The View Up Here

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

A Lack of Noise Measurements

Posted by Tom Benedict on 10/11/2016

When I built the Alice / TSB-2555B microphone, my intention was to get out with my normal recording gear, bring the Alice along, and get some side-by-side nature recordings to demonstrate how good or how bad the self-noise on the Alice is.

I have yet to pull this off. But in the process I became utterly paranoid about my gear, and feared I’d somehow broken it. In the process I made a number of completely useless recordings in the Upper Waiakea Forest Reserve and out on the Cane Haul Road north of Honokaa. I got some really interesting bird calls and some halfway decent coqui frogs, but most of all what I recorded was what sounded for all the world like… noise… (I’m pretty sure it was wind, but I’m giving away the punch line.)

This finally led me to do a series of recordings in the back seat of my car, parked in the relative un-quiet of my yard, under what is finally a wind-free clear night. In the process I think I demonstrated the excellent noise characteristics of the Alice microphone, and I’m pretty sure I recorded my own heartbeat from the seat next to my gear. (Yes, I had the gain cranked up that high.)

I found that the Alice edges out the Primo EM-172 capsules for noise, potentially by more than -3dB. I also found that even at ten o’clock at night there are very few windows of time when cars aren’t going by on the nearby highway. (I also found there’s a limit to how long I can sit still, not tic, and breathe shallow.)

With my faith in my gear restored and an expectation of the performance of the Alice microphone, I’m planning to head out Friday to record inside a lava tube. It’s not the rain recording I’ve been trying to nail for the past several months, but if I manage to get water dripping from the roof into underground pools, I’ll count that as something pretty cool.

This also gave me the confidence to order all the parts to build Jules Ryckenbusch’s MS Alice microphone. I still have mixed feelings about how mid-side is likely to fit into how I record in the field, but I know I’ll have a good mic to try it with.


Posted in Audio | Leave a Comment »

Alice and Behringer Sitting In A Tree – Part 1

Posted by Tom Benedict on 03/11/2016

The field tests on my BM-800 Alice conversion will have to wait. Late last week I handed it over to a friend for tests I’m not equipped to make, including several mic comparisons. I’m eager to see (and hear!) his results.

Meanwhile my Behringer C-2 mics showed up. These are the ones I placed a bid on over at Ebay before I realized they were coming from Haifa, Israel. Despite the distance the shipping was actually less than FedEx charges to ship a letter-shaped package from the mainland US to Hawaii. (Go figure.) It still ramped the price of the mics up almost to market value, which on Amazon with its super saver free shipping basically means I could’ve ordered them new and had them weeks ago.

But they’re here. And they’re mine. And… to be honest they’re in pretty ratty shape. One of them had something loose in the capsule. If you pointed the mic up and shook it, it made a hellish noise and clipped constantly. Turn it upside down and shake, and you hear something rattling around. The other mic has something massively wrong with its circuit board. It sounds for all the world like a Huey is hovering overhead. Bup bup bup bup bup bup bup… It never ends.

So long story short, I don’t mind gutting these things and building something new. In the short term I put the good capsule on the good mic body so I have one working mic to play with. The other one I started taking apart.

Behringer C-2 Capsule Removed

These have interchangeable capsules, though I don’t know if Behringer (or anyone else) makes any other capsules for it. The one that came with my mics is a hypercardioid. (At least that’s what the icon on the side of the capsule looks like.) Given the size of the vents at the back, I can believe it.

Underneath the capsule is a white plastic plug with a pogo pin centered in it. Not much to look at. And no real clue how to open things up past there.

To gain access to the innards of the mic, peel back the the “Behringer Condenser Microphone” name tape at the base. This reveals a small set screw that should be familiar to anyone with Switchcraft XLR connectors. Screw the set screw all the way in. This releases the XLR connector from the body. Next, center the pad/high-pass filter switch and pull the switch button out with needle nose pliers. Finally, push on the white plastic plug to expose the circuit board.

Behringer C-2 Stereo Pair - Partially Disassembled

Here’s what’s inside:

Behringer C-2 PCB Top

Since one of my mics has a damaged board, rather than figure out how to tweak what’s already here, I went ahead and tried to figure out how to pack a Pimped Alice into the same board space. I started by taking measurements.

The board is 15.5mm wide x 52.35 mm long, and is 1mm thick. The thickness is important because the board slots into the white plastic insert. One nice thing about this method of mounting the board is that there are no screw holes, and except for the humongous XLR pins and the 2mm area that slots into the plastic insert the rest of the real estate on the board is free. The board is mounted just below the centerline of the mic, so there’s vertical room as well. Up to a point, anyway. Those capacitors are 6.5mm diameter x 8mm tall. Nothing bigger than that will fit, even centered on the board.

There’s really not enough room to use through-hole components everywhere, so I converted most of the Pimped Alice circuit to 0805 SMT components. The exceptions are the filter capacitors, the 1Gohm resistor, and the FET.

There seems to be some resistance to using surface mount technology for DIY mics, but SMT has been used for over a decade for DIY robotics and electronics. I’ve built AVR processor boards using SMT components, and figured this wouldn’t be much different. With the exception of the big filter caps and the FET, that’s how Behringer built the original board for the C-2, so I figured it was a safe way to go. As soon as I have a new PCB layout, I’ll send it out for fab.

Meanwhile I started taking apart the capsule. Just looking through the grille, it seems like the C-2 uses a Transsound capsule similar to the TSB-165A Scott Helmke used in the original Alice.

Behringer C-2 Capsule Front

I started by removing the back plate. This is just pressed into place, but it’s a bear to get out. I eventually removed it by gripping it by an inside edge with needle nose pliers (pushing outward), and spinning it out. It took a couple of attempts, but it came apart.

The rear side of the capsule has an open cell foam washer in it, presumably to provide wind protection and to act as a delay plate to shape the hypercardioid pickup pattern. With the washer removed, the back side of the capsule is visible, held in place by a brass retaining ring

Behringer C-2 Capsule Back - Baffle Removed

The holes in the ring are really tiny. My existing pin wrench didn’t work, so I used an old divider with dull points as a pin wrench. There’s a bit of red enamel to prevent the ring from backing out, which took a little force to crack. Once that was done, though, the ring backed out easily. (I’ll need to be sure to apply a fresh bit of enamel when I get the new capsule installed.)

Behringer C-2 Capsule Disassembled Back

Behringer C-2 Capsule Disassembled Front

I was hoping the capsule was a Transsound TSB-165A, the same one Scott Helmke used in his original Alice microphone. Unfortunately it’s not. The capsule in the C-2 is 16mm in diameter x 6mm thick. The TSB-165A is 16.5mm x 8mm. But after some poking around on the JLI Electronics web site I think I found a match: the TSB-160A. The specs are almost identical to the TSB-165A, so it should play nicely with the Alice circuit (yay!), but the form factor matches what’s in the C-2. I’ll order a pair of these when I place the order for the 165A capsules for my MS Alice.

Behringer C-2 Capsule Mesh Outside

Another concern with the C-2 capsule holder is how the capsule is recessed, and how close the edges of the holder come to the input ports on the capsule. From my experiences with my first rev of mic bodies, I know that can color the sound enough to hear it. I’d like to open this up, if possible. It’s a simple enough job on a lathe as long as I can get the grill out.

Behringer C-2 Capsule Mesh Inside

The grill looks like it’s a two-layer mesh that’s either glued or soldered into the capsule holder. That should be easy enough to remove with heat, one way or the other. I might even be able to re-use it if I’m not too rough getting it out.

The grill serves two purposes. First and foremost, it’s an RF shield to keep stray electromagnetic radiation from getting into the signal path. Second, it helps to keep the capsule free of debris. Third, some manufacturers will stick enough mesh in front of the capsule to act as a rudimentary pop filter, and at least reduce the effect of wind. The problem with that third purpose is that you need a lot of tight mesh to pull that off. Enough so that it colors the sound of the mic. Not surprisingly, one of the more obvious mic mods is to remove a layer of mesh from the capsule housing.

But given how open the outer mesh is, I’m afraid it will make the mic prone to RF interference. For now I’ll leave it alone.

The next steps are to finalize the design of the new board, send it out for fab, and source all the components and capsules. But before I can finalize the board design I want to see if I can add in one of the features of the C-2: The switch on the side of the mic lets you select a high pass filter or a -10dB pad. If I can find the real estate on the board to accommodate the switch and the components necessary to add these into the Alice circuit, I will.


Posted in Audio, Electronics, Engineering | Tagged: , , , , , , , , , , , , , , | Leave a Comment »

BM-800 Microphone Conversion Part 2

Posted by Tom Benedict on 22/10/2016

This is the second half of a two-part article describing my conversion of a BM-800 microphone to an Alice microphone using a Transsound TSB-2555B cardioid capsule. All of this is based off of a pair of Instructables written by Jules Ryckebusch: Modify a cheap LDC Condenser microphone and Build the MS Alice Stereo Microphone.

Part 1 of this article showed pretty pictures of the donor mic (a Neewer NW-800 with an excess of bling), a description of the cable that came with the mic (which I don’t intend to use), photos of the mic in various stages of disassembly, and a CAD drawing of the salient features inside the microphone to help others lay out circuit boards for their own conversions.

Since writing part 1, all of the bits and pieces I ordered to do the conversion arrived: enough electronics to build three Alice boards, and a TSB-2555B capsule to put in the first one.

Everything for an Alice Conversion

Before populating the boards I did a test fit to make sure they would actually fit. I was pleased to see how well the screw holes lined up, and I came pretty close with the taper.

NW-800 With Alice PCB

The next step was to populate the boards. Opinions differ on how to wire the high-impedance (high-Z) end of the board, so I started with all of the low-Z components.

The circuit used in Jules’s first article had zener diodes on the output stage to protect it against over-voltage on the XLR pins. The circuit as-built in his second article omits the zeners since the 2N5087 transistors are rated for more than the 48V likely to be seen on an XLR connector. I ordered the zeners, but left them out for now.

Alice Trio with Low-Z Components

After I’d already wired all the boards I installed one in the mic and ran into my first problem: With the board installed right-side-up, the 47uF capacitor pokes up high enough that it interferes with the body tube. For my first mic I’m planning to install the board up-side-down to give the capacitor more room. But if I wind up building the MS mic from Jules’s second Instructable, I’ll need to install new capacitors that lay flat against the circuit board.

BM-800 Alice Board Placement

The reason for the difference of opinions on the high-Z end of the circuit is that it’s sensitive to contamination: leftover solder flux, dirt, dirt combined with humidity, oxidization, etc. on the high-Z end can all cause unwanted noise in the mic. Jules soldered his components to the board without issue. Others have used Teflon standoffs to float that part of the circuit above the PCB. Homero Leal built his Charis mic by point-to-point soldering the high-Z components, letting them float above the board without standoffs. Scott Helmke, the original designer of the Alice circuit, solders the high-Z components directly to the back of the mic capsule. For my first pass at this I soldered the low-Z legs of the FET to the board, but floated the high-Z circuit without stand-offs, similar to Homero’s Charis mic. I can always change my mind later and re-wire them.

High-Z Components Air-Floated

With the board built, the next step was to add 22nF capacitors between pins 1 and 3 and pins 1 and 2 on the XLR connector to provide additional RF noise filtering. After that I installed the modified connector and the board in the mic body.

Alice Board and XLR with RF Filter Caps

The rest of the action takes place inside the headbasket.

It’s possible to cut away the original mic capsule to leave a saddle for mounting the TSB-2555B, but I wanted to make an entirely new saddle. Chalk some of this up to not wanting to make a modification I can’t back out. Chalk some of it up to my wanting a machining project to go along with the electronics project. Either way it needlessly complicates an otherwise pretty simple project.

Space inside the headbasket is tight, so rather than run into more interference issues I fleshed out the 2D CAD drawing and turned it into a 3D model. The space constraints almost entirely dictated the shape of the new saddle and post. The mic frame is drilled and tapped for M2.5 screws on a 10mmx15mm rectangular pattern, only two of which are used on the original saddle. I chose to use all four. The mic wires pass through holes spaced 20mm apart, centered on the long axis of the bolt pattern. In the CAD model I indicated these with 3.13mm holes, but in the final part I cut them as slots to make installing and removing the capsule easier.

Mic Saddle - CAD vs. As-Built

I attached the TSB-2555B capsule to the saddle with E-6000 silicone adhesive. A better method for the saddle shape I used would’ve been a polyurethane adhesive like Gorilla Glue, but I wanted to be able to remove the capsule in case I decide to add shock isolation inside the mic to cut down on handling noise. As-built the capsule can be removed by passing a fine wire between the capsule and the saddle, cutting the silicone bond.

EDIT: The first time through, I missed an important step: One of the charms of the Pimped Alice circuit is the potentiometer next to the 1Gohm resistor. It allows you to bias the FET properly, regardless of which FET you use. The catch is that by definition, if you don’t do anything with the potentiometer it will not be properly biased! In all ignorance I soldered everything up, closed up the mic, and went testing. Even with an improperly biased FET it still performed beautifully. I did go back and do a proper job of it, though.

In Jules’s first Instructable, toward the end, there’s a nice write-up for how to bias the FET. The catch is that this step must be done before the capsule is soldered to the board.

With the FET properly biased and the capsule attached to the saddle and post, all that was left was to put it all together and close it up.

Finished BM-800 / TSB-2555B Alice

I did a quick side-by-side against one of my Primo-EM184 cardioid mics. The Alice runs a little hotter, but not by too much. I’m reserving further judgement on the new mic until I have a chance to get it out in the field and try it on some quiet sources.


Posted in Audio, Electronics, Engineering | Tagged: , , , , , , , , , | 3 Comments »

BM-800 Microphone Conversion Part 1

Posted by Tom Benedict on 09/10/2016

Several posts ago I mentioned a plan to build an MS mic by following an Instructable written by Jules Ryckebusch. Jules used a BM-800 microphone as a donor mic and replaced its guts with two Pimped Alice circuits and three cardioid capsules. After several Ebay vendors whose listings indicated they would ship to Hawaii later changed their story and said they wouldn’t, I finally picked up a BM-800 microphone off of Amazon. The one I got is a Neewer NW-800. (I liked its shock mount better than the other one I found.) It arrived, and I started poking and prodding at it.

Along the way I discovered another reason to use a windscreen on a microphone. This thing is bling central. To be fair some of the other BM-800 mics I found on Ebay didn’t have nearly as much… presence… but this is the one I could get.

Neewer NW-800 Bling

Unless I’m recording birds that are drawn to shiny objects the windscreen will probably become a permanent fixture on this mic, just to keep me from going blind.

Neewer NW-800 Windscreen

Before tearing into the thing I decided to try it as-is. On the face of it it’s a phantom powered cardioid condenser mic. This means plugging it into a device that doesn’t provide some kind of power (aka my laptop, my phone, even my kids’ desktop computer) won’t work.

The mic has a male XLR jack at the back, and came with an XLR-to-3.5mm cable. 3.5mm inputs that provide power typically provide plug-in-power (2.3V to 5V, depending on the device). XLR inputs provide phantom power (typically 12V, 24V, or 48V). That discrepancy made me a little leery of just plugging this into whatever and cranking volts through it. I started by ringing out the cable to see what it was actually doing.

XLR to 3.5mm Cable

Up to this point all the XLR plugs I’ve dealt with have been for balanced signals. That is to say that one pin of the 3-pin XLR is ground (pin 1), another is the positive signal (pin 2), and the third is the inverse or negative signal (pin 3).

3.5mm inputs typically use a TRS connector and unbalanced signals. In the case of the 3.5mm stereo input on my recorders the tip is the left positive channel, the ring is the right positive channel, and the sleeve is ground.

The cable supplied with the BM-800 ties XLR pins 1 and 3 together and routes them to the sleeve of the 3.5mm plug, and routes XLR pin 2 to both the tip and ring of the 3.5mm plug. This effectively turns the balanced output of the mic into two channel mono unbalanced output on the 3.5mm plug, meaning it should be able to be plugged into any 3.5mm stereo input and drive both left and right channels with the same signal. Neat!

What this also means is that as long as the mic can run on a wide range of voltages, the plug-in-power on any recorder should be able to drive this thing. So should the battery box I got from Church Audio. Or by removing the XLR-to-3.5mm cable and plugging in an XLR-to-XLR cable, I should be able to power it with phantom power (12V or 48v – the only two options on my recorder) and use it as a single channel balanced input.

Still leery of running such a wide range of voltages through it, I tried all three configurations anyway. I’m planning to gut this mic, after all, so if I burned it out the loss would be minimal. To my surprise all three worked! The plug-in-power on my DR-70D puts out a little under 3V, and my battery box from Church Audio puts out a little over 9V with a fresh battery. The 48V phantom power on the XLR inputs on the DR-70D put out right around 48V. I noticed a gain difference between the PiP and battery box, but because of the different gains on the XLR vs. 3.5mm inputs on the DR-70D I wasn’t able to tell if the additional voltage was doing anything to the mic itself. (My guess is it doesn’t. To survive that wide a range of voltages I’m guessing the mic has a voltage regulator on board. Past a certain point it’s just dissipating as heat.)

So how does it sound?


How to put this…

I’ve seen the shock mount it came with listed for more than what I paid for the mic. I don’t think this is too far out of line with how it sounds. It’s not bad, mind you. It’s just not anything I’d write home about. A little creative EQing would probably make it a decent podcast microphone. But as for making ambient nature recordings? Mmmm… no.

So without further ado I tore into it to see what I was going to have to deal with.

Neewer NW-800 Disassembly: Assembled

The first step in disassembling the microphone is to unscrew the butt cap. This also releases the shell, which simply slips off to expose the circuit board. The shell is keyed to a tab just under the headbasket which fixes the orientation of the logo on the mic. This is important since the mic is a side-entry rather than end-entry, meaning sound must enter from the side and not the end. Added to that, it’s a directional microphone so it’s only sensitive on one side. Can you guess which side? (Answer: The one with the logo.)

Neewer NW-800 Disassembly: Shell Removed

Some nice features on the inside of the thing: First, there’s a ton of room. Second, there’s a nice frame with mounting holes tapped for M2.5 screws. (More about those in a sec.) The only weird part is the taper on the frame and the circuit board. I like the look of the tapered board, so I decided to taper the boards for my Alice conversion, too, and put mounting holes in the boards to make use of the holes in the NW-800 frame.

Neewer NW-800 Disassembly: Headbasket Removed

Two M2.5 flat head Phillips screws hold the headbasket in place. They’re located just under the headbasket, above the circuit board. Once the screws are removed the headbasket lifts off, exposing the capsule.

Despite the appearance, the capsule in this mic is the same size as the EM-172 and the EM-184 capsules from Primo: 10mm diameter. At this point I was sorely tempted to gut the mic, drop an EM-184 capsule in the mic saddle, and call it quits. But the whole purpose of this exercise is to move beyond Primo all-in-one capsules and try my hand at building more complicated (and better performing!) microphones.

Neewer NW-800 Disassembly: Circuit Board Closeup

All of this starts with the circuit board.

Simple stuff first: The screws are M2.5, spaced 30mm apart. They’re biased a couple of millimeters above the centerline of the cavity. If you’re planning to make a rectangular circuit board to fit inside this mic, that’s probably all you’ll need. (The tube with the logo has vertical walls, so a rectangular board will fit fine.)

Since I wanted to make a tapered board I measured the whole cavity and threw it into CAD. At some point I’ll draw it in 3D, but for now a 2D representation is plenty for me to design the new board outlines. I’m building the Alice boards using through-hole components, so I needed a little more real estate than the original board provided. The 2D drawing of the cavity and the new board outline looks like this:

2D CAD - NW-800 Cavity and Board Outline

I sent the boards out for fab and ordered enough components from Mouser to build out three of them. One is destined to receive the TSB-2555B capsule I ordered from JLI. The other two will eventually be used to build a copy of Jules’s MS mic using three TSB-165A capsules, but that’s a project for another time. Once all the bits arrive I’ll write the second half of this article, which will cover the construction of the TSB-2555B mic.


Posted in Audio, Electronics, Engineering | Tagged: , , , , , , , , | 5 Comments »

Job Satisfaction

Posted by Tom Benedict on 20/09/2016

One of the questions I’ve asked myself over the years is whether I would be proud to tell my kids what I do for a living.

Much of the time the answer has been yes. When I worked for Academic Computing at the University of Texas I spent close to a year working in the Student Health Center. We took care of the servers that maintained medical records, handled customer satisfaction surveys, and made sure the desktop and handheld computers of the doctors, nurses, and administrators all worked. We helped them help the student body of UT stay healthy. Feel good about it? You bet!

Some of the time the answer has been a resounding no. The last year I was with IBM my job was to spy on my co-workers. I was the weenie who read all the logs from all of our servers and flagged “security risks”. During that time we never had an actual outside attack, and I think we had fewer than ten internal “ethical hacks”, all of which we caught. But I lost count of how many times my co-workers had a typo or tried to do something as root out of innocent ignorance. I had to report them all. Not one was a malicious act, and yet my job was to ding them for it anyway. Feel good about it? You gotta be kidding me…

Working at CFHT is solidly in the yes category. I’ve had downer days. Heck, I’ve had downer months, if not years. But at the root of it all we’re in the business of exploring the universe to better understand how the whole thing works. When people ask what I do for a living I tell them that my job description basically amounts to doing whatever is required so we can collect science-grade photons at night. Sometimes this means designing and building new instruments; sometimes it means sweeping the floors. When things get floor-sweepy it’s easy to lose sight of the fact that all of the things we do here contribute to our understanding of the universe.

Yesterday I brought in my camera bag and lighting gear so I could photograph a set of filters for customs paperwork. We’re shipping the filters to France to be scanned on a better spectrophotometer than the one we have here. Customs had a set of requirements for the photographs, so I was taking my time to make sure I got everything right. Toward the tail end one of our resident astronomers came in to see what I was doing. I explained about customs, about their need for documentation and serial numbers, etc. Not exactly sweeping floors, but documentation photography is pretty mundane stuff.

He listened patiently, then said, “You understand the importance of what you’re doing?”

“What do you mean?” I asked as I moved the last of the filters from the lighting scoop, back to its packing crate.

“Right now the tightest constraint on the cosmological constant is the SNLS survey, made with these filters. The scans they’re planning to do will further refine our understanding of the cosmological constant.” He pointed to the filter I was holding, the r’ filter. “That filter is key.”


Yeah. Even the act of photographing these filters so the customs agents can identify them and their serial numbers was helping to contribute to our understanding of the universe. No “Eureka!” moment. No lone genius. Just a lot of people doing a lot of seemingly mundane tasks, all of which is further refining our knowledge of how the universe works.

Today I’m processing the pictures, putting together documentation packets for customs, and packing the filters lovingly in their cases for shipment to France. Am I proud to tell my kids what I’m doing for a living?

You bet your ass.


Posted in Astronomy | Tagged: , , , , , , | Leave a Comment »

Building New Mics

Posted by Tom Benedict on 17/09/2016

When you’re faced with a dilemma like choosing the next step to improve your recording gear, instead of finding the right answer to the question, sometimes it’s more fun to dodge the issue completely and go off on a tangent.

So I went off on a tangent! Building new microphones!

I’ve currently got two projects in the works. A parabolic mic and a self-contained mid-side mic.

The Parabolic

I’m basing my parabolic mic off of the family of parabolic mics  from  Telinga Microphones. The mics from Telinga offer all kinds of neat options. One contains a cardioid mic facing the parabola and an omni facing away from the parabola. This lets you record a distant sound and the ambient sound field at the same time on two different channels. Another contains two omnis on either side of a baffle plate so you can record a distant sound in stereo.

But at its simplest, a parabolic microphone uses a parabolic reflector to direct pressure waves at a single microphone located at or near its focus. That’s where I’m starting.

I picked up a 22″ parabolic dish from sdill471 on Ebay. He sells them for around 50USD and ships all over the place, including Hawaii (yay!), using USPS shipping.

The microphone for this project is the first EM-172 lavalier mic I built back when I started building external mics for my DR-05. It’s since been converted to XLR and received the full shielded treatment the rest of my EM-172 mics got when I did that conversion.

The rest of the project will be to make all the mechanical bits to place the EM-172 mic at the focus of the dish. I’m drawing a good bit of inspiration from WW Knapp’s Homemade Parabolic Mic page, though I’m making two big departures Knapp’s design: The first is to think more in terms of parts I can make in a machine shop rather than what I can find at the hardware store. (This departure is called “needlessly complicating a good, simple design”.) The other is to take a tip from Klas Strandberg at Telinga: You don’t always want the mic to be at the exact focus of the paraboloid. Having the ability to rack the microphone through focus gives you some much needed flexibility in the field to widen or narrow the pickup pattern of the mic, or even to tune which frequencies are focused on the mic by the dish.

I’ll post the design and build articles once I’ve finished the mic.

Mid-Side Microphone

The entire idea for the self-contained mid-side microphone comes from an Instructables article written by Jules Ryckebusch. Jules took a BM-800 mic – about ~20USD off of Ebay depending on the seller – gutted it, and replaced its innards with two Pimped Alice amplifier boards and three TSB-165 capsules. The really clever part is how he did it, but for any of that to make sense it helps to understand how mid-side microphones work.

The easiest way to understand mid-side recording is to read a really good article about it. What I wrote below won’t be nearly as good, so I urge you to follow that link. That being said, here’s my take on mid-side:

Back when recording was in its infancy no one even thought in terms of stereo recordings, quadrophonic, 5.1, 7.1, or any of the other immersive formats we’ve since come up with. Mid-side was one of the earliest stereo techniques, patented by Alan Blumlein in 1933.

Mid-side uses two microphones: one to pick up the center part of the sound field (the “mid” mic) and another to pick up the sound on either side (the “side” mic). In most cases the mid microphone is a cardioid, which preferentially picks up sound in front of the mic. In all cases the “side” mic is a figure-eight – a microphone that picks up sound in two opposite directions, but nowhere else.

To create what we consider a conventional Left-Right stereo image from a Mid-Side (M/S) recording requires a little math. The equations look like this:

Left = Mid + (+Side)

Right = Mid + (-Side)

In the equations the Mid channel is taken as-is. The Side channel is used twice: first it’s used as-is (+Side) and the second time it’s used inverted (-Side).

As wonky as that sounds, and as convoluted as the post-processing sounds, it offers some distinct advantages when mixing the tracks afterward. Want a wider stereo sound? Mix in a little more of the Side channel and a little less Mid. Want to focus the listener’s attention on the bird in front of the mic and down-play the forest full of frogs chirping in the background? Bump up the Mid and turn down the Side. Want to mix a mono track to go with an accompanying video on Youtube? Use only the Mid channel for clean mono without any phasing issues. The real strength of mid-side is the flexibility and versatility it offers after the fact.

The one catch with mid-side, as with all stereo techniques, is that it requires two distinct microphones. ORTF requires two cardioid mics and a bar to mount them on. A/B requires two widely spaced omnis. Even my SASS consists of two omni mics mounted in an admittedly rather large baffle. M/S is no different, requiring a cardioid and a figure-eight.

What makes M/S special is that you want the microphones to be as close to each other as you can get them. By its very design it’s inherently physically compact. (Side note: This is true of X/Y as well, which uses two cardioids pointing 90 degrees to each other, and of the Blumlein arrangement, which replaces the cardioids with figure-eights.)

Which leads us back to Jules’s M/S microphone, which takes “compact” to a new level by cramming multiple microphones into just one mic body. That makes for a light, portable recording kit that’s quick to set up and tear down; perfect for traveling, or for recording subjects that require substantial hiking to reach.

So why three capsules instead of two? Jules realized that if he took two of the TSB-165 cardioid capsules, faced them in opposite directions, and wired them 180 degrees out of phase with each other in series, they act like a single microphone with a figure-eight pickup pattern. Add a third TSB-165 capsule in the center and you have all the makings of a well matched mid-side microphone.

Where Things Stand Now

My parabolic reflector arrived last week. The mic for the parabolic project is already in-hand, though I may have to (yet again) cut it out of its housing and install it in a new one. I’m in the process of designing the mechanical bits, and should be able to start making them in the next couple of weeks.

I ordered the BM-800 donor mic for my mid-side mic just this morning. Jules posted a link to download the Pimped Alice PCB files that Homero Leal designed based off of Scott Helmke’s original Alice design. Once I have the board mounting hole pattern off of the BM-800 microphone, I’ll add those to Homero’s PCB layout and send the files off to OSH Park for fab.

Work on both of these is contingent on my getting a number of other gotta-do’s off my plate, but I hope to make some progress on both in the next couple of weeks.


Posted in Audio, Engineering | Tagged: , , , , , , , , , , , , , , , | 1 Comment »

Microphone Self-Noise vs. Recorder Equivalent Input Noise

Posted by Tom Benedict on 25/08/2016

Yet another attempt at combining math and sound recording… Ye have been warned!

A number of threads on a number of field recording forums revolve around a simple question: I have X amount of money. Where do I throw it to improve my recording?

An obvious and common answer is, “upgrade your pre-amps!” This can be done a couple of different ways: The first is to trade out your recorder for one with better pre-amps. The second is to send your recorder to a shop to have the pre-amps changed out for better ones. The third is to buy an external pre-amp like the Sound Devices MicPre or MicPre-D, and plug it into the Line-In jack on your existing recorder.

But is that always the right approach?

A bunch of head-scratching, web-searching, and number-crunching led me to the conclusion that it’s not as obvious as it might appear. A number of factors come into play: noise level, sound quality, build quality, ergonomics and convenience, useful features of the gear, battery life, etc. Of these, the easiest to tackle from a quantitative standpoint is noise, so that’s where I’m starting.

Most of the calculations I’m doing are spelled out in an article on the RANE web site titled, Selecting Mic Preamps. The first set of calculations help you determine the maximum pressure levels a particular mic/pre-amp combination can handle. Since the field recording I’m doing involves quiet sources I skipped that bit and went to the second set of calculations. These help you determine the level of self-noise a given combination of mic and pre-amp will have.  To run the calculations you need information about the mics as well as the pre-amps.

(If you’re recording loud sources that first set of calculations may be of use to you! You don’t have to skip them just because I did.)

Right now all of the mics I own are based off of Primo capsules: BT-EM172, BT-EM158, and BT-EM184. The data I used for the mics all comes entirely from the Primo datasheets.

I currently own two recorders: a Tascam DR-05 and a Tascam DR-70D. In the spirit of this question I’m looking at two competing solutions: one is to buy a new recorder, a Tascam DR-680 MkII, and the other is to buy a used Sound Devices MicPre to use as an external pre-amp. The data I used for the recorders comes from a mix of sources, the most important being the Avisoft Bioacoustics Microphone Input Noise Comparison website. The rest came from the manufacturer’s datasheets.

The RANE calculations require the self-noise and sensitivity of the mics in question. From these you can use Table 3 in their article to calculate the mic output noise. For all of these I’m using A-weighted noise values for the mics and recorders. A-weighted noise levels are scaled for the auditory response of a normal human. They tend to be about 5dB more optimistic than their non-weighted counterparts. So long as I stick to A-weighted for both, I’m comparing apples to apples. The numbers for my mics and for the DPA 4060 omni by way of comparison are:

  • DPA 4060
    • Self Noise 23dBA
    • Sensitivity -34dB
    • Mic Output Noise -103dBu A-weighted
  • EM172
    • Self Noise 14dBA
    • Sensitivity -28dB
    • Mic Output Noise -106dBu A-weighted
  • EM158
    • Self Noise 20dBA
    • Sensitivity -32dB
    • Mic Output Noise -106dBu A-weighted
  • EM184 Cardioid
    • Self Noise 22dBA
    • Sensitivity -39dB
    • Mic Output Noise -110dBu A-weighted

The RANE article says that when you compare the output noise of the mic to the equivalent input noise of the pre-amp, you really want to see a factor of -10dB lower noise in the pre-amp or better. A -10dB lower noise in the pre-amp means it’s only contributing 0.4dB of noise to the final signal. Looking at the recorders I’m using, along with the two I’m considering, their EIN levels are:

  • Tascam DR-05 EIN -109dB A-weighted
  • Tascam DR-70D EIN -120dB A-weighted
  • Tascam DR-680 MkII EIN -127dB A-weighted
  • Sound Devices MixPre -126dB A-weighted

Here’s how I’m reading this:

If I plug any of these mics into my DR-05, the noise from the recorder’s pre-amps will be the limiting factor. Getting a better mic won’t improve my sound with that recorder.

My DR-70D is -12dB lower noise than the EM172 that my go-to mics are built around. In this case the mic’s own self-noise is the limiting factor. Switching to a DPA 4060 won’t help from the standpoint of noise, either. (I’m not mentioning any improvements in the character of the sound, mind you.) This does imply that I’m coming up on the limits of my pre-amps with the EM184 cardioid mics.

Switching to either a DR-680 MkII or a MixPre certainly wouldn’t hurt, and the higher quality amplifiers on either device may improve the sound in other ways, but it probably wouldn’t help the noise much overall because the mics would still be the limiting factor. At most I could improve my noise levels by a tenth of a dB.


Unfortunately what this means is that to make any substantial improvement in the noise level of my recordings, I need to upgrade both my recorder and my microphones. Upgrading either one without the other really won’t buy me that much.

The Real Conclusion:

This leads to the next obvious question: Have I reached a point from which the only way to improve my gear is to throw orders of magnitude more money at it than I already have? (Or to word that only slightly differently, more money than I have at all?)

In short, is this it?

(Or is this the excuse I need to stop improving the gear I’m using and start building parabolic mics?)


Posted in Audio, Engineering | Tagged: , , , , , , , , , , , , , , | 2 Comments »

SASS and ORTF Side-by-Side

Posted by Tom Benedict on 09/08/2016

Or top and bottom, rather.

I had the opportunity to stick my SASS rig and my newly minted ORTF bar on the same mount, one right over the other, and use them to record coqui frogs in a eucalyptus forest on the Big Island of Hawaii.

Of course whenever you record in a forest here you also get insects.

And if you happen to be within a hundred yards of a bunch of… dinosaurs? You also get them.

And the rain.

Ok, just a bunch of stuff. Anyway, here’s the recording. It’s an A-B test, switching between SASS and ORTF at thirty second intervals with a two second cross-fade.

My take: The two are different. (Well duh!) They provide different sounds. Neither one is “right” to my ear, just… different. But I’ll let you decide for yourself.


P.S. No I didn’t say which is which in the recording. What would be the fun of that?

Posted in Audio | Tagged: , , , , , , , , | 2 Comments »

Clippy EM184 Cardioid Mics and ORTF

Posted by Tom Benedict on 05/08/2016

I’d planned to write an article describing my trip to Edinburgh for SPIE 2016, but I got side-tracked. That article is yet to come.

I did some audio recording while I was there, but not nearly as much as I’d have liked. I wound up packing all of my sound gear, including my SASS, but the few times I pulled it out it rained. The one time I thought I’d get to use it for sure – poking it out of my hotel room window to record traffic sounds – I found it was too big to fit through the window. I wound up using spaced omnis to record traffic sounds, but the SASS didn’t get used even once. I found myself wishing I had other options.

A number of common stereo techniques require the use of cardioid microphones. Up until my trip to Scotland I only had omni microphones in my bag. There are still some stereo techniques that use omnis that I haven’t tried, but I’ve been wanting to play with cardioid mics for some time. Step one was to buy or make some cardioids.

The same circuit I used to make my EM172 omni mics can be used with other FET-enabled Primo capsules, including the EM184 cardioid capsule. FEL Communications (micboosters) sells these on their site either as individual caps or as matched pairs. I picked up a matched pair along with a pair of Clippy mic bodies, clips, and windscreens. I still had some Mogami cable and Neutrik connectors on hand, so I just drew from that stock to build out the new mics.

The Clippy mic bodies work nicely with the cardioid capsules, and the resulting mics have very little pickup at the back. It’s not zero, though, so you do have to be aware of everything that’s not directly in front of the mic. I’d been warned that cardioids are more sensitive to wind than omnis, and these mics bear that out. They’re stupid sensitive to wind. Even with the foam windscreens and some furries I got from Cat Ears, the slightest bit of wind kills them. I need to figure out some other solution for wind protection.

Step two was to come up with a way to hold the mics so they record a clean, well separated stereo image. There are plenty of choices for this, but the one I chose was ORTF, a technique designed around 1960 by Office de Radiodiffusion Télévision Française (ORTF) at Radio France. (See? Astronomers aren’t the only ones to recycle their acronyms!)

ORTF requires the microphones to be separated by 170mm and angled away from each other at a 110 degree angle. It’s a bit of a pain to set up in the field without some way to gauge the angle, so many people favor other setups such as NOS (Nederlandse Omroep Stichting) in which the mics are separated by 300mm and are angled out by 90 degrees. I wanted to play with ORTF, though, so I decided to solve the setup problems with a fixture.

Clippy ORTF Bar

Since the Clippy mic bodies register nicely with their lapel clips, I used the clips to orient the mics both in location and rotation. The clips have a tab on top that’s just over 6.2mm wide. I made 6.5mm wide slots at either end of a bar to receive the clips.

Clippy ORTF Bar With Mics

I wanted to keep things simple so I didn’t have to fuss with stuff in the field, and this lets me do that. With the clips fully seated in the slots the mics are angled out at a 110 degree angle and are 170mm apart. It takes more time to unroll the cables than it does to install the mics on the fixture. And the flat bar packs down a lot smaller than my SASS.

Clippy ORTF Bar Slot Detail

The bar I used was just over 4mm thick. I cut the slots to leave 2mm of material for the mic to clip to. This wound up being a little thin, but it made for a nice, deep slot to register the clip in.

Clippy ORTF Bar Velcro

The bare metal of the bar was too slick for the clip to get any real grip, so I put a tab of Industrial Velcro on the bottom of the bar under each of the slots so the clips would have something to grab onto.

I’m pleased with how easy it is to use this setup, and it’s tough to beat how compact it is. But I’m not 100% satisfied with how it works in the field just yet. I already mentioned the wind issue. Even with double protection the mics saturate when almost any amount of wind touches them. They’ll probably fare better inside  a Rycote or a Rode blimp, but for now I’ll have to save them for wind-free environments.

The sound is also significantly different from that of my SASS. (Sorry, no side-by-side comparisons yet.) The SASS picks up more reverberation than the ORTF setup, so there’s more of a sense of the space with the SASS than with the ORTF. But you don’t always want that sense of space. During an earlier test I had one of my omnis and one of the cardioids in a car. The omni picked up so much of the car noise, it was difficult to hear the people in the car speaking. The recording from the cardioids was much cleaner.

Needless to say there’s still plenty of testing to be done. Once I learn the strengths and weaknesses of this setup and have a better handle on wind protection, I’m sure it’ll see plenty of use.


Posted in Audio, Engineering | Tagged: , , , , , , , , , , , | 4 Comments »

DIY Wind Protection for Custom Microphones

Posted by Tom Benedict on 19/06/2016

With the exception of some heavy-duty rain protection that I’ll need to build in order to get a set of sounds I’m after, I think I’m zeroing in on a field recording setup I’m happy to use for the foreseeable future.

DIY-SASS Wind Protection

The last(ish) step was to add real wind protection for the microphones. In the past I’ve used whatever I had on-hand to protect the microphones from wind: my t-shirt, my fleece, the headrest covers that came with the seat covers for my car, etc. They worked, but they weren’t pretty and they were a little frustrating to use. When tying a shirt onto a microphone it’s easy to leave gaps that wind can get through. After adding the anti-vibration mount I figured enough was enough. Time to make proper wind protection.

I made this out of the thinnest “wetsuit” material I could find. (Real wetsuit material uses closed-cell Neoprene foam. The core in this fabric is open-cell foam that bears a strong resemblance to foam microphone covers.) It costs some high frequency response, but the EM172 microphones are already pretty bright. I consider it heavy-duty in that it’s tough to breathe through this fabric, but most of the recording I’ve been doing has been in areas I’ve flown kites in the past. Heavy duty isn’t necessarily a bad idea. It’s removable, so if I record in an area that doesn’t need this level of protection, it’s easy enough to remove.

I’d do a whole write-up on how I did this, but it’s basically sewing. There’s not much point in posting the pattern, either, since it’s designed around this particular microphone array. This whole setup started life as a 3D CAD model. That provided me with a cut list for building the microphone array out of plywood, foam, and sheet metal. That same CAD model provided me with a pattern for making the wind protection. This material is pretty easy to work with, though hems tend to be a little fat. It doesn’t respond well to ironing, so all of the hems were done using pins. Lots and lots of pins. You do what you have to do to work with the material at hand. At some point I’ll make a fuzzy to go with this for when it’s really howling. But for now I think I’m done.

Last night I took the whole kit ‘n kiboodle down to Kua Bay to record the summer surf. Kua Bay is a white sand beach that’s exposed to open ocean. There is a reef, but it’s deep enough that waves break on the sand rather than out on the reef. When the waves come out of the right quarter they can break left-to-right, right-to-left, and across the entire beach one right after the other. The conditions last night were perfect. The gates close at 7pm, so by the time I got there at midnight the place was completely deserted. I set up my gear, grabbed my book, and walked off to enjoy the moonlit landscape while the recording gear did its work.

I’m pleased by how things turned out. And you can’t beat a deserted beach on a moonlit night. I had a blast.


Posted in Audio, Engineering | Leave a Comment »