IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Using Wavpack for wireless transmission
GoEk
post Jan 27 2014, 16:39
Post #1





Group: Members
Posts: 4
Joined: 27-January 14
Member No.: 114234



Hi,
does WavPack contain functions suitable for transmitting and receiving primarily speach but also general audio over a wireless data link?
The stream must be "quick" in syncing to a new stream with low latency, as low as possible but less than 20ms delay between input and output signal so it can be used live in the same room.

Regards
Goran
Go to the top of the page
+Quote Post
bryant
post Jan 27 2014, 19:47
Post #2


WavPack Developer


Group: Developer (Donating)
Posts: 1287
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



WavPack might be suitable for such an application. The scalability of the lossy mode in particular might be useful. WavPack has been used in low-latency streaming applications before as described here. There is also a low-latency version available, although it relies on a robust communication link.

Interestingly, Broadcom has recently announced that WavPack is part of their WICED SDK for tranmitting HD audio over wifi, although I have no idea what that support entails (or whether it might even be a mistake/joke). I've always thought, however, that the lossy mode of WavPack would be an ideal transport for HD audio under varying communication conditions.

Feel free to contact me if you have further questions (or just post here), and good luck with your project!
Go to the top of the page
+Quote Post
GoEk
post Jan 27 2014, 23:17
Post #3





Group: Members
Posts: 4
Joined: 27-January 14
Member No.: 114234



QUOTE (bryant @ Jan 27 2014, 19:47) *
WavPack might be suitable for such an application...


Many thanks! Do you have any "feeling" for how high bitrate you would need for a "high quality" voice transmission? Would the math be possible to run on a "simple", low cost fixed point DSP?

Regards
Goran

This post has been edited by GoEk: Jan 27 2014, 23:20
Go to the top of the page
+Quote Post
saratoga
post Jan 28 2014, 00:03
Post #4





Group: Members
Posts: 4718
Joined: 2-September 02
Member No.: 3264



We use bryant's excellent fixed point wavpack encoder in rockbox if I'm not mistaken, and it is able to record to wavpack on very simple ARM processors. A real DSP should be more than enough I think, although I don't know how easy it will be to port to your specific DSP system.
Go to the top of the page
+Quote Post
bryant
post Jan 28 2014, 18:37
Post #5


WavPack Developer


Group: Developer (Donating)
Posts: 1287
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



QUOTE (GoEk @ Jan 27 2014, 14:17) *
Do you have any "feeling" for how high bitrate you would need for a "high quality" voice transmission?

In the WavPack test suite there is a folder called hybrid_bitrates that contains sample lossy files from 24 kbps to 1024 kbps that will give you a good idea of the quality potential of the lossy mode. Specifically, there are speech samples at 24 and 32 kbps. You may get slightly worse results than those depending on how you address your low-latency requirement, but they will give you an idea. It's obviously no where near as good as a dedicated speech codec (or Opus), but should be easier for a less powerful CPU.
Go to the top of the page
+Quote Post
GoEk
post Jan 29 2014, 12:15
Post #6





Group: Members
Posts: 4
Joined: 27-January 14
Member No.: 114234



QUOTE (bryant @ Jan 28 2014, 18:37) *
QUOTE (GoEk @ Jan 27 2014, 14:17) *
Do you have any "feeling" for how high bitrate you would need for a "high quality" voice transmission?

In the WavPack test suite there is a folder called hybrid_bitrates that contains sample lossy files from 24 kbps to 1024 kbps that will give you a good idea of the quality potential of the lossy mode. Specifically, there are speech samples at 24 and 32 kbps. You may get slightly worse results than those depending on how you address your low-latency requirement, but they will give you an idea. It's obviously no where near as good as a dedicated speech codec (or Opus), but should be easier for a less powerful CPU.


Thanks! Do you need to be a "DSP math expert" to use the lib's?
Go to the top of the page
+Quote Post
bryant
post Jan 30 2014, 21:02
Post #7


WavPack Developer


Group: Developer (Donating)
Posts: 1287
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



QUOTE (GoEk @ Jan 29 2014, 03:15) *
Thanks! Do you need to be a "DSP math expert" to use the lib's?

Absolutely not. Obviously being a proficient C programmer would be useful.
Go to the top of the page
+Quote Post
GoEk
post Feb 6 2014, 14:11
Post #8





Group: Members
Posts: 4
Joined: 27-January 14
Member No.: 114234



QUOTE (bryant @ Jan 30 2014, 21:02) *
QUOTE (GoEk @ Jan 29 2014, 03:15) *
Thanks! Do you need to be a "DSP math expert" to use the lib's?

Absolutely not. Obviously being a proficient C programmer would be useful.


Good! Programming C/C++ is no problems. I am thinking about lab evaluations, would a PC Windows application be suitable to "feed" WavPack with audio and then send it out on the audio port? Are they written in "plain/ANSI C" or do they require specific compilers/dev. environments?

This will be for a commercial product, are there any legal problems or requirements for that?

Regards
Goran

This post has been edited by GoEk: Feb 6 2014, 14:12
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 21st April 2014 - 13:12