[Matt] likes to make videos (and he’s pretty good at it judging by the quality of his videos). But video isn’t much without audio. Handheld recorders with small built-in microphones have a fairly high noise floor so [Matt] has a Rode NT1-A — a pricey but very quiet microphone. However, for field work, it isn’t handy since it requires a power supply and preamp to go along with it.
Another problem is that for stereo recording you need two and because they are quiet, they tend to pick up handling noise so you probably need to mount them on tripods. That’s all too much to carry around, especially on a hike. So [Matt] cannibalized two microphones. He repackaged them in a shock mount (made from a bird feeder and elastic), and added a battery pack and a custom preamp. The shock mount eliminates the handling noise and the custom PC boards mean you don’t have to carry a lot of extra gear.
The end result (see the video below) looks like someone made a purse out of a tribble, but it does sound good. If you hang on through most of the video (of fast forward to about 7:25), you can hear the microphones picking up thunderstorms, the ocean, the wind, and even [Matt’s] heartbeat.
[Matt] claims the batteries provided can keep the device running for 24 hours of continuous run time. There’s a strap to make it easy to carry, and a handle made from a 24-bit digital recorder rounds out the design.The positioning of the microphones makes a sort of binaural effect, although the body of the device probably doesn’t model the human head like a proper setup would.
If you are interested in microphones, you might want to tear up an old hard drive. Or you can see how ribbon mics are made.
The eighth letter of the alphabet is “aitch” not “haitch”.
In ‘ertford ‘ereford and ‘ampshire ‘urricanes ‘ardly hever ‘appen.
… and by the way it’s “Hotel” :-p
That depends on who you ask. Proper pronunciation of words and letter varies from place to place.
I would have thought that anyone on an internationally popular site like this one would have understood that already.
+1
It is not dependant on anything. Look it up in a dictionary if you have one.
And that depends on which dictionary you’re reading, if it doesn’t bother including multiple pronunciations. ;)
“H (named aitch /ˈeɪtʃ/ or haitch /ˈheɪtʃ/ plural aitches or haitches)[1] is the eighth letter in the ISO basic Latin alphabet.”
“H” Oxford English Dictionary, 2nd edition (1989); Merriam-Webster’s Third New International Dictionary of the English Language, Unabridged (1993); “aitch”, op. cit.
Congratulations, you’re illiterate.
Looking at the pic with the fluffy thing, I thought it was something strapped to his dog. Then I read the article.
I’m not convinced low noise microphones require you to spend more than $220
I mean noise is the pre-amp, pre-amps can be made good and cheap with modern technology. So I think you can find relatively cheap mic capsules that don’t have that much noise.
Of course as always it’s hard to find the wheat amongst the chaff. And unfortunately the ‘reviews’ people put up are often nonsense too, so that doesn’t help as well as you might hope.
Anyway it’s a nice setup he made for himself and kudos to him for just doing it and being successful.
That would be an article I’d like to read. A variety of mic’s using the same low noise floor pre-amp. Guess I should start checking the audiophile forums.
Audiophiles don’t really use microphones, they mostly sit and listen to vintage vinyl while filling the room with hot air complaining about oxygen levels in their wiring.
I didn’t know that, thanxz.
And yet, if you cruise any audiophile forum, you’ll find microphone reviews scattered amongst the speaker/headphone reviews and builds.
I guess it’s a physical limit and a trade off between noise floor, sensitivity, frequency response, dynamic range.
Small capsule are in the -60dB.noise floor.
I would like to know where the cheap electret capsules get replaced by something better. Until that dividing line, you might spend more money without improvement.
They do have an FET in there, small but an actual plastic encased FET. That might be the weak point, though the time I took a capsule apart, I’m not sure one can get them apart and back together.
Michael
That’s the thing with electret, they almost always have a FET pre-amp right in the capsule, and that is the thing that is the source of any noise. Now how much noise is obviously dependent on the quality of that FET, and on the purity of the power you feed to it of course. The actual size of the capsule should only affect sensitivity and frequency response not the noise floor I wold think.
I wonder if you could get a better quality electret if you took one from a high-end phone. Although those might be frequency limited.since you don’t need full range for a phone really, I just can picture Apple under Jobs having done some good effort to get the perfect capsule, they used to be like that when he was alive. I think they lost some of that these days though.
But as the size (of the actual vibrating element) gets bigger you need less amplification to get a good signal. Hence the noise is amplified less as well.
But larger means more cost, and if you spend that on the diaphragm, then a better FET is in order too. It all adds (or doesn’t!) together.
That is why quality microphones are always large.
a small diaphragm has less mass and so should move more easily though.
Even when in reality that doesn’t seem to play.
Yes he always does very professional builds. However you do raise an interesting point, what if you also got multiple cheap capsules and averaged their output, on the assumption that noise is self cancelling over multiple samples?
Check out the specs for this ADC from TI, http://www.ti.com/lit/ds/symlink/ads1256.pdf
Do a search on ebay for “ADS1256 24Bits ADC AD Module Converter Communication Modules Module Kit” and you will see they are about USD $30, just add the capsules, and note the differential input option that a multi-capsule noise reduction hack could potentially use. Also note that it has programmable gain. So you’d need two, one for each channel and then perhaps a USD $8 NanoPi NEO to record the digital audio streams over SPI.
So there you go as much (or much more) functionality for about $100, all depending on if you can pull off the hack with the capsules.
30KSPS means an nyquist frequency of 15kHz, which is probably not the ideal ADC for high quality audio applications.
I don’t think noise is self-canceling in the case of sound, it works with image sensor as we all know, but I don’t think multiple microphones can do that because you don’t have the same dots at the same spot like with image sensors.
Although you could do location of the capsules and then calculate the phase differences and maybe do some reconstruction that way? But I think it’s probably much better cheaper and more reliable to get a good single source in the case of sound.
Not that sort of noise! Noise as in a single circuit vs multiple identical ones or pair that are the inverse of each other.
You said multiple capsules and averaging their output, that’s what I think I was responding to.
Shrug.
Well it looks like a Polar Bear’s penis (I would imagine) but it sounds good (listens good?) and that’s the point.
hm….LM2577 as powersupply? A 52kHz boost converter is not quite the best option for a “low noise” audio solution, isn’t it?
Hmm… 52khz is well beyond the useful audio range, so probably won’t affect it too much?
Have I just fed a troll?
If you sample a 52khz signal at 48khz, you get a 4khz signal. On the other hand current in this case should be pretty low, so any 52khz interference or supply ripple is likely to be miniscule already, and the recorder should have an appropriate LPF stage for its sampling frequency, so I doubt it’s causing any detectable interference.
Specs would be nice, the blurb on here did not go into any tech detail at all. What capsules did he use? What did he use for preamps?
He used the rode mics and this pre-amp http://sound.westhost.com/project66.htm
Sometimes you gotta go read the build log (or in this case, youtube description) to get the details.
I thought that sound of his fingers touching the touchscreen and his heartbeat from a slight distance thing was a good demonstration of what low noise can achieve.
But the CIA must have smaller mics that can do it expect (and fear)right?
Oh and incidentally, that link is dead, but he shows a nice fullscreen of the schematic in the video though.
Works fine for me?
Both the one I posted & the one in the video description.
I get a security error, mismatch certificate, and when I add a temp exception I get “Sorry, this page was not found”
For completeness, the error I get is:
“www.sound.westhost.com uses an invalid security certificate. The certificate is only valid for the following names: westhost.com, http://www.westhost.com ” And yes, it adds te www first, it also changes to a secured https.
Does this work?
Opps.
link
I still struggle with this type of linking.
Just put this in address bar without brackets.
[sound.westhost.com/project66.htm]
Nope, putting that in the address bar did not work. It’s the damnedest thing.
I even tried using a proxy and an alternative DNS and even another browser.
And I saw pone comment on that youtube where a person also said he could not get to the site. So I will have to ponder on this one.
I also don’t get why it insists on redirecting to a HTTPS version.
And incidentally, where the other day I got that error I described today is different and it says I can’t reach the site at all “”can’t establish a connection to the server at http://www.sound.westhost.com.”(still added the www in front for some reason, probably because the ”sound. didn’t connect and it tries ‘fixed”alternatives automatically).
And another try says connection refused. I guess the site helps me out by blocking me completely so I dont have to worry anymore :)
Would you trust a guy who can’t even correctly pronounce words? He actually missed using an arduino.
Don’t you mean “Harduino”?
There are worse shortcomings. And some of those don’t even impede you from being a viable presidential candidate.
So meh, let him screw up the H.
hm…low noise and LM2577 as powersupply? A 52kHz boost converter seems to be not quite the best option for a “low noise” audio application… ;)
Maybe because there’s a huge amount of filtering on the power input? It’s not like there’s any signal at that point you need to be careful about.
And, considering the results obtained, it’s blindingly obvious the powersupply was perfectly fine.
I would wonder if it could beat or cause sampling birdies in a DAC. Otherwise with filtering and hearing, not a problem.
I had to plug the phones straight into the computer (multi channel monster) but it is worth it. Perhaps a little less depth than dummy head, but clean fur sure. I assume that the real gain here is with the large area mics (Mick’s) vs. the tiny but capable electrets.
I have pondered how to put the large mics into a dummy head, but wonder if there is a loss of image clarity because of the larger area of pickup with the large diaphragm? Time smear.
I will pass on the being put on hold music, or our candidate believes in… music under.
Well with a dummy head you’re deliberately introducing psychoacoustic effects (an HRTF) into the signal. If all you want are binaural recordings for their own sake then fine, but if you’re going to mix the recordings into something else, it’s not generally what you would want.
A big, furry microphone cover for shielding against wind noise used to be called a dead cat.
And it still is. Even RØDE calls their windscreens Deadcat: http://www.rode.com/accessories/deadcat