Post Processing Guide
Back Page Decoding RF Data
Sub Page's Digital Tape Guide & Conventional Digitisation Guide
This doc goes over what and how you can post-process your video files.
This is a in-drafting doc not a final workflow doc as the export tool could integrate most common modern workflows.
You will most likely want to start at the Deinterlacing doc which is the first thing you need to know after producing some decoded data.
Note
Highlights information that users should take into account, even when skimming.
Tip
Optional information to help a user be more successful.
Important
Crucial information necessary for users to succeed.
Warning
Critical content demanding immediate user attention due to potential risks.
Caution
Negative potential consequences of an action.
This guide assumes you are using FFV1 or V210 (10-bit 4:2:2) files that are uncompressed or lossless compressed data, as you never want to in any capture chain go to lossy formats while in post production only when mastering files to lower quality "disro" or distributable copies should that be done for fixed standard systems or limited support systems such as web streaming or optical formats.
Important
Native camera media i.g ENG/Consumer/Sports you will want to use deinterlacing filters like QTGMC, ware as animated or film media you will want to use IVTC filters.
Playback tools
Media Server Tools
Metadata Tools
- MediaInfo
- MediaInfo Lite (Windows context menu quicktool)
- DVanalyse (Lossless Cut Supports Extracted Timecode Data from this tool)
Editing Tools (NLE - Non-Linear Editors)
- Davinchi Resolve (Full Production Editor)
- Lossless Cut (Basic Quick Editor, Handy for cutting and chapters)
Why do we love it?
- FFV1 In/Out support
- MKV Container/Wrapper support
- FLAC audio support (mostly, its buggy with it inside mkv files but fine on the timeline)
- It's 100% free for SD interlaced or progressive editing
StaxRip, Hybrid, Resolve
StaxRip - Your new best media friend
Pre-Made StaxRip Bundle V0.3 - Windows
StaxRip has FFmpeg, Avisynth+, Vapoursynth and over 40 tools for muxing, subtitling, and audio timing but one thing it has always provided is a decent visual cropping tool and simple in-cut and out-cut mode for the most basic of chop together editing.
It's not a 1-click encoder it's a 5 click and your ready for production encoder.
Crop - This lets you crop using a mouse wheel or dragging margin (to the pixel level of control)
Cutting - Preview --> Set your playing position --> HOME for the start point and END for stop point.
If you set multiple cut points it will chop the video up into one cut with those scenes.
There are templates for drag and drop de-interlacing & upscaling.
Profiles for NTSC & PAL with FFV1/ProRes/V210/V410 progressive & interlaced export are available.
PAL FFV1 to Progressive FFV1:
NTSC FFV1 to ProRes HQ Progressive:
Top Left - Input file & info about file
Top Right - Output file & info about file (Turns red on render finish)
Bottom Left - Your filters (manipulators for video)
Centre - Resolution & Aspect ratio control
Bottom - First 2 Audio tracks (4 Channels in 2 stereo tracks is all analogue has anyways)
If using profiles you only need to flag the aspect ratio correctly before export as this can change sometimes.
For 4:3 this will in most cases be DAR: 1.363636 / SAR: 1.25 / PAR: 12:11 inside StaxRip.
In FFmpeg directly this is -aspect 1.333333 for 4:3 for example
Breakdown:
1.333333 = 4:3 (Full-Frame with 4:3 as well)
1.777777 = 16:9 (Letterbox Crop or Anamorphic)
1.255555 = 5:4
1.555555 = 3:2 (Full Frame with 16:9 Media)
Hybrid, Resolve
Hybrid, Vapoursynth Direct, Resolve
Install Hybrid
First add the repository to your package manager
add-apt-repository ppa:djcj/hybrid
Update
apt-get update
Install Vapoursynth Editor
apt-get install vapoursynth-editor
Deinterlacing
25i & 29.97i
50p | 59.94p | 24p | 23.97p
Tools
Automated:
StaxRip, Hybrid
Manual/Advanced Scripting
AvsPMod
Topas Labs?
Just No it damages original details, and if you have watched the Stargate SG1 re-mastered Blu-Rays you wonder why people want to get there hands on the original Betacam tapes intended for broadcast use, they ruined the source material, there is a clear line between cleaning up sensor noise and utterly destroying the native look of actual film grain.
Vurnk11's personal denoiser project, that's fairly effective for NTSC.
| 3D decoding from ld-decode | AI Y/C separation + 2d chroma decoding | AI Y/C separation + 3D decoding on luma + 2D chroma decoding |
|---|---|---|
 |
 |
 |
For direct Blu-Ray & DVD mastering ProRes HQ is recommended and supported directly by Sony DVD Architect.
tbc-video-export --profile prores_hq example_decoded_media
ProRes HQ --> DVD Architect --> Master.iso --> dvdisaser Embed ECC Data --> Burn to Disc --> Verify --> Physically Store
DVD is limited to 6-9mbps, with M-Discs being really your only "high end" disc option left as DataLifePlus uses dyes sadly.
Blu-Ray is limited to 35-40mbps in standard mode, BDXL/UHD goes from 60-80mbps, with DataLifePlus being standard inorganic discs.
We are also working on manual encode DVD/BluRay profiles for both NTSC & PAL regions.
Sony DVD Architect is the last easy to get ''it works'' tool for creating compliant Blu-Rays takes ProRes HQ that is properly encoded with FFmpeg directly and does not make any compliancy errors that players or software players made in the last 10 years or so should run into.
We recommend use of genuine DataLifePlus/M-Disc/DM-Archive/Sony 128GB optical discs which are archival grade using high quality moulding and in-organic layers you can tell the quality by the perfectly smooth rims of the discs.
Firstly, the correct way to word a video involving the projects is to state its an FM RF Capture that is decoded with VHS-Decode, LD-Decode and or HiFi-Decode etc for example best not to conflate wording this causes no end of issues for new users.
Its always best to separate the hardware archival method from the current software generated video files when representing the projects.
RF capture files are the archive, not the decode results as they can always be improved so never be considered a "final".
Now that's all out of the way
To upload anything on YouTube you will need to upscale to the 2160p or "4k" bracket personally I have been using a 2880x2176 scale avoid horrible macro blocking due to forced re-compression.
There is 2 ways to do this Spline16 or Spline64 filters in StaxRip/Hybrid
Two FFmpeg scripts for this (copy paste into text document then change extension to .bat)
PAL
:: Upscale PAL video to the 4k bracket for YT use.
pushd %~dp0
echo Upscaling to 4k bracket
ffmpeg.exe -hwaccel auto -i "%~1" -pix_fmt yuv422p10le -vf scale=2880:2176 -c:v ffv1 -aspect 4:3 -vf bdwif auto -c:a copy -color_range tv -color_primaries bt470bg -colorspace bt470bg -color_trc bt709 "%~n1_YT_4k_Scaled.mkv"
echo Done.
PAUSE
NTSC
:: Upscale NTSC video to the 4k bracket for YT use.
pushd %~dp0
echo Upscaling to 4k bracket
ffmpeg.exe -hwaccel auto -i "%~1" -pix_fmt yuv422p10le -vf scale=2880:2176 -c:v ffv1 -aspect 4:3 -vf bdwif auto -c:a copy -color_range tv -color_primaries smpte170m -colorspace smpte170m -color_trc bt709 "%~n1_YT_4k_Scaled.mkv"
echo Done.
PAUSE
Note
Odysee uses direct streaming, no transcode/re-encoding so care has to be taken with client side playback issues, i.g non phone users.
If just doing quick or already synchronised source encodes then you can just use the tbc-video-export web profile in most cases.
You can pick either x264_web or x265_web depending on how much system/client support of streaming you want.
tbc-video-export --profile x264_web example_decoded_media
Some useful free tools to note for post processing for upscaling and deinterlacing.
*Note Hybrid has MacOS/Linux builds
DaVinci Resolve & Lossless Cut
These provide NLE or Non-Linear-Editing environments for quick cutting or full production mastering.
These give you basic editing to quickly handle uncompressed files across operating systems, and for Windows users an easy FFmpeg/AviSynth/Vapoursynth encoding and
QTGMC de-interlacing experience, and full colour grading and post-production ability.
Today de-interlacing is not a daunting task hard or complex just requires correct handling.
Interlaced video is 2 fields of information that are timed together to make a solid picture to the human eye.
Progressive video is a solid single frame followed by the next single frame, every single frame is its own picture.
Instead of wasting time like most YouTube videos will do, 90% of the time today you will use QTGMC, on Windows/MacOS use a dedicated tool, Linux your going to have to play with scripting.
Most people used to use dedicated scripts, but today script loading software like StaxRip combine multiple tools and have a very simple and easy to edit workflow alongside telling you what's broken when mistakes are made in a GUI manner.
If you have film media i.g somthing shot on 16-35mm and or cell frame animated you want IVTC in most cases.
Analogue is almost always top field first or TFF, ware as digital tapes are almost always bottom field first or BFF.
Camcorder Tapes are 25i PAL / 29.97i NTSC ---> Deinterlacing --> 50p PAL / 59.94p NTSC
TV Recordings & Mastered Media is 24fps telecine'ed to 25i/29.97i.
(60i or 30i etc does not exist for standard SD analogue colour format media, it's always 25i PAL 29.97i NTSC)
Movies, Anime etc are 12p and 24p content and can use the IVTC/Anime IVTC filters.
De-interlacing can be done via FFmpeg, StaxRip, Hybrid, or any Vapursynth/Avisynth based workflow.
FFmpeg has: W3DIF, BDWIF, NEDDI
Using only FFmpeg filters is considered very limited but a a good exercise in knowing what filters can be used
For digital video sources bwdif/w3fdif do a wonderful job on HD era content and there is also NEDDI
bwdif stands for "Bob Weaver Deinterlacing Filter"
Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic interpolation algorithms.
A high speed modern deinterlacer ideal for digital content
BDWIF FFmpeg Command Options
'mode' The interlacing mode to adopt. It accepts one of the following values:
'0, send_frame' Output one frame for each frame.
'1, send_field' Output one frame for each field.
The default value is send_field.
'parity' The picture field parity assumed for the input interlaced video. It accepts one of the following values:
'0, tff' Assume the top field is first.
'1, bff' Assume the bottom field is first.
'-1, auto' Enable automatic detection of field parity.
The default value is auto. If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
'deint' Specify which frames to deinterlace. Accept one of the following values:
'0, all' Deinterlace all frames.
'1, interlaced' Only deinterlace frames marked as interlaced.
The default value is all.
4:2:2 covers the entire colour bandwidth of Composite video
4:2:0 has universal device support pre-2016 era hardware.
HDR or BT-2020 supporting hardware 2016 onwards supports 10-bit 4:2:0 and higher end devices support 4:2:2.
-pix_fmt yuv422p - 8-bit 4:2:2
-pix_fmt yuv422p10le - 10-bit 4:2:2
-pix_fmt p010le - 10-bit 4:2:0
-pix_fmt yuv420p - 8-bit 4:2:0
Colour Spaces for NTSC/PAL and a TRC flag to Rec 709 is handled via tbc-video-export.
There are a couple of issues in NTSC-J to Rec 709 conversion:
The YCbCr<->RGB matrix transforms use different co-efficients
Kraz3yIvan — Today at 09:19
Yeah i read that Japanese NTSC never changed primaries and whitepoint to SMPTE "C", continuing to use the 1953 NTSC primaries and whitepoint.
Stephen_Neal — Today at 09:20
The RGB primaries used in NTSC-J and Rec 709 are different - so you need to map or tone-map between them.
You need to know what the actual NTSC-J RGB primaries are in CIE space - as the primaries in the spec may not be the ones actually used in the telecine that transferred the film in reality (and going to this level of detail may be impossible and send you mad)
Gamma... (Not really that well defined or stuck to)
NTSC-J lack of set-up (and whether it's been followed on the NTSC content you have) plus film stuff can often be given a 'look' (crushed blacks, raised blacks, gamma tweaked, colour cast, saturation change etc.) in the telecine intentionally by a director or colourist wanting a specific look. Without being in their heads 'correct' is difficult to define.
Back Page Decoding RF Data
Sub Page's Digital Tape Guide & Conventional Digitisation Guide
- FAQ - Frequently Asked Questions
- Diagram Breakdowns
- Visual-Comparisons
- VCR Reports / RF Tap Examples
- Download & Contribute Data
- Speed Testing
- Visual VBI Data Guide
- Closed Captioning
- Teletext
- WSS Wide - Screen Signalling
- VITC Timecode
- VITS Signals
- XDS Data (PBS)
- Video ID IEC 61880
- Vapoursynth TBC Median Stacking Guide
- Ruxpin-Decode & TV Teddy Tapes
- Tony's GNU Radio For Dummies Guide
- Tony's GNU Radio Scripts
- DomesDay Duplicator Utilities
- ld-decode Utilities