The Strøm open source speaker project have been started with the main goal of creating a fully functional "wifi speaker", but as an open sourced product.
Creative Commons Attribution Share Alike 4.0 International
103
stars
2
forks
source link
Dynamically load different - and custom - filter configurations #11
With the current design, the filter (DSP) configuration of the speaker are hardcoded within the software platform. This makes for compile time validation of the filters as well as very a performant implementation.
However it also makes changes to these filters hard, or at least cumbersome.
Solution
Move the filter configuration into a configuration file, just as it is done with the Camilla DSP (see example).
It should be possible to store several different filter configurations and to swap between them without recompiling the application. A small downtime/reboot of the software platform would not be a deal breaker.
A bonus feature would be to allow for custom filter configuration. Each user should be able to active their own filter configuration.
A rather important question still remains unanswered: how to ensure the integrity of the drivers and the AMP.
Alternatively the end user will need to recompile the application if they want to change the filtering. That isn't for everbody.
For information on the voicing process take a look at issue #6
Problem
With the current design, the filter (DSP) configuration of the speaker are hardcoded within the software platform. This makes for compile time validation of the filters as well as very a performant implementation. However it also makes changes to these filters hard, or at least cumbersome.
Solution
Move the filter configuration into a configuration file, just as it is done with the Camilla DSP (see example). It should be possible to store several different filter configurations and to swap between them without recompiling the application. A small downtime/reboot of the software platform would not be a deal breaker.
A bonus feature would be to allow for custom filter configuration. Each user should be able to active their own filter configuration.
A rather important question still remains unanswered: how to ensure the integrity of the drivers and the AMP.
Alternatively the end user will need to recompile the application if they want to change the filtering. That isn't for everbody.
For information on the voicing process take a look at issue #6