free software

[gordoncalder]

Re: free software

Thanks all for your advice. I will get the Dayton mic and the bringer mixer. I have computers with adequate sound cards. My plan was to purchase some of the cheaper PE buy-outs to experiment with and learn speaker building. With-out frequency response curves how can I have a clue which speakers might work together until I purchase and measure them?

If the goal is to end up with one pair of (decent) inexpensive speakers then it's cheaper to spend a little more on drivers that provide frequency and impedance graphs or have been measured by others, and then trace them with SPL trace and model it with Jeff B's PCD program. On the other hand, if the goal is to learn to build speakers, then definitely an inexpensive mic/pre-amp and the demo version of ARTA will take you a long way. [handy link here for measurement set-up http://www.soundhobby.com/measurementsystem.htm]

Either way, you've come to the right place; lots of knowlegable people on this site.

Gordon

[billfitzmaurice]

Re: free software

Just as with a woofer use the modeled response through the range of pure pistonic function, switching over to the data chart SPL at the frequency where the modeling predicts high end roll off.

[dano]

Go

Thanks all for your advice. I will get the Dayton mic and the bringer mixer. I have computers with adequate sound cards. My plan was to purchase some of the cheaper PE buy-outs to experiment with and learn speaker building. With-out frequency response curves how can I have a clue which speakers might work together until I purchase and measure them?
________
Medical Marijuana Dispensary Santa Clarita

[Taterworks]

Re: free software

Dano, what you're asking about will only work for frequency response anomalies that are electrical in origin, such as a decrease in top-end extension due to inductance, or mechanical anomalies that are transmitted mechanically to the voice coil, such as impedance peaks and blips at mechanical resonance frequencies. You won't see effects of cone breakup on an impedance plot, and you won't see distortion effects either.

Sorry, but you really can't get around having a measurement setup to measure the driver's response. Download Audua's SpeakerWorkshop, build a Wallin Jig, and buy an inexpensive measurement mic (Dayton or Behringer) and an inexpensive USB audio interface like the M-Audio MobilePre and you're set. You won't break $250. Model the system as best you can with Bagby's tools (requires Excel 2000 or later). Then build your speaker enclosures, load the drivers, and measure driver responses individually to take into account the real-world effects of the cabinet, and design your crossover based on what you measure. Finally, tweak by ear.

If it's absolutely necessary to spend no money on a measurement setup, then do the best you can based on the measured responses on Zaph's web site, or the manufacturer's curves, and use Bagby's tools to model as much system behavior as you can, and then cross your fingers and hope for the best. That's really all you can do.

Happy building.

[joekraska]

Re: free software

I would say. It's quite likely that it is in fact not close. As you say, you model up to 200hz or so, but everything else really ought to be tracings or someone's actual measurement data.

What you didn't say is what you'd call "good enough". At what +-dB do you consider your approximation okay?

Joe.

[Jeff B.]

Re: free software

I need a program to model mid-range frequency response using Thiele small data. I read Thiele small data shouldn't be used to model past a few hundred cycles. I don't know but if it’s close that would be enough for my purposes. I would like to model some of the jamo buyouts, trace the curves and create frd and zma files.

You can model a tweeter's roll-off accurately with Thiele Small data (If you know it), however you can't actually model the frequency response of any driver with these parameters alone.

[dano]

I need a program to model mid-range frequency response using Thiele small data. I read Thiele small data shouldn't be used to model past a few hundred cycles. I don't know but if it?s close that would be enough for my purposes. I would like to model some of the jamo buyouts, trace the curves and create frd and zma files.
________
Wage insurance dicussion