It's a decoder ("plug-in") for AVC (H.264) formats, that mostly replaces compoundplug.dll.
But something I read (that I now can't find) makes me think it's not just a self-contained decoder. I think it has the ability to pass streams off to other decoders depending on their specific content.
It doesn't replace QT but it does decode AVC streams within QT .mov containers.
My GUESS is that compoundplug may entail costs (e.g. from MainConcept?) in terms of royalties/on-going licensing and that so4compoundplug.dll avoids these costs because it has been developed in house? This is speculation and might be nonsense. Or it might be that compoundplug was somehow terminally limited, for example in terms of hardware acceleration.
Would love someone from the VEGAS team to confirm or deny that speculation, and also tell us what "so4" stands for, cuz I have no idea either.
If you have an established, successful workflow using QT Animation then for an easy life you could just enable Quicktime in VEGAS File I/O preferences and carry on. I doubt VEGAS will produce another native decoder that will read that format like they did for Quicktime AVC and ProRes (and HEVC?).
If you want to use a different codec for alpha, I'm afraid I'm not sure which to recommend. I rarely use alpha and can't remember which intermediate-worthy codecs support it in VEGAS.
Personally, I've not experienced audio sync issues with so4compoundplug.
Q1 : I’ve been using QT Animation to produce ALPHA. What’s the “informed” thinking, going forward?
I'm not a big fan. I tested QT Animation thoroughly back when I ran these tests in 2011, and panned it quickly as a candidate for true lossless encoders.
The RLE compression is terribly inefficient, except with low/no motion source. With full motion scenes, it is essentially no more efficient than uncompressed.
The encode is full of rounding errors, making the blue ramp vectorscope line "fat," meaning noisy with increased small-scale banding.
As an early attempt at "lossless" compression, it was a good idea in its day, but has been obsoleted by advanced intraframe encoders such as Magic (RGB), UT, and even Huffy. And of course it is no match for interframe encoders, although they usually don't support Alpha.
So in my workflow, it sits on the shelf as a dusty but well-earned trophy. No charge for the editorial.
I like the animation codec, but mostly because I use it for graphic animation, so rounding errors and fat vectorscopes don't show. Usually it is a text graphic or manipulating a logo. So don't write it off completely. The file size is small relatively, but the quality, especially text, holds up very well.
@Former user - Yes, me too. Unless @Musicvid has some concerning input regarding this too, I shall keep using it. No, my initial concerns was to do with the Windows 10 and QT support falling away and what alternatives I could use - easily😉 .
@Former user - Yes, me too. Unless @Musicvid has some concerning input regarding this too, I shall keep using it. No, my initial concerns was to do with the Windows 10 and QT support falling away and what alternatives I could use - easily😉 .
I got to thinking this morning (a rarity), and decided to rerun my tests that were almost a decade old. Using my generated banding test media, I was only able to see a little of the rounding errors I reported before. Certainly not enough to compromise the output in any significant way. So until I've had a chance to dive deeper, I hope you'll take those earlier observations with a grain of salt.
My personal choices for RGBA intermediates are either Magic or UT AVI, and when encoding camera video, UT 422, which comes in at only 1-2% of uncompressed size. Thanks for cueing me to take a second look.