[pulseaudio-discuss] Volume ramping APIs

[Tanu Kaskinen]

On Fri, 2016-06-24 at 11:37 +0900, 김정연 wrote:
> Dear all,
> 
> Recently, I noticed volume ramping patches are currently staging at below
> link:
> https://lists.freedesktop.org/archives/pulseaudio-discuss/2012-August/014351.html
> 
> I looks nice feature but, they didn't accepted to upstream.
> Anyone who share us any information or history? Or, any plans to apply them
> in future?

I talked about this topic with Sangchul Lee a while ago in IRC. I'll
copy the IRC log here, it should answer your questions:

<sangchul1011> dear all
<sangchul1011> I'm thinking of adding feature for fade in/out to module-role-ducking by using ramp APIs.
<sangchul1011> I think it would be good option for graceful movements when ducking and unducking.
<sangchul1011> how do you think about it?
<tanuk> sangchul1011: You mention "ramp APIs". Are you referring specifically to the Tizen volume ramping implementation?
<sangchul1011> tanuk: I meant pa_sink_input_set_volume_ramp() in sink-input.h
<tanuk> sangchul1011: That's not in upstream.
<sangchul1011> oh... is that?..
<tanuk> sangchul1011: Do you mean that does upstream have a volume ramping implementation? The answer to that is no.
<sangchul1011> tanuk: hm..ok, could you tell me why that feature is not merged to upstream? any reason?
<tanuk> sangchul1011: Jaska Uimonen submitted the patches years ago, and they happened to get ignored for no particular reason.
<tanuk> sangchul1011: One issue with the patches is that they add a client api, and I don't think that's a good idea (or at least necessary at first).
<sangchul1011> tanuk: ok.. how about not adding client APIs, but adding the core APIs.. if it is not harmful. I'm not gonna use the client APIs..
<tanuk> sangchul1011: Resubmitting the patches without the client API part might be a good starting point.
<sangchul1011> tanuk: if then, I can use the core API for adding new feature to module-role-ducking.. do I prepare for resubmitting? or some other guy?
<tanuk> sangchul1011: I doubt Jaska works on pulseaudio any more, so you'll have to do the resubmission. If you manage to get the code accepted, then yes, there's no problem with role-ducking using the ramping functionality.
<tanuk> sangchul1011: There are some complications related to ramping.
<tanuk> I'm not sure if the ramping patches handle rewinding correctly. If not, then that's one hairy problem to solve.
<tanuk> Also, previously the plan kind of has been to add a generic filter framework first, and implement the ramping using that.
<tanuk> I'm not sure that is really required, though.
<tanuk> Resubmitting the ramping patches should clarify the situation.
<sangchul1011> tanuk: hm..it seems having some issues....although it is good feature from functional point of view...
<sangchul1011> tanuk: surely, i'm pleased to resumit the patch.. but i'm not sure if I could figure out the problem.. could you tell me any hints for that? or is there anyone who can look into the issue?
<tanuk> sangchul1011: What do you mean by "the problem"? The possible rewinding issue?
<sangchul1011> tanuk: yes right, the possible issue you said..
<tanuk> sangchul1011: I don't remember the content of the patches exactly, and I don't know what the exact problem with rewinding is, if there even is any problem.
<tanuk> As I said, resubmitting should clarify the situation.
<tanuk> If there's some difficult problem, I can't point to anyone who would be willing to help, if you have trouble with it, unfortunately.
<tanuk> I can of course answer questions, but I'm not willing to take it on me to implement the feature.
<sangchul1011> tanuk: ok, is it ok to resumit the patch first? and if there's issue, i'll try to manage that. what do you say?
<tanuk> sangchul1011: Yes, that sounds good.
<sangchul1011> tanuk: ok, first of all, I'll prepare resubmit the patches except client APIs in april, after then adding the fade in/out feature to ducking module.
<sangchul1011> tanuk: thanks for your opinion.
<tanuk> sangchul1011: You're welcome :)

-- 
Tanu