Erik,
> Notice that step where you throw away the middle N/2 samples?
> That is a low pass filter applied in the frequency domain.
Not if the amplitudes are all already zero above N/2. In that case, the
input and output frequency spectra are absolutely identical. If nothing
was removed or even altered, then no filtering has actually occurred.
A similar situation is what I have previously stated as my starting
assumption.
> Less accurate how? Is this measurable or is this just hand waving?
I have already addressed this in previous postings where I described the
fact that I had merely listened to the recordings repeatedly until I
could distinguish the original from the sndfile-resample one and could
not distinguish the FFT-overlap one from the original --- ergo less accurate.
My intention was not to "measure" accuracy, but to listen for the alleged
distortions. You had posted a sort of celebration of the fact that I
thought the sinc-based resampler produced a better-sounding version of
the recording than my FFT-overlap resampler, but apparently had also
neglected the fact that I was also saying that it sounded better than
the original --- ergo was inaccurate (albeit very, very slightly so).
I was merely attempting to correct possible misunderstandings.
Now I could take time to measure my resampler's performance, but I think
we both know the expected results, don't we? In terms of absolute
accuracy, any FFT-overlap resampler which utilizes large windows (hundreds
of thousands or millions of samples) and which is properly implemented
should, in fixed-rate conversions, outperform a sinc-based resampler that
has been localized in the manner that Smith has developed for the very
useful purpose of allowing variable-rate conversions. However, the
difference may be measurable but not audible, hence of questionable utility.
Best regards,
Dave.
P.S. I'll be "sans computer" for the next few days, heading towards Ivan
so won't be able to correspond on this for a while....
Dave (Phillips),
I had to alter a file for libgig:
[davec]libgig $ diff src/gig.h.org src/gig.h
588c588
< class Instrument : protected DLS::Instrument {
---
> class Instrument : public DLS::Instrument {
Crude but effective for compilation (but see below).
Also, even though libgig is installed, qsampler config appears not to find
it. It's installed in /usr/local/lib/libgig. Another problem is that
qsampler cannot actually load a *.gig file:
gig::Engine error: Failed to load instrument, cause: Can't open "NON_MODAL
'/usr/local/dosh/Gigs/etc...'"
I have been assuming that this is libgig-related, or maybe I should say
libgig-qsampler interface-related. I have not tested libgig by itself.
After reading Rui's message, I'm not so sure what the problem is.
Best regards,
Dave.
P.S. I'm going to be sans computer for a few days, so won't be able to
respond to followups.... but I hope this helps someone.
I got about half way through an ecasound set up for controlling many of
hermes' 52 parameters with midi controllers and noticed the "input gain"
control. I was under the impression hermes could run as a stand alone
synth. But, I guess the name it bears, "hermes filter", might suggest
otherwise. I tried to catch Steve on irc today, but kept missing him.
Does anyone know if the input is optional for hermes?
If input is required, can I feed it an ecasound null input and still
get sounds out of it?
I'll use it however it's supposed to work. I just want to understand my
options.
Thanks,
Eric Rz.
Hello,
I'm having a problem with Linuxsampler. It's compiling ok, but when i try to
run it I have the following display:
magoo@darkstar:/home/magooslk$ linuxsampler --gig ns_kit7free.gig
LinuxSampler 0.2
Copyright (C) 2003, 2004 by Benno Senoner and Christian Schoenebeck
Creating Sampler...OK
Registered MIDI input drivers: ALSA
Registered audio output drivers: ALSA,JACK
Starting LSCP network server (on TCP port 8888)...OK
LinuxSampler initialization completed.
And the program stop. There is no alsa midi connections or jack ports.
I'm running Slackware 10 with kernel 2.6.7+voluntary preemption patch, alsa
1.0.6a, jack 0.98.8 and linuxsampler CVS version.
I realy thanks for any help.
André Magoo
PS: Sorry for my very bad english...
Hi All,
I'm wondering if there is an app which can scan through audio files to
identify ones with similar characteristics to a reference file? A bit
like the apps which can identify duplicate images.
Cheers
Dylan
--
"I see your Schwartz is as big as mine"
-Dark Helmet
Hi.
I released ZynAddSubFX software synth todays
(standalone linux/windows and VST).
For those who don't know about it: zynaddsubfx is a
powerful opensource synth with many features (require
few pages just to make a list of them ) and the
homepage is at http://zynaddsubfx.sourceforge.net
News: - Solved some problems regarding VST
- added an advanced clipboard/preset module. On each
windows/object that support this, will appear 2
buttons "C" for copy and "P" for paste (from
clipboard/preset). You can make a collection of
waveforms, LFOs, ressonance functions, envelope,etc,
and share to others or use as favorites
- Completely removed the old format support (you can
use the prereleases
- bugs removed
- others
Also a very good news is that it comes with about 250
instruments (from usual organ/strings to the most
weird sounds ever heard )
Hope you like it.
I have intention to stop coding for a while and to
start to write the documentation and make some real
good sound examples.
Paul
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail
I'm trying to get a sequencer like rosegarden working
under knoppix 3.4 (a debian flavor linux live CD
distro). I am unable to get a midiport like
/dev/sequencer to work.
knoppix loads the OSS driver by default, and it
recognizes the soundcard and I'm able to hear audio
waves thru the soundcard, but no midi. I can
alternatively boot with ALSA but then I don't even
hear audio, so it seems that it may be easier with OSS
since I'm halfway there.
I'd like to get the external mpu401 midiport working,
and also access the onboard FM sounds. I am using an
ALS100 soundcard (and can resort to another ESS1869 if
I need to).
Here is my system info:
"lsmod" shows uart401, sound, and soundcore
cat /dev/sndstat displays nothing.
/etc/modules.conf doesnt seem to have anything
relevant (grep didnt find midi or sound anywhere)
What is the next step to enable my midiport? I can
promptly post back any requested info. Thanks!
__________________________________
Do you Yahoo!?
Take Yahoo! Mail with you! Get it on your mobile phone.
http://mobile.yahoo.com/maildemo
So after much pondering, I started writing this thing a few hours ago. As I
wrote the first paragraph, I figured I'd do it to provide something I could
never seem to find....Lot's of simple detail for Noobs! Damn!!! Now I know
why no one ever want's to write that basically; there's seventy kajillion
things you need to esplain!!! :)
Oh well....It's on the rails and headed out of the shed! I belive Mark
bellied up as a proof reader and some others...who were they now? I need to
put this in front of a few folks to make sure I'm not losing my mind!
R~
Hello,
It has turned out that the output of 44100 Hz files on my Intel 865 AC97
is only distorted on SPDIF. The analog output is all right!
Looks like it's not rate conversion at all, but SPDIF status. I saw
something like that in ALSA 1.0.7rc1 changelogs (but can't find these
changelogs again; I wonder why even ALSA tarballs don't include
changelogs).
I am now trying to compile the 2.6.8.1 kernel with alsa 1.0.7 rc1. It
may be strange idea, but I'm somewhat reluctant to to go to an -rc
kernel as yet. I'll report whether the problem is gone.
Yours, Mikhail Ramendik
Hi Erik,
Although I did say that sndfile-resample sounded ever so slightly better,
I also indicated that it was also ever so slightly less accurate.
The sndfile-resample sounded ever so slightly better than either the
original or the FFT-overlap resampled version (which sounded identical
to each other). I detected almost-imperceptible smoothing. It was
very difficult to hear any difference at all, so I think it did a very
good job on the voice. I have not attempted to evaluate the clicks and
pops wherein one might expect to hear more differences.
Best regards,
Dave.