Getting a chart recorder trace?

[bentoronto]

I see nice frequency response chart traces all the time. How can I do it on a DIYer budget without buying a trick audio oscillator and XY plotter?

Challenges:

1. My special interest these days is acoustics (speakers and rooms) and so I need pretty fine-grained traces, not just the kind of gradual smooth curves for crossovers. That's not feasible with a hand-switched audio oscillator and Microsoft Excel.

2. Any room responses need to be done with a mic in one swept pass, or better, averaged over a few listening locations.

3. The input has to sweep across a band on its own. I recently get a very helpful download of one-Hz notes lasting 10 seconds, 10-300 Hz and a pink noise reference, that i made into a CD.

4. The output starts as AC from a Radio Shack sound pressure level meter, mic, or AC voltmeter, and then to a sound card or a mic pre-amp using a USB input (the Griffin iMic USB device). It has to be displayed in the end, as always, in dB.

5. Maybe too optimistic, but Macintosh is preferred.

Soooo, I have that nice CD sweeping 10-300 Hz (or part) over 48 minutes (or part). And I like to end up with a chart or at least a long row or column of numbers in Excel.

How can I do that? Help... I haven't had a good nights' sleep while pondering this --- or maybe I just need to spend the coin to buy the software?

 

[Robh3606]

Why not just look at freeware like Holm Impulse, if that won't do it try others. You can adjust your plot window in your software as well as smoothing. Look up Fuzzmeasure and check this Forum

Room Measurement and Acoustic Treatment series: Part 5 - FuzzMeasure Pro 3 for OSX

Rob

 

[bentoronto]

Rob -
Thanks for suggestions.

I gather those programs produce the automatic sweep signals which ultimately drive the speakers?

And at the other end, programs like the Holm require inputs. Holm software works with special Holm mic pre-amps but it also seems to work with generic sound cards that take audio inputs?

My old Radio Shack SLM puts out an AC signal coming from its mic pre-amp and kind of 1 volt AC out for 100 dB at the mic. And that output is upstream/independent of the dB range setting knob that controls the range on the SLM meter. So if I plugged that into a sound card, would the Holm software read it properly?

On the web, you can find mic low frequency dB correction factors for that popular old Radio Shack SLM. Working in Excel, easy to factor those factors in... but would be nice to be able to do so on dedicated sound measurement software too.

 

[Robh3606]

I gather those programs produce the automatic sweep signals which ultimately drive the speakers?

Yes just make sure they can do what you want. I don't know how many can do a Sine or Gated Sine sweeps if that what you are looking for.

So if I plugged that into a sound card, would the Holm software read it properly?

It should the levels might be too high for the Mic Input though.

On the web, you can find mic low frequency dB correction factors for that popular old Radio Shack SLM.

Might want to look at a different Microphone where you don't have to do that. Save you a lot of work and curve manipulation. I got a battery powered Crown PH-1 as my phantom power source and then got one of the Berhringer Measurement Microphones. Parts Express sells one as well.

That way you can do it all in the software and not have to worry about using Excel

Rob

 

[Conrad Hoffman]

Swept sines for rooms and speakers are, IMO, either useless or will drive you nuts trying to interpret them. Typically one uses a pink noise signal and a swept filter. Or, sometimes a swept multi-tone signal and swept filter. Third octave bands are, again IMO, kind of coarse. Tenth octave bands or continuous SA of the noise signal make more sense. I think you can do this with the free Visual Analyser software, though I use it for other things.

 

[bentoronto]

Swept sines for rooms and speakers are, IMO, either useless or will drive you nuts trying to interpret them. Typically one uses a pink noise signal and a swept filter. Or, sometimes a swept multi-tone signal and swept filter. Third octave bands are, again IMO, kind of coarse. Tenth octave bands or continuous SA of the noise signal make more sense. I think you can do this with the free Visual Analyser software, though I use it for other things.

A distinction has to be made between speaker response and room or total response. Your good suggestions apply to discerning the performance of speakers.

My impression is that playing music in rooms does strongly include tones sustained enough to trigger room modes on top of speaker response. Actually, I'd guess that it is an empirical question just to what degree say, classical music content engages room modes. If it is much, then sustained or slowly swept frequency ranges are helpful in addressing issues of room modes.

I presently am interested in total sound, particularly in the woofer range and think traditional frequency response is helpful.

You comments about driving the tester nuts - yes, absolutely. In the late 60s, I worked with pulses and in the world's largest anechoic chamber. Subsequently as a DIYer, I completely abandoned acoustic measurement because I couldn't do it and get interpretable results. Now, with this thread and modern software out there, I am going to give it another try.

Your views on that?