W3C home > Mailing lists > Public > public-webrtc-logs@w3.org > May 2019

Re: [mediacapture-main] Echo Cancellation (#597)

From: henbos via GitHub <sysbot+gh@w3.org>
Date: Thu, 23 May 2019 10:46:12 +0000
To: public-webrtc-logs@w3.org
Message-ID: <issue_comment.created-495167996-1558608371-sysbot+gh@w3.org>
It's debatable if it should be a "recommends", a "SHOULD" or a "MUST" to have echo cancellation apply to everything. But even if it's a "MUST" to _attempt to remove all the sound_, you can easily imagine a browser that _attempts_ to remove echo from WebAudio but it just isn't very good at doing so, and you end up with echo anyway. This relates to the problem that _how to do echo cancellation_ is outside the scope of the spec, and without any objective, testable requirements about what is or is not "good enough", it would be hard to tell a bad implementation from a spec-non-compliant implementation. The spec is vague here on purpose. We _could_ upgrade the language to a hard SHOULD, but SHOULDs are optional to implement.

Without major revisions to the spec, what this really boils down to is if the **Type** field on the chromium bug should say "Bug" or "Feature". Which does not necessarily dictate the **Pri** field.

I would focus my efforts on why it's a good idea to improve Chrome's implementation or how many would benefit from it rather than revisiting the spec's definition. In any case, it's up to the Chromium team how the Chromium team prioritizes bugs and features.

I'm suggesting to close this spec issue.

-- 
GitHub Notification of comment by henbos
Please view or discuss this issue at https://github.com/w3c/mediacapture-main/issues/597#issuecomment-495167996 using your GitHub account
Received on Thursday, 23 May 2019 10:46:13 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:22:23 UTC