similar to: questions concerning adaptation rate

> When you count the adaptation step size in case where the filter has already had > minimal adaptation, you use the following expression: > > r = (0.7*r + 0.3*15*RER*e)/e*(power[i] + 10) = (0.7*leak_estimate*Yf[i] + > 0.3*15*RER*(Rf[i] + 1))/((Rf[i] + 1)*(power[i] + 10)). Actually, the QCONST16(.3,15) doesn't mean (.3*15), but the value .3 in Q15 format (fixed-point). >

Hi I just upgraded to http://svn.xiph.org/trunk/speex r10357 and got this build warning: alfredh@io:$ make -s mdf.o libspeex/mdf.c: In function 'speex_echo_cancel': libspeex/mdf.c:321: warning: statement with no effect libspeex/mdf.c:317: warning: `adapt_rate' might be used uninitialized in this function Is this intentional? In any case here is a simple fix: Index:

Jean Marc & list, Following up on this- there seems to be an issue with this calculation in mdf.c (Smooth echo energy estimate over time) Pey += Eh*Yh; Pyy += Yh*Yh; it goes too large for floating point, which seems to be caused by extremely large values in st->Rf[j], st->Yf[j] and smaller values in st->Eh[j], st->Yh[j]- things were relatively quiet

Hi I have a question about the AEC and preprocessor. I have seen that when near end is talking for a longer time (about 10 s) without being interrupted by the far end the residual echo (or rather leak_estimate) increases making the preprocessor suppress the near end talk when it shouldn't. Is there a way to make the leakage estimate only update when far-end is present or similar in order to

Did a bit of tuning in SVN. Could you try that? What frame size, filter length, sampling rate are you using? All the process you describe is actually the tip of the iceberg I think. There is something making the adaptation diverge, which causes Eh and Yh to be large and so on. The part I changed is probably the one causing divergence. If it doesn't work, reducing the coefficients even more

Hi all, I've got a problem with the nls function. I have an adjustment which works when I fix one of the argument of my function (Xo=150) : *Xo*=150 f<- function (tt*,Xo*,a,b) ifelse(tt<*Xo*,a*exp(-b**Xo*),a*exp(-b*tt)) ajust<-nls(RER~f(tt,*Xo*,a,b),data=data.frame(tt=Ph2[,2*k],RER=Ph2[,2*k+1]),start=list(a=0.5,b=0.014)) But, when I use it as a "normal" parameter (and

Jean-Marc, I have traced the second infinite loop further. When st->adapted becomes true (mdf.c line 623), the first Yf[i] value is 4, the leak_estimate is 0xd4e, the resulting r is 3. The first value in st->Rf is 0, so e is 1, and r is set to e>>1, or 0. A little later there is a divide by r, and there is the hang. It seems that the 0 in Rf[0] is the problem, but I am not

yop, http://rubyonrails.org/api/classes/ActionView/Helpers/FormHelper.html ici j''ai vu qu''on pouvait g?n?rer des ?l?ment d''un formulaire gr?ce des m?thodes. Apres quelque petite recherches j''ai trouv? comment g?n?rer un select <%= select(''category'', ''category.id'', Category.find_all.collect {|category| category.nom})

Dear R Users, If a variable, say CO2(ppm), is varying with time. Then how to calculate CO2 (ppm) growth rate /a-1 I have CO2 time series (1991-2000), as: time, year, month, day, hour, min, sec, lat, long, height, CO2 1991.476722 1991 6 24 0 5 0 -38.93 145.15 4270 353.680 1991.476741 1991 6 24 0 15 0 -39.20 145.22 4270 353.950 1991.476747 1991 6 24 0 18 0 -39.43 145.28 4270 353.510

Hi Jean Marc & List, So I have been fiddling with the denoiser (again). While poking around I noticed that nb_preprocess is basically a counter that is mod-ded with 100 (the default), which causes Smin[] to be re- seeded with the value in Stmp[] (min of the previous adaptation period). Smin[] is then used to update the noise probability, which is (probably) less likely when adaptation

I get this error while computing partial correlation. *Error in solve.default(Szz) : system is computationally singular: reciprocal condition number = 4.90109e-18* Why is it?Can anyone give me some idea ,how do i get rid it it? This is the function i use for calculating partial correlation. pcor.mat <- function(x,y,z,method="p",na.rm=T){ x <- c(x) y <- c(y)

Gents, In Version 8.1, QCONST16 is defined as follows in arch.h within the FLOATING_POINT section: #define QCONST16(x,bits) (x) However, in both celt.c and quant_bands.c, QCONST16 is referenced without a conditional FLOATING_POINT define. So when I compile as FIXED_POINT I get an "undefined identifier" for QCONST16. Am I doing something wrong? Thx, MikeH

Dear friends i am trying to fit an Ornstein-Uhlenbeck process by MAXIMUM LIKELYHOOD method. i found these formulas on http://www.sitmo.com/article/calibrating-the-ornstein-uhlenbeck-model/ this is the mean-reverting process http://r.789695.n4.nabble.com/file/n4637271/process.txt process.txt and this is the script that i am using....... ouFit.ML=function(spread) { n=length(spread)

Hello, I'm doing some linear regression: >lm<-lm(osas~alp,data) >summary(lm) However, the Rsquared in the output of summary() is not the same as the "standard" Rsquared calculated by spreadsheets, and outlined in statistical guidebooks, being SSR/SSTO. The output says "multiple Rsquared", but it is no multiple regression... What's the difference? Thanks,

bonjour, je voudrais savoir s'il serait possible de sugg?rer aux d?veloppeurs de R de proposer une option suppl?mentaire pour les moustaches, ? savoir les placer sur d1 et d9 comme c'est pr?conis? dans les programmes du secondaire en France, option du style boxplot (serie,range=91) par exemple... i would like to know if it is possible to have (with the agrement of developpers)

Hi, This is most likely only a minor technicality, but I saw the following: On page 6 of the 'parallel' vignette (http://stat.ethz.ch/R-manual/R-devel/library/parallel/doc/parallel.pdf), the random-number generator "L'Ecuyer-CMRG" is said to have seed "(x_n, x_{n-1}, x_{n-2}, y_n, y_{n-1}, y_{n-2})". However, in L'Ecuyer et al. (2002), the seed is given with

Sorry for re-posting the question, I did not get any reply. Kindly reply please if any one can. ### Hello R users, I have a montly time series over a several year period. It's easy to compute a monthly climatology (12 values), Now, I would like to calculate the corresponding standard deviation, ie the 12 values calculated from the january values, february values, etc. What's the best way

Sorry that I wrote in german : Ich benutze asterisk mit dem zaphfc Treiber. Jetzt hab ich folgendes Problem, habe 2 ISDN-Telefone angeschlossen. zaphfc im nt-mode. Anrufe von ausserhalb per sip (sipgate.de) kommen an. Wenn ich aber intern zwischen den zwei Telefonen (Ascom Eurit 30) sprechen m?chte geht das nur wie folgt : Erst die Nebenstelle w?hlen und dann den H?rer am Telefon abnehmen.

> I am writing to gain a better understanding of the limitations of speex echo > cancellation, esp. with respect to the fixed point implementation. > If these limitations have been documented elsewhere already, please let me > know! Nothing officially documented, sorry. > I observe experimentally that when one or both of the echo or ref data for > speex_echo_cancel() have

dear all, i'm trying to reproduce an spss-anova in R. It is an 2x3x3 repeated measures desingn with repeated contrasts. In R i've coded a contrast matrix for all factors and made a split in the aov summary - but I can't get the repeated interaction contrasts. The output from SPSS looks like this: TaskSw * CongNow * CongBefore: SS df Mean Square F Sig. 1 vs. 2 1 vs. 2 1 vs. 2