
VBV buffer size: 112
Constrained param. flag: No
Chroma format: 4:2:0
DCT precision: 10
Pic. structure: Frame
Field topfirst: Yes
DCT type: Field
Quantscale: Nonlinear
Scan type: Alternate
Frame type: Interlaced

But after checking it with Vdub+dvd2avi it seems to me that the Movie is Progressive...any thoughts?

Matthew
3rd December 2003, 03:56
Originally posted by KeyserSoze
But after checking it with Vdub+dvd2avi it seems to me that the Movie is Progressive...any thoughts?

The alternate scan order bit is not especially common, but it is very common for PAL progressive material to be marked as interlaced (but is usually encoded using zigzag).

KeyserSoze
3rd December 2003, 20:13
Originally posted by Matthew
The alternate scan order bit is not especially common, but it is very common for PAL progressive material to be marked as interlaced (but is usually encoded using zigzag).

Thanks for your reply, Matthew:)

Can you also tell me if its a good idea to lower the Intra DC Precision from 10 in LOTR2 TTT to 9 or use the Auto option(LOTR2 TTT PAL uses 10).LOTR2 has a lot of parts with high movement, thats why i'm asking;)
LOTR 1 PAL has Intra DC 9(but uses higher Nom. bitrate: 8000000 Bit/Sec instaid of 75000000 bit/sec in LOTR TTT)!

Using settings:
Bitrate 1000/3242/6500 for 246873 video frames(LOTR2 TTT minus ending credits ofcourse;)) with CCE 2.5 IQP 19+bias 22.

Matthew
4th December 2003, 00:04
I have little idea, except to say that the CCE 2.66 manual does recommend a figure of 10 for DVD video and going by bitrate viewer, this seems to be most common on DVDs.

However, as LOTR2 is a long film the bitrate is somewhat low and the 2.50 manual does state:

"A low number of bits may be specified for use at a low bitrate up to 4Mbps, and a high number of bits may be specified for use at a high bitrate."

KeyserSoze
4th December 2003, 00:22
Originally posted by Matthew

However, as LOTR2 is a long film the bitrate is somewhat low and the 2.50 manual does state:

"A low number of bits may be specified for use at a low bitrate up to 4Mbps, and a high number of bits may be specified for use at a high bitrate."

Thats exactly the place where i got my idea for the lower Intra Dc Precision:)
I got to test it;)