I'm using the tutorial that came with Csound. Xanadu.csd renders as it should and plays as it should but tut2 crashes when I try to render. I tried the newest version first and it didn't crash but didn't render either and so I reverted to 5.06 and it just crashes.
This is my first week using Csound but I'm pretty sure it is not syntax. Also, I am using windows xp. Any help would be appreciated.
haven't even finished the tutorial and I am already stuck
Submitted by MMiller on Sun, 04/27/2008 - 01:18.
I'm going to take a guess.
I'm going to take a guess. I run Csound mainly on Linux, but since I'm a habitual experimenter, I just had to load it on to a Windows machine to see if there were any differences.
My first time on Windows, Csound kept on getting hit by segmentation faults. You need to set certain environment variables. There's a guide, but I believe the ones you want to make absolutely certain are set are either OPCODEDIR or OPCODEDIR64 (depending on whether you have float or double precision internals).
At any rate, every problem I ever found I tracked by looking at the command line (or terminal) output. I believe you can see it in Csound5gui in the 'console messages' window. If you get a chance, next time it crashes, paste the output you see in the console messages window here, as a reply to this message. Lots of times the smoking gun is right there. (Note: it helps if you place the output between 'code' tags, so it isn't formatted as a paragraph).
Hope this helps. I'm out of my neighborhood, talking about Windows.
--
http://www.jamendo.com/en/artist/bruce.h.mccosar
http://bmccosar.wordpress.com/
Windows/Linux Lego's don't click?
The gui crashed with not so much as a word in the console messages window and the command line returns to c:\csound. This may be yet another language that I give up on at the hello world stage.
which tutorial?
I'm not a Windows afficionado either, so I don't know if I can be of any help, but I could take a look -- *if* I could find the 'tut2' you refer to. (I have Xanadu.csd, but no sign of the other.) I even downloaded Winsound.zip, and peeked inside, but it's not there either. Where's it from -- a tutorial downloadable from here?
My mistake
Actually, the name is tutorial2 but yes this a tutorial available on the tutorials page of this site. I am referring to "An Introductory Csound5 Tutorial". Also, I am using the double precision float version and opcode64 seems to be set unless I am supposed to edit it in some way. I have a feeling I wouldn't be having this problem if I were not using windows. Thank you for your help.
Can you post it?
Oh, yeah. In fact I have that tutorial, but it's too Windows oriented to be much use to me. Reading it now, I realize that you had to type in 'tutorial2.csd' yourself before trying to run it. Can you post the file here? There might be something simple that we can spot.
Hopefully it is just a syntax error I cut off the <
CsoundSynthesizer>
CsOptions>
-W -f -R -o tutorial2.wav
/CsOptions>
CsInstruments>
sr = 88200
ksmps = 1
nchnls = 2
gisine ftgen 0, 0, 65537, 10, 1
instr 1
ikey = p4
ivelocity = p5
ioctave = ikey / 12 + 3
idb = ivelocity / 127 * 84
ifrequency = cpsoct (ioctave)
iamplitude = ampbd (idb)
asignal poscil iamplitude, ifrequency, gisine
outs asignal, asignal
endin
/CsInstruments>
CsScore>
i 1 1 3 60 100
/CsScore>
/CsoundSynthesizer>
I think it's minor
Well, the good news is that it only needed a minor typo fix (below) [after adding the less-than characters back, of course] to run.
However, I was able to see what was wrong quickly because -- running it from the command line -- I got a page of errors. Can you not do that in Windows, rather than going straight into the GUI? Really strange if you can't.
Anyway, you just had "ampbd" instead of "ampdb"! (The gisine argument was on the wrong line, but I presume that was just the pasted-in version.)
One more 'however', though... You had the '-f' flag set, which for me creates an unplayable wav file. Don't see any good reason for it.
[Added later:] I just went back to the tutorial, and see that, yes, he includes that '-f' flag too, but gives no logic for it. Maybe WAV files can have float format on some systems, but certainly not on mine. Personally, I'd treat that tutorial with a grain of salt [never did me any good... (:-/)]. Better to look at some of the simple examples included with the manual, and develop your expertise that way. (And I think most would agree that Csound is not an easy language! A lot of trial and error is involved, whatever route you take. Getting some error messages helps a lot, though!)
Good luck!
pffft
I can not believe it was just a typo! I supposed that is why there is a beginner section. To answer your question the command line gave me no errors it just crashed which still has me wondering why. I had the -f set because the tutorial told me to. Maybe, chalk it up to using different versions? Thank you very much I will be sure to check more closely for typos in the future. Fixing the typo did the trick by the way.
Excellent!
Glad that fixed it. That you didn't get error messages even when running from the command line is NOT good, though. [I could spew for hours on my opinion of Microsoft, but I won't... (:-/)] Hard to believe that this is normally the case, though. A lot of people seem to use the Windows version, so maybe there's some trick. Anybody care to chime in?
Command-line output
I run Csound 5.05 (float samples) on WinXP without any trouble (that is, without any trouble that isn't my own damn fault). It always gives me more-or-less comprehensible errors when run from the command line. Not sure what's up with MMiller's version.
I've just tried the (latest) Windows version too
Had some spare time and a nearby Windows machine at the lab, so I installed the package from SourceForge. Aside from it always complaining that I don't have Python installed [which I couldn't be bothered to do] it seems to run.
Interestingly, though, (running from the command prompt) when I reinsert MMiller's typo, I get the error messages OK, but Csound does then crash with a segmentation violation. I wonder if you (MMiller) are running it in some fashion that causes it to close the window when it crashes? I opened a command prompt window and then simply typed "csound tut2.csd".
Regarding that '-f' switch, it seems to me very bad advice. As far as I can tell, the 'float WAV' doesn't officially exist. Some apps deign to play it (Audacity on Linux and VLC on anything), but RealPlayer on the Windows machine dosen't for example. For almost any purpose, using a float sample seems pretty pointless, because it has to be integer for the hardware. (Rare mixing uses might theoretically benefit, but I'd bet 24 bits are in reality ample.)
You're right...
You're right, even 16 bits is ample. The main use for -f is archival, high quality .wav files, generally 32 bit float at some fantastic resolution (96000 kHz). When performing heavy audio processing, certain effects can eat away at the resolution; when you renormalize, you get noise. The extra precision is like the "guard digits" in scientific notation.
In recording, the second use is that the 32 bits represent additional 'headroom' for the signal. Recording into a 16-bit ADC means running the signal in pretty hot. Recording to 24 means even if your signal lurks down 30 dB below full scale, you're still getting decent resolution, and not some sort of digital Aztec pyramids.
---
http://www.jamendo.com/en/artist/bruce.h.mccosar
http://bmccosar.wordpress.com/
I've got it figured out
I found out why the gui and prompt were crashing. It seems there is something wonky with the new fluid opcodes in windows so I just had to delete fluidopcodes.dll and it works fine now.