i have the VR working ... how cool!

I'm now back at home in Canada, and have a suitable mic with me here. All ya have to do is drop by to try it out. I also have one of the disc's here..

Cheers

Ah, I didn't know about the 29.4KHz rate. I don't believe it was mentioned in the old kernels and I've never read the full specs for the CS4231A. I'm guessing that makes a big difference. There's a lot of audible information between 11.05KHz and 14.7KHz.

If you're looking for a way to regenerate sine waves of a known frequency then I may be able to help. About a year ago I created a hijack patch to use the DSP's tone generators to produce any tone on demand. Unfortunately the patch never made it into the hijack tree, and since then some of the existing code that the patch relied on has been removed (I used the calculator UI for numerical entry), so it's not a completely trivial merge. The patch included ioctls and also included a fixed point library for the kernel which could be useful for doing other DSP related coding, athough I guess you've got such a library in the player codebase anyway.

Anyway, if you are interested then once I get the new alpha then I can create a patch against it.

Can i drop by and try out the disc?

By all means, feel free!

Hm. Wonder how far of a drive it is.

Mmmm.. about 7-8 hours I should imagine -- it takes me about 6.5 hours to get here from the Mohonk Preserve (aka "gunks"), which is 70miles this side of NYC.

Cheers

Thanks Mike.

Hm, I wonder if I should buy like 10 of these mic's (no, not possessive, just using an apostrophie to pluralize an abbreviated word). Since I'd have to ship from the UK, I might as well make it worth the cost.

.... scratch that. A quick attempt to register with Maplins shows that they don't ship to the US. Hm, ok. No "Build Brothers Certified" mic for me.

I'm sure some one here will do the honours of mailing over a bundle of them to someone for forward to others.

I'm sure that we could find a local equivalent.

apostrophe*

Quote:
I'm sure some one here will do the honours of mailing over a bundle of them to someone for forward to others

I will gladly help out.

Don't worry too much about which mic element to use - I'll try and calibrate mine this weekend but I expect to find it's flatish from about 100Hz to 10KHz and just droops a bit outside this range. Pretty much any small condenser element should behave similarly.

Maybe someone over there in the US could choose an easily obtainable element and calibrate it (or you could post a sample to me and I'll do it, but my setup isn't exactly professional ).

One with a published curve would be better of course.

Cheers, John

Does this work for you? I can get this and send it over asap.

That mic. is unidirectional, not omnidirectional.

A more likely RatShack candidate is this one.

If that's not suitable then digikey have a long list of mic's. I'd be willing to do a bulk order.

It has to be a condensing microphone -- dunno if that one is or not.

On a related note -- running AutoEQ with NO MICROPHONE attached currently does not do what one might expect of it.. needs some fine tuning, I guess.

Cheers

Wow... And I was looking in the entirely opposite direction when it came to a microphone...

Hmm. I thought that all electret microphones were condensers...

What we do need to know is which parameters are important for this use. We know that we don't need a response beyond 15KHz as the sampling rate is 29.4KHz, although a microphone with an extended range may be flatter for longer. What what about the low end? The Maplin model goes down to 50Hz. We know the DSP is capable of EQing lower than that (17Hz IIRC), although it was limited in software (to 50HZ IIRC). How low does the autoEQ measure?

Is the sensitivity important? Presumably, because the software doesn't know anything about amp gains, it cannot make any assumptions about the audible volume vs the empeg's logical volume. As such, I would also presume that the software has to be fairly smart and would level the frequency response to the average volume, without caring what that volume actually was, hence making the sensitivity of the microphone non-critical.

Voltage and impedance requirements? The Maplin one is 4.5V (optimum) and 1Kohm. How much variance is acceptable?

Hi

Both Radio Shack microphones are condensor types and will be compatible with the car player's input (I think the phantom power supply is about 3V in a car player). There's a data sheet for the KUC1515-01-0850 at http://support.radioshack.com/support_supplies/doc9/9798.htm which specifies a wide voltage range of 2-10Volts. From this, and thinking about how these mics work, I believe the exact voltage isn't critical in general.

We do however need to use an omnidirectional mic, since omni mics measure the pressure due to the sound not the air velocity (hence we get a true SPL).

I also prefer the KUC1515-01-0850 as a "standard" unit because it has wires on. Soldering onto those mics is a bit tedious for someone with "normal" soldering skills like me.

I love the Behringer ECM8000 measurement mic - in fact the idea of AutoEQ came from a rainy day in Somerset with Roger, John Ripley and me in Roger's car with my ECM800 and UltraCurve2496 basically trying to determine how Roger's EQ should be set for a flat response. Sadly the ECM8000 specifies 15-42 volts DC supply (typical for pro equipment). It might just work anyway at 3V, but I couldn't guarantee it would still be flat, making the excercise a bit pointless. Another option might be to pre-amp the mic properly and use that signal.

The frequency ranges specified for the microphones is likely to be a "reasonably flat in this range" type thing. We should be able to compensate as long as the microphone doesn't lose so much sensitivity out of range that its signal drops below the noise floor.

I just looked at the EQ code, and can't find a trace of the limits that I added based on some research by someone on this BBS into which low-frequency settings make the EQ go wrong. It's in there somewhere, I'm sure . Basically it replaced the old 50Hz limit with a combination of limited boost/cut limits at frequencies below 50Hz.

AutoEQ bands are centred on:

• 25
• 40
• 63
• 100
• 160
• 320
• 800
• 2k
• 5k
• 12.5k

In the absence of calibration data you should only rely on the results for bands comfortably inside the mic's flat range. Other bands can be set to 0dB or set by ear.

Best microphone positioning is near your head, because that's where your ears are. You also need to keep background noise as quiet as possible (including handling noise). While AutoEQ is running, the level should be within the range of the cute little VU meter and ideally >=10dB higher than the background noise level. AutoEQ won't do this for you, you have to do it yourself and that may take a few trial runs.

Cheers, John

I love the Behringer ECM8000 measurement mic - in fact the idea of AutoEQ came from a rainy day in Somerset with Roger, John Ripley and me in Roger's car with my ECM800 and UltraCurve2496 basically trying to determine how Roger's EQ should be set for a flat response. Sadly the ECM8000 specifies 15-42 volts DC supply (typical for pro equipment).

I can think of no better recommendation. And, because I have a 9V battery operated phantom power supply (as linked to in the other thread), I think this will be a good choice. Thanks!