Skip to content

Commit

Permalink
README: version: Update README and version -> v1.1
Browse files Browse the repository at this point in the history
Update the README to note we have moved to decimating the audio path.
Also bump the version, as this is quite a large change.

Add a version history table to help track changes.

Signed-off-by: Graham Whaley <graham.whaley@gmail.com>
  • Loading branch information
grahamwhaley committed Feb 16, 2021
1 parent a4529fd commit 955095b
Show file tree
Hide file tree
Showing 2 changed files with 18 additions and 9 deletions.
25 changes: 17 additions & 8 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -47,6 +47,10 @@ Honorable mentions then to:

and many others.

## Versions

A [history of versions](#version-history) can be found at the end of this document.

## Hardware

The basic hardware is built around a [Teensy 4.0][9] board along with its matching
Expand Down Expand Up @@ -79,14 +83,10 @@ A key feature, and a powerful feature of the audio library, is that inbetween th
can, and do, apply other processing to the data flow, such as using the [ARM CMSIS][11] functions or
open coding algorithms (as is done for some of the noise reduction code).

Presently the whole flow runs at the native audio library sample rate of 44.1kHz. This costs us
processing and data bandwidth overhead, but gives us a little simplicity, skipping any decimation
and interpolation steps. Later it may be necessary or desirable to move to a decimated datapath,
for reasons including:

- will take less processing time and data flow overheads
- will give us better filtering (FIR filters for instance) for the same number of 'taps'
- may improve the noise reduction, by allowing finer grained 'bins'
From Version 1.1 the audio path is run with a decimation of 4, bringing the path down to 11kHz sample
rate. Although this saves us processing overhead, the main goal was to 'focus' the noise reduction
algorithms to process just the audio spectrum (0-5.5kHz), rather than trying to process the whole
CD quality spectrum (0-22kHz), if we'd not decimated the native 44kHz sample rate.

### Display

Expand Down Expand Up @@ -221,6 +221,15 @@ dismissed. No point duplicating work.
All submissions should come via Github pull requests. Requests, bugs etc. should be reported via
Github issues.

## Version History

This section details the version 'releases', and what changed.

| Version | Date | Changes |
| ------- | ---------- | ------- |
| v1.0 | 2020-02-10 | Initial release |
| v1.1 | 2020-02-16 | Normalise FIRs. Decimate audio x4 |

[1]: https://github.com/DD4WH/Teensy-ConvolutionSDR "Teensy-ConvolutionSDR"
[2]: https://github.com/df8oe/UHSDR "UHSDR"
[3]: https://github.com/NR0V/wdsp "wdsp library"
Expand Down
2 changes: 1 addition & 1 deletion global.h
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@
#include <arm_math.h>

#define VERSION_MAJOR 1
#define VERSION_MINOR 0
#define VERSION_MINOR 1

#define VERSION_UINT16 ((uint16_t)((VERSION_MAJOR<<8)|VERSION_MINOR))

Expand Down

0 comments on commit 955095b

Please sign in to comment.