Engineering/DIY Question for builders

PhilGood

PhilGood

Juice box hero
If you were to build the ultimate tube microphone today, what parts/qualities would you use?

Generic, I know, but I wanna see what people have dreamed about.

Any combination/suggestion welcome.
 
I don't know about ultimate. But I would not attempt to recreate designs of the past. That's not because those aren't great designs, but within say ten years or so there will be no more extra functional VF14s, no?

So I'd have to start with a tube in production, which is a huge barrier, since most of those types are not ideal and NOS versions are better. I don't really get why there are still tons of NOS tubes available, but anyway.
 
Id start with the best capsule...then the cleanest opamps...then design an analog circut that would give me the tube sound without the tube...I think that with what they can do with digital modeling these days it is possible that they could make a modeling mic with a USB design that could sound like any vintage mic.
 
An opamp eh? I've used that approach, but not directly off the capsule. Sounds interesting. The cleanest FET-input opamps aren't the lowest current (you also need low noise), so you start to run into difficulties with phantom power. At a minimum, you would drop the voltage available to the capsule (and the opamp), which would then require a converter to get the voltage back up. But there's still the problem of reduced headroom in the opamp, although that should be manageable.

You might decide that you want an external supply for your opamp mic, but an external supply for a solid-state microphone could be a tough sell.
 
mshilarious said:
You might decide that you want an external supply for your opamp mic, but an external supply for a solid-state microphone could be a tough sell.
Click to expand...

He did say USB, which IIRC provides a lot more power than 48V phantom would. Of course, USB mics are notoriously train wrecks with crappy converters and no useful way of using more than one at a time, which pretty much shoots that idea in the head for most people, but....
 
dgatwood said:
He did say USB, which IIRC provides a lot more power than 48V phantom would. Of course, USB mics are notoriously train wrecks with crappy converters and no useful way of using more than one at a time, which pretty much shoots that idea in the head for most people, but....
Click to expand...

I think there were two ideas, first an analog tube circuit model, then the digital-USB option.

The great question that would need to be addressed before considering the modeling option is whether the capsule-tube interaction contain signal elements that do not exist and cannot be recreated in a capsule-FET circuit. The obvious example is something like mic modeling using an SM58--it's impossible to recreate the transients that the SM58 missed. I do not know if there is some phenomenon that exists with a tube mic that would frustrate a modeling approach, but I wouldn't be surprised.

As an aside, you know there is actually little difference between a USB interface and a USB mic. In fact, if you open them up, you might even find they are using the same IC. That means a USB mic probably has the capability to be stereo input and stereo output, just that those pins on the converter IC would not be connected.
 
mshilarious said:
I don't know about ultimate. But I would not attempt to recreate designs of the past. That's not because those aren't great designs, but within say ten years or so there will be no more extra functional VF14s, no?

So I'd have to start with a tube in production, which is a huge barrier, since most of those types are not ideal and NOS versions are better. I don't really get why there are still tons of NOS tubes available, but anyway.
Click to expand...

It could be with NOS tubes, plus there are tube options being manufactured today. EH is making EF86 tubes, which are known to be very linear and quiet. Russia is making tubes that are their own version of the EF86

6AU6 types are easily found. No short supply there. Run 'em as triodes.

I'm looking at a C3M or C3G NOS tube as an idea.

I'm not really asking for a production idea. Just to think about building one and see what we could come up with.

Transformer/No transformer? From who? What FET if not.

What king of capsule? What sound would you want? How much proximity effect?

PCB or point to point? Silver wire, copper wire?

Yadayadayada, blah blah blah.

Just throwing it out there...
 
USB is off topic, BTW.

Well, maybe not. I re-read the idea, but that wasn't part of my original question.

I'm not opposed to modeling or tube simulation, but I'm a bit of a purist and don't think you'll ever achieve the same sound.
 
dgatwood said:
He did say USB, which IIRC provides a lot more power than 48V phantom would. Of course, USB mics are notoriously train wrecks with crappy converters and no useful way of using more than one at a time, which pretty much shoots that idea in the head for most people, but....
Click to expand...

The whole thing holding back the USB mics are the cheapest parts have been used to make them...now if you had quality stuff...some capsule that is neutral sounding enough...you could get it to model any mic...Line6 with the GearBox software does a brilliant job of modeling the API...Neve...and Avalon preamps...mics that model the C12...U87...etc is the next natural step.
 
mshilarious said:
As an aside, you know there is actually little difference between a USB interface and a USB mic. In fact, if you open them up, you might even find they are using the same IC. That means a USB mic probably has the capability to be stereo input and stereo output, just that those pins on the converter IC would not be connected.
Click to expand...

I would note that they're mostly the same chips as the bottom end of USB interfaces. Most of them are limited to 44.1/48 kHz, for example. But yes, you're right about that. A few of them do have headphone jacks for monitoring. The problems with using them in a real-world environment are:

  • Monitoring: the headphone jack is on the mic. That's okay for somebody monitoring himself/herself for a podcast, but it basically precludes any useful monitoring by an engineer without running a splitter cable from the mic (which would really suck).
  • Multiple signals: you can't synchronize USB devices with each other very well, so you can't easily just hook up multiple USB mics.
  • Multiple signals: there's no real room on a microphone to add an XLR jack to connect a second microphone. Adding one would probably be a colossal hack.
  • Multiple signals: Adding a bunch of other jacks for multiple inputs on a microphone would be insane.
 
You can't really model a microphone with another microphone's signal unless the two mics are of the same operating principle and polar pattern. I guess if you stipulate an on-axis source in a free-field, you can ignore polar pattern. That might approximate a vocal recording technique, except that you still have to account for proximity effect.

Anyway, you would have to design an ambisonic-style microphone with excellent transient response and very low distortion. From that, you could derive any polar pattern and apply DSP to alter frequency response according to angle of incidence, transient response, etc. But by that time, you might as well just buy another microphone . . . especially since a good ambisonic mic is not exactly cheap!

In comparison, modeling compressors or nonlinear aspects of microphone amplifiers is trivial. Even so, I would question whether the analog and A/D stages of the Line 6 are adequate to provide a pristine enough signal for a good model. How good is its Grace algorithm, for example?


Phil: some years ago here I remember somebody built an analog optical mic with a 5" (!) diaphragm. Interesting project. That wasn't you, was it? :o
 
mshilarious said:
Phil: some years ago here I remember somebody built an analog optical mic with a 5" (!) diaphragm. Interesting project. That wasn't you, was it? :o
Click to expand...

That was Crazydoc. I still see him here, every ones in awhile.

Phil,

If I were to design a tube (or any other) mic I'd do something completely different, completely new and unique. I don't really see a good reason of coming up with yet another mic based on yet another variation of the same capsules, used over and over again in the same old and boring mics.

At this point and stage the tube would be my least worry. I'd concentrate on a capsule design, as something what defines the sound in the biggest way.

Best, M
 
Got any good books on capsule design? Any resources at all?

I know there is acoustic modeling software for room analysis. I wonder if it could be adapted for transducer design and analysis. I'd rather do it in software as opposed to trial and error machining.

I know what you mean. It seems everything is using the same 3 types. K67, K47, CK12 etc.

I have a Sony C38 capsule, but the output is so dang low. I like the whole idea David Bock is working on with someone, An eliptical capsule, but the hole pattern looks to be an adaptation of the K67.
 
PhilGood said:
Got any good books on capsule design? Any resources at all?
Click to expand...

I am afraid, not. You see, it is easy to clone, but to design a good capsule is a very long, time consuming, and frustrating process of hard work. Those who figured things out (incl. myself) are not in a rush to give those secrets away that easily, or publish in a book.

For example, to figure out my 3 chamber MXL603 capsule design took me about a year of intense work.

I am afraid, here you are whether on your own, or else might consider licensing it.

Best, M

P.S. Just send that capsule whenever you can.
 
Somebody does rectangular capsules; I forget exactly (in case you can't tell, I have a bad memory). It might be interesting to have nonplanar diaphragms, but I have no idea how you'd tension them . . . :o Something maybe like the carbon nanotubes C&T used (and just sold to Shure) might work, if they could be used to form a resilient shape on that scale. I guess if they can permanently corrugate a ribbon, why not any other shape? I wonder if they tried that--if they did, maybe it didn't work as they might have done that instead of 5 or 6 flavors of the ribbon.
 
Marik said:
I am afraid, not. You see, it is easy to clone, but to design a good capsule is a very long, time consuming, and frustrating process of hard work. Those who figured things out (incl. myself) are not in a rush to give those secrets away that easily, or publish in a book.

For example, to figure out my 3 chamber MXL603 capsule design took me about a year of intense work.

I am afraid, here you are whether on your own, or else might consider licensing it.

Best, M

P.S. Just send that capsule whenever you can.
Click to expand...

What sort of science are we talking about here?

There are certainly principles used by the originators on say, the M7.
 
PhilGood said:
I also have an idea for an LDC optical mic, straight to digital, but that's for another topic.
Click to expand...

Hmmm... so do I!!! But I wasn't going to bring it up because this thread is about a tube mic design. (I also do not want to give my idea away. :) )
 
You must log in or register to reply here.
Back
Top