[Devel] port-props extension (rescued "extended port information" extension)

Stefano D'Angelo zanga.mail at gmail.com
Fri Nov 4 05:54:44 PDT 2011


2011/11/4 David Robillard <d at drobilla.net>:
> On Thu, 2011-11-03 at 00:11 +0200, Stefano D'Angelo wrote:
>> * pprops:absoluteValue predicate: when you use lv2:sampleRate you have
>> no way to specify absolute values. E.g., a sine wave oscillator might
>> have [ lv2:portProperty lv2:sampleRate ; lv2:minimum 0.0 ; lv2:maximum
>> 0.5 ; lv2:default [ pprops:absoluteValue 440.0 ] ] ; - I understand
>> this has type-related implications in RDF, so if you have better
>> ideas...
>
> I don't understand the purpose of this one.

When you use lv2:sampleRate you cannot specify values that are
independent of the sample rate, obviously... this would be just an
aesthetical thing to allow to have e.g., an oscillator with minimum
frequency 0, maximum frequency 0.5 * sample rate, default frequency
440 (independent of sample rate).

Another option could be to have pprops:absoluteMinimum,
pprops:absoluteMaximum, pprops:absoluteDefault .

>> * pprops:mute property - applied to at most one toggled control input
>> port to indicate a "don't produce output" parameter
>>
>> * pprops:bypass property - as pprops:mute, but for bypass (audio
>> output = audio input)
>
> Seems reasonable.
>
>> * pprops:dryWetBalance property - applied to one (or more?) control
>> input port(s) to specify that is a dry/wet balance port (range 0 to
>> 100 in %)... to make it generic we maybe need some kind of
>> this-audio-input-correpsonds-to-this-audio-output rdf thing (e.g.,
>> multi-channel mixer)?
>
> Seems reasonable.  We do, it's in port-groups

Well, upon further thinking, I think there are better ways to handle
mute and bypass in such ways that they actually can both avoid part of
the computation and allow for less buffering... so maybe more thinking
needed. Maybe the same is true for dry/wet.

>> * pprops:enableSmoothing - this is a very tricky one, it would be like
>> the mute/bypass properties and would indicate whether some smoothing
>> is applied to input parameters to avoid artifacts - maybe conflicts
>> with pprops:causesArtifacts?
>
> Not sure about this one...

... me neither actually.

Stefano



More information about the Devel mailing list