VideoHelp Forum
+ Reply to Thread
Page 1 of 2
1 2 LastLast
Results 1 to 30 of 43
Thread
  1. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Guys, I need some help, I haven't been to the forum in a while, so sorry if I'm in the wrong topic.

    I have an NTSC video, part of it, and I would like to know what is the best way to deinterlace it, that is, a 480i to 480p, so I can use it in Davinci Resolve 18 and do an Upscaling to FullHD and/or 4K.

    I used the direct interlaced video, and there was a jittery artifact at the top of the video, and even using the Davinci deinterlacer it didn't work.

    What would be the best software to do this? I have Megui and Hybrid installed here, I saw comments about AviSynth and FFMpeg, can anyone guide me about Scripts and how to use them so I can get the best deinterlacing please.

    Here is the part of the file I mentioned, as it is showing artifacts on the nose of the second trumpet player behind the main one in an Upscaling to 1080p in Topaz Video AI v.4.0.4, using the DIONE TV preset.


    Topaz Video AI v.4.0.4 video:

    https://mega.nz/file/h6gVwIBA#52EvqamAHQl2zZaZLMqKNBV6FJ1kdRc21AuiLvqU588

    Images:

    Image
    [Attachment 75122 - Click to enlarge]


    Image
    [Attachment 75123 - Click to enlarge]


    Interlaced Original:

    https://mega.nz/file/N751SQgD#sfj6sMfBfbG-4OBsV9Vqx8uRwSQoXL92xvNisEZ0iEk

    Att.

    Druid®.
    Last edited by DruidCtba; 30th Nov 2023 at 11:23.
    Quote Quote  
  2. Member
    Join Date
    Mar 2008
    Location
    United States
    Search Comp PM
    unable to download your file ....
    Image Attached Thumbnails Click image for larger version

Name:	i1.jpg
Views:	15
Size:	153.2 KB
ID:	75111  

    Quote Quote  
  3. Member DB83's Avatar
    Join Date
    Jul 2007
    Location
    United Kingdom
    Search Comp PM
    +1

    I assume, if I read you correctly, the sample is the upscaled attempt.


    Just post the original, interlaced, one and direct as an attachment to this topic. (as long as it does not exceed 500 mb) Or use a host that down not restrict downloads.
    Quote Quote  
  4. Member
    Join Date
    May 2005
    Location
    Australia-PAL Land
    Search Comp PM
    Post removed due new attachment below.
    Last edited by Alwyn; 28th Nov 2023 at 18:32.
    Quote Quote  
  5. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by davexnet View Post
    unable to download your file ....
    https://mega.nz/file/h6gVwIBA#52EvqamAHQl2zZaZLMqKNBV6FJ1kdRc21AuiLvqU588

    Att.

    Druid®.
    Quote Quote  
  6. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by DB83 View Post
    +1

    I assume, if I read you correctly, the sample is the upscaled attempt.


    Just post the original, interlaced, one and direct as an attachment to this topic. (as long as it does not exceed 500 mb) Or use a host that down not restrict downloads.
    OK, JUST DO IT .
    Quote Quote  
  7. Member
    Join Date
    May 2005
    Location
    Australia-PAL Land
    Search Comp PM
    The nose has it.

    The AVISynth protagonists can describe how to use QTGMC to get a nice nose.
    Last edited by Alwyn; 28th Nov 2023 at 19:03.
    Quote Quote  
  8. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by Alwyn View Post
    The nose has it.

    The AVISynth protagonists can describe how to use QTGMC to get a nice nose.
    OK, I will be waiting .

    Att.

    Druid®.
    Quote Quote  
  9. Any AviSynth deinterlacer seems to do a better job than the Topaz AI crap. One does not even need to jump onto QTGMC
    Code:
    ffms2("Yanni - Tribute (1997 DVD-9 720 x480)__00012.mkv")
    BWDIF()
    And b.t.w. why always jump onto the bandwagon of deinterlacing and upscaling? Too many good video is just getting spoiled....
    Last edited by Sharc; 29th Nov 2023 at 02:49.
    Quote Quote  
  10. Member
    Join Date
    May 2005
    Location
    Australia-PAL Land
    Search Comp PM
    Originally Posted by Sharc
    And b.t.w. why always jump onto the bandwagon of deinterlacing
    Maybe not for this particular video, but analogue captures? 50fps is much smoother than 25fps.
    Quote Quote  
  11. Originally Posted by Alwyn View Post
    Originally Posted by Sharc
    And b.t.w. why always jump onto the bandwagon of deinterlacing
    Maybe not for this particular video, but analogue captures? 50fps is much smoother than 25fps.
    Sure. For decent interlaced footage it may just be better to let a TV do the bob-deinterlacing and scaling rather than making things (often) worse by processing it by all means ...
    Quote Quote  
  12. Member
    Join Date
    May 2005
    Location
    Australia-PAL Land
    Search Comp PM
    Another issue is playing these files in VLC Player, a popular computer player (as you are aware). It's deinterlacer in AUTO is poor; to properly get rid of the jaggies, one needs to turn the deinterlacer ON. =nuisance.
    Quote Quote  
  13. I would not dispute that there are better deinterlacers than you get in VLC, but VLC is correct to not automatically deinterlace the source, since it's not flagged as interlaced. Auto deinterlace normally only means that frames(/macro blocks) that are flagged as interlaced get deinterlaced, not that the source gets somehow analyzed.

    Just for fun, I attached something (script: https://pastebin.com/iQfevL8S) I would have hoped tools like Topaz would produce.

    Cu Selur
    Image Attached Files
    users currently on my ignore list: deadrats, Stears555
    Quote Quote  
  14. Banned
    Join Date
    Nov 2022
    Search PM
    Originally Posted by Sharc View Post
    Originally Posted by Alwyn View Post
    Originally Posted by Sharc
    And b.t.w. why always jump onto the bandwagon of deinterlacing
    Maybe not for this particular video, but analogue captures? 50fps is much smoother than 25fps.
    Sure. For decent interlaced footage it may just be better to let a TV do the bob-deinterlacing and scaling rather than making things (often) worse by processing it by all means ...
    I don't think that software bobber to 50p can look worse that a dumb bobber in a TV set. Maybe a Faroudja or HQV chip... still, I don't think 20-year old realtime hardware can compete with modern software.

    I would also upscale to at least 720p for youtube.
    Quote Quote  
  15. Member
    Join Date
    Mar 2008
    Location
    United States
    Search Comp PM
    I'm hardware limited. My PC is too slow to smoothly play Selur's AV01 video.
    I have not gotten into Vapoursynth yet, but I would have hoped NNedi3 in Avisynth would have been close to this ...
    Quote Quote  
  16. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by Sharc View Post
    Any AviSynth deinterlacer seems to do a better job than the Topaz AI crap. One does not even need to jump onto QTGMC
    Code:
    ffms2("Yanni - Tribute (1997 DVD-9 720 x480)__00012.mkv")
    BWDIF()
    And b.t.w. why always jump onto the bandwagon of deinterlacing and upscaling? Too many good video is just getting spoiled....
    Well my friend @Sharc, I partially agree with you, this Yanni show, as it was well recorded, and using the current Topaz Video AI v.4.0.4 preset, DIONE TV, gave a great result in Upscaling up to 1080p , looking like an HD video, it was not perfect, but considering the limitations of the older versions in which the human faces of the audience and the distant artists on the stage looked blurry or like a gouache painting, now they are almost perfect, my only complaint was these artifacts on the musician's nose, otherwise it was very good, not perfect, but very good, really looking like an HD show.
    It is visibly clear that these AI work much better for HD and FHD content than SD content, but they don't say this, and I know that there is a lot of content on DVD, VHS, TV, in SD that currently most software is not capable of to recover.

    I noticed that Davinci Resolve 18 Upscaling does a good job for HD and FHD content, but SD, by far Topaz Video AI is light years better, with stable and more pleasant images.

    Att.

    Druid®.

    P.S. Just one question: as I use Avisynth, this command you gave, as I said above I only have the MEGUI and HYBRID GUI, and I don't know how to use the Windows prompt command, could you guide me step by step on how to use and obtain the 480p file?
    Last edited by DruidCtba; 30th Nov 2023 at 11:23.
    Quote Quote  
  17. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by Selur View Post
    I would not dispute that there are better deinterlacers than you get in VLC, but VLC is correct to not automatically deinterlace the source, since it's not flagged as interlaced. Auto deinterlace normally only means that frames(/macro blocks) that are flagged as interlaced get deinterlaced, not that the source gets somehow analyzed.

    Just for fun, I attached something (script: https://pastebin.com/iQfevL8S) I would have hoped tools like Topaz would produce.

    Cu Selur
    Very good my friend @Selur, how did you achieve this Upscaling, using what software, or what commands for this, could you show us step by step?

    Thanks in advance for your help :P.

    Att.

    Druid®.

    P.S. Now I saw my friend @Selur, the script you made to UpScale the 480i video to 1080p, I just don't know where to apply it, could you guide me please.

    Funny that the result in mediainfo doesn't say that the result was a progressive video, but I saw the command:

    # set output frame rate to 59.94fps (progressive)
    clip = core.std.AssumeFPS(clip=clip, fpsnum=60000, fpsden=1001)
    Last edited by DruidCtba; 29th Nov 2023 at 22:49.
    Quote Quote  
  18. Originally Posted by Bwaak View Post
    Originally Posted by Sharc View Post
    Originally Posted by Alwyn View Post
    Originally Posted by Sharc
    And b.t.w. why always jump onto the bandwagon of deinterlacing
    Maybe not for this particular video, but analogue captures? 50fps is much smoother than 25fps.
    Sure. For decent interlaced footage it may just be better to let a TV do the bob-deinterlacing and scaling rather than making things (often) worse by processing it by all means ...
    I don't think that software bobber to 50p can look worse that a dumb bobber in a TV set. Maybe a Faroudja or HQV chip... still, I don't think 20-year old realtime hardware can compete with modern software.
    Well, I had some doubts looking at the OP's Topaz example .
    And yes, I remember Farudja. It is/was a nearly perfect deinterlacer. Unfortunately TV manufacturers hardly ever (or never) specify their (proprietary) deinterlacing principle.

    I would also upscale to at least 720p for youtube.
    There are reasons for upscaling, sure. But there are pitfalls as well e.g. with the colormatrix (601 vs 709). If one is careless there is a 50% chance that things go wrong.
    Quote Quote  
  19. Now I saw my friend @Selur, the script you made to UpScale the 480i video to 1080p, I just don't know where to apply it, could you guide me please.
    The script I used is a Vapoursynth script, which I created in Hybrid and it's torch addon.
    I used QTGMC for deinterlacing, cropped, then used CodeFormer followed by DPIR denoising which I applied with a limit mask to only filter suff with a luma value <=80, then I appplied CAS and used NNEDI3 to resize.

    Funny that the result in mediainfo doesn't say that the result was a progressive video
    If mediainfo does not report that the source is interlaced, its progressive.
    Also av1 does not support interlaced encoding.
    Code:
    MediaInfo.exe --full g:\Output\CodeFormer_DPIR_NNEDI3_CAS.mp4
    General
    Count                                    : 349
    Count of stream of this kind             : 1
    Kind of stream                           : General
    Kind of stream                           : General
    Stream identifier                        : 0
    Count of video streams                   : 1
    Count of audio streams                   : 2
    Video_Format_List                        : AV1
    Video_Format_WithHint_List               : AV1
    Codecs Video                             : AV1
    Video_Language_List                      : English
    Audio_Format_List                        : AC-3 / AC-3
    Audio_Format_WithHint_List               : AC-3 / AC-3
    Audio codecs                             : AC-3 / AC-3
    Audio_Language_List                      : English / English
    Audio_Channels_Total                     : 8
    Complete name                            : g:\Output\CodeFormer_DPIR_NNEDI3_CAS.mp4
    Folder name                              : g:\Output
    File name extension                      : CodeFormer_DPIR_NNEDI3_CAS.mp4
    File name                                : CodeFormer_DPIR_NNEDI3_CAS
    File extension                           : mp4
    Format                                   : MPEG-4
    Format                                   : MPEG-4
    Format/Extensions usually used           : braw mov mp4 m4v m4a m4b m4p m4r 3ga 3gpa 3gpp 3gp 3gpp2 3g2 k3g jpm jpx mqv ismv isma ismt f4a f4b f4v
    Commercial name                          : MPEG-4
    Format profile                           : Base Media
    Internet media type                      : video/mp4
    Codec ID                                 : iso4
    Codec ID                                 : iso4 (iso4/av01)
    Codec ID/Url                             : http://www.apple.com/quicktime/download/standalone.html
    CodecID_Compatible                       : iso4/av01
    File size                                : 183527643
    File size                                : 175 MiB
    File size                                : 175 MiB
    File size                                : 175 MiB
    File size                                : 175 MiB
    File size                                : 175.0 MiB
    Duration                                 : 57920
    Duration                                 : 57 s 920 ms
    Duration                                 : 57 s 920 ms
    Duration                                 : 57 s 920 ms
    Duration                                 : 00:00:57.920
    Duration                                 : 00:00:57;50
    Duration                                 : 00:00:57.920 (00:00:57;50)
    Overall bit rate                         : 25349122
    Overall bit rate                         : 25.3 Mb/s
    Frame rate                               : 59.940
    Frame rate                               : 59.940 FPS
    Frame count                              : 3470
    Stream size                              : 17777
    Stream size                              : 17.4 KiB (0%)
    Stream size                              : 17 KiB
    Stream size                              : 17 KiB
    Stream size                              : 17.4 KiB
    Stream size                              : 17.36 KiB
    Stream size                              : 17.4 KiB (0%)
    Proportion of this stream                : 0.00010
    HeaderSize                               : 17693
    DataSize                                 : 183509874
    FooterSize                               : 76
    IsStreamable                             : Yes
    Encoded date                             : 2023-11-29 14:30:12 UTC
    Tagged date                              : 2023-11-29 14:30:12 UTC
    File creation date                       : 2023-11-29 14:30:13.319 UTC
    File creation date (local)               : 2023-11-29 15:30:13.319
    File last modification date              : 2023-11-29 14:30:13.392 UTC
    File last modification date (local)      : 2023-11-29 15:30:13.392
    Writing application                      : Hybrid 2023.11.28.1
    Writing application                      : Hybrid 2023.11.28.1
    
    Video
    Count                                    : 382
    Count of stream of this kind             : 1
    Kind of stream                           : Video
    Kind of stream                           : Video
    Stream identifier                        : 0
    StreamOrder                              : 0
    ID                                       : 1
    ID                                       : 1
    Format                                   : AV1
    Format                                   : AV1
    Format/Info                              : AOMedia Video 1
    Format/Url                               : http://aomedia.org/
    Commercial name                          : AV1
    Format profile                           : Main@L5.0
    Codec ID                                 : av01
    Duration                                 : 57890
    Duration                                 : 57 s 890 ms
    Duration                                 : 57 s 890 ms
    Duration                                 : 57 s 890 ms
    Duration                                 : 00:00:57.890
    Duration                                 : 00:00:57;50
    Duration                                 : 00:00:57.890 (00:00:57;50)
    Bit rate                                 : 24718947
    Bit rate                                 : 24.7 Mb/s
    Maximum bit rate                         : 31198592
    Maximum bit rate                         : 31.2 Mb/s
    Width                                    : 1920
    Width                                    : 1 920 pixels
    Height                                   : 1478
    Height                                   : 1 478 pixels
    Sampled_Width                            : 1920
    Sampled_Height                           : 1478
    Pixel aspect ratio                       : 1.000
    Display aspect ratio                     : 1.299
    Display aspect ratio                     : 1.299
    Rotation                                 : 0.000
    Frame rate mode                          : CFR
    Frame rate mode                          : Constant
    Frame rate                               : 59.940
    Frame rate                               : 59.940 (60000/1001) FPS
    FrameRate_Num                            : 60000
    FrameRate_Den                            : 1001
    Frame count                              : 3470
    Color space                              : YUV
    Chroma subsampling                       : 4:2:0
    Chroma subsampling                       : 4:2:0
    Bit depth                                : 10
    Bit depth                                : 10 bits
    Bits/(Pixel*Frame)                       : 0.145
    Stream size                              : 178876266
    Stream size                              : 171 MiB (97%)
    Stream size                              : 171 MiB
    Stream size                              : 171 MiB
    Stream size                              : 171 MiB
    Stream size                              : 170.6 MiB
    Stream size                              : 171 MiB (97%)
    Proportion of this stream                : 0.97466
    Title                                    : ?@GPAC2.3-DEV-rev623-gc08973cc8-ab-suite
    Language                                 : en
    Language                                 : English
    Language                                 : English
    Language                                 : en
    Language                                 : eng
    Language                                 : en
    Encoded date                             : 2023-11-29 14:30:12 UTC
    Tagged date                              : 2023-11-29 14:30:13 UTC
    colour_description_present               : Yes
    colour_description_present_Source        : Container / Stream
    Color range                              : Limited
    colour_range_Source                      : Container / Stream
    colour_primaries_Source                  : Container / Stream
    transfer_characteristics_Source          : Container / Stream
    Matrix coefficients                      : BT.470 System B/G
    matrix_coefficients_Source               : Container / Stream
    Codec configuration box                  : av1C
    
    Audio #1
    Count                                    : 313
    Count of stream of this kind             : 2
    Kind of stream                           : Audio
    Kind of stream                           : Audio
    Stream identifier                        : 0
    Stream identifier                        : 1
    StreamOrder                              : 1
    ID                                       : 2
    ID                                       : 2
    Format                                   : AC-3
    Format                                   : AC-3
    Format/Info                              : Audio Coding 3
    Format/Url                               : https://en.wikipedia.org/wiki/AC3
    Commercial name                          : Dolby Digital
    Commercial name                          : Dolby Digital
    Format settings, Endianness              : Big
    Codec ID                                 : ac-3
    Duration                                 : 57920
    Duration                                 : 57 s 920 ms
    Duration                                 : 57 s 920 ms
    Duration                                 : 57 s 920 ms
    Duration                                 : 00:00:57.920
    Duration                                 : 00:00:57.920
    Bit rate mode                            : CBR
    Bit rate mode                            : Constant
    Bit rate                                 : 448000
    Bit rate                                 : 448 kb/s
    Maximum bit rate                         : 473088
    Maximum bit rate                         : 473 kb/s
    Channel(s)                               : 6
    Channel(s)                               : 6 channels
    Channel positions                        : Front: L C R, Side: L R, LFE
    Channel positions                        : 3/2/0.1
    Channel layout                           : L R C LFE Ls Rs
    Samples per frame                        : 1536
    Sampling rate                            : 48000
    Sampling rate                            : 48.0 kHz
    Samples count                            : 2780160
    Frame rate                               : 31.250
    Frame rate                               : 31.250 FPS (1536 SPF)
    Frame count                              : 1810
    Compression mode                         : Lossy
    Compression mode                         : Lossy
    Stream size                              : 3243520
    Stream size                              : 3.09 MiB (2%)
    Stream size                              : 3 MiB
    Stream size                              : 3.1 MiB
    Stream size                              : 3.09 MiB
    Stream size                              : 3.093 MiB
    Stream size                              : 3.09 MiB (2%)
    Proportion of this stream                : 0.01767
    Title                                    : Surround 5.1
    Language                                 : en
    Language                                 : English
    Language                                 : English
    Language                                 : en
    Language                                 : eng
    Language                                 : en
    Service kind                             : CM
    Service kind                             : Complete Main
    Default                                  : Yes
    Default                                  : Yes
    Alternate group                          : 1
    Alternate group                          : 1
    Encoded date                             : 2023-11-29 14:30:13 UTC
    Tagged date                              : 2023-11-29 14:30:13 UTC
    bsid                                     : 8
    Dialog Normalization                     : -31
    Dialog Normalization                     : -31 dB
    compr                                    : -10.10
    compr                                    : -10.10 dB
    acmod                                    : 7
    lfeon                                    : 1
    dialnorm_Average                         : -31
    dialnorm_Average                         : -31 dB
    dialnorm_Minimum                         : -31
    dialnorm_Minimum                         : -31 dB
    dialnorm_Maximum                         : -31
    dialnorm_Maximum                         : -31 dB
    dialnorm_Count                           : 32
    compr_Average                            : -11.62
    compr_Average                            : -11.62 dB
    compr_Minimum                            : -13.20
    compr_Minimum                            : -13.20 dB
    compr_Maximum                            : -9.68
    compr_Maximum                            : -9.68 dB
    compr_Count                              : 32
    dynrng_Average                           : -1.16
    dynrng_Average                           : -1.16 dB
    dynrng_Minimum                           : -2.14
    dynrng_Minimum                           : -2.14 dB
    dynrng_Maximum                           : 0.00
    dynrng_Maximum                           : 0.00 dB
    dynrng_Count                             : 32
    
    Audio #2
    Count                                    : 307
    Count of stream of this kind             : 2
    Kind of stream                           : Audio
    Kind of stream                           : Audio
    Stream identifier                        : 1
    Stream identifier                        : 2
    StreamOrder                              : 2
    ID                                       : 3
    ID                                       : 3
    Format                                   : AC-3
    Format                                   : AC-3
    Format/Info                              : Audio Coding 3
    Format/Url                               : https://en.wikipedia.org/wiki/AC3
    Commercial name                          : Dolby Digital
    Commercial name                          : Dolby Digital
    Format settings, Endianness              : Big
    Codec ID                                 : ac-3
    Duration                                 : 57920
    Duration                                 : 57 s 920 ms
    Duration                                 : 57 s 920 ms
    Duration                                 : 57 s 920 ms
    Duration                                 : 00:00:57.920
    Duration                                 : 00:00:57.920
    Bit rate mode                            : CBR
    Bit rate mode                            : Constant
    Bit rate                                 : 192000
    Bit rate                                 : 192 kb/s
    Maximum bit rate                         : 202752
    Maximum bit rate                         : 203 kb/s
    Channel(s)                               : 2
    Channel(s)                               : 2 channels
    Channel positions                        : Front: L R
    Channel positions                        : 2/0/0
    Channel layout                           : L R
    Samples per frame                        : 1536
    Sampling rate                            : 48000
    Sampling rate                            : 48.0 kHz
    Samples count                            : 2780160
    Frame rate                               : 31.250
    Frame rate                               : 31.250 FPS (1536 SPF)
    Frame count                              : 1810
    Compression mode                         : Lossy
    Compression mode                         : Lossy
    Stream size                              : 1390080
    Stream size                              : 1.33 MiB (1%)
    Stream size                              : 1 MiB
    Stream size                              : 1.3 MiB
    Stream size                              : 1.33 MiB
    Stream size                              : 1.326 MiB
    Stream size                              : 1.33 MiB (1%)
    Proportion of this stream                : 0.00757
    Title                                    : Stereo
    Language                                 : en
    Language                                 : English
    Language                                 : English
    Language                                 : en
    Language                                 : eng
    Language                                 : en
    Service kind                             : CM
    Service kind                             : Complete Main
    Default                                  : Yes
    Default                                  : Yes
    Alternate group                          : 1
    Alternate group                          : 1
    Encoded date                             : 2023-11-29 14:30:13 UTC
    Tagged date                              : 2023-11-29 14:30:13 UTC
    bsid                                     : 8
    Dialog Normalization                     : -31
    Dialog Normalization                     : -31 dB
    compr                                    : -3.25
    compr                                    : -3.25 dB
    dsurmod                                  : 0
    acmod                                    : 2
    lfeon                                    : 0
    dialnorm_Average                         : -31
    dialnorm_Average                         : -31 dB
    dialnorm_Minimum                         : -31
    dialnorm_Minimum                         : -31 dB
    dialnorm_Maximum                         : -31
    dialnorm_Maximum                         : -31 dB
    dialnorm_Count                           : 32
    compr_Average                            : -4.60
    compr_Average                            : -4.60 dB
    compr_Minimum                            : -5.49
    compr_Minimum                            : -5.49 dB
    compr_Maximum                            : -3.25
    compr_Maximum                            : -3.25 dB
    compr_Count                              : 32
    If one is careless there is a 50% chance that things go wrong.
    Only if the tools you use do not know how to respect color matrices. And if that is the assumtion you could also encounter a player/whatever that doesn't use 709 but always something else,...

    Cu Selur

    Ps.: I agree that adding a color matrix change to 709 might be a good idea when encoding for distribution.
    Last edited by Selur; 30th Nov 2023 at 08:52.
    users currently on my ignore list: deadrats, Stears555
    Quote Quote  
  20. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by Selur View Post
    Now I saw my friend @Selur, the script you made to UpScale the 480i video to 1080p, I just don't know where to apply it, could you guide me please.
    The script I used is a Vapoursynth script, which I created in Hybrid and it's torch addon.
    I used QTGMC for deinterlacing, cropped, then used CodeFormer followed by DPIR denoising which I applied with a limit mask to only filter suff with a luma value <=80, then I appplied CAS and used NNEDI3 to resize.
    Friend @Selur, I sincerely thank you for all the answers and help for me here on the forum, but as I'm a complete noobie, I use GUI tools and even so, for me HYBRID and MEGUI are complex applications for my level of knowledge , which I think is in relation to you and the people here of just 1% .

    Anyway, taking into consideration everything I explained above, I count on your patience and again I ask you to please do a step by step guide for someone who is a complete noobie, if you can, because I think you have reached the heart of my need and that The result you obtained for me was satisfactory.

    The only thing I remember, many months ago, when I had problems with the audience's faces being deformed by Topaz, was that they mentioned NNEDI3, which was an excellent UpScaling, and according to people, Topaz, uses everything you explained above in your software, but not with this advanced knowledge, which is why the musician's nose fails, and that paying almost US$400, when you have the same for free and better, is not worth it, only if you don't have no knowledge on how to use these commands, which is my case.

    In any case, I reiterate that I need you to tell me in detail, and step by step, how you did, so that I can use it here, both for this Yanni DVD, and for others that I have here to try to improve the image, now that record companies refuse to produce such an improvement of their productions to HD, FHD and/or 4K content from their original analogue tapes, which is very sad, as many memorable shows and films deserve this, and I can name one, apart from Tribute by Yanni, which made me very upset that record companies didn't do it, which is Jean Michel Jarre's show - Live In Beijing, and Loreena McKennitt's - Nights From The Alhambra .

    From what I understand, you did something very complex, you used not just HYBRID, anyway, I could be wrong, but there were several steps that I didn't understand very well and I can't even reproduce it here :/, sorry.

    Att.

    Druid®.

    P.S. In the case of the show Loreena McKennitt - Nights From The Alhambra, a user of rutracker, did a wonderful job, like the one you did here, which was 99% perfect, but he didn't want to say how he did it, only that he didn't use any application like Topaz or another and yes all these open source tools with ffmpeg and avisynth (you certainly used NNEDI3, QTGMC, CAS, Vapoursynth, etc.)
    Last edited by DruidCtba; 30th Nov 2023 at 11:24.
    Quote Quote  
  21. Hybrid is meant for advanced users and yes, I not simply threw a source into Hybrid and pressed a button and there was the output.
    (I didn't try to tweak the options and only used some values that looked 'okay' to do what I had in mind when I saw the source.)

    To be frank, writing a step-by-step seems to be too much trouble, for someone who doesn't really care enough to write even the name of the tool correct.
    That said what I did to end up with the script I posted was, I:
    • Installed Hybrid + addons.
    • Started Hybrid, set default output paths. (save defaults)
    • Loaded the source (Base->Open File).
    • Overwrote the scan order, which wrongly was indicated as progressive to 'Top Field First' (by enabling 'Filtering->(De-)Interlace/Telecine->Deinterlace/Telecine Settings->Overwrite scan type to' and setting it to 'top field first').
    • Enabled bobbed deinterlacing by the selected deinterlacer ('QTGMC (Vapoursynth)') (by enabling 'Filtering->(De-)Interlace/Telecine->QTGMC Vapoursynth->Bob'),
    • Cropped the source to '702x480' (by enabling 'Crop/Resize->Base->Picture Crop' and setting left=12, right=6, top=0, bottom=0).
    • Set the output PAR to 1:1 (by enabling 'Crop/Resize->Base->Pixel Aspect Ratio (PAR)->Convert output to PAR' and setting it to 'Square Pixel (1:1)').
    • Set the output width to 1920 (by setting 'Crop/Resize->Base->Picture Resize->Target resolution->Width' to 1920)
    • Then I enabled and configured CodeFormer (by enabling 'Filtering->Vapoursynth->Other->CodeFormer' and setting 'Filtering->Vapoursynth->Other->CodeFormer->Weight' to 1).
    • Moved the CodeFormer-call directly behind the Crop-call. (by adjusting the order in 'Filtering->Vapoursynth->Misc->Filter Order/Queue').
    • Enabled the masking UI elements (by enabling 'Filtering->Vapoursynth->Misc->UI->Show 'Masked' controls').
    • Enabled and configured DPIR denoising (by enabling 'Filtering->Vapoursynth->DeNoise->DPIR' and setting it's strength to 15)
    • Enabled and configured the masking for DPIR (by enabling 'Filtering->Vapoursynth->DeNoise->DPIR->Masked' - scroll to the right - and setting the Limit to 80).
    • Enabled and configured contrast sharpening through CAS (by enabling 'Filtering->Vapoursynth->Sharpen->CAS' and setting 'Sharpness' to 0.6)
    • Enabled the use of NNEDI3 (by enabling 'Filtering->Vapoursynth->Frame->Resizer' and setting it to 'NNEDI3').
    • What I did not do, but what might help in further in playback compatibility, is to convert the color matrix to BT 709 or 2020ncl (by enabling 'Filtering->Vapoursynth->Color->Matrix->ColorMatrix' an adjusting the output; by default Hybrid will use here 2020ncl since the output resolution > 1080p)
    • I also told Hybrid to use DGDecNV as source filter (by enabling 'Filtering->Vapoursynth->Source->Use DGDecNV when available), but that is not really needed.
    • Then I configured the output video encoder, container, set the output name, let Hybrid create a job and after that started the job.
    I looked at the Vapoursynth Script (Filtering->Vapoursynth->Show Vapoursynth Script) and the Vapoursynth Preview (Filtering->Vapoursynth->Show Filter Preview).
    The script I created with these steps looks like this:
    Code:
    # Imports
    import vapoursynth as vs
    # getting Vapoursynth core
    import site
    import ctypes
    import sys
    import os
    core = vs.core
    # Import scripts folder
    scriptPath = 'F:/Hybrid/64bit/vsscripts'
    sys.path.insert(0, os.path.abspath(scriptPath))
    # Loading Support Files
    Dllref = ctypes.windll.LoadLibrary("F:/Hybrid/64bit/vsfilters/Support/libfftw3f-3.dll")
    # Adding torch dependencies to PATH
    path = site.getsitepackages()[0]+'/torch_dependencies/bin/'
    ctypes.windll.kernel32.SetDllDirectoryW(path)
    path = path.replace('\\', '/')
    os.environ["PATH"] = path + os.pathsep + os.environ["PATH"]
    # Loading Plugins
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/SharpenFilter/CAS/CAS.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/GrainFilter/RemoveGrain/RemoveGrainVS.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/GrainFilter/AddGrain/AddGrain.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/DenoiseFilter/DFTTest/DFTTest.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/DenoiseFilter/NEO_FFT3DFilter/neo-fft3d.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/Support/EEDI3m_opencl.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/ResizeFilter/nnedi3/NNEDI3CL.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/Support/libmvtools.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/Support/scenechange.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/Support/fmtconv.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/MiscFilter/MiscFilters/MiscFilters.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/DeinterlaceFilter/Bwdif/Bwdif.dll")
    core.std.LoadPlugin(path="F:/Hybrid/64bit/vsfilters/SourceFilter/DGDecNV/DGDecodeNV.dll")
    # Import scripts
    import havsfunc
    # source: 'C:\Users\Selur\Desktop\Yanni - Tribute (1997 DVD-9 720 x480)__00012.mkv'
    # current color space: YUV420P8, bit depth: 8, resolution: 720x480, fps: 29.97, color matrix: 470bg, yuv luminance scale: limited, scanorder: top field first
    # Loading C:\Users\Selur\Desktop\Yanni - Tribute (1997 DVD-9 720 x480)__00012.mkv using DGSource
    clip = core.dgdecodenv.DGSource("J:/tmp/mkv_d603b0d1a8552802aa433b7391d38442_853323747.dgi",fieldop=0)# 29.97 fps, scanorder: top field first
    # Setting detected color matrix (470bg).
    clip = core.std.SetFrameProps(clip, _Matrix=5)
    # Setting color transfer info (470bg), when it is not set
    clip = clip if not core.text.FrameProps(clip,'_Transfer') else core.std.SetFrameProps(clip, _Transfer=5)
    # Setting color primaries info (), when it is not set
    clip = clip if not core.text.FrameProps(clip,'_Primaries') else core.std.SetFrameProps(clip, _Primaries=5)
    # Setting color range to TV (limited) range.
    clip = core.std.SetFrameProp(clip=clip, prop="_ColorRange", intval=1)
    # making sure frame rate is set to 29.97
    clip = core.std.AssumeFPS(clip=clip, fpsnum=30000, fpsden=1001)
    clip = core.std.SetFrameProp(clip=clip, prop="_FieldBased", intval=2) # tff
    # Deinterlacing using QTGMC
    clip = havsfunc.QTGMC(Input=clip, Preset="Faster", TFF=True, opencl=True) # new fps: 59.94
    # Making sure content is preceived as frame based
    clip = core.std.SetFrameProp(clip=clip, prop="_FieldBased", intval=0) # progressive
    # ColorMatrix: adjusting color matrix from fcc to 2020ncl using transfer BT.709
    clip = core.std.SetFrameProp(clip,prop="_Transfer",intval=1)
    # adjusting luma range to 'limited' due to post clipping
    clip = core.resize.Bicubic(clip=clip, matrix_in_s="fcc", matrix_s="2020ncl", range_in=0, range=0)
    clip = core.std.CropRel(clip=clip, left=12, right=6, top=0, bottom=0)# cropping to 702x480
    # changing range from limited to full range
    clip = core.resize.Bicubic(clip, range_in_s="limited", range_s="full")
    # Setting color range to PC (full) range.
    clip = core.std.SetFrameProp(clip=clip, prop="_ColorRange", intval=0)
    # adjusting color space from YUV420P8 to RGB24 for vsCodeFormer
    clip = core.resize.Bicubic(clip=clip, format=vs.RGB24, matrix_in_s="2020ncl", range_s="full")
    # Blind Face Restoration using CodeFormer
    from vscodeformer import codeformer as CodeFormer
    clip = CodeFormer(clip=clip, upscale=1, weight=1.000) # 702x480
    # changing range from full to limited range
    clip = core.resize.Bicubic(clip, range_in_s="full", range_s="limited")
    from vsdpir import dpir as DPIR
    # adjusting color space from RGB24 to RGBH for vsDPIRDenoise
    clip = core.resize.Bicubic(clip=clip, format=vs.RGBH, range_s="full")
    # denoising using DPIRDenoise
    clip = DPIR(clip=clip, strength=5.000, task="denoise", device_index=0)
    # adjusting color space from RGBH to RGB48 for vsCAS
    clip = core.resize.Bicubic(clip=clip, format=vs.RGB48, range_s="full")
    # contrast sharpening using CAS
    clip = core.cas.CAS(clip=clip, sharpness=0.600)
    # Resizing using 10 - bicubic spline
    clip = core.fmtc.resample(clip=clip, kernel="spline16", w=1920, h=1478, interlaced=False, interlacedd=False) # resolution 1920x1478 before RGB48 after RGB48
    # adjusting output color from: RGB48 to YUV420P10 for x265Model
    clip = core.resize.Bicubic(clip=clip, format=vs.YUV420P10, matrix_s="2020ncl", range_s="full", dither_type="error_diffusion")
    # set output frame rate to 59.94fps (progressive)
    clip = core.std.AssumeFPS(clip=clip, fpsnum=60000, fpsden=1001)
    # Output
    clip.set_output()
    Which aside from the color matrix change, which I didn't do before, should be the same as I posted before.

    Keep in mind that to do this, you will need Hybrid, the torch-addon and a rather powerful NVIDIA GPU. CodeFormer only works with an NVIDIA gpu.
    (for DPIR can also be used through vs-mlrt to not require an NVIDIA gpu)
    Also, without understanding what I did there you will probably not be able to adjust this in case the settings cause problems in other clips/sections,...
    => This is the main difference between using Vapoursynth/Avisynth against Topaz tools. Yes, with enough knowledge and drive you can often produce better results than Topaz&Co, but without proper understanding you might not get any decent result. So depending on the source and what the goal is for tons of users using something like Topaz might be a good choice.

    Cu Selur
    users currently on my ignore list: deadrats, Stears555
    Quote Quote  
  22. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Thank you my friend @Selur, for the prompt response, and please apologize for the error in the application name, which I have already corrected .

    Do you mean to say that without an NVIDIA GPU I won't be able to achieve the same results that you achieved on your machine?

    Well, I have a modest home computer for home use, a Rizen 3 3220G with an ASROCK A320M-HD MOBO and 32GB of DD4 3200KHz KINGSTON FURY RAM.

    I'll try to see if I can follow your steps, or do you think that without an NVIDIA GPU I shouldn't even try?

    Anyway, thanks for the tips, and your patience and answers to my noobie questions, and may God bless you always for your altruism and time spent here with us.

    Att.

    Druid®.

    P.S. You encouraged me to study these commands and scripts, and if you have any tutorials to recommend, I would appreciate it in advance .
    Quote Quote  
  23. Do you mean to say that without an NVIDIA GPU I won't be able to achieve the same results that you achieved on your machine?
    Yes.
    I'll try to see if I can follow your steps, or do you think that without an NVIDIA GPU I shouldn't even try?
    CodeFormer is for the face enhancement/replacement.
    https://imgsli.com/MjIzMzAy

    Seeing your (rather low) computation power and the fact that you don't have a (powerful) NVIDIA GPU, I would not recommend trying this approach, since it will not be so great and also will be !terribly! slow.

    Cu Selur

    Ps.: no clue about a good Vapoursynth or Avisynth tutorial (the ones at the project pages are not that good)
    stuff like:
    https://www.l33tmeatwad.com/vapoursynth101
    https://iamscum.wordpress.com/_test1/_test2/
    https://guide.encode.moe/encoding/basics-and-workflow.html
    seems useful, but no clue how good they are for newbies.
    Last edited by Selur; 30th Nov 2023 at 12:00.
    users currently on my ignore list: deadrats, Stears555
    Quote Quote  
  24. Banned
    Join Date
    Nov 2022
    Search PM
    Originally Posted by Sharc View Post
    But there are pitfalls as well e.g. with the colormatrix (601 vs 709). If one is careless there is a 50% chance that things go wrong.
    Indeed, I still am not 100% what exactly I need to do when up-rezzing video in VDub: should I select Rec. 709 in the compressor, or should I use Convert Format filter, or something else.
    Image Attached Thumbnails Click image for larger version

Name:	compressor format.png
Views:	10
Size:	20.4 KB
ID:	75155  

    Click image for larger version

Name:	Convert Format filter.png
Views:	12
Size:	13.5 KB
ID:	75156  

    Quote Quote  
  25. Attached an example of using conventional (no NVIDIA card or ml stuff aside from NNEDI3 + can be done with Avisynth&Vapoursynth) filtering. (script: https://pastebin.com/vKLn0cRJ)

    Cu Selur
    Image Attached Files
    users currently on my ignore list: deadrats, Stears555
    Quote Quote  
  26. Member
    Join Date
    Mar 2008
    Location
    United States
    Search Comp PM
    Originally Posted by Selur View Post
    Attached an example of using conventional (no NVIDIA card or ml stuff aside from NNEDI3 + can be done with Avisynth&Vapoursynth) filtering. (script: https://pastebin.com/vKLn0cRJ)

    Cu Selur
    That's gone too far, it's clearly degraded the quality. (haloes, no real detail, etc)
    Perhaps 720p would be bettter - unless the 1080p takes priority based on the resolution alone?
    Quote Quote  
  27. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by Selur View Post
    Do you mean to say that without an NVIDIA GPU I won't be able to achieve the same results that you achieved on your machine?
    Yes.
    I'll try to see if I can follow your steps, or do you think that without an NVIDIA GPU I shouldn't even try?
    CodeFormer is for the face enhancement/replacement.
    https://imgsli.com/MjIzMzAy

    Seeing your (rather low) computation power and the fact that you don't have a (powerful) NVIDIA GPU, I would not recommend trying this approach, since it will not be so great and also will be !terribly! slow.

    Cu Selur

    Ps.: no clue about a good Vapoursynth or Avisynth tutorial (the ones at the project pages are not that good)
    stuff like:
    https://www.l33tmeatwad.com/vapoursynth101
    https://iamscum.wordpress.com/_test1/_test2/
    https://guide.encode.moe/encoding/basics-and-workflow.html
    seems useful, but no clue how good they are for newbies.
    I understand friend @Selur, although the time factor wouldn't be that stressful for me, for example, with my machine, in Topaz Video AI v.4.0.4, using DIONE TV, 1 h 24 min of Yanni's show - Tribute, it took 14 hours, 2.7 fps per second, do you think with everything you did this would take 24 hours or more?

    I look for the quality of these operations at all costs, with an UpScaling to 1080p, it would be very good, and I will have to pay the price of time, because of my modest configuration.

    Well, I'm continuing my journey and I'm going to take a look at the tutorials you gave me.

    May God always bless you for everything and thank you very much for your understanding and time in responding promptly.

    Att.

    Druid®.
    Last edited by DruidCtba; 30th Nov 2023 at 17:03.
    Quote Quote  
  28. Member
    Join Date
    Sep 2009
    Location
    Brazil
    Search Comp PM
    Originally Posted by davexnet View Post
    Originally Posted by Selur View Post
    Attached an example of using conventional (no NVIDIA card or ml stuff aside from NNEDI3 + can be done with Avisynth&Vapoursynth) filtering. (script: https://pastebin.com/vKLn0cRJ)

    Cu Selur
    That's gone too far, it's clearly degraded the quality. (haloes, no real detail, etc)
    Perhaps 720p would be bettter - unless the 1080p takes priority based on the resolution alone?
    Apparently yes, my friend @davexnet, as the degradation is visible, you could even notice some tremors, anyway, in the previous case my friend @Selur used specific commands for the NVIDIA GPU, if I'm not mistaken, and also the CodeFormer software, which adds considerable value to the faces, which for me made all the difference, anyway, apart from the problem with the musician's nose, generated by Topaz's DIONE TV, the rest of the show performed reasonably well for 1080p, which is the definition that look.

    Att.

    Druid®.
    Quote Quote  
  29. Member
    Join Date
    Mar 2008
    Location
    United States
    Search Comp PM
    Originally Posted by DruidCtba View Post
    Originally Posted by davexnet View Post
    Originally Posted by Selur View Post
    Attached an example of using conventional (no NVIDIA card or ml stuff aside from NNEDI3 + can be done with Avisynth&Vapoursynth) filtering. (script: https://pastebin.com/vKLn0cRJ)

    Cu Selur
    That's gone too far, it's clearly degraded the quality. (haloes, no real detail, etc)
    Perhaps 720p would be bettter - unless the 1080p takes priority based on the resolution alone?
    Apparently yes, my friend @davexnet, as the degradation is visible, you could even notice some tremors, anyway, in the previous case my friend @Selur used specific commands for the NVIDIA GPU, if I'm not mistaken, and also the CodeFormer software, which adds considerable value to the faces, which for me made all the difference, anyway, apart from the problem with the musician's nose, generated by Topaz's DIONE TV, the rest of the show performed reasonably well for 1080p, which is the definition that look.

    Att.

    Druid®.
    I agree, I thought Selurs previous version was very good. What is the casue of Topaz causing such damage to the musicians nose?
    Is it an aggressive setting somewhere?
    Quote Quote  
  30. @davenext: used TemporalDegrain2, using SMDegrain or simply adjusting the strenght of the filtering would probably be less destructive.
    @DruidCtba: did a quick testing on a Ryzen 9 9850 with an NVIDIA 4080 and got around ~2.5fps, so my system it would probably take a similar time that it took you on yours to run Topaz.
    users currently on my ignore list: deadrats, Stears555
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!