Generating test tones

[Hondaman]

Can I generate some test tones with my computer? Let me explain.

I have the BSR SA3-X spectrum analyzer hooked up to my stereo. It can generate pink noise -- equal amounts of all frequencies at once. Yes, it sounds like water running from a faucet. The included microphone gives a signal to the BSR SA3-X based on the sound that reaches the microphone.

Adjusting my regular graphic equalizer, I can boost 250Hz, cut the 4KHz band, etc. so the microphone receives equal amounts of all frequencies.

So the SA3-X is putting out equal amounts of all frequencies, and now the microphone is receiving equal amounts of all frequencies. I have overcome the imperfections in my speakers and the sounds absorbed by my furniture. So my ear will hear the music or movie exactly as it was recorded.

This is very nice, but adjusting is proving to be primitive, and isn't working out so well.

I would like to generate, say, 30 seconds of 125 Hz, then 30 seconds of 250 Hz, then 30 seconds of 500 Hz, and so forth. Can I do this at, say, -20 dB, -10 dB, or zero dB? It should be easy, but how? This would help me adjust the EQ better than pink noise I think.

(I can always write a MIDI file with the note closest to those frequencies -- piano, trombone, whatever, and make lots of whole notes. Then convert from MIDI to WAV and record it onto CD-R. But that won't let me specify the decibel level, which is kind of important here.)

Any ideas?

[Sparkey55]

Re: Generating test tones

http://www.ringbell.co.uk/software/audio.htm
Try this link above. Many other software if you use Google search.

[budm]

Re: Generating test tones

http://www.electronics-lab.com/downloads/pc/005/

[Hondaman]

Re: Generating test tones

I've heard there are some $100 USB microphones and software. But I'm kind of broke right now, so I'm looking for a cheaper solution. (Plus I would have to set up a PC at the new house, and I'm a little busy.)

(Setting up the new house, getting a bed, landscaping, A/V rack, kitchen table/chairs, bookshelves, food and soap/shampoo, it is more expensive than I thought!)

I know it is possible to specify decibel level when you convert into ADPCM ("store-bought audio CD format", "red book" I think), but that would be more technical than I know how to do right now. (And unlike the old tape decks, I think you cannot go over "0" because of the digital nature of CDs.)

Thanks, I'll try these. If anyone else has ideas, please do post them. Thanks!

[tom66]

Re: Generating test tones

Audacity+PC.
http://audacity.sourceforge.net/

Generate > Tone

[Hondaman]

Re: Generating test tones

You were so kind to help, I thought I should write this up like a proper science experiment. So here we go:

The software from sparkey55 proved very useful. If "repeatable results = scientific", then I don't think this method was scientific. See below.

First of all, thanks to the KX Project for software to get the Sound Blaster Live! 5.1 digital working under Win 7 Ultimate. Sorry, Creative Labs, I cannot build every computer around which card and which version of Windows you decide to support. Problems with sound in the Wintel world make Apple's computers look very attractive!

A few things we learned:

1. Moving the slider on one band (1K for example) changed amount of 500 and 2000 Hz heard as well. Maybe it is better to adjust 500 Hz, 2K, and 8K to the same volume, then adjust 250, 1024 and 4096 Hz to that same level. But when you do this, the original sliders (500 Hz, 2K and 8K) will be affected, so they must be checked again. The process is "iterative".

2. Either the microphone or the SA3-X were giving inconsistent results. The fluorescent display in the SA3-X showed that the microphone hardly picked up 2000Hz, then at other times 2000 Hz was too loud. But it heard healthy sound levels at 1700, 1800, 2100 Hz, 2250, etc. From this, I see that this is not perfect, but I have confidence that the 2K slider (and other sliders) is as good as it can be on the main equalizer (for this level of equipment).

3. More testing (test tone discs or other, better equipment) would be helpful. We'll see if there is time and money later on. AVS Forum can help with equipment ant methods.


Thanks!

[Khron]

Re: Generating test tones

My hunch is, Creative would've rather you used (as in "bought") one of their new X-Fi sound cards

But on to more serious stuff, the effect of the 1k slider affecting the surrounding ones is down to the Q ("quality factor") of the band-pass filter corresponding to each slider. Low Q = the bell-shaped peak/notch is wider; high Q = the peak/notch is narrower.

http://en.wikipedia.org/wiki/Q_facto...trical_systems

For more precise adjustments, a parametric EQ would be near-mandatory. That would allow you to adjust the center-frequency freely, as well as the boost/cut (dB) and the Q ("width" of the boost/cut) independently.

Regarding inconsistencies... 2k is high-enough a frequency that it could be affected by small physical changes in the space between the speakers and the mic, i'd say. If you moved a chair or something, or if you yourself were in the "measurement area", that could easily lead to varying results. Averaging might help a lot

And then there's another detail (that the devil's in ) - good sound has two aspects: the frequency response, and the "time response" (for lack of a better term). What i mean by that is the sound signature of the room (reflections, reverberation time). "Well-timed" reflections from the walls / ceiling / floor can easily lead to "comb filtering" (which basically "sucks out" series of frequencies, depending on the time relationship between direct sound and the reflections - http://www.prosoundweb.com/images/up..._Filtering.jpg ).