ISSUE-47: (Bugzilla 17340) DelayNode smoothing
DelayNodeSmoothing
(Bugzilla 17340) DelayNode smoothing
- State:
- CLOSED
- Product:
- Web Audio API
- Raised by:
- Philip Jägenstedt
- Opened on:
- 2012-05-15
- Description:
- The spec requires that "the implementation must make the transition smoothly" but does not define how.
For example, if the delay value is a sinusoid (AudioNode connected to AudioParam) or an even a linear ramp (linearRampToValueAtTime), what must happen? If the value changes fast enough, does the sound play backwards?
Related to https://www.w3.org/2011/audio/track/issues/42 - Related Actions Items:
- No related actions
- Related emails:
- [web-audio-api] (DelayNodeSmoothing): DelayNode smoothing (#79) (from notifications@github.com on 2013-09-11)
- Re: Limit in sample or ms for circular routing (Re: Minutes of Audio WG Teleconference, 2012-05-30) (from philipj@opera.com on 2012-06-04)
- Re: Summary of issues (from crogers@google.com on 2012-05-22)
- Summary of issues (from mage@opera.com on 2012-05-22)
- Audio-ISSUE-47 (DelayNodeSmoothing): DelayNode smoothing [Web Audio API] (from sysbot+tracker@w3.org on 2012-05-15)
Related notes:
Agreed, there are several nuances relating to this. There are several interpolation approaches (linear, cubic, sinc) and also whether to interpolate delay changes using a "doppler" approach where the delay times are interpolated, or if the interpolation is a "cross-fade" interpolation. EA has asked for the "cross-fade" technique in addition to what I consider to be the standard "doppler" technique.
I have no strong opinion if we should include the "cross-fade" approach right now, or hold that technique off for a later version of the spec (where we can add an .interpolationTechnique attribute (which will default to the standard "doppler") thus remaining backwards-compatible
Setting as OPEN.
Olivier Thereaux, 30 May 2012, 20:22:19[MikeSmith]: https://www.w3.org/Bugs/Public/show_bug.cgi?id=17340
5 Jun 2012, 13:53:51Display change log
