On 15/04/13 06:53, Kristoff Bonne wrote:
> Stuart,
> 
> On 14-04-13 06:52, Stuart Longland (VK4MSL) wrote:
>> On 14/04/13 13:48, Albert Cahalan wrote:
>>>> I don't understand this. Both the GMSK signal and the voice signal are 
>>>> mono.
>>> Both signals are mono, so they can fit on a single stereo device.
>>> One gets the left, and the other gets the right. This means you
>>> don't need to power two devices and have room for two devices.
>> In a perfect world, yes, I'd agree with you.  It can be done all on one
>> audio device.
>> In fact, it can be done with a single mono full-duplex device.  Just add
>> a couple of CD4053s which switch the input and output between headset
>> and radio along with the PTT.  This is extra complexity in the audio
>> interface.
> Interesting idea.
> 
> Once we have the whole setup under control, we can look for this as a 
> next step.
> 
> However, do keep in mind that the mixer-settings of the two ports (radio 
> and voice) can be quite different; so it might require quite a bit of 
> "magic" needed to get this all running correctly.

Another way to work around this would be hardware gain controls, just
put potentiometers on the radio and headset outputs.

Plus, there's `alsactl`, we can just tell it to store the mixer settings
to a temporary file, and ask it to recall the settings for transmit...
or we introduce portmixer to FreeDV which could be advantageous as well.

> In the end, it might even be better to get somebody to do a open source 
> design of a raspi / beaglebone / pcduino / ...  addon board as radio 
> interface + mic interface. What we need is that that we have a device 
> where people can just plug in a handmic, plug in the other side into the 
> radio (discriminator port) and get going.
> 
> So, if somebody with better hardware skills then me feel an urge to give 
> this a try, please do drop a message in the list.

The Raspberry Pi model B mk 2 has a port which exposes all I²S signals.
 You could hook it to one of these:

http://littlebirdelectronics.com/products/codec-shield
which is intended for Arduino, but I see no reason why it wouldn't work.

>> USB audio devices with stereo inputs are not nearly as common.  Do you
>> really expect the average user to be able to spot the difference from
>> the packaging alone?
> True.
> 
> When doing experiments, I found out a number of odd things about USB 
> audio-dongles:
> - some do not support stereo capture
> - some say they support mono playback, but -in fact- do not

No, they don't, it's Windows that supports up-conversion.  i.e. we need
to use the plug: interface in ALSA.

> - quite a lot of them only support 44k1 and 48k sampling. Dongles 
> supporting 8 Khz sampling are in fact pretty rare

Yep, once again, they "support" it by leaning on the OS driver to do the
rate conversion.  Plus in David Rowe's experience, many are pretty
rubbish at it even when they do support 8kHz.

> As albert has indicated, it would be interesting to create a "supported 
> hardware" page. But I think it is up to "the community" to do this. What 
> I will do is set up a wiki so that everybody can update this information 
> him/herself without my involvement.

Indeed, it's a good thing to do.  Gives the experimenters a starting
point.  But we shouldn't stop people doing their own thing if they
choose.  The good news is they'll have the code if they want to beat
their own path through the bush.

>> I don't see a problem with using two USB audio devices provided the
>> underlying hardware can support it.
> Well, today I've done an end-to-end test with two raspberry pi's which 
> did work pretty good, each using two usb audio-dongles.
> The combination of forceing the speed of the USB bus to 12 Mbps (USB 
> 1.1) and using half-duplex on the audio-dongles seams to do it. 
> PTT-switching is done via GPIO (not ideal but good enough for a start).

How difficult is it to achieve the force to 12Mbps?  I'm looking at
putting together a transceiver using the Rpi's GPIO4 pin as the
reference for a local oscillator, using the aforementioned "CODEC
shield" as the I/Q interface to the analogue frontend and USB audio for
microphones/headsets.

It'd be nice to be able to plug more than one headset in and have them
both "playback" the received audio.  Do we know why full-duplex
saturates the USB?

> I just need a couple of days to incorperate the changes I did today into 
> the code on github and write a "installation howto".

This is great news.  I look forward to seeing it. :-)
-- 
Stuart Longland (aka Redhatter, VK4MSL)

I haven't lost my mind...
  ...it's backed up on a tape somewhere.

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
_______________________________________________
Freetel-codec2 mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

Reply via email to