similar to: fir_mem_up filter question

Hi JM, At line 221 in celt_lpc.c (the celt_iir function) I think you really want the RESTORE_STACK statement to be before the #endif instead of after it. Also, I couldn't help notice that your SSE code for xcorr_kernel reads more than "len" elements of "_x". I don't know if that's really a problem when running the codec, but a tool like valgrind will have a

Dear useRs, Following the given codes below, I generated a plot that has 6 regions around a center point (IL), with 5 regions containing a point (L1, L2 to L5) and sixth vacant region. I want background of all the filled regions turned "green", while "red" for the vacant region. Can it be done through a quicker way? Thanks in advance ###Codes start from

Used for celt_pitch_xcorr on aarch64, and celt_fir and celt_iir on both armv7 and aarch64. --- celt/arm/arm_celt_map.c | 17 +++++++++++++ celt/arm/celt_neon_intr.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++- celt/arm/pitch_arm.h | 31 +++++++++++++++++++++++- 3 files changed, 107 insertions(+), 2 deletions(-) diff --git a/celt/arm/arm_celt_map.c b/celt/arm/arm_celt_map.c index

Because a wideband signal has both a narrowband part and a wideband part the first part of the wideband decode is to call the narrowband decode. In the fixed point version the signal generated from the narrowband decode is downshifted at the end. In this same version the wideband then upshifts the signal by the same amount. I would like to do away with the downshift on the narrowband side

Unfortunately I don't have a setup that lets me easily profile ARM code, so I really can't tell which method is faster (though I suspect Mr. Zanelli's code is). Let me offer up another intrinsic version of the NEON xcorr_kernel that is almost identical to the SSE version, and more in line with Mr. Zanelli's code: static inline void xcorr_kernel_neon(const opus_val16 *x, const

> But chroma subsampling? no. And this is where much of the complexity > comes. > > If we were to combine them, we would be, essentially, doing it something > like > this: > Value Meaning > 0 RGB > 1 YUV444 > 2 YUV422 > 3 YUV420 > 4 YUV411 > ..... Yes. > And then spend an additional field on bits/channel, whereas both chroma

> But chroma subsampling? no. And this is where much of the complexity > comes. > > If we were to combine them, we would be, essentially, doing it something > like > this: > Value Meaning > 0 RGB > 1 YUV444 > 2 YUV422 > 3 YUV420 > 4 YUV411 > ..... Yes. > And then spend an additional field on bits/channel, whereas both chroma

Hi John, Thanks for the two fixes. They're in git now. Your SSE version seems to also be slightly faster than mine -- probably due the the partial sums. As for the NEON code, it would be good to compare the performance with the code Aur?lien Zanelli posted at http://darkosphere.fr/public/0002-Add-optimized-NEON-version-of-celt_fir-celt_iir-and-.patch Cheers, Jean-Marc On 06/06/2013 08:07

Hello Theora developers, I'm doing some cursory research into Theora's IP status in preparation for asking Apple to reconsider the possibility of shipping an implementation. I have a few questions and I'm hoping knowledgeable people can help out. 1) What are the terms of any patent licenses or disclaimers, and do they have field of use restrictions or limitations on code for

Hi All, Regarding cycle measurements for ARM/NEON, ARM no longer provide cycle accurate simulators. The method we use is to to make measurements on hardware via the PMU unit on the core itself. Note that if your running under Linux you may be 'allowed' to access the PMU directly but can access via it system calls. Typically you will need to make a series of measurements and average them.

FYI, Google is acquiring On2: <http://www.betanews.com/article/Future-of-open-Web-video-may-change-with-Google-acquisition-of-On2/1249481107> Also on Slashdot: <http://tech.slashdot.org/article.pl?sid=09/08/05/1742249> Anyone knows what, if anything, this means for Theora, open video and/or the Xiph.Org Foundation? Wild speculation: considering that they complained about the

hello, i'am is new in R software.i have try to make a function but it can't give what it should.i dunno what have to do next. Can somebody help me to solve it.i'll very appreciate... ##GEV simulation(Non-stationary) dsim<-function(n, alpha, beta,sca,sha){ t <- 1:n location <- alpha + beta*t inv.df<-function(x) location + -(sca/sha)+(sca/sha)*(-log(x))^(-sha) u<-runif(n)

Dear all, I would like to draw a dendrogram and mark some parts/branches (by using "segments") including their labels. If I draw it without specifying the length of x axix, I am able to do that (as in My dendrogram 1 of the following codes). However, if I want to specify the x axix, I am not able to draw marking line (by using "segments") including labels (as in My dendrogram

Using the PPC frsqrte and fres instructions, I got the percentage time take in the smoothing code in _vp_compute_mask() down from 13.64% to 1.88% of the running time. In my local copy of Vorbis I have a fast_math.[hc] in vorbis/lib and have a _fast_sqrt() inline in fast_math.h. If anyone else wants to try it out, it follows. I can currently encode my test file (the first 15 seconds of

Hi! I have some questions about R function. I try to write a function for multi-returns. The function code is as attachment. dgp.par<-function(ai, bi, t, n) { t0<-t+20 y0<-matrix(0, nr=t0, nc=n) y0[1,]<-ai/(1-bi) for(tt in 2:t0) { y0[tt,]<-ai+bi*y0[tt-1,]+rnorm(n, 0, 1) } y<-y0[21:t0,] x<-y0[20:t0-1,] z<-y0[19:t0-2,] z<-z[2:t,] dy<-y[2:t,]-y[1:t-1,]

Hello dear members of R-help and R-core mailing list, I am not sure if this request is a "ticket" that should be filled somewhere outside the mailing list. If so, I apologize for not doing and would like to know where I should have filled it. And to the subject matter: I would like to use a command like this: plot(c(1,8), 1:2, type="n") polygon(1:7, c(2,1,2,NA,2,1,2),

Hello dear members of R-help and R-core mailing list, I am not sure if this request is a "ticket" that should be filled somewhere outside the mailing list. If so, I apologize for not doing and would like to know where I should have filled it. And to the subject matter: I would like to use a command like this: plot(c(1,8), 1:2, type="n") polygon(1:7, c(2,1,2,NA,2,1,2),

I suggest a simple enhancement to segments() and arrows() to facilitate drawing horizontal and vertical segments -- set default values for the second x and y arguments equal to the first set. This is handy, especially when the expressions for coordinates are long. Compare: Segments: < function (x0, y0, x1 = x0, y1 = y0, col = par("fg"), lty = par("lty"), --- > function

I have a function in R^2, say f <- function(x,y) { ...skipped } I want to plot this function using contour, persp. wireframe, etc. I know that the function has a global minimum at (x0, y0) The naive approach is to evaluate the function on the outer product of two arrays, like this: sx <- c(seq(-3, x0, len = 100), seq(x0, 3, len = 100)[-1]) sy <- c(seq(-3, y0, len = 100), seq(y0, 3,

Is there a way to tell during the NB decode if it was kicked off by the SB decode? I would like to avoid saturating and packing the output of the NB decode, only to unpack it for the SB. Thanks. -Fritz