- From: jan-ivar via GitHub <sysbot+gh@w3.org>
- Date: Wed, 27 Jan 2016 21:58:29 +0000
- To: public-media-capture-logs@w3.org
It's hard to infer a general rule from the `echoCancellation` constraint, given how complicated it is, conflating both feature vs. value (boolean), and inherent-property vs. control-setting (there are mics with controllable hardware aec, mics with hardware aec that cannot be turned off, mics without aec, and browsers that do software aec), but I'll try. Say my system has two mics, `micA` with hardware aec ***that cannot be turned off***, and `micB` without aec. If a browser claims to support the `echoCancellation` constraint, and the browser 1. does ***not*** support software aec, then I would expect `devInfoMicA.getCapabilities().echoCancellation` to yield `[ true ]`, and `devInfoMicB.getCapabilities().echoCancellation` to yield `[ false ]`. 2. If it ***does*** support software aec, then I might expect `devInfoMicA.getCapabilities().echoCancellation` to yield `[ true ]`, and `devInfoMicB.getCapabilities().echoCancellation` to yield `[ false, true ]`. 3. If it ***does*** support software aec, but ***cannot glean hardware capabilities***, then I might settle for `devInfoMicA.getCapabilities().echoCancellation` to yield `[ false, true ]`, and `devInfoMicB.getCapabilities().echoCancellation` to yield `[ false, true ]`. (The third one is arguably a bug, but is where browsers find themselves atm, and some control is better than no control.) The fourth case, where the browser neither supports software aec, nor can tell the difference between my two mics, seems identical to not supporting the constraint. Claiming support here and returning all `[ false ]` values would make the fourth case is indistinguishable from (1) where neither `micA` nor `micB` support aec, undermining the value of the information for no good reason. -- GitHub Notification of comment by jan-ivar Please view or discuss this issue at https://github.com/w3c/mediacapture-main/issues/296#issuecomment-175880778 using your GitHub account
Received on Wednesday, 27 January 2016 21:58:35 UTC