Assessing VSFilter for ASS Rendering #445
Comments
|
Pretty sure everything directshow is not compatible. Maybe we can use it for reverse engineering. |
|
Yes, or just use parts of the code if the license is compatible. Without having ever seen the code, what might stand on the positive side is that it's probably very Windows oriented code and API usage and things might be done in ways that can benefit from Windows APIs and acceleration techniques than libass. But I don't know how it renders and whether that would be compatible in a more elegant way than libass. |
|
The Windows API part is very important, because we cannot use APIs that's are not available in UWP. |
Oh..right! |
|
Just checked for libass. Here they are explicitly mentioning UWP, so the lights is green in that direction: |
|
Yes iirc they were supporting uwp for quite some time but the build system was meh so we waited until meson was supported. |
I just had a kind of Epiphany.... 😉
Seems I've done too much x-platform work in the past years. I notice that I've even assimilated the habit to rule out platform-specific or worse: Windows-specific solutions - would have never happened earlier.
So - the point is: there's another ASS rendering implementation which is generally considered equal to libass (even though there are some differences and some may disagree on both sides). The de-facto-standard authoring tool for ASS subs is Aegisub. And Aegisub allows to switch rendering betwen the two, while VSFilter is the default.
What is is? It's a DirectShow filter, development has ceased long ago, but there are several forks of it still alive.
Is it suitable for us? I absolutely can't say that. Haven't thought about it neither looked at it.
But it's definitely worth to take a quick look and think about it once.
The text was updated successfully, but these errors were encountered: