You can't really do that with current software. However, it should be partially possible in theory. The way that I would attempt to do this would be with the help of an FFT display software. A good FFT spectrogram/oscilloscope type program should give you a good moving graph of time versus frequency. The colours denote intensity or amplitude. On the PC, there are a few freeware programs like that, but i forget what they are called. Anyway, by monitoring a known piece with the same kind of reverb, you should be able to see and hear where the reverb is.

The FFT display will show you where the sonic energy of the reverb is, and how long it lasts. With that information, you should have an idea of the main frequencies involved and the decay time of the reverb. Of course, the more advanced the reverb, the more complex this will be. However, some reverbs are nothing more than layered delays with feedback, so finding the dominant resonant frequencies won't be so hard.

Once you know these things, you can either try to manually edit all reverb tails out (they tend to be the same length). Or you could try using a noise gate. If you have a keyable noise gate, they you can feed the key of the noise gate with material EQ'd to pass only the same frequencies as the reverb. You could also use a keyable compressor the same way. If you don't have keyable dynamics processors, then you will need to duplicate your audio files and then
EQ them separately, invert their polarity, and experiment with mixing them back with the original content.

Overall, mixing processed and inverse polarity duplicates is how to do unusual kinds of spectral edits like that. It's not likely to result in the full removal of reverb, but it could be an interesting effect if you get rid of a lot of it.

You might try looking up some JAES or JASA (Journal of the Audio Engineering Society/ Journal of the Acoustical Society of the Americas) for some of the really serious professional research on this kind of thing. I bet some nordic or Japanese chap has already created something of this sort for forensic audio. Other than that, ask trent Reznor.

Hold on, that reminds me. Some interesting vocal changes have been done with time expansion algorithms. I have done some of this myself. You might actually achieve a minor bit of reverb removal by using an editor to Time Compress the recording. You might have to try time compressing in reverse too. Similarly, some FFT-based time compression/expansion softwares might also be helpful with this kind of experiment. The standalone compress/expanders often have more detailed FFT settings such as overlap, and window type and size. There might be some particular setting that overlaps small windows properly enough and often enough to reduce reverb.
Since some FFT expansion actually CAUSES reverb as a side effect, trying reversed audio might help a little. I think I may have actually done this once myself. You'll want to use as many bands as your computer can handle though, if you want clearer results. 8192, or 16384, or 32768, etc. 1024 probably won't be enough. Sometimes these bands are called bins.

If you' ve ever worked with a vocoder, some of this might start to seem familiar. I suggest trying HOG. It's free, and it has some great tools for this kind of thing.

Now that i'm thinking better, I'm remembering a great tool called D-Noise, or D\Noise or D/Noise. (I forget which). It is a bit hard to understand for me. I did'nt take Calculus or higher math. However, it's great at clearing up convoluted audio. It uses some interesting time-varying statistics to do its work. I'll be it could be adapted to remove reverb for sure, since reverb has a particular statistical distribution.

That is too much of a tall order. The sound of an instrument or vocal with the reverb is 'one' sound and is no longer a send or insert effect to be seperated. The on chance you might have is with a noise gate and that would only generally work with one word or one instrumental phrase at a time. Not a singer as such... I would forget about it if I were you... ';/

Rimmer

I've found it :P hehe
want the site

???
www.driversguide.com

Hey people , after weeks of looking for the problem , I found it , a bad system.ini file , so my advise is to open system.ini in a notepad and look for the following lines:
(they should be exactly like this, if you don't have them, add them)

[boot]
drivers=mmsystem.dll

[mci]
cdaudio=mcicda.drv
sequencer=mciseq.drv
waveaudio=mciwave.drv
avivideo=mciavi.drv
videodisc=mcipionr.drv
vcr=mcivisca.drv

GOOD LUCK , IT WORKED FOR ME

hey everytime i try to play my mp3s it says waveout format not supported... can anyone help?

Also the Sound Volume 32 Icon that should display on everyones system isnt loading at startup,

I never changed OSes Ive been at win ME.

Actually I didnt do anything I was chatting on Trillian (Combination MSN/Yahoo/AIM/ICQ Instant Messanger), when all of a sudden my computer Froze, I restarted the System and I lost all sounds, only the Sys Speaker worked. SndVol32 stopped showing up in the SysTray.

If anyone has any Ideas please let me know.. (and no there isnt a Red Exlamation point letting me know there is any probs with the SoundCard, Controller, Codec, or any other device)

Music Match gives me the same error message. "WavOut format not supported" All I did was upgrade to Windows 2000. Is there something I can do?

: hey everytime i try to play my mp3s it says waveout format not supported... can anyone help?

Either through a hardware/software conflict, or who knows what else, your sound card was likely unrecognized in the system. Unlike Win95 that used to have a red bang(!), the system will show nothing.

Easy fix. Remvoe your sound card and start up. Then shut down and put it back in. Dust it off, change slots, whatever. Next time it will likely come up but won't show you a new hardware found dialog (gee isn't XP great?). Everything should work now.