Re: [LAD] Better lossless compressions?

From: A. C. Censi <accensi@email-addr-hidden>
Date: Wed Jun 23 2010 - 00:18:17 EEST

On Tue, Jun 22, 2010 at 5:31 PM, <fons@email-addr-hidden> wrote:
>
> Does it decompress to the original ?

The algorithm, patented by the way, is oriented to lossless
compression of HIGH SPEED signals, "New data compression technologies
can save power and cost in applications with sampling rates above 10
Msamples/sec." A lot above usual digital frequencies.

See:
http://www.dspdesignline.com/208808676;jsessionid=CEV13G2LQ4VWDQE1GHPSKHWATMY32JVN?pgno=1
Data compression for high-speed DSP, part 1

and
http://www.embedded.com/columns/technicalinsights/209400089?_requestid=198348
Data compression for high-speed DSP, part 2

In Part 2 it is compared with LZ algorithms, of PKZIP fame. Somewhat
better, around 50% better, probably at the same level of the
algorithms already used for sound.

It advantage, is efficient implementation, in power, instructions and
pin count of DPS/FPGA chips, used for real time compression of those
high-speed images.

Also see, if interested:
http://www.dsp-fpga.com/articles/id/?4458
Compressing ADCs defuses data rate explosion in Data Acquisition Systems

There is some use for lossy compression, like ogg and mp3, but it not
well explained in the article

Final words of Part 2:
"By integrating compression into high-speed signal processing systems,
end-to-end system performance is maintained with reduced pin count,
memory, power, and cost. Just as speech, audio, image, and video
compression are now ubiquitous and integrated components of consumer
media, we believe that compression for high-speed DSP systems will
someday become as ubiquitous as these low-speed compression algorithms
have become."

-- 
A. C. Censi
accensi [em] gmail [ponto] com
accensi [em] montreal [ponto] com [ponto] br
_______________________________________________
Linux-audio-dev mailing list
Linux-audio-dev@email-addr-hidden
http://lists.linuxaudio.org/listinfo/linux-audio-dev
Received on Wed Jun 23 04:15:02 2010

This archive was generated by hypermail 2.1.8 : Wed Jun 23 2010 - 04:15:02 EEST