- From: Daniel Weck <daniel.weck@gmail.com>
- Date: Thu, 7 Jul 2011 00:53:20 +0100
- To: "www-style@w3.org style" <www-style@w3.org>, fantasai <fantasai.lists@inkedblade.net>
On 6 Jul 2011, at 23:49, fantasai wrote: > I think you need to drop the 'normal' value. At least, I have no idea > what it would mean. 'normal' is equivalent to +0.0dB applied to the default/intrinsic voice volume (i.e. it doesn't apply to the inherited value). This is equivalent to SSML's: -- A child prosody element may use the label "default" to reset the current volume level. -- http://www.w3.org/TR/speech-synthesis11/#edef_prosody However I can see a different interpretation of "reset", if we consider that initial is "medium" (see below). In other words, we don't really have to care about the intrinsic voice volume, what matters is exclusively what the user wants (based on the keyword values that map to user preferences). > The initial value should be 'medium'. I agree that the initial value should be the user's preferred volume level (not the intrinsic voice level). http://dev.w3.org/csswg/css3-speech/#mixing-props-voice-volume > Then make the Computed Value line read something like > > Computed Value: a keyword and, if it is not zero, a decibel value > > Then the computed value is always a keyword plus a <decibel>, which > will represent any changes. Then make the value allow both a keyword > and a <decibel> to be specified simultaneously (using ||). What about changing the volume relatively to the inherited value? (i.e. no keyword specified)
Received on Wednesday, 6 July 2011 23:53:59 UTC