[DAPM question] DAPM support for single input multiple outputs

[DAPM question] DAPM support for single input multiple outputs

[Qiao Zhou]

Hi Mark,

I have one question and need your suggestion. Thanks in advance.

In our DSP, one stream path is like this: There is one bit controlling 
whether a signal A goes to A_out or B_out. That's the reverse case of 
DAPM_MUX(multiple inputs, single output).

One way is to use two switches but it seems not good since two switches 
control the same bit. It's strange. Do you have any opinion about such 
requirement support?
-- 

Best Regards
Qiao

Re: [DAPM question] DAPM support for single input multiple outputs

[Lars-Peter Clausen]

On 03/09/2015 09:38 AM, Qiao Zhou wrote:
> Hi Mark,
>
> I have one question and need your suggestion. Thanks in advance.
>
> In our DSP, one stream path is like this: There is one bit controlling
> whether a signal A goes to A_out or B_out. That's the reverse case of
> DAPM_MUX(multiple inputs, single output).
>
> One way is to use two switches but it seems not good since two switches
> control the same bit. It's strange. Do you have any opinion about such
> requirement support?

There is currently no support for a demux. But it shouldn't be too hard to 
add one in the DAPM core, if you want to take a try at it.

A demux would behave very similar to a mux, except that you'd reverse 
sources and sinks. The only thing that is a bit tricky is that we only have 
one control element per path and currently it always belongs to the sink. 
For a demux it needs to be long to the source. This means you can't have a 
mixer directly connected to the same path as a demux and the core needs to 
make sure that that doesn't happen.

- Lars

Re: [DAPM question] DAPM support for single input multiple outputs

[Qiao Zhou]

On 03/09/2015 04:53 PM, Lars-Peter Clausen wrote:
> On 03/09/2015 09:38 AM, Qiao Zhou wrote:
>> Hi Mark,
>>
>> I have one question and need your suggestion. Thanks in advance.
>>
>> In our DSP, one stream path is like this: There is one bit controlling
>> whether a signal A goes to A_out or B_out. That's the reverse case of
>> DAPM_MUX(multiple inputs, single output).
>>
>> One way is to use two switches but it seems not good since two switches
>> control the same bit. It's strange. Do you have any opinion about such
>> requirement support?
>
> There is currently no support for a demux. But it shouldn't be too hard to
> add one in the DAPM core, if you want to take a try at it.
>
> A demux would behave very similar to a mux, except that you'd reverse
> sources and sinks. The only thing that is a bit tricky is that we only have
> one control element per path and currently it always belongs to the sink.
> For a demux it needs to be long to the source. This means you can't have a
> mixer directly connected to the same path as a demux and the core needs to
> make sure that that doesn't happen.
I'll have a try to specially handle demux type. Thanks for the 
suggestions. BTW, is there plan to add demux support?
>
> - Lars
>


-- 

Best Regards
Qiao

Re: [DAPM question] DAPM support for single input multiple outputs

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 459 bytes --]

On Mon, Mar 09, 2015 at 07:16:18PM +0800, Qiao Zhou wrote:

> I'll have a try to specially handle demux type. Thanks for the suggestions.
> BTW, is there plan to add demux support?

Not really - it's not that there's any reason not to do it, it's just
that nobody is doing the work as far as I know.  It's quite unusual to
see it in hardware (usually you see mixers in the places where you might
see a demux) so few people have the motivation to do the work.

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: [DAPM question] DAPM support for single input multiple outputs

[Qiao Zhou]

On 03/09/2015 08:20 PM, Mark Brown wrote:
> Not really - it's not that there's any reason not to do it, it's just
> that nobody is doing the work as far as I know.  It's quite unusual to
> see it in hardware (usually you see mixers in the places where you might
> see a demux) so few people have the motivation to do the work.
Understood and I agree that it is rare. Thanks for clarifying.

-- 

Best Regards
Qiao