[PR #13] [CLOSED] lavf/rkrga: format map support #238

New issue

Closed

opened 2025-12-23 10:36:33 +01:00 by backuprepo · 0 comments

backuprepo commented

2025-12-23 10:36:33 +01:00

Owner

📋 Pull Request Information

Original PR: https://github.com/nyanmisaka/ffmpeg-rockchip/pull/13
Author: @hbiyik
Created: 1/21/2024
Status: ❌ Closed

Base: master ← Head: ffmpeg-rockchip-rga-formatmap

📝 Commits (1)

1426d33 lavf/rkrga: format map support

📊 Changes

1 file changed (+56 additions, -4 deletions)

View changed files

📝 libavfilter/vf_vpp_rkrga.c (+56 -4)

📄 Description

With this option, user can set dynamically the the output format of rkrga_scale & rkrga_vpp filters, according to the input format. User gives a pixel format map separated with + and and coupled with - sign. Ie: formatmap=nv15-nv12+nv20-nv16 means that if the in input pixel format is nv15, the output must be nv21, or if the input format is nv20 then the output format must be nv16. The mapping can be extended as wished.

fmt1=fmt2:fmt3=fmt4... s specifically because players already serialize the filter inputs with the same/similar syntax, therefore parsing of individual avoptions together with dict type avoption is not always possible due this restriction.

i have tested and good to go from my pov.

Some side note:

I think vpp scale and overlay filter can be seperated cleaner in the code.

Specificly:

seperate scale filter to a new c file.
Merge common context stuff scale,vpp,overlay to rgacontext
Use one single filtercontext containing rgacontext, vppcontext, scalecontext, and overlay structs in it.
And so, on. I did not get in to this task not no annoy you with more refactoring things, so this format map thing is only for vpp+scale, overlay has its own thing atm.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/nyanmisaka/ffmpeg-rockchip/pull/13 **Author:** [@hbiyik](https://github.com/hbiyik) **Created:** 1/21/2024 **Status:** ❌ Closed **Base:** `master` ← **Head:** `ffmpeg-rockchip-rga-formatmap` --- ### 📝 Commits (1) - [`1426d33`](https://github.com/nyanmisaka/ffmpeg-rockchip/commit/1426d337afd36548b307755d4803b619623e7842) lavf/rkrga: format map support ### 📊 Changes **1 file changed** (+56 additions, -4 deletions) <details> <summary>View changed files</summary> 📝 `libavfilter/vf_vpp_rkrga.c` (+56 -4) </details> ### 📄 Description With this option, user can set dynamically the the output format of rkrga_scale & rkrga_vpp filters, according to the input format. User gives a pixel format map separated with + and and coupled with - sign. Ie: formatmap=nv15-nv12+nv20-nv16 means that if the in input pixel format is nv15, the output must be nv21, or if the input format is nv20 then the output format must be nv16. The mapping can be extended as wished. fmt1=fmt2:fmt3=fmt4... s specifically because players already serialize the filter inputs with the same/similar syntax, therefore parsing of individual avoptions together with dict type avoption is not always possible due this restriction. i have tested and good to go from my pov. Some side note: I think vpp scale and overlay filter can be seperated cleaner in the code. Specificly: seperate scale filter to a new c file. Merge common context stuff scale,vpp,overlay to rgacontext Use one single filtercontext containing rgacontext, vppcontext, scalecontext, and overlay structs in it. And so, on. I did not get in to this task not no annoy you with more refactoring things, so this format map thing is only for vpp+scale, overlay has its own thing atm. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>