similar to: Regain play analysis patches

Mike, I'm not an expert at DSP work, and I'd gladly defer to any one with any reasonable amount of DSP experience. I followed the link, but I'm not sure how to interpret the graph exactly. I believe it's showing filter behaviour. The two curves look similar, but are clearly not the same. At the 1kHz mark, the blue shows -23, and the green shows perhaps 17.? Are they supposed to

Hi all, In the last couple of months, there have been two proposed patches for the regain play analysis code. The first by Nathan Rennie-Waldock: http://lists.xiph.org/pipermail/flac-dev/2011-December/003070.html simply adds some more higher sample rates. The second by Earl Chew: http://lists.xiph.org/pipermail/flac-dev/2012-January/003067.html is more comprehensive. Neither of the

Earl Chew wrote: > That being said, I think my patch leaves us better off than before ! I agree. We will add it as it is and then tweak futher as needed. Earl, would you be able to update your patch so that it applies against the current git master? Cheers, Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/

Earl Chew wrote: > Erik, > > Ok. I've updated the patches, and made some additions to the test harness. > > For the test harness, I used sox to generate 1kHz wav files over a range of > sampling rates. The test harness runs the replaygain analysis, and compares > the resulting metadata. > > Are you ok with this approach ?? It means that the wav files need to be

Earl Chew wrote: > I'm a little reluctant to introduce another compiled program when there are > so many other options that will work well enough out of the box. > > Here are two ideas: > > 1. Use bc(1) to compute the raw samples > 2. Use perl(1) to compute the raw samples > > To generate raw unsigned samples using bc(1) for example: > > samplerate = 1000;

Earl Chew wrote: > One per filter table entry. That's 12 files right now. Thats probably a bit much. > The files could be generated on the fly, but that would either require > another program or script (eg Perl), Anything but Perl. I can write Perl but I can't read it :-). > or a dependence on an external toolkit such as sox. I'd be ok with requiring sox to run the

Brian Willoughby wrote: > What about using the C library sin() and cos() functions to generate > the test audio instead of sox? I did not see a description of how > the test files are generated, so maybe this is easy or maybe it is > hard. The benefit of shipping the test audio generation source code > around with the FLAC sources is that the tests won't break when

Hi, I recently switched from courier imap to dovecot. With courier I had a working IDLE setup that informed me immediately when new mail arrived. With Dovecot it is different, sometimes i get an immediate result but most of the time it takes a rather long time for the notification to return to the client. For testing purposes I set mailbox_idle_check_interval = 1 and i now get the same

Erik, Ok. I've updated the patches, and made some additions to the test harness. For the test harness, I used sox to generate 1kHz wav files over a range of sampling rates. The test harness runs the replaygain analysis, and compares the resulting metadata. Are you ok with this approach ?? It means that the wav files need to be checked into the repository. (There are already some small image

Erik, It turns out bc(1) is too accurate, and a little slow, for this purpose. I've switched to using awk(1) which uses floating point. Do you feel I need to test for the presence of awk(1) ? It is specified as one of the standard commands in the LSB : http://refspecs.linuxfoundation.org/LSB_1.0.0/gLSB/command.html Earl ??? awk -- ' ??? BEGIN { ??????????? samplerate = 8000;

Erik, One per filter table entry. That's 12 files right now. The files could be generated on the fly, but that would either require another program or script (eg Perl), or a dependence on an external toolkit such as sox. The tests increase the runtime of test_metaflac.sh from 11s to 15s on my machine. What are your thoughts ? Earl ----- Original Message ----- From: Erik de Castro Lopo

I'm a little reluctant to introduce another compiled program when there are so many other options that will work well enough out of the box. Here are two ideas: 1. Use bc(1) to compute the raw samples 2. Use perl(1) to compute the raw samples To generate raw unsigned samples using bc(1) for example: samplerate = 1000; duration = 2; bitspersample = 24; samplerange = 2 ^ (bitspersample-1) -

What about using the C library sin() and cos() functions to generate the test audio instead of sox? I did not see a description of how the test files are generated, so maybe this is easy or maybe it is hard. The benefit of shipping the test audio generation source code around with the FLAC sources is that the tests won't break when sox is modified. Brian Willoughby Sound

Application is working on Windows but when I run on RedHat10 under Wine 1.1.15. It looses focus and underlying Linux applications to get the keystrokes from the keyboard. Clicking on the application we're trying to run or switching with ALT+TAB doesn't not give it the focus back. The parameters we're running Wine are: Allow the window manager to decorate the windows : OFF Allow the

We have still having problems with Ferret dying on us regularly with the error message: >> ferret server error IO Error occured at <except.c>:93 in xraiseError occured in fs_store.c:127 - fs_each doing ''each'' in /var/www/web1/oms/current/script/../config/../index/production/band/20070805130005: <Too many open files> << We are running Ferret as a

Copy additional filter values from Foobar2000 as found in <http://code.google.com/p/sirens2/source/browse/trunk/libwavpack-4.32.psp/wvgain.c?r=32> to allow metaflac to perform gain analysis on high sample rate audio. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/flac-dev/attachments/20120107/018c821b/attachment-0001.htm

Hey everyone, I'm new to the world of audio, so can you please drop some very basic audio knowledge on me? What's the correct way to convert from float audio samples to integer ones like FLAC uses? If I'm converting to 8-bit, that has a range of -128 to 127, right? Do I multiply my float input by 127 or 128? Brendan

The Partek package (www.partek.com) allows only two selections for Multiple Test Correction: Bonferroni and Dunn-Sidak. Can anyone suggest why Partek implemented Dunn-Sidak and not the other methods that R has? Is there any particular advantage to the Dunn-Sidak method? R knows about these methods (in R 2.1.1): > p.adjust.methods [1] "holm" "hochberg" "hommel"

I know acroread is not part of CentOS but many people use it. Thought forwarding this post on the SciLinux mailing list might help those who use acroread on CentOS-4. === excerpt === The "latest" version officially compatible with RHEL4 is acroread-7.0.9, but this currently has open security holes (CVE-2007-5663 et al). http://www.adobe.com/support/security/advisories/apsa08-01.html

The test is rather simple, and only compares the computed output values for a 1kHz signal at the known filter table sampling frequencies with previously obained golden values. As such, the test only verifies that the performance of the replay gain analyser has not changed. A better test would be to compare the computed value with an independently obtain theorectical value. This is possible by