search for: mdct_backward

Can anybody point me at any resources that would explain how to optimize mdct_backward for a cpu with a fused multiply-accumute unit? >From what I understand from responses to my older postings, Tremor's mdct_backward could be rewritten to take advantage of a muladd. My target machine can do either two-wide 32x32 + Accum(64) -> Accum(64) integer muladd or eight-wide 16x16...

Hi all, I'm attempting to convert all the pointers to arrays the mdct_backward function so it can be partitioned off for a hardware implementation. Although this code is quite short I'm finding it a little tricky. As it stands, mdct_backward is passed values by reference i.e. void mdct_backward(mdct_lookup *init, DATA_TYPE *in, DATA_TYPE *out) o my modified version s...

...other topic, this is actually a reply to Segher's post on Sun Oct 22 on the 'mdct question' topic. I wasn't subscribed properly and so I didn't get email confirmation and thus can't add to that thread. So Segher, if the equation is indeed what you say it is, then replacing mdct_backward with this version should work, but it doesn't. Am I applying it incorrectly? void mdct_backward(mdct_lookup *init, double *in, double *out) // I call this with in != out { // For 256 size blocks, coef are in[0..127] // For 2048 size blocks, coef are in[0..1023] int N = init->n...

Hi vorbis-dev, I'm investigating the mdct* function in libvorbis and writing a small client program to test it. I found that I have to set ARRAYSIZE when mdct_init(lookup, ARRAYSIZE) to minimum value of 62 otherwise it will segfault with mdct_backward. Is there any lower limit or it is more likely that my code has bug? Thank you very much, Pattara -- Please avoid sending me Word or PowerPoint attachments. See http://www.fsf.org/philosophy/no-word-attachments.html ---------------------------------------------------------------------- Ott Patta...

...r kicks, I re-enabled _LOW_ACCURACY_ and lo and behold it was still substantially faster. I also got some gains out of tremor by changing the longs in cookbook and sharedbook to ogg_int32_t's like I did for vorbis. I think _LOW_ACCURACY_ is a win mostly because I'm entirely cache-bound in mdct_backward. But my question is, is it good enough for production work? I diffed a few output files and it looked like the largest differences in sixteen bit data were not more than one or two lsb's. Are there any situations where it fails spectacularly I should watch out for? Looking at the lowmem-bra...

...D is set, then it begins 16384 0 16364 -803 16305 -1605 which is simply a left shift of TRIGBITS (14 in this case), or a multiply by 16384 if you prefer to think of it like that. So far, so good. However if I display the data blocks (*in and *out) before and after transformation: mapping0.c: mdct_backward(b->transform[vb->W][0],pcm,pcm); in floats: 'in' begins (before transformation) -0.003689 -0.017678 -0.041495 -0.044482 -0.011444 (although it goes a bit odd later on, with long strings of zeros and then long strings of -431602080.000000) 'out' begins (after transformation...

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi! I want to encode wav file into fixed window length vorbis file. I don't know if "window" is good name for that, but i mean MDCT buffer size. This buffer is passed to mdct_backward() as "in" and it's size is determined by "init->n" variable which is also passed to this function. So, when I say "window size" I mean "init->n" value. I noticed, that when decoding ogg vorbis file, window size is changing and it's values for no...

Viswanath Puttagunta wrote: > Can we please have review on RFCv2? We have quite a few optimizations > (Eg: ifft/mdct_backwards, fixed point fft/ifft mdct_forward/backward > etc) that are in my pipeline that depend on this patch series being > accepted. So, trying to make progress on this... On an armv7l board running Ubuntu, you've broken the build with just --enable-intrinsics (without the NE10 options):...

...nt > optimizations? Also, are any of your patches using NE10 library? > > So far, what I submitted is to optimize opus_fft using NE10 (float > only). I also modified mdct_forward to take advantage of this > optimization. > > My next action plan: > - use opus_ifft to optimize mdct_backwards (float via Ne10) > - Re-work/add opus_fft and opus_ifft for fixed (again via Ne10) > - Enable for Aarch64 and test (I'm hoping your patch already does this) > - float-to-int and int-to-float conversions using neon > - etc. > > > > > >> On Mar 3, 2015, at 5:18 P...

...- _book_maptype1_quantvals _book_unquantize __bss_start drft_backward drft_clear drft_forward drft_init _edata _end _fini _float32_pack _float32_unpack floor0_exportbundle floor1_encode floor1_exportbundle floor1_fit floor1_interpolate_fit _floor_P _init _make_words mapping0_exportbundle _mapping_P mdct_backward mdct_clear mdct_forward mdct_init ov_ilog res0_free_info res0_free_look res0_inverse res0_look res0_pack res0_unpack res1_class res1_forward res1_inverse res2_class res2_forward res2_inverse residue0_exportbundle residue1_exportbundle residue2_exportbundle _residue_P _ve_envelope_clear _ve_envelope...

Lancer 20060807 (based on aoTuV 4.51) http://homepage3.nifty.com/blacksword/ 2006/08/07 Lancer 20060807 Changes: * fix SSE optimization code in mdct_forward, mdct_backward. * reintroduce vorbis_oggpack_write for only static version. * fix problem in local_book_besterror_dim1x4. 2006/08/06 Lancer 20060807(test version for memory problem) Changes: * Since heap memory error occurs, remove vorbis_oggpack_write. that's optimization move to oggpack_write. * oggpack_...

Hi Vorbis, Now I get libvorbis-rc3 + integerized mdct_backward + integerized window running. The sound output is ok. Now we have to go further on possible optimization of the code in order to have the music played realtime on our chip (LEON-Sparc). There are 2 possible ways: 1.Make everything integer, remove all floating point usage. Then we can remove FPU an...

Hello, I have used "MDCT_INTEGERIZED" version of mdct and the resulting bitstream is absolutely silent. Do I have to make some other changes to the source code to get it working? I mean some int->float or float->int conversion before/after call a mdct function. Thanks Rado --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage:

...--host=arm-linux-gnueabihf --enable-intrinsics \ --with-NE10-libraries=$NE10_DIR/build/modules \ --with-NE10-includes=$NE10_DIR/inc $ make I have also tried $ ./configure --host=arm-linux-gnueabihf --enable-intrinsics It works fine to me. > I will start working on mdct_backward and fixed point fft for ARMv7. Glad to hear that. > Regards, > Vish Regards, Phil Wang -- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not dis...

mdct.c, lines 436 onward (part of mdct_backward) state { DATA_TYPE *oX1=out+n2+n4; DATA_TYPE *oX2=out+n2+n4; DATA_TYPE *iX =out; T =init->trig+n2; do{ oX1-=4; oX1[3] = MULT_NORM (iX[0] * T[1] - iX[1] * T[0]); oX2[0] = -MULT_NORM (iX[0] * T[0] + iX[1] * T[1]); oX1[2] = MULT_N...

Hello opus-dev, Can we please have review on RFCv2? We have quite a few optimizations (Eg: ifft/mdct_backwards, fixed point fft/ifft mdct_forward/backward etc) that are in my pipeline that depend on this patch series being accepted. Regards, Vish On 4 February 2015 at 12:21, Viswanath Puttagunta <viswanath.puttagunta at linaro.org> wrote: > Changes from RFC PATCH v1: > - passing arch paramet...

...anybody interested in cleanly separating encode and decode in vorbis? I'm trying to take a stab at it myself from the head of cvs as of this morning, but it looks like a lot of files contain both encode-only and decode-only functions. An obvious one is mdct.c, which contains mdct_forward and mdct_backward (which isn't that big of a deal, since both are pretty well-contained, but it could be easily fixed by moving all of the helper functions into a header file). A more complicated one is block.c, which contains several calls into psy.c, which I would presume should be mostly encode-only. I'...

Hello all, I have recently ported Vorbis to a Texas instruments DSP (TMS320C5510), The problem I have is that it is not decoding the audio in real time( about 50% of real time). Could any one give me a clue in which area I could start to optimize the code. Regards Kevin <p><p><p>--- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage:

On Mar 3, 2015, at 11:08 PM, Viswanath Puttagunta <viswanath.puttagunta at linaro.org<mailto:viswanath.puttagunta at linaro.org>> wrote: On 3 March 2015 at 21:59, Jonathan Lennox <jonathan at vidyo.com<mailto:jonathan at vidyo.com>> wrote: Viswenath, My patch should be against the tip, but it?s the very recent tip, including some changes this past Friday (27 Feb). I

On 25 February 2015 at 19:54, Timothy B. Terriberry <tterribe at xiph.org> wrote: > Viswanath Puttagunta wrote: >> >> Can we please have review on RFCv2? We have quite a few optimizations >> (Eg: ifft/mdct_backwards, fixed point fft/ifft mdct_forward/backward >> etc) that are in my pipeline that depend on this patch series being >> accepted. > > > So, trying to make progress on this... > > On an armv7l board running Ubuntu, you've broken the build with just > --enable-intrins...