Author Topic: Audio Streamer  (Read 3380 times)

john k2ox

  • Full Member
  • ***
  • Posts: 110
Audio Streamer
« on: September 25, 2013, 10:26:20 AM »
There is a very high quality audio shield sold by Sparkf## that has a 24bit audio codect.  It uses SPI to communicate and there is already a sketch for it.

So,  I'm tinking steaming 2ch high quality audio from any source (internet radio) to anyplace on your propery.

Any takers?  It will be a while before I can get to it.

My main interest is that it could be used for a very high dynamic range remote A/D to transmit measurement data.

LMK,
john

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5938
  • Country: us
    • LowPowerLab
Re: Audio Streamer
« Reply #1 on: September 25, 2013, 12:06:57 PM »
Audio streamer over Moteino/RFM69?
So would you just use it to measure A/D and transmit periodic packets or actually stream audio?
For streaming I'd be concerned that it would be difficult with the current packet mode of the lib.
RFM69 has a continuous mode which I think could be more suitable for streaming...

john k2ox

  • Full Member
  • ***
  • Posts: 110
Re: Audio Streamer
« Reply #2 on: September 25, 2013, 12:34:47 PM »
Yes it would have to be in Cont mode.  But, that's easy.  You did the hard part w/ handshaking. 

It's normal use is to take 10s of Ksps.  For measuring a voltage you could average a ton of samples or just take one.

They are CODECS, essentially a sound card on a chip.  Apple uses the same chip in the I phone.

Think of it.  A baby monitor with AES encription.  No Google or snoopy neighbors listening!

A secure walkie talkie. ...

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5938
  • Country: us
    • LowPowerLab
Re: Audio Streamer
« Reply #3 on: September 25, 2013, 12:50:49 PM »
Yeah I thought about using cont mode for streaming images from a JPEG camera that detects movement.
Also thought about cont mode for using an indoor wireless intercom - actually this one would be sweet. I guess cont mode could be pretty easy for this one since dropped packets could be tolerated a lot better than in packet-mode.... you just get some noise or lost bits of sound. Doesn't even need to be high quality, I think land phone is some low baud figure right? Im sure cont mode could handle that...
I'm not sure about streaming continuous video though ... it's really hard to say/guess/hope when it hasn't been done. But at least the sound idea could actually work pretty well. Maybe that's a first step to see if streaming a decent sound line is feasible...

john k2ox

  • Full Member
  • ***
  • Posts: 110
Re: Audio Streamer
« Reply #4 on: September 25, 2013, 01:35:20 PM »
Yep you'd just get a some distortion. Just like a cphone.  For decent voice you can get by with 3Kbps. With compression about 1Kbps.

Thats one of the reasons I'm trying to solve the freq stability issue.  I want to have a lot of radios doing different things all at the same time.  That means only using the BW thats neccessary and that means the tx has to remain in the middle of the rx BW.

As for the camera, I don't need realtime video.  It could take a snapshot due to some event and then take it's time sending it to me at whatever rate I choose.  If compression is used, it only sends the difference in the scene.  If the 'door' doesn't move no new data.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 5938
  • Country: us
    • LowPowerLab
Re: Audio Streamer
« Reply #5 on: September 25, 2013, 02:20:20 PM »
Right, well I'll keep an eye for your stability fix. I can't change the hardware. And doing a lot of calibration is also not very ideal. So the simplest solution is always the most elegant and transparent to the user. If a manual trigger is necessary from time to time that's great also.
Ultimately I think about 20-30 nodes in a house is a decent number. Reducing the BW would be beneficial, as you mentioned.

bstott

  • Newbie
  • *
  • Posts: 37
Re: Audio Streamer
« Reply #6 on: September 26, 2013, 08:36:28 AM »


I am seeing a trend? I think you may be overly critical about some of the problems (potentially theoretical) rather than continuing with the Hacker/Maker mentality. Do not start limiting yourself too much by being too smart with experience and maybe some edumacation. Remember - Education comes AFTER the Innovator Explores, Invents, Creates, Builds and IT all becomes REAL. Then, there comes education to teach the masses the story and give us the How-Tos.  --- See Apple in the early 80's. Paragliding in the late 80's. Cell phones were several pounds and wired to the car or a carry around box. Hang Gliding in the early 70's. Internet in the Mid-90's and all the other cool things we have.

What am I saying to me actually? I'm saying John's idea works. How do I know? The chip exists and is sold. The TV works in my home - with an antenna. My AM/FM radios work, so too the cell phones and more.

I'd suggest that the chip be hooked up. The code be written and the devices connected. Then, as in the development of all early explorations and technology, fix and refine. Oh, so what if packets get dropped in the audio/video streaming. We are not expecting HD for a Theater. So what if it looks like BW from the 50's at first. Actually it WILL be better. Why? COLOR! LOL....

Be a Maker! GO FOR IT!  ;)