The View Up Here

Random scribblings about kites, photography, machining, and anything else

  • Flickr Gallery

  • Advertisements

Archive for the ‘Uncategorized’ Category

DIY Stereo Parabolic Microphone

Posted by Tom Benedict on 21/12/2017

I’ve had a parabolic dish I bought off Ebay sitting around for the better part of a year. The whole idea was to build a parabolic mic out of it based on the dishes from Telinga, but for one reason or another I got stuck in the design-and-abandon phase and never got to the just-cut-metal phase. So the dish sat, collected dust, and didn’t do much except confuse my cats and annoy my family.

At some point in there I picked up eight EM-172 capsules in two matched sets of four from FEL Communications aka Mic Booster along with a bunch of yoga blocks. My plan with these was to make a pair of quad capsule foam SASS arrays similar to Vicki Powys’s for recording surround sound. I got stuck in design-and-abandon on this project, too.

In the end, though, it all worked out for the best because of yet another project. Only, this one actually reached completion, mostly because I was doing it for someone else. (There’s nothing like a deadline for getting a designer off top dead center.) The project involved designing and 3D printing some mating threaded parts. I modeled the threads in CAD, put the clearance into the model, and printed the parts. The threads meshed first try. W00t!

That was reason enough to celebrate, but it was also reason to scratch my chin and go, “Hmmm!” Being able to 3D print threads solved one of my issues with the parabolic mic: how to attach the handle to the dish. While reading this article from Avesrares on a DIY parabolic dish I had my second epiphany: Their stereo parabolic mic used four matched EM-172 capsules… just like the ones I had sitting in an envelope. Thank goodness I hadn’t built those SASS arrays, either!

So with technique and components in hand, I set about making my own stereo parabolic mic.

The Dish

One difference between my DIY parabolic mic project and many others is that I made no attempt to fabricate my own dish. There’s an Ebay seller who makes 22″ Lexan dishes for under $50USD. Shipping to Hawaii was expensive, but the dish itself was pretty affordable. This was where the whole project started.

The Handle

I like how the Telinga handles attach to their mics. The Avesrares article mentioned using auxiliary drill handles in a similar fashion, so I searched Ebay and found one of a suitable size for under $10US. The next step was to model a corresponding threaded tube and two screw plates to attach the handle to the dish.

DIY Parabolic - Clamp Tube and Collars Render

I wanted to be able to build interchangeable microphones, so I sized the inside diameter to take 3/4″ schedule 40 PVC pipe. The screw plates are 4″ in diameter – about the maximum my 3D printer can handle. The screw pitch is 14 TPI (sorry, metric system), and was sized to fit the outside diameter of the tube. If I remember right the threads left 10% flats top and bottom, and I left 0.020″ clearance between the tube and the screw plates. (They meshed great, too. W00t!)

Since these are sandwiching a parabolic form, the inner surfaces of the screw plates are curved to match. I added a layer of craft foam inside and out in an effort to provide some damping for handling noise (which it didn’t) and to make a nice snug fit (which it did.) Everything went together fine.

DIY Parabolic - Clamp Tube and Collars

The Microphone

In an article written by Klas Strandberg from Telinga Microphones, he explains that the focus of an acoustic paraboloid isn’t a precise affair. The mathematical focus of the paraboloid will be the point of best focus, but the size of the region of focus is frequency-dependent. I’d try to paraphrase here, but the article puts it so well there’s really no point. If you’re interested, read his article.

The upshot is that you can, to some degree, adjust the EQ of a parabolic mic by shifting where the capsules are with respect to the point of best focus. I knew I wanted some latitude, but I didn’t know how much range I might need. So I cut the 3/4″ PVC pipe a little long, reasoning I could always cut it down later.

One bonus to using 3/4″ PVC pipe is that Neutrik XLR connectors are a nice snug press fit into the tube. Since I was planning to make a stereo microphone, and since I’d already built an XLR-5 to twin XLR-3 splitter cable for my MS Alice microphone, I used an XLR-5 connector for this mic as well. While not the most elegant setup for this connector, it was an expedient solution that let me get on with the business of building the rest of the microphone. I can always revisit this later.

(As a quick aside, “I can revisit this later” is designer code for “You’re going to see this in the final version.”)

DIY Parabolic - Handle, XLR5 Connector, and Reflex Sight

My first design for the business end of the microphone followed the logic in the Avesrares article: a baffle that splits the dish left-from-right and a second baffle facing the dish. Unfortunately my dish has a shorter focal length than its depth, so adding the second baffle would’ve cut off about half the surface area of the dish.

Instead I went with a baffle similar to the one used on the Telinga microphones, except that instead of using the microphone capsules as pressure zone transducers (PZT), I stuck to familiar territory and arranged them as boundary mics. I’m not 100% sure this was an ideal approach since it moves the diaphragms of the capsules away from the axis of the dish, limiting how close they can get to the point of best focus in the radial direction. Still, it worked.

DIY Parabolic - The Giant Lollipop - Render

Originally I’d intended to make solid baffle plates on the 3D printer, but these would’ve been costly in terms of material and time. Instead I took a tip from the SASS I built and made them only 1mm thick. This made for a very thin, unfortunately flexible plate with a high resonant frequency. I backed it with Dynamat, which does an excellent job of damping vibrations, and sandwiched dense latex foam between the two plates to provide further damping. The whole stack wound up being acoustically dead to handling noise. (But more on the handling noise later.)

Unfortunately, before I took any pictures of the finished parts I stuck open celled foam on either plate to provide wind protection. So the only photos I have look like a big foam lollipop.

DIY Parabolic - The Giant Lollipop

I wired the capsules using the same pattern as Vicki Powys’s SASS, except instead of wiring them into a 3.5mm TRS plug, I wired them into the XLR-5 connector using David McGriffy’s and Ricardo Lee’s Simple P48 circuit. Tuning the resistor for the Simple P48 circuit took some experimentation since each circuit was now driving two capsules instead of one. I’d quote the value of the resistors I wound up using, but there’s really no point. If you choose to build one of these, best to tune the resistors to get the bias voltage you want for your capsules.

With everything closed up, the microphone slides into the clamp tube on the dish and locks into place by twisting the drill handle. Voila, the DIY stereo parabolic microphone.

DIY Parabolic - Finished and In the Field

The Bling:

With such a narrow pattern microphone, I erroneously thought I’d need some way to aim it. (I’ve since used it in the field enough to know you aim these things by ear and not by eye.) One of the remnants of a previous project of mine, the camera helmet, was the small reflex sight I’d used for aiming the cameras. I mounted a short length of Weaver rail to the drill handle using a 3D printed adapter, and clamped the reflex sight to the rail. Voila, a neat looking, but ultimately useless addition to the parabolic mic.

This was rendered even more useless by the next bit of bling.

DIY Parabolic - Useless Reflex Sight
Wind Protection:

After some initial testing in the field, I found I wanted the option of another layer of wind protection. The foam on the lollipop was nice, but it wasn’t really enough to stop the kind of wind you get outdoors. (Foam is never enough to stop real wind. Like ever. Eventually I will learn.)

DIY Parabolic - Wind Protection

Chris Owens wrote a really good article describing the wind protection he made for his Telinga parabolic mic. I followed his directions to the letter and wound up with a very serviceable (and machine-washable!) cover for my mic.

DIY Parabolic - Wind Protection Detail

He used his dish to measure a circle of cloth, and rolled a hem at the edge that holds a length of shock cord. Once the hem is finished, the shock cord is pulled tight(ish) and is tied off. Chris Owens cautioned not to pull the shock cord too tight or the dish will warp. I made mine just snug enough to make sure the cloth stays on unless I’m ready to pull it off.


Post Script:

Now that I’ve had the chance to use my dish in the field I’ve found a number of shortcomings I’m planning to address in the future:

  • Anything you can’t put a shoulder strap on is lame – My dish is incredibly awkward to carry on a trail. Klas at Telinga solved this by making 1mm thick dishes that can be rolled up and carried in a shoulder bag. I’m sticking with my dish, but I have to add a shoulder strap. This is irritating.
  • Handles that can twist bother me – The drill handle I chose tightens (and loosens!) by twisting the handle. I’ve had it try to come undone in the field a couple of times when turning the dish. I’m planning to replace mine with a Bosch handle that uses a thumbscrew.
  • Handling noise is lame – I’ve been chasing handling noise on the BM-*00 conversion mics I’ve been building with some success, but this thing has it twice as bad. You can’t just shock-isolate the capsules. You have to shock-isolate the dish as well because it basically acts like a great big sounding board. Anything that’ll conduct handling noise to the dish shows up in the recording. A cushioned handle like the ones on the Telinga dishes would be a step in the right direction. For now I’m wearing gloves any time I use it.
  • I’m not 100% happy with the mic setup – The first prototype mock-up I did with this dish involved mounting my Alice M-S at the focal point of the dish. This worked way better than I suspected, and provided a better stereo image and a better mono image than the mic I’ve got in there right now. I’m thinking of recovering my four EM-172 capsules so I can go ahead and build that SASS surround mic, and building a dedicated mid-side mic to live in the dish.


And now for the rest…

First, A Sob Story:

I like to include audio samples in my build articles. Unfortunately I can’t for this one. Not yet, anyway. A little over a week ago I got some kind of outer ear infection in both ears. The tinnitus I’ve experienced for the last twelve years or so is about a bazillion times worse at the moment, and my right ear has lost about half its frequency range. I tried to edit a recording I’d made of a katydid and realized I couldn’t even hear the katydid in that ear. (And it’s not like they’re subtle or anything.) Even worse, my two ears currently hear broad-spectrum noise as two completely different sounds. Until this infection clears up, my ears are screwed. I can record, but I can’t edit. Sorry.

Second, A Note on the Photos and a Thank You:

No, that’s not me holding the mic in the field. I was the one tripping the shutter. (I know which end of the camera I like to stay on.)

About half the pictures in this set involve wet gear. This is one of the hazards of recording in Hawaii: it rains. Rain is not ideal for recording with a parabolic mic. Each rain drop on the dish was picked up by the mic, and there were plenty of rain drops. I did make some recordings during the photo session, but because of the rain none of them really showed off the mic very well. Everything survived fine, though.

The black audio bag worn by the recordist in the photos is the brainchild of Andrew Jones, described in an article he wrote on WAV.REPORT, Tutorial: “DIY” Audio Drop Bag. Mine’s only about half finished (it still zips shut) but it’s a really good article and an unbeatable bag for the price. Thanks, Andrew!

Third, An Apology:

Normally at the end of my build articles I like to link to all the necessary bits and pieces, including any parts I designed. I’m not 100% comfortable doing that for this project because it’s still such a work in progress and there’s so much work still to be done to make it what I want it to be. But if you’re dead set on building one of these, feel free to contact me.



Posted in Uncategorized | 2 Comments »

Useful Little Clampy Things

Posted by Tom Benedict on 20/12/2017

A while back I needed the ability to put a 1/4″-20 threaded stud… Somewhere. Anywhere. Everywhere. I tried some things on the 3D printer, but realized none of them would be as robust as I needed these to be. So…

I turned to Ebay!

A Useful Clamp

Turns out you can get insanely cheap clamps by searching on “super clamp”. Some come with T-bar handles, some with offset handles like this one. Some come with articulated arms (which I used for mounting cameras on our vacuum chamber for work). And some just come with 1/4″-20 and 3/8″-16 threaded holes. Mine came with threaded holes.

I added the 1/4″-20 threaded stud by chopping the heads off some bolts and cleaning them up with a grinder and a die. The thumb screws came off of a camera hotshoe to 1/4″-20 adapter (also courtesy of Ebay.)

All in all these things came together for less than $6 apiece.

A Useful Clamp with Shock Mount

My original purpose for these was to be able to space mics anywhere I wanted across a bar for trying different stereo recording techniques. To do that I needed to be able to mount any of my self-contained stereo setups, all of which take a 1/4″-20 thread, or to mount individual shock mounts like this one.

A Useful Clamp with Ballhead

But there’s no reason to stop there. Pop a small ballhead on one of these and you can mount a camera, a handheld sound recorder, an off-camera strobe, or whatever else strikes your fancy.

The only problem I’ve found with these is that the handles like to come off, leaving the clamp firmly attached to whatever you cranked it onto. If you decide to get any of these, be sure to test them first, and if any handles come off, re-install them using red or green Loctite. (I had to re-install three of mine!)

I wound up building five of these, but now I wish I’d ordered more.

P.S. Since writing this, I’ve been told this is all pretty standard fare for lighting and grip gear on a film set, and that re-purposing lighting gear is a cheap and easy way to build solid recording setups. My take-away: 1 – There’s nothing new under the sun with this article. 2 – My butt’s gonna go numb the next time I sit down with the B&H catalog or start surfing their web site! 3 – Neither of these is a bad thing.

Posted in Uncategorized | Leave a Comment »

Clean Clean Clean!

Posted by Tom Benedict on 13/12/2017

I’ve spent the past year or so at work helping prepare for the arrival of a new instrument. One of the requirements is for the room to meet ISO 8 standards with a goal to reach ISO 7.

The reason for the requirement is that the instrument is one big cryostat – about 2m in diameter and about 3m long – and will be bolted to the floor. Once it’s installed, the only way to service it is to open it up wherever it is. It can’t go in the clean room, so we brought the clean room to it.

ISO 8 roughly translates to Class 100,000. It allows for 29,300 >=5µm particles per cubic meter, 832,000 >=1µm, and 3,520,000 >=0.5µm particles. ISO 7 , our goal, knocks that down by a factor of ten (2,9300 >=5µm, 83,200 >=1µm, 352,000 >=0.5µm, Class 10,000).


Coude Room Particle Counts 2017-12-13

As of this week we’re hitting under 175,000 for 0.5µm, 75,000 for 1µm, and under 3,500 for 5µm with upwards of ten people in the room, two of them installing weather stripping on the doors. That’s well inside ISO 7 with a shot at ISO 6 once we finish all the work, provided we can demonstrate counts at 0.3µm and smaller. (That instrument is in use in another room at the moment.)

When no one is in the room, the particle counter doesn’t register any particles at all. W00t!!

Just a happy point in a long work schedule. Details of the work on the room will be presented in a paper at the SPIE Astronomical Telescopes + Instrumentation conference in Austin, Texas in June, 2018. The instrument itself will show up at the end of January, 2018, so we’ll have plenty of time to put the room through its paces during integration and commissioning.

Posted in Uncategorized | Leave a Comment »

Clippy ORTF / X-Y Bar

Posted by Tom Benedict on 25/11/2017

One of the real pains about setting up some stereo mic arrangements is positioning everything juuuust right. The ORTF setup is the classic example. It uses two cardioid microphones positioned so the diaphragms are 170mm apart and angled outward at 110 degrees. Possible to do with a stereo bar? Sure. Easy to do? Not so much. Not without some way to verify the angle and spacing.

My first cardioid mics were a pair of Primo EM-184 cardioid capsules mounted in Clippy lavalier bodies from FEL Communications ( Take all the normal complications of trying to set things up for ORTF and complicate them by doing everything with lapel clips. Possible to do? Sure. Easy to do? You gotta be kidding me.

Back in May of 2016 I wrote about a nifty little bar I made in the shop to hold everything in just the right place. I machined it out of a chunk from the scrap box that wound up being 7075 aluminum. Massive overkill, all things considering, but it worked. I profiled the bar and machined slots in it so the clips would hold the mics juuuust right. It worked, and it made the job of setting up for ORTF a snap.

Since then I machined another set of slots and cut-outs in the bar so I could also set up for X-Y. (Almost X-Y… the capsules aren’t quite coincident, but they’re quite close.) This also worked great, though I still have problems remembering to swap the mics. (ORTF mics face outward, X-Y mics face inward, so the left and right mics are on opposite sides for the two setups.)

My plan was always to make this design available for others trying to do stereo on a budget. I dropped the ball and never got around to putting the design out there, but I finally fixed that.

I put the STL file up on Thingiverse for those who want to grow their own, and on Shapeways for those who want to get one pre-made (or whose printers don’t have enough volume to print the whole bar (like mine!))


Posted in Uncategorized | Leave a Comment »

BigMic With a Side of Cheese Sticks

Posted by Tom Benedict on 15/11/2017

Some projects just never stop being projects. And sometimes the resulting feature creep isn’t necessarily a bad thing.

A couple of months ago I was trying to find a good acoustic space for doing voice acting, and after pulling my hair out trying to find a quiet spot in my house (which happens to be within twenty feet of a highway) I finally settled on using my Civic. It’s mobile, it’s sound-proofed…ish, and it’s actually a pretty comfy place to sit. I added some acoustic treatment to reduce the remaining standing waves in the car and moved on to the next problem: The electronics.

To make the Civic work as a mobile sound booth I needed a way to mount my scissor arm while still letting it act like a scissor arm. So I machined a bar that would clamp to the headrest uprights on one of the front seats. The bar had a row of threaded holes so I could get coarse left-right adjustment to center the pivot point of the scissor arm, and an upright with a 1/2″ bore for the scissor arm to socket into.

"Cheese Stick" with scissor arm and vocal mic

While I had the thing on the mill I added some 1/4″-20 threaded holes through the top of bar so I could stick a ball head on the opposite side from the scissor arm socket. This let me flip the bar over to use as a mic mount for doing in-car recordings. It’s shown here with my SASS bolted on top.

(As a side note, this isn’t actually the best setup unless you’re trying to record the sound of the car itself. The SASS uses omni capsules, so it picks up every sound in the car.)

"Cheese Stick" with SASS for recording in-car stereo sound

Since the threaded holes were so useful I went back and had a mad drilling session on the mill to pepper the thing with threaded and clear holes for doing… whatever…

That’s when I found out there’s a name for this thing. For the drilled and tapped end of this thing, anyway: It’s a “cheese bar”.

But when I looked up “cheese bar” on Google all I got were these specialty cheese shops that are set up like bars. Go figure. (It’s worth searching on “cheese bar” just to see what these places look like!) Searching on “video cheese bar” got me closer to the mark.

Since mine started life as something else entirely, I hesitate to call it a cheese bar. There’s the headrest clamp at one end and the scissor arm socket at the other. Rather than add confusion to confusion, I decided to call mine something a lot more descriptive: the “cheese stick”.

It’s probably just as bad a term as “cheese bar” when it comes to Google searches, but now I can truthfully say “I’ll have a big mic with a side of cheese sticks!”


I’ve used the cheesy side of the thing a number of times, now. While doing EQ testing on a bunch of microphones I clamped it to a stand and lined the mics up across the top. Over the last weekend I used it to build something a little more ambitious:

Double-MS setup using the "cheese stick" for fixturing

This is a double-MS setup entirely built using Alice microphones. The center mic is my self-contained MS Alice, but in this case I’m only using the figure-8 for the side channel. The other two mics are a pair of the TSB-25AX Alice mics I built, set up as the front and rear cardioid channels of the double-MS.

I haven’t taken it out in the field yet, but it’s a pretty straightforward setup to put together. I’m hoping to do field testing with it in the next couple of weeks.

Meanwhile I took the test files I made in my house and tested the post processing toolchain I’m planning to use. I started off doing the mid-side decoding by hand, but in the end ran them through the Schoeps Double-MS VST plugin in Reaper, which did an excellent job of generating 5.0 surround sound. If I add an omni on the fourth input of my recorder, I can use it for the LF necessary to do full-blown 5.1 surround.


And all because of the feature-creepy cheese stick!


Posted in Uncategorized | Leave a Comment »

Contact Mics – Four Channels Good to Go

Posted by Tom Benedict on 05/11/2017

I finally finished installing the other two channels in the impedance-matching preamp box I built back in July for plugging contact mics into my recorder. The amplifiers I used came from Stompville in the UK. They’re compact, low noise, and let plenty of the low frequencies through.

Contact Mic Preamp Box

Side note: I just checked the Stompville web site and they’re running out of the JFETs they used in this design. They’re going to offer a new product using a new JFET, so if you build something like this you may want to order all of yours at once rather than piecemeal the way I did. Otherwise there’s no guarantee of a match between channels.

I also finally packaged my contact mics so they’re not just copper tape wrapped piezo discs like the one I used for my kite line recording. (I eventually broke this contact mic through rough handling.)

Contact Mic on Line

Coming up with a housing for the mics took some thought. You don’t want to do anything that changes the sensitivity or frequency response (an impossible task, but you try to minimize those effects) but you also want a housing that works with how you intend to use the mics.

Most of the time I’ve attached my contacts using double-sided tape or Blue Tak (thanks to Tim Prebble and his excellent post, The First Rule of Contact Mic Club for that tip!) Other times I’ve used big honkin’ magnets (thanks to Richard M and his post on his Megalithia site for that tip!) But sometimes the easiest way to attach a contact mic to a thing is to clamp it.

In all the testing I’ve done with piezo contact mics, the biggest gotcha I’ve seen is that they really don’t like to have pressure applied directly to the top of the piezo element. It introduces a big DC offset that scales with pressure. What this means is that clamping directly to the center of the disc is not just bad for the piezo disc. It’s also bad for anything it’s plugged into. So clamping directly to the disc is a big no-no. (That’s also how I cracked my tape-wrapped contact mic!)

I wanted to find a housing for my contact mics that didn’t change the characteristics of the piezo more than strictly necessary, was made of something a magnet would stick to in case I wanted to use a magnet mount, and let me use spring clamps to attach the mic without applying any pressure to the piezo disc. Here’s what I made:

I started with the piezo disc recommended on the Stompville site, the Murata 7BB-35-3. I got mine from Mouser, but you can get these from a number of sources. They’re unfortunately quite large, with the outside diameter a whopping 35mm.

Murata 7BB-35-3 Size

After a couple of months of poking around and not coming up with much, I started searching on metal cans just to see what was out there. I found that the 1/2 oz screw-top tins used for lip balm were a perfect fit. Rather than caking on the lip balm for a couple of months to get empties, I found a supplier who sold them in bulk for DIY folks like me. (Though I think they meant them for people who make their own lip balm rather than people trying to cram electronics in them.) The Murata piezos fit fine.

It Fits!

One of the things I like about the Stompville preamp is that it uses three wires for the piezo: positive, negative, and ground. The ground is intended to be used as a shield. Since the lip balm tins are actually made of steel, this provided a nice way to provide a continuous shield all the way from the piezo element to the recorder. This seriously helps minimize RF interference. The only catch is that the tins are coated to keep them from rusting when used for lip balm, so I had to sand some clear patches for soldering the shield as well as for making good contact between the container and the lid.

The good thing about this coating is that it keeps the brass disc from shorting out against the shield. No special treatment is required. I glued the discs using the same E6000 silicone I use for mounting microphone capsules to their posts. This probably loses me some high frequencies, and epoxy would probably be a better choice, but I kept the layer as thin as possible so the effect should be minimal.

As it turns out, the same servo grommets I used for providing isolation on the shock-mounted capsule posts for the Alice mics are a really good fit on Mogami lavalier cable. I drilled a hole in the side of each tin to take the cable, installed the grommet, and glued it in place with E6000 as well.

Contact Mic in a Tin

With everything installed, all that was left was to solder all the wires and close it up.

Four Contact Mics Ready to Run

Not quite true. All that was left was to solder all the wires, close it up, and go play.

So why four channels?

Initially I knew I wanted four because my recorder has four inputs. I hate having inputs I can’t use simply because I didn’t plan ahead. But after playing with the two channel version for a while I realized something about contact mics: Depending on the medium, sound may not travel all that far. Having widely spaced contact mics may mean that an activator traveling across the object will move out of the range of one mic before it really starts getting picked up by another. With four contact mics you have all the same choices available that you would setting up the mics for a concert or for field recording. In this case, setting up a mic for hard left, center left, center right, and hard right can fill in the gaps on an object that naturally damps sound, like a chalk board.

Or you can use them to pick up four parts of an object simultaneously, like a coffee maker or a vacuum pump or a helium compressor.

Or maybe I just had four inputs and wanted to stuff them all full of contact mics. Who knows? The cost was minimal, and now I have the option to record four channels of contact mic.

One last thing: In addition to being good mics to have in your arsenal, contact mics are useful for other projects as well. June last year, I wrote an article about a shock mount for my SASS I’d made from a re-purposed multi-rotor anti-vibration mount. I used contact mics to characterize the shock mount and arrive at the -21dB of attenuation figure for that mount.

More recently I’ve been trying to solve a handling noise issue on the Alice mics I’ve been building. Having four channels to work with may finally let me tackle that in a more systematic way, or at least characterize the work I’ve already done.

But I may need some smaller discs to work with. The big Murata discs I got for this project are beasts.


Posted in Uncategorized | Leave a Comment »

LDC Alice – Take Two

Posted by Tom Benedict on 16/09/2017

I decided to build another set of Alice microphones along the lines of Jules Ryckebusch’s Instructable, using the circuit originated by Scott Helmke, modified by Jules, and put in PCB form by Homero Leal. For these, though, I made a couple of changes:

When I built my first Alice microphone, I’d intended to use it for field recording. Unfortunately I found that the very qualities that give a vocal mic a lot of clarity and presence wind up making it extremely sensitive to the sound of wind in the treetops. It did a beautiful job, but was overly sensitive to wind hiss.

Homero came to my rescue and told me about a document written by another member of the Yahoo! micbuilder forum, Ricardo Lee, describing a modification I could make to the Alice circuit to tame the microphone’s presence peak. I rolled this change on my first Alice and on the Mid-Side Alice I built shortly after. The modification worked, and tamed the microphone’s over-sensitivity to wind hiss. I eventually put the modification on a switch so I could choose how the mic was voiced.

For this set of Alice mics I put those components on the PCB with a dedicated set of pads to wire in the switch, visible just to the left of the filter cap on the back of the board.

New PCBs

The next change I made was to use the Transound TSB-25AX capsule rather than the TSB-2555B capsule I’d used in my first Alice. From the datasheets on the Transound web page the two should behave almost identically, with the S/N performance of the 25AX just barely edging out that of the 2555B. To my ear the two sound identical.

Probably the biggest change was how I mounted the capsules. When I built my first Alice I didn’t own a 3D printer and still thought entirely in terms of subtractive machining (taking a chunk of stuff and cutting away whatever doesn’t look like what you want). This time I started over and designed a completely new saddle and post with 3D printing in mind. I wanted at least rudimentary shock mounting and a cleaner run for the wires.

Shock-Isolated Capsule Mount

I tried a number of approaches on shock mounting, but most of them strayed from something else I wanted out of this design: I wanted it to be something anyone could print and use. So I stuck with easily accessible materials for the shock mount itself, in the form of anti-vibration servo grommets available at practically any hobby shop.

Shock-Isolated Capsule Mount with Grommets

Shock-Isolated Capsule Mount Grommets Installed

As with all of the capsule saddles I’ve designed thus far, the capsule is attached using E6000 adhesive. This stuff is at least somewhat compliant, extremely sticky, and can be removed after the fact by passing a fine wire between the capsule and saddle. So far I haven’t needed to remove one, though.

I printed the final parts using Proto-Pasta Carbon Fiber PLA filament. It’s not the cheapest filament out there so I try to save it for lightweight structural parts, but it just looks so danged good, I couldn’t resist.

TSB-25AX Installed Front Quarter View

One of the things I like about 3D printing is that you can add features that are difficult to get any other way, except possibly by casting. In the case of this saddle and post, a 3mm tunnel runs from the back facet of the post down the center to take the capsule wires to the PCB inside the mic body. The tunnel has a 6mm radius curve inside the post, which makes wire installation easier.

TSB-25AX Installed Rear Quarter View

The final change I made was to switch from the admittedly gaudy BM-800 mic bodies I’d used on the previous two microphones to a slightly less gaudy un-branded BM-700 body I picked up off of Ebay for under $20USD. This had the added benefit of switching to a different headbasket geometry that doesn’t contribute as much ringing to the handling noise.

Here are all the bits and pieces going together. (I haven’t installed the voicing switch on this mic yet.)

Microphone Innards Front

Microphone Innards Rear

And a closer look at the wire routing, which is a lot more satisfying than the routing on my previous mics:

Headbasket Innards Rear

Here’s what it looks like closed up and ready to roll:

Microphone Front

For anyone else building their own condenser microphones, I’ve made this and a saddle for the TSB-2555B capsule available on Shapeways and Thingiverse. Be sure to print the right one for the capsule you’re using:

TSB-25AXZ3 Shapeways

TSB-25AXZ3 Thingiverse

TSB-2555BXZ3 Shapeways

TSB-2555BZ3 Thingiverse


Happy recording!


Posted in Uncategorized | Leave a Comment »

Vacuum Chamber Camera Mounts

Posted by Tom Benedict on 26/07/2017

August 1st we’re taking our telescope apart so we can re-coat the mirror. To date I’ve helped coat mirrors from four telescopes on the mountain including those from CFHT (M1 and M2), Gemini North (M2), IRTF (M1), and UKIRT (M1). Later this year we’re coating the 2m M1 from the Las Cumbres Observatory Global Telescope Network’s Faulkes telescope on Maui. Next year (we think) we’ll add a fifth Big Island telescope to the list when we coat the M1, and two M2 mirrors from the University of Hawaii 88″.

All of these are done in the basement of the CFHT Observatory, my home away from home.

CFHT Aluminizing Room

Mirror coatings offer a unique opportunity to photograph a really cool process: going from a bare piece of glass to a mirror in under 60 seconds. The only problem? The coating chamber itself is one huge, opaque piece of steel.

The Chamber

Luckily, the chamber has two windows installed in it. One lets you view the mirror. The other lets you view the filaments that evaporate the aluminum onto the glass. They’re relatively small, with a clear aperture of seven inches, but with some careful arrangement it’s possible to position cameras to peer through them and photograph the flash.

But it’s not easy. In years past I’ve balanced tripods on relatively small platforms, bolted makeshift 80/20 frames above and below the windows to try to support cameras, and even resorted to using tape. At times my solutions have been downright ugly, but they’ve worked well enough to let me photograph the transition from glass to mirror.

Before and After - Telescope Mirror

This year I wised up. I asked my boss if I could spend a little money and solve this problem once and for all. “Sure,” was his reply. That one word solved everything.

The windows are held in place by a ten inch diameter ring and six 3/8″ bolts. I couldn’t modify the rings, but I could install longer bolts and add tabs that let me attach additional hardware.

With the tabs in place I was able to install four 11″ articulated arms that I ordered from Amazon (two per window). If I had the option I’d install more, but even with two the windows become crowded quickly.

Because the two windows offer such different views of the chamber, the photographic requirements differ as well:

The window on the top of the chamber – the one that offers the view of the filaments – is used for diagnostics: Did all of the filaments fire? Did the aluminum wick onto the filaments without dropping? Did the aluminum vaporize without spatter? These require a relatively wide angle view, so I mounted a GoPro on one of the arms. On the other I mounted a Pulnix low-light video camera that’s wired into a monitor down at the controls to the coating chamber.

Cameras on the top porthole

Because outside light will reflect off the chamber window, causing glare, when we’re using the cameras we have to keep them covered.

Top porthole black-bagged

The window on the side of the chamber – the one that offers a view of the mirror being coated – is used to observe the coating process. Was there arcing during the glow discharge stage? Did the glow discharge generate a usable plasma? Did all of the filaments fire? Did the glass get a coating of aluminum?

This window is a little more complicated. To observe the glow discharge and the filaments, we installed a convex mirror at the base of the window. A Gopro aimed at this mirror can see, albeit in a very distorted way, most of the filaments and almost all of the glow discharge cathode. To observe the mirror, I used a DSLR with an 11mm lens pressed flat against the window.

Cameras on the side porthole

The side window also needs to be protected from stray room light, and will be black-bagged the same way the upper window was. Before the next coating I’d like to make proper blackout bags to cover the windows, but for now the cloth-and-bungee method will work fine.

This year I hope we get some really excellent video and stills of our mirror coating without having to resort to duct tape and bailing wire!

Posted in Uncategorized | Leave a Comment »

Camouflage, Paint, and Hiding in Plain Sight

Posted by Tom Benedict on 12/06/2017

I’m planning on using my weatherproof recording box to record wind in the rocks and grass at a remote lake up on a mountain, weather permitting. (Weather not permitting, I’ll record something else!) But the place I’m planning to record is primarily browns and reds rather than the greens I used to camouflage my weatherproof gear. It’s too late to change my gear now, but in the future I’ll need to think more about the camouflage I use to hide my stuff in plain sight.

Meanwhile an interesting discussion about camouflage took place on one of the field recording discussion groups on Facebook. In the discussion, someone raised the point that paints, fabrics, etc. that look fine to us in the spectrum we can see may be overly visible or downright jarring in the spectrum visible to some wildlife. The discussion centered around deer, which can see into the UV, but also applies to birds, insects, etc.

Since I’m already in the business of characterizing the reflected spectra of materials, surface treatments, paints, etc. I scanned the paints under discussion to see what their reflected spectra look like. I prepped the samples the same way I did for the SPIE paper: four coats, applied at roughly a 45 degree angle, coming in from the four cardinal directions.

Krylon Paint Reflectivity 250nm - 750nm

All of these are from the Krylon Ultra Flat Camo Paint series, one of which, their flat black, was part of the sample set I scanned for the paper I presented at the 2016 SPIE Astronomical Telescopes and Instrumentation conference in Edinburgh.

From 250nm to 750nm things look relatively normal. The behavior short of 350nm is more a function of the binders than the pigments, and all are relatively similar to the flat black, some even performing a little better in the near-UV.

From 350nm out to 650nm you see the action of the pigments in each of the paints. Khaki is fairly broad-spectrum, reflecting a good amount of reds, greens and blues, tending toward the redder end of the spectrum. The brown (which is quite dark) peaks in the red, with lower reflectivity in the greens and blues. The olive and light green peak at shorter wavelengths, favoring more greens than reds. It all makes sense until you look at the near-infrared.

Krylon Paint Reflectivity 250nm - 2500nm

That’s where it gets really weird. The Krylon flat black paints use carbon as a pigment, so they tend to stay low well into the NIR. The long tails on the brown, olive, and khaki paints aren’t surprising, especially given what I saw with some of the other samples I measured for the SPIE paper.

What’s weird is the light green paint. It’s more reflective in the NIR than any of the other pigments in the visible. I have no idea what they use for a pigment, but it’s got one heckuva NIR signature.

None of which may matter much when it comes to camouflage. Photosynthesizing vegetation is quite reflective in the NIR, so having one out of five colors reflect strongly at NIR wavelengths may actually help the disguised object blend better in plant settings.

In reading further about deer and their ability to see into the near-UV, I learned that there are two things at work: First, deer really can see further into the UV than we can. In the case of reindeer it helps them find the lichens that are one of their primary food sources and it helps deer spot UV-absorbing urine markers from predators.

The other factor at work is that the sensitivity of the blue receptors in deer eyes peak around 400nm, right around the center wavelength at which fabric and paper brighteners fluoresce when exposed to UV light. It’s this effect that makes white cotton shirts, shoestrings, and paper glow under black light.

The Shimadzu spectrophotometer I use to measure samples registers light of any wavelength, so it’s sensitive to fluorescence as well as directly reflected light. But just to be on the safe side I photographed the paint samples under UV illumination, as well as with an IR-converted camera, the same one I used for the SPIE paper.

Paint Three Ways

All in all, I think the Krylon Ultra Flat Camo paints are good to use as camouflage, both for humans and for wildlife. But I need to come up with a better color combination for blending with lava rock than the green scheme I’ve currently got on my gear.

Posted in Uncategorized | 4 Comments »

Weatherproof Recording Box – Part 2

Posted by Tom Benedict on 02/06/2017

A little while back I wrote about a weatherproof recording box I’d built so that I could do drop-and-recover field recording without having to care quite so much about the weather. The design had a couple of shortcomings, most of which I’ve been able to address. Since then I had the chance to take the revised box out into the field and put it through its paces. I’m happy with how it performed.

The two biggest issues I had with the first revision of the box were the heating problem I mentioned at the tail end of that article and the fact that with the connectors I was using, I couldn’t actually plug all four channels into my recorder and still have it fit in the box. I had a couple of minor issues as well, including the fact that the lid on the weatherproof outlet cover rattles. Here are all the changes I made to address these problems:

I fixed the rattle in the outlet cover with small snippets of weatherstripping. Nothing fancy, just adding enough spring to take out any tendency to rattle in the wind.

I also added strips of industrial strength Velcro to either side of the box. This makes it easy to use the box as a baffle and mount omnis on either side like tree ears. I’m not entirely happy with how this sounds, though, since the box represents a much harder boundary than tree bark does.

The way I solved the connector issue was to swap out all the Neutrik XLR plugs for low profile connectors from Cable Techniques. Redco Audio sells these for about $15US. They come with black caps, but Cable Techniques makes color caps for the connectors as well, so I was able to maintain the color coordination between the outside panel and the inside connectors.

Low Profile Plugs

While I was at it I bought a right-angle USB cable for the battery pack. In 20/20 hindsight I wish I’d bought one with a right angle connector at both ends instead of just at the micro-USB end. But at this point it’s just a nit-pick. The cable I got works fine.

Right Angle USB

With those two changes, everything now fits in the box with all the foam panels installed.

Everything Fits

Which brings me to the second major modification: In an edit to the last article I said that I’d tested everything with a temperature logger, and in a 21C room the air temperature inside the box had risen to 34C after eight hours of operation. What I didn’t take into account is that the battery and recorder were considerably warmer than the air in the box. The operating range on my DR-70D is 0C-40C. I wasn’t comfortable running it that close to the limit.

Thermal Pad
The foam panels in the box are all removable. They just wedge into place. So I pulled the two side panels and installed thermal pads. Thermal pads are made out of silicone that’s had thermally conductive additives mixed in. They’re graded by their rate of heat transfer. The pads I bought have a decent transfer rate, but aren’t up into exotic territory. (I got mine off of Amazon.) As it turns out they’re probably overkill for the job, but I’d rather err on the side of too cool than not cool enough.

Everything Heat Sinked

With both pads installed and everything loaded into the box, a small block of closed cell foam acts like a spring to keep the battery and the recorder pressed up against their respective pads.

Weatherproof Recording Box in the Field

Last weekend I took it out to an old cane haul road on the Hamakua Coast to record coqui frogs. Unfortunately the road was blocked off about a mile from where I wanted to record, so I had to set in a different location along the road. It looks like logging operations are about to start out there, so this may be the last time I get to record at that location. (The box is weatherproof, but it won’t stop a bulldozer!)

I had some issues with moisture causing noise on the mics early on in the evening, but overall the session went well. I used my SASS with EM-172 capsules for this. EM-172s aren’t typically all that sensitive to humidity, so I’m looking into why the mics glitched. So far I haven’t been able to reproduce it, but I might spend some time monitoring the next time I use this box just so I can hear if it’s happening again.

Meanwhile, here’s a long(ish) track from the test. It covers the sunset transition from primarily birdsong and insects to the coqui frogs dominating the soundscape. And there’s the occasional cow. (I really love the way cow calls reverberate through eucalyptus trees.)

Posted in Uncategorized | 1 Comment »