Tuesday, November 20, 2007

Typical "20 to 20" response curves

You see a lot of curves shown so the response looks fairly flat. Here is one of them. I see that the numbers are a little hard to read. The bright vertical bar on the right corresponds to 10kHz. As you can see, both mics tend to drop off a bit after that point. There are other ups and downs at the lower end of the spectrum too. What we are looking at is the compared response curves of the Proscenium vs a Cardioid microphone. The red curve is the Proscenium, our most rolled off mic, and the blue is the cardioid. (more about that) (even more about that) The point of this post, however, is to show the "typical" Y axis steps, which are compressed. We have in our printed specifications chosen to "uncompress" them a little, in order to show the detail as we see it. That makes the mics look even less "flat". But it doesn't change their sound.

Amplitude and frequency plots tell too little about the way a mic or any transducer sounds. The process of receiving, converting and sending acoustic signals as electrical signals is surprisingly complex, even for the most simple forms of transducers, and
tone color,
field patterns,
impulse response,
acoustic impedance matching,
acoustic resistance,
electrical load,
transformer design,
housing design,
transducer orientation,
electrical contact quality,
noise shielding,
ohmic resistance,
magnetic shielding,
flux leakage
and many other factors predominate tone and affect the application of a particular device.

Here is a link to a gearslutz thread relating to this post.


Gordon said...

What polar patterns (all of them?) are you be able to create w/ ribbons? Any idea when you'll have a product on the market that isn't fig8?

Bob Crowley said...

Yes, all of the patterns plus some more.

Naked Eye is so asymmetrical that I hardly call it figure 8. A hypercardioid pattern is available as an SPLx and we even used one at AES. Special order, about $3500 I'd guess, depending on what you specifically want.

Herb said...

"Amplitude and frequency plots tell too little about the way a mic or any transducer sounds. "

There are way too many people out there that just don't understand that. Frequency response plots can be powerful tools for those that understand the plots and understand what they want. Unfortunately people have the impression of "give me a flat mic/speaker/etc from 20 to 20 and I'll be happy."

You can go to Bruel and Kjaer right now and buy a measurement microphone that measures with +/- 0.5 dB from 5 Hz to 30 kHz for about $1,000. And I'll guarantee that when you listen to the recording you made with it, you'll hate it.

Microphones are instruments. Just like you wouldn't depend primarily/solely on a frequency response plot of a guitar to choose your guitar, neither should you put too much stock in a mic's freq response unless you're someone who clearly understands FR, phase, impedance, diffraction, polar response, etc.

If you're not in that group, I'd highly suggest using a simpler tool: your ears.