Hi,
I have been doing some work on a fixed-point port and it is now in CVS.
All of the CPU-intensive parts should now be using only integer
operations. Now, it's in a state where even non-coders can help: find
bugs. With all these changes, there are probably several bugs and some
may show up only under specific circumstances. Please test to see if the
fixed-point behaves the same as the floating point, under all kinds of
modes, all kinds of files, mixing versions for encoder and decoder, and
so on. Please report any case where the new version behaves badly
compared to other versions.
Last thing: to enable the fixed-point, you need to compile the
FIXED_POINT defined (sorry, no configure option yet).
Jean-Marc
P.S. I'm also interested on speed reports on CPU's without an FPU.
(expect the fixed-point to run slower on x86)
--
Jean-Marc Valin, M.Sc.A., ing. jr.
LABORIUS (http://www.gel.usherb.ca/laborius)
Université de Sherbrooke, Québec, Canada
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 190 bytes
Desc: Ceci est une partie de message numériquement signée
Url :
http://lists.xiph.org/pipermail/speex-dev/attachments/20031008/bf13fc38/signature-0001.pgp
Hi, Is there any speed advantage to using the fixed point version on a PC with an Intel or AMD processor? An advantage would be that you don't have to convert the audio input from short to float but what about encoding/decoding itself? BTW, thanks for a great codec :-) Best regards, Chris <p>>> ----- Original Message ----->> From: "Jean-Marc Valin" <Jean-Marc.Valin@USherbrooke.ca> >> To: "speex" <speex-dev@xiph.org> >> Sent: Wednesday, October 08, 2003 8:48 AM >> Subject: [speex-dev] Fixed-point in CVS >> >> Hi, >> >> I have been doing some work on a fixed-point port and it is now in CVS. >> All of the CPU-intensive parts should now be using only integer >> operations. Now, it's in a state where even non-coders can help: find >> bugs. With all these changes, there are probably several bugs and some >> may show up only under specific circumstances. Please test to see if the >> fixed-point behaves the same as the floating point, under all kinds of >> modes, all kinds of files, mixing versions for encoder and decoder, and >> so on. Please report any case where the new version behaves badly >> compared to other versions. >> >> Last thing: to enable the fixed-point, you need to compile the >> FIXED_POINT defined (sorry, no configure option yet). >> >> Jean-Marc >> >> P.S. I'm also interested on speed reports on CPU's without an FPU. >> (expect the fixed-point to run slower on x86) >> >> -- >> Jean-Marc Valin, M.Sc.A., ing. jr. >> LABORIUS (http://www.gel.usherb.ca/laborius) >> Université de Sherbrooke, Québec, Canada--- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'speex-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> Is there any speed advantage to using the fixed point version on a PC with > an Intel or AMD processor? > An advantage would be that you don't have to convert the audio input from > short to float but what about > encoding/decoding itself?With the speed of x86 FPU's, I don't see how the fixed-point would be faster. Also, speaking of the short vs. float input, these changes (that will eventually go in 1.2) intruduce a small change in the API in that signals are how sent to functions (encode, decode, ...) as short instead of float. Jean-Marc -- Jean-Marc Valin, M.Sc.A., ing. jr. LABORIUS (http://www.gel.usherb.ca/laborius) Université de Sherbrooke, Québec, Canada -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 190 bytes Desc: Ceci est une partie de message numériquement signée Url : http://lists.xiph.org/pipermail/speex-dev/attachments/20031008/d897873c/signature-0001.pgp