Audio Spectrum Analyzer

July 16, 2009 at 12:38 AM ·

It depends on what you mean by analyzer doesn't it?

www.trueaudio.com/rta_abt1.htm

www.spectran.com/Handheld_Spectrum_Analyzer.htm

www.dewresearch.com/fftp-main.html

The FFT spectrum analyzer I used to use in a previous job was about $6,000 back in 1995. That's like going back to Mesopotamia.

What are you looking for or what are you doing? 

July 16, 2009 at 02:03 AM ·

I would agree with Bill. I have five or six different programs on my computer which all do slightly different things. Some want recorded input, some live, some don't care. Some measure peaks, some sum up the whole experience, and my favorite gives a static waterfall plot taken from a selection from a recording (http://darntonviolins.com/images2/148.jpg for instance). Some are designed specifically for an impact reading.

What were you thinking of doing?

Both of my favorites, Spectrogram and SoundLandscape are defunct, but Spectrogram was free and the older versions are easy to find, and work fine. That might be a place to start. I had not seen the last one Bill links to, but it would be a good candidate for the type of thing I do the most..

If you search "FFT" you'll find that there are a whole lot of things out there. Don't bother with one that doesn't give you some sort of precise frequency read-out, in my opinion (some of the simple and free ones don't).

July 16, 2009 at 04:13 PM ·

Michael,

My hope is to be able to compare the spectra of my various instruments, over time, and including the effect of different strings, soundpost adjustments, bows, and other variables. I've delayed this for a number of years because I've only used Mac computers (since 1984) but I just upgraded my newest Mac with virtual windows so I can do this and also run "Tonelab" for tuning pianos.

Andy

July 16, 2009 at 10:38 PM ·

Andrew, there are two basic ways to accomplish what you want. The current fave in the testing crowd is to damp the strings and smack the bridge with a tiny, standardized hammer. This gives a one-shot reading of all of the things that are happening, all at once. The equipment and software to do it properly costs big bucks.

The other method, developed by Oliver Rodgers, is to play a one-octave glissando on each string, and run it through an FFT program giving a waterfall display output; something like the link I gave, above. This is what I do. I feel like I get a better view of what's going on because you can see what each string is doing, and get a real idea of how the violin sounds, easily. There are theoretical problems with it (of course--aren't there always?), but in the real world it turns out to be very consistent.

July 16, 2009 at 11:43 PM ·

Thanks Michael, I see what you mean. But I'm going to start out with the True Audio recommendation of Bill. Since I just had success with the 1-octave free version, I have decided to order the 24-steps/octave version for sale (they have 1, 3, 6, and 24 steps/octave versions for sale). This will allow me to take the baby steps with a few tones on some of my instruments that have had me puzzled.

Andy

July 17, 2009 at 12:00 AM ·

You might want to check out this one. I don't use it a lot, but it will do what you want, for free, I think:
http://freenet-homepage.de/dl4yhf/speclab/index.htm
 

July 17, 2009 at 12:16 AM ·

Michael, thanks for the SpecLab suggestion - looks good. Now I "own" that one too. I prefer I logarithmic display to the linear one I see there now, do you know if I can get that with SpecLab?

Andy

July 17, 2009 at 12:25 AM ·

I didn't find any freeware/cheepware which would give the resolution I wanted, so I ended  up with "SpecraPlus", which is still what I use today.

Understand that multiiple very good minds working on spectral maps haven't  solved the mysteries of good sound. This isn't to say that it won't  happen.

July 17, 2009 at 01:28 AM ·

Yes, I think you can change the scale. I haven't gotten all parts of the program together yet, myself. There's quite a bit to it.

As for David's comment: basically I agree with that. I found that the best use of software was to help my ear know what it was hearing, but a lot of the qualities I look for in an instrument aren't revealed by FFTs, and those qualities which are aren't all that important to me. Mostly, knowing how the stuff works makes what others are saying comprehensible.