Jean-Marc Valin
Jean-Marc Valin
Just curious, does that fix or improve a particular behaviour. I was wondering whether 400 ms is still appropriate for longer packets or if we might want to make the...
The real question is why do you need the version at compile time in the first place?
There's someone working on CMake support. That being said, the idea is that getting the version from the header should *not* be needed. Instead, it's best to query actual features....
In release tarballs, the version is set without any script, so it shouldn't be a problem. When cloning, the version scripts should work. If someone makes a tar.gz of the...
You don't understand, the contents of the git repo does not include the version number *anywhere*. The version is always extracted from the git tags (and commits), which is the...
OK, back to the start. It seems like I don't actually understand your problem. What is it that a header version would solve? Especially considering that Opus 1.0 through 1.3...
OK, I think I get the problem. Now, the issue I see with your solution is how you handle older versions. You won't ever be able to see query version...
Is this an old version of Opus or does it occur with master?
Can you contact the original authors of the patch to see what they have to say about it. I don't really have any way to test that arch.
The values that get loaded with _mm_load1_ps() are of type opus_val16. When Opus if compiled as floating-point, then that type is defined as float, so there should be no problem....