Any free MVC encoder yet?

[eddman]

Originally Posted by Blue_MiSfit

Also, @eddman:

Absolutely. That's the only proper way to do 3D for live action.

Derek

Thanks. Sorry for deleting my post just before your reply. I thought it's off-topic.

[kieranrk]

Originally Posted by Dark Shikari

If you're cutting down resolution, that's different. I meant if you weren't cutting resolution.

Ah, that makes more sense.

[mariner]

Originally Posted by shon3i

@Dark Shikari, tell me how is possible to use two inputs? Left eye and Right eye, and pack that into one stream?

Originally Posted by Dark Shikari

Using any of the 6 methods listed in --frame-packing. All could be done with dead-trivial Avisynth scripts.

Greetings. I'm trying to convert two 1920x1080/60p .mts streams into single 3840x1080/60p side-by-side video without success using avs scripts suggested by Limobar and crl2007. Would appreciate if you could point out what I did wrong.

Originally Posted by crl2007

For 3840x1080 you must use Limobar's avs script. Just that instead of mkv you use m2ts. Directshowsource is capable of opening m2ts streams. Only that you will have a mkv file as a final result.

Originally Posted by mariner

Not working. The output was about 100KB. This is what I did:

1. Created a 3D700.avs file with following:

VideoLeft = directshowsource("D:\3D\left.mts", audio=false)
VideoRight = directshowsource("D:\3D\right.mts", audio=false)
VideoStacked = StackHorizontal(VideoRight,VideoLeft)
ConvertToYV12(VideoStacked)

2. Ran batch file

D:\3D>x264 -o 700.mkv 3d700.avs

avs [info]: 3840x1080p 0:0 @ 10000000/166833 fps (cfr)
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.1 Cach
e64
x264 [info]: profile High, level 5.1
x264 [info]: frame I:4     Avg QP:28.00  size:   799
x264 [info]: frame P:211   Avg QP:31.00  size:   131
x264 [info]: frame B:622   Avg QP:31.00  size:   125
x264 [info]: consecutive B-frames:  0.4%  0.2%  0.0% 99.4%
x264 [info]: mb I  I16..4: 100.0%  0.0%  0.0%
x264 [info]: mb P  I16..4:  0.0%  0.0%  0.0%  P16..4:  0.0%  0.0%  0.0%  0.0%  0
.0%    skip:100.0%
x264 [info]: mb B  I16..4:  0.0%  0.0%  0.0%  B16..8:  0.0%  0.0%  0.0%  direct:
 0.0%  skip:100.0%  L0:-1.$% L1:-1.$% BI:-1.$%
x264 [info]: 8x8 transform intra:0.0%
x264 [info]: coded y,uvDC,uvAC intra: 0.0% 0.0% 0.0% inter: 0.0% 0.0% 0.0%
x264 [info]: i16 v,h,dc,p: 99%  0%  1%  0%
x264 [info]: i8c dc,h,v,p: 100%  0%  0%  0%
x264 [info]: Weighted P-Frames: Y:0.0% UV:0.0%
x264 [info]: kb/s:62.07

encoded 837 frames, 34.24 fps, 62.46 kb/s

D:\3D>pause
Press any key to continue . . .

This is input file properties:

General
ID                               : 1 (0x1)
Complete name                    : D:\3D\left.mts
Format                           : BDAV
Format/Info                      : Advanced Video Codec
File size                        : 41.6 MiB
Duration                         : 13s 938ms
Overall bit rate                 : 25.0 Mbps
Maximum Overall bit rate         : 28.0 Mbps

Video
ID                               : 4113 (0x1011)
Menu ID                          : 1 (0x1)
Format                           : AVC
Format/Info                      : Advanced Video Codec
Format profile                   : High@L4.2
Format settings, CABAC           : Yes
Format settings, ReFrames        : 4 frames
Format settings, GOP             : M=3, N=30
Codec ID                         : 27
Duration                         : 13s 963ms
Bit rate mode                    : Variable
Bit rate                         : 23.6 Mbps
Maximum bit rate                 : 26.0 Mbps
Width                            : 1 920 pixels
Height                           : 1 080 pixels
Display aspect ratio             : 16:9
Frame rate                       : 59.940 fps
Color space                      : YUV
Chroma subsampling               : 4:2:0
Bit depth                        : 8 bits
Scan type                        : Progressive
Bits/(Pixel*Frame)               : 0.190
Stream size                      : 39.3 MiB (95%)

Audio
ID                               : 4352 (0x1100)
Menu ID                          : 1 (0x1)
Format                           : AC-3
Format/Info                      : Audio Coding 3
Mode extension                   : CM (complete main)
Codec ID                         : 129
Duration                         : 14s 16ms
Bit rate mode                    : Constant
Bit rate                         : 384 Kbps
Channel(s)                       : 6 channels
Channel positions                : Front: L C R, Side: L R, LFE
Sampling rate                    : 48.0 KHz
Bit depth                        : 16 bits
Compression mode                 : Lossy
Video delay                      : -34ms
Stream size                      : 657 KiB (2%)

Text
ID                               : 4608 (0x1200)
Menu ID                          : 1 (0x1)
Format                           : PGS
Codec ID                         : 144
Duration                         : 13s 449ms
Video delay                      : -34ms

[sneaker_ger]

1. Correct your script:
VideoStacked = StackHorizontal(VideoLeft, VideoRight)
ConvertToYV12(VideoStacked)
AssumeFPS(60000, 1001)

2. Define frame packing
x264 --frame-packing 3 -o 700.mkv 3d700.avs

Whether your player supports this kind of file is a hole different story, though.

[mariner]

Originally Posted by sneaker_ger

1. Correct your script:
VideoStacked = StackHorizontal(VideoLeft, VideoRight)
ConvertToYV12(VideoStacked)
AssumeFPS(60000, 1001)

2. Define frame packing
x264 --frame-packing 3 -o 700.mkv 3d700.avs

Whether your player supports this kind of file is a hole different story, though.

Thanks for your kind reply, sneak_gear.

Same problem, output is 113KB and took just a few seconds.

VideoLeft = directshowsource("D:\3D\left.mts", audio=false)
VideoRight = directshowsource("D:\3D\right.mts", audio=false)
VideoStacked = StackHorizontal(VideoLeft, VideoRight)
ConvertToYV12(VideoStacked)
AssumeFPS(60000, 1001)

D:\3D>x264 --frame-packing 3 -o 700.mkv 3d700.avs
avs [info]: 3840x1080p 0:0 @ 60000/1001 fps (cfr)
x264 [info]: using cpu capabilities: MMX2 SSE2Fast SSSE3 FastShuffle SSE4.1 Cach
e64
x264 [info]: profile High, level 5.1
x264 [info]: frame I:4     Avg QP:28.00  size:   814
x264 [info]: frame P:211   Avg QP:31.00  size:   131
x264 [info]: frame B:622   Avg QP:31.00  size:   125
x264 [info]: consecutive B-frames:  0.4%  0.2%  0.0% 99.4%
x264 [info]: mb I  I16..4: 100.0%  0.0%  0.0%
x264 [info]: mb P  I16..4:  0.0%  0.0%  0.0%  P16..4:  0.0%  0.0%  0.0%  0.0%  0
.0%    skip:100.0%
x264 [info]: mb B  I16..4:  0.0%  0.0%  0.0%  B16..8:  0.0%  0.0%  0.0%  direct:
 0.0%  skip:100.0%  L0:-1.$% L1:-1.$% BI:-1.$%
x264 [info]: 8x8 transform intra:0.0%
x264 [info]: coded y,uvDC,uvAC intra: 0.0% 0.0% 0.0% inter: 0.0% 0.0% 0.0%
x264 [info]: i16 v,h,dc,p: 99%  0%  1%  0%
x264 [info]: i8c dc,h,v,p: 100%  0%  0%  0%
x264 [info]: Weighted P-Frames: Y:0.0% UV:0.0%
x264 [info]: kb/s:62.11

encoded 837 frames, 34.12 fps, 62.50 kb/s

[sneaker_ger]

Then the problem is probably in the source filters. You could try to install ffmpegsource or one of neuron's tools - test the AviSynth script in a player first to see if it's OK. If you can't succeed search/open a thread in the AviSynth usage section of this forum.

[mariner]

Originally Posted by sneaker_ger

Then the problem is probably in the source filters. You could try to install ffmpegsource or one of neuron's tools - test the AviSynth script in a player first to see if it's OK. If you can't succeed search/open a thread in the AviSynth usage section of this forum.

Downloaded ffmpegsource-2.14.7z.

What do I do with ffms2.dll? Copy it to the AviSynth/plugins directory or register it?

Apologize for being so thick, but this is first attempt at encoding.

Many thanks.

[sneaker_ger]

Originally Posted by mariner

Downloaded ffmpegsource-2.14.7z.

What do I do with ffms2.dll? Copy it to the AviSynth/plugins directory or register it?

Apologize for being so thick, but this is first attempt at encoding.

Many thanks.

Copy to AviSynth/plugins.

Then change script to:
VideoLeft = FFVideoSource("D:\3D\left.mts")
VideoRight = FFVideoSource("D:\3D\right.mts")

[mariner]

Originally Posted by sneaker_ger

Copy to AviSynth/plugins.

Then change script to:
VideoLeft = FFVideoSource("D:\3D\left.mts")
VideoRight = FFVideoSource("D:\3D\right.mts")

Success. Many thanks.

It seems that further compression has been applied. Is there a way to keep this to a minimal to preserve quality?

The new 3840x1080/60p file properties:

General
Complete name                    : D:\3D\700.mkv
Format                           : Matroska
File size                        : 36.3 MiB
Duration                         : 13s 981ms
Overall bit rate                 : 21.8 Mbps
Writing application              : x264 r1834 a51816a
Writing library                  : Haali Matroska Writer b0

Video
ID                               : 1
Format                           : AVC
Format/Info                      : Advanced Video Codec
Format profile                   : High@L5.1
Format settings, CABAC           : Yes
Format settings, ReFrames        : 4 frames
Codec ID                         : V_MPEG4/ISO/AVC
Duration                         : 13s 981ms
Bit rate                         : 21.4 Mbps
Width                            : 3 840 pixels
Height                           : 1 080 pixels
Display aspect ratio             : 3.556
Frame rate                       : 59.940 fps
Color space                      : YUV
Chroma subsampling               : 4:2:0
Bit depth                        : 8 bits
Scan type                        : Progressive
Bits/(Pixel*Frame)               : 0.086
Stream size                      : 35.6 MiB (98%)
Writing library                  : x264 core 112 r1834 a51816a
Encoding settings                : cabac=1 / ref=3 / deblock=1:0:0 / analyse=0x3:0x113 / me=hex / subme=7 / psy=1 / psy_rd=1.00:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=1 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 / threads=6 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / constrained_intra=0 / bframes=3 / b_pyramid=2 / b_adapt=1 / b_bias=0 / direct=1 / weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=40 / rc=crf / mbtree=1 / crf=23.0 / qcomp=0.60 / qpmin=0 / qpmax=51 / qpstep=4 / ip_ratio=1.40 / aq=1:1.00
Language                         : English

The original 1920x1080/60p file properties:

General
ID                               : 1 (0x1)
Complete name                    : D:\3D\left.mts
Format                           : BDAV
Format/Info                      : Advanced Video Codec
File size                        : 41.6 MiB
Duration                         : 13s 938ms
Overall bit rate                 : 25.0 Mbps
Maximum Overall bit rate         : 28.0 Mbps

Video
ID                               : 4113 (0x1011)
Menu ID                          : 1 (0x1)
Format                           : AVC
Format/Info                      : Advanced Video Codec
Format profile                   : High@L4.2
Format settings, CABAC           : Yes
Format settings, ReFrames        : 4 frames
Format settings, GOP             : M=3, N=30
Codec ID                         : 27
Duration                         : 13s 963ms
Bit rate mode                    : Variable
Bit rate                         : 23.6 Mbps
Maximum bit rate                 : 26.0 Mbps
Width                            : 1 920 pixels
Height                           : 1 080 pixels
Display aspect ratio             : 16:9
Frame rate                       : 59.940 fps
Color space                      : YUV
Chroma subsampling               : 4:2:0
Bit depth                        : 8 bits
Scan type                        : Progressive
Bits/(Pixel*Frame)               : 0.190
Stream size                      : 39.3 MiB (95%)

[nm]

Originally Posted by mariner

It seems that further compression has been applied. Is there a way to keep this to a minimal to preserve quality?

You used the default CRF 23. Try --crf 20 or maybe 18 if file size doesn't matter.

If the video is live action (not cel animation), also use --tune film

[mariner]

Originally Posted by nm

You used the default CRF 23. Try --crf 20 or maybe 18 if file size doesn't matter.

If the video is live action (not cel animation), also use --tune film

Thanks nm for the kind reply. Will try that.

Best regards.

[rica]

Dear coders, MVC and VLC encoder open source codes are waiting for you and we are waiting for you.
Thanks in advance.

[invincible]

What's source filter can index right.h264 (raw h264 stereo)?

[Zerofool]

Originally Posted by Stephen R. Savage

Dark Shikari: Is one stereoscopic packing mode more effective compression-wise than the others? Column interlacing and checkboard look pretty awful to me in this aspect, but I'm not sure how top/bottom compares with left/right.

I did some testing few months ago, and both side-by-side and over/under a.k.a. top/bottom (full-rez per eye), produced files ~99% bigger than just one of the views (in CRF mode). Frame sequential on the other hand, produced noticeably smaller file, saving 20-25% (if my memory serves me well), but I had doubts if all these psycho-visual optimizations stripped some of the details compared to sbs and o/u (thinking the human eye can't see the difference between the very fast switching images). I didn't really make in-depth image quality comparison, so my question is if there is something we should take in consideration when encoding frame-sequential 3D videos (regarding psy optimizations for example, or other x264 options)?

Otherwise, as an experienced Stereo-3D (PC) gamer (and 3D-video-watcher ) I personally think side-by-side is better for viewing than over/under (top/bottom) and dual-file encodes, because in the first case there's often vertical misalignment (yes, it's ridiculous, there shouldn't be, each view occupies exactly the half of the image, and still, one of the views appears to be higher or lower than the other, by let's say one pixel; if that was Stereoscopic Player issue, it would have been resolved in the many years this software has existed), and in the second case - synchronization is needed and also decoding and rendering are heavier tasks with two files.

Originally Posted by Dark Shikari

Frame interleaving may be better for compression, especially once x264 gets awareness of the frame interleaving occurring. I will probably be working on this after the end of this semester.

That's great, I can't wait to see some results. I've been wishing for such option for quite some time!

Originally Posted by Zerofool

...it's about efficient storage and playback of 3D videos on the PC.

[kolak]

What about 3D BD spec- my tests show quite often 50% saving on 2nd view. R eye is just a difference and requires much less bits than L (main) view. Very common setting can be 24Mbit for L and 12Mbit average for R eye.


Andrew

[Zerofool]

Hi, Andrew. It really depends on the content. As the stereo separation increases, the savings will get lower. Other factors may be camera misalignment, difference in focus, gamma and so on.
May I ask what MVC encoder are you using for your tests?

[kolak]

Originally Posted by Zerofool

Hi, Andrew. It really depends on the content. As the stereo separation increases, the savings will get lower. Other factors may be camera misalignment, difference in focus, gamma and so on.
May I ask what MVC encoder are you using for your tests?

Cinemacraft MVC- few different movies show that R eye needs 70%- 40% of L, depending on the source. This is for movies eg. like Avatar, Step Up.

Andrew

[shon3i]

I am playing with Mainconcept MVC (not cinevision) and usualy get same results as kolak. Right eye definitly need less bitrate, but depends from many factors.

[mariner]

Originally Posted by nm

You used the default CRF 23. Try --crf 20 or maybe 18 if file size doesn't matter.

If the video is live action (not cel animation), also use --tune film

Tried --crf 18 --tune film. Bit rate is bumped up to 55 mbps. Is this the highest it will go?

What is the advantage of profile 5.1 over 4,2 or 4.1?

Many thanks and best regards.

[nm]

Originally Posted by mariner

Tried --crf 18 --tune film. Bit rate is bumped up to 55 mbps. Is this the highest it will go?

No. You can go up to lossless (CRF 0) if necessary. But CRF 18 is already visually transparent in most cases, and realtime decoding will get increasingly difficult when you raise the bitrate. Or is that not a concern for you?

What is the advantage of profile 5.1 over 4,2 or 4.1?

You mean level, not profile. Higher levels allow higher bitrates and more reference frames at high resolutions. See: http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Levels