Three new SDRs in the shack

[Terry N4TLF]

Sort of agree on both points.

The BooyaSDR works on raw RF samples from a single ADC, not audio samples.  However, without compression (lossless or lossy), any file of raw data samples will be of the same size, independent of what values of those samples.  Whether an audio file has NO sound, or full volume sound (to the ADC’s limits), the file size will be the same – given the bit resolution, sample rate, and length of recording.  Number of channels does also affect the file size as a multiplier, but I was assuming raw RF samples from an SDR, since that was the subject of my original email.

I did mention that NO (fancy) compression was assumed in my comments, again as that’s what the SDR discussed produces.  Compression can indeed affect file size.  Lossless compression implies that every bit of the original material can be totally reproduced.  Lossy compression provides a more compressed final product, but every single bit cannot be faithfully reproduced.  For example, with audio file lossy compression, physco-acoustic (sp) techniques are usually implemented.  Such as if there is a real loud sound, the ear/brain cannot detect low-level changes immediately before or after that loud sound.  There are a lot of other techniques.

I’m not familiar with similar compression techniques applying to a single-channel stream of raw RF data samples, but there may be.  Lossless compression, such as used with zip, rar, and other binary file techniques would certainly work, post recording.

Keep in mind that with the BooyaSDR, a common off-the-shelf computer will have trouble processing the raw data samples.  100 million samples per second, 16-bits per second, results in 200 million bytes per second, or 1.6 BILLION (giga?) bits per second.  Gigabit Ethernet is not good enough at that point.  I think.  My math is not great.  Hence the use of USB-3 (USBss).
73, Terry, N4TLF



From: Phil via Tacos 
Sent: Tuesday, May 30, 2017 1:29 PM
To: tacos 
Subject: Re: Three new SDRs in the shack

Fourth factor - data compression? 

Phil M1GWZ


  On 30 May 2017, at 18:19, Joe palsa via Tacos <tacos at amrad.org> wrote:

  Food for thought------- 
  The size of an audio file is determined by three factors: sampling rate, bit resolution and channel count.

  Let's take an example audio file which was recorded at 44.1kHz, mono, and 16 bit resolution. The bit rate for this file is calculated as: 44100 x 1 x 16 = 705,600 bits per second. Divide that 705,600 by 8 and you've got 88,200 bytes per second. Divide that 88,200 by 1024 and you end up with 86 kilobytes per second. Multiple that by 60 and you get 5167 kilobytes per MINUTE. Or 5.047 MB per minute.

  The audio with more activity I would expect to require more MB.


  73's
  Dr Joseph Palsa k3wry
  ARRL-Virginia Section Manager
  Virginia State Government Liaison
  K3WRY at ARRL.ORG
  804-350-2665

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.amrad.org/pipermail/tacos/attachments/20170530/46ed6637/attachment.html>