Re: [mediacapture-main] Mandate implementations to support video frame subsampling (#723)

> It seems the spec should ask implementations to do a best effort to support the whole [min, max ] range.

I feel that it would be much better if the specific resolutions or ranges supported were provided.  As a web developer, I have to make a great deal of effort now to try to get the most raw original form of what is captured.  I don't want the user agent resizing, resampling, or doing anything to the video.

> If 30fps is natively supported, but 5fps is not by the camera, the User-Agent can reasonably implement frame rate adjustment instead of throwing an OverConstrained error.

I think having an OverConstrained error is much better.  Again, any sort of media processing happening transparently is harmful.  We want the cleanest form of whatever the video is, as close to the original as we can get it.

Of course, I have no objection to opting in to some sort of dynamic resize or frame interpolation, but I don't think it should be automatic.  Minimally, there needs to be a way to opt out.

-- 
GitHub Notification of comment by bradisbell
Please view or discuss this issue at https://github.com/w3c/mediacapture-main/issues/723#issuecomment-690391604 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Thursday, 10 September 2020 15:53:42 UTC