Why do people still build tube microphones?

[Rossi]

I like clarity. I find clarity in FET mics and I find clarity in some classic tube mics. I rarely find clarity in modern tube mics, because these days people want to hear the tube - which usually means distortion. Classic tube mics are not distortion generators, they were designed for low distortion.

That said, I don't particularly like supermodern squeaky-clean solid state gear. Stuff like that has its uses, but you often find that avoiding distortion at all costs is just overdoing it and the means taken may be worse than the the actual cause.

P48 current limitation may actually be a good thing. It has kept designers from freely using opamps for stuff that a few transistors can do just fine. I don't have a problem with opamps and I don't have a problem with NFB as such. But using opamps often leads to *massive* amounts of NFB that can make a circuit sound lifeless. I don't think it's a good idea to use lots and lots of NFB just because you can. P48 current limitation is a challenge, and a challenge often is a good thing. It makes you more creative.

I like stuff that works and is easy to set up. I don't like external power supplies, and I don't like having to worry about tube life, especially when it comes unobtanium tubes. I have a U47, and I worry every time I use it. In fact, sometimes I worry about it, without even using it.

I think tube circuits are popular here, because they're easy to build. Many popular tube circuits are just a handful of components, you don't even need a circuit board and you don't need a dc/dc converter for proper polarization. That said, a simple 1-FET circuit such as a KM84 or Oktava MK-219/319 isn't too difficult, either. The Oktava MC012 circuit is underrated, maybe because it is transformerless. You hear a lot about transformer magic and tube magic, but sometimes the lack of any (obvious) magic can be magic as well.

 

[Gus]

I agree the 012 circuit is a underrated circuit.  I have used it with a LD capsule build.

 

[Rossi]

I like what abbey road d enfer posted      "Studio acoustics, mic placement, echo chambers, players/producer/sound engineer interaction, moderate processing. Tubes, transformers, class A are far down in the list".

Yes, spot on!

Why will people have no problem with a power supply for a tube microphone but not want a external power supply solid state microphone?

Well, they do (I do), but they know there's just no other way to power a tube mic (except that P48-powered Gefell one). But basically, people dislike external power supplies. Tube mics were replaced by FET mics for that very reason. People often think it was because of noise or distortion, but many early FET mics were noisier than good tube mics and had lower SPL handling, too. And while tube mics are popular for tracking pop/rock stuff, you don't see them in broadcast applications anymore and apart from a very few such as the M50/M150, you hardly see them used for classical music.

 

[Marik]

Why do people still build tube microphones?

Good question! I record on location classical only and found no use in tube mics--bulky, with heavy PSU (same for preamps--too heavy to take on the plane), high noise, unreliable, etc. etc. etc. I did not find any sonic advantages, either.

One of the nice things about tubes is they take overloading well. Another advantage, they usually have superior transformers, but in the end of the day it is not about topology, but implementation.

The question is: What is it that gave the old stereo-tracked RCA Living Stereo recordings their 3D-ness and musicality?

Sound esthetics and knowledge of sound engineers as for what to do. You know, I find it quite ridiculous--MOST of the modern sound engineers who records classical music don't even know how to read the music. IMHO, if a sound engineer goes into the hall for recording and does not know the score in and out, then how one would expect from that recording the 3D-ness, let alone musicality?

Speaking of technical secrets, those would be good miking and minimum of mixing.

Best, M

 

[abbey road d enfer]

I feel that people think that because SS circuitry exists that can be powered by P48 all SS mics should be powered by P48. This yields a power budget which makes it very hard to use, say, modern low-distortion op-amps as part of a mic head amp. I find myself falling in that trap from time to time.

Exactly. I find myself compromising significantly to accomodate the weak phantom current capability.
Makes me want to design some SS mics with external supplies. An attenuator switch for high spl would no longer be needed!

Les
L M Watts Technology

OTOH, what is the efficiency budget?
We need to output about 1mW of power max (that would be about 1.5Vrms into the recommended 2kohms load! - seriously hot). With all the losses due to class A operation, power regulators, phantom powering and miscellany, I reckon the typical current consumption of a static mic should be a fraction of milliamp. The part of the circuit that should be allowed to draw the most current is the output stage, and at 4mA per leg, that allows a hefty output level (1Vrms at least). I don't see the need for more than that.
Audio designers have always been concerned by the waste of energy in low power tube electronics, because the heaters consume more energy than the "noble" supply. For most of the designers who started with tubes and discovered solid-state, there's been no coming back.

 

[Rossi]

Given the fact that there are mics out there that are almost impossible to overload and have decent output, P48 seems to be sufficient. Take for instance a Neumann TLM103. Self noise is very very low at 7 dB-A. Its sensitivity is 23 mV/Pa (about -33 dB), which IMHO is about as hot as is sensible, and its max SPL is rated at 138 dB. In practice, I have yet to find a source that overloads the mic itself. Current draw is about 3 mA. It's not my favorite mic, but it proves that you can fulfil just about any technical requirement within the limits of P48.

What is annoying is that often enough preamps don't entirely fulfil the P48 requirements (+/-4V, 10 mA max current draw). Which means it's wise to stay below say 5 mA and not expect full voltage, either.

 

[riggler]

Sound esthetics and knowledge of sound engineers as for what to do. You know, I find it quite ridiculous--MOST of the modern sound engineers who records classical music don't even know how to read the music. IMHO, if a sound engineer goes into the hall for recording and does not know the score in and out, then how one would expect from that recording the 3D-ness, let alone musicality?

Speaking of technical secrets, those would be good miking and minimum of mixing.

Best, M

I play trombone and have played in several orchestras and professional groups. Not in the world's finest orchestras mind you, but I've "been around the block" in that regard. I don't think that almost any of the modern day recordings capture the musicality present in the old RCA recordings. I've studied this extensively. I'm very lucky in that one of the original Chicago Symphony trumpet players lives about 30 minutes from me!

The Bartok recording in question (RCA, Reiner conducting Chicago) was recorded on a two track Ampex 350, after they used RCA RT-21 30ips decks -- but before Layton got into 3-track recording. Even so, the 3-track recordings were really used as separate stereo (LR) and mono (C) recordings initially.

Anyway, that entire album was a 2-track stereo recording done with a spaced pair of Neumann M50 mics. Mics were spaced between 12-16 feet apart, about 10 feet back from the front of the stage, "up high"... RCA tube mic preamps (BA? OP?) to the recorder. Unsure if they continued to use the RCA preamps when they transitioned to the Ampex machines... Most likely they used the 350 electronics at that time.

But from there the sound went to Westrex cutting heads and into the vinyl.

So that whole noisy process just sounds great to me. And my empirical experience is that using that DW Fearn VT-2 in place of my normal preamps gave me some of that "sound", even with solid-state transformer-less mics.

But has anyone considered that the space-charge itself in a tube could contribute to the "sound"? I've always wondered that...

 

[tv]

But has anyone considered that the space-charge itself in a tube could contribute to the "sound"? I've always wondered that...

Crossed my mind in past a couple of times - but not related to mics. All guitar "starved tube" boxes lack that magic-touch compared to hi-voltage tube boxes. But I am not a tube-tone-freak myself. But it made me curious for sure.

 

[lassoharp]

Sound esthetics and knowledge of sound engineers as for what to do. You know, I find it quite ridiculous--MOST of the modern sound engineers who records classical music don't even know how to read the music. IMHO, if a sound engineer goes into the hall for recording and does not know the score in and out, then how one would expect from that recording the 3D-ness, let alone musicality?

Speaking of technical secrets, those would be good miking and minimum of mixing.

Best, M

I play trombone and have played in several orchestras and professional groups. Not in the world's finest orchestras mind you, but I've "been around the block" in that regard. I don't think that almost any of the modern day recordings capture the musicality present in the old RCA recordings. I've studied this extensively. I'm very lucky in that one of the original Chicago Symphony trumpet players lives about 30 minutes from me!

The Bartok recording in question (RCA, Reiner conducting Chicago) was recorded on a two track Ampex 350, after they used RCA RT-21 30ips decks -- but before Layton got into 3-track recording. Even so, the 3-track recordings were really used as separate stereo (LR) and mono (C) recordings initially.

Anyway, that entire album was a 2-track stereo recording done with a spaced pair of Neumann M50 mics. Mics were spaced between 12-16 feet apart, about 10 feet back from the front of the stage, "up high"... RCA tube mic preamps (BA? OP?) to the recorder. Unsure if they continued to use the RCA preamps when they transitioned to the Ampex machines... Most likely they used the 350 electronics at that time.

But from there the sound went to Westrex cutting heads and into the vinyl.

So that whole noisy process just sounds great to me. And my empirical experience is that using that DW Fearn VT-2 in place of my normal preamps gave me some of that "sound", even with solid-state transformer-less mics.

But has anyone considered that the space-charge itself in a tube could contribute to the "sound"? I've always wondered that...

Nice post.  That's a great description of a straightforward no frills minimal signal path for the job at hand.  Your comment that today's recordings "lack the musicality of the older RCA . . . "  has me a bit puzzled as to exactly what aspect of the recordings you were referring to.  If it's strictly gear related - assuming an equally talented orchestra performing in the same (good sounding)space - I would look towards the tracking ability/characteristics of analog vs digital over the presence of tubes. Then add to that the sound characteristics of vinyl compared to a CD - simply different.  I've had virtually zero experience recording orchestras so I may be missing some other aspect you're speaking of.

original Q:

Commercially:  The market for bottled-gear has been going steady since the mid-90s.  Why stop if the product is still selling?

DIY:  Sound, curiosity, Luddite lunacy, the challenge of replicating $12,000 classics, or simply the love of an honorable profession.

 

[pstamler]

Unfortunately, the wonderful acoustics of Orchestra Hall, where the great Reiner/CSO recordings were made, vanished in the 1960s due to a disastrous refurbishing of the room. When the CSO records now, they do it at the Medinah Temple (where I used to go see the Shrine circus). Good acoustics, but not the magic of Orchestra Hall.

Peace,
Paul

 

[mad.ax]

I think we need first to make a distinction between “clean”, and “amazing sounding” microphones.
Depending on the application, do we want a microphone that captures exactly what happens in the room? Or do we want a microphone that delivers that “magic” sound straight out of the plug?

In the first case, solid state seems to be the obvious choice as successfully demonstrated by references named B&K, DPA, Schoeps, and the like...

In the second case, even if we forget all the marketing, mojo, legend, and vintage trend, the fact remains that companies like Neumann, AKG, and Schoeps, did their best to come up with a convincing solid state version of their classics, and failed to reach that “WOW!” factor...
Compare a 47FET to an U47, a 87 to a 67, a 414 to a C12, a CMC to a 221... Despite using the same capsule, none of these can stand up against their predecessor...

From there, it is not so surprising that people believes “tubes are better”.

A solid state design would have to be seriously attractive in order to compare to the classics, and then one would still have to fight the vintage mojo marketing brigade!

Axel

 

[tv]

heh, vintage mojo marketing brigade

V.M.M.B. GmbH

 

[pasarski]

Sorry for going slightly ot, but as I wrote earlier it seems to be harder to find info about solid state mics than tube ones in internet and here. While Rossi, Gus and some others have done their share, unfortunately many of the links to the schematics are not up any more, like the ones in threads discussing Gus's solid state mic ideas. It would be great to have those back up! (zebra50?)

And like it's been said, solid state circuits seem to be a little bit more complex than the tube ones in general. If somebody knows good/obligatory reading (web, books...) concerning solid state amp design please link me up! I found this http://www.electronics-tutorials.ws/amplifier/amp_1.html

I'm especially interested in simple single stage single ended
class A designs like the Zen amplifier http://www.passdiy.com/articles.htm and how are they implemented in mics.

Whoa, much to learn as my valve know-how is also still minor, but I think the future is in the transistor

 

[pstamler]

Mics like he Neumann U 87 and KM 84 used simple single-ended class-A FET amplifiers, with output transformers -- the circuits were very similar to single-ended class-A tube circuits. Making one that delivers a balanced output without a transformer is a little harder but not that much harder. Schoeps, I believe, uses a single-ended FET plus two bipolars for the output.

As for solid-state mics with magic mojo, try a Microtech Gefell M930.

Peace,
Paul

 

[zebra50]

It would be great to have those back up! (zebra50?)

Hi!

We've moved server twice, and domain name at least once since those went up, and it's hard to go back and fix every old link. Most of the commercial mic schematics are now at.... http://www.xaudia.co.uk

I'll check with gus and see if he wants his old designs back up on the web somewhere.

Cheers!

Stewart

 

[Martin B. Kantola]

Why do people still build tube microphones?

Great thread! Will spend some more time studying all the interesting replies when I can, but some early thoughts: IMHO the simplicity of a tube circuit is appealing. Also, the combination of tube and transformer almost automatically gives you suitable input and output impedances with a balanced output. In a classic U87, it doesn't work as well if you ask me. A good tube circuit handles the dynamic range in a way that pleases the ear; noise spectrum is pleasant and gradually increasing distortion makes sense to us.

I think solid state can be better.

There are several benefits, and have to agree that there is untapped potential in solid state circuits!

Martin

 

[Rossi]

Well, in a single FET circuit, distortion rises almost as gradually as in a tube mic. The problem is, the onset of distotion is much earlier, so headroom is lower. The difficulty in designing a (subjectively) pleasing solid state circuit is in improving headroom while at the same time retaining the soft saturation characteristics. Which is why pointed out the Oktava MK012 circuit. That's a pretty simple circuit that still sounds pleasing and has a bit more headroom than a single FET circuit.

You might also want to check out the various C414 circuits. They go from simple with low headroom to ever more complex and distortion-free. Somewhere in the EB-Line there seems to be a good compromise.

 

[Martin B. Kantola]

Given the fact that there are mics out there that are almost impossible to overload and have decent output, P48 seems to be sufficient. Take for instance a Neumann TLM103. Self noise is very very low at 7 dB-A. Its sensitivity is 23 mV/Pa (about -33 dB), which IMHO is about as hot as is sensible, and its max SPL is rated at 138 dB. In practice, I have yet to find a source that overloads the mic itself. Current draw is about 3 mA. It's not my favorite mic, but it proves that you can fulfil just about any technical requirement within the limits of P48.

This is a perfect example of the issue, with numbers that look really good and a microphone that's far from great sounding. Technical requirements are one thing, sound is another it seems. Unfortunately we sometimes have to balance the two.

Martin

 

[Wagtbtoobz]

I think we need first to make a distinction between “clean”, and “amazing sounding” microphones.
Depending on the application, do we want a microphone that captures exactly what happens in the room? Or do we want a microphone that delivers that “magic” sound straight out of the plug?

In the first case, solid state seems to be the obvious choice as successfully demonstrated by references named B&K, DPA, Schoeps, and the like...

In the second case, even if we forget all the marketing, mojo, legend, and vintage trend, the fact remains that companies like Neumann, AKG, and Schoeps, did their best to come up with a convincing solid state version of their classics, and failed to reach that “WOW!” factor...
Compare a 47FET to an U47, a 87 to a 67, a 414 to a C12, a CMC to a 221... Despite using the same capsule, none of these can stand up against their predecessor...

I think we need first to make a distinction between “clean”, and “amazing sounding” microphones.
Depending on the application, do we want a microphone that captures exactly what happens in the room? Or do we want a microphone that delivers that “magic” sound straight out of the plug?

In the first case, solid state seems to be the obvious choice as successfully demonstrated by references named B&K, DPA, Schoeps, and the like...

In the second case, even if we forget all the marketing, mojo, legend, and vintage trend, the fact remains that companies like Neumann, AKG, and Schoeps, did their best to come up with a convincing solid state version of their classics, and failed to reach that “WOW!” factor...
Compare a 47FET to an U47, a 87 to a 67, a 414 to a C12, a CMC to a 221... Despite using the same capsule, none of these can stand up against their predecessor...

From there, it is not so surprising that people believes “tubes are better”.

A solid state design would have to be seriously attractive in order to compare to the classics, and then one would still have to fight the vintage mojo marketing brigade!

Axel

I met Dr Wuttke three years ago and asked him about reasons of release a new 221 tubed version . His reponse was clear : Market asked for although studies showed that most of the musicians couldn't hear any difference between FET and TUBE version !  (with a blind test made with a black box..) He admitted that iron transformers gave better WOW sound than op-amps.. Radio France(France musique ) purchased some new tubed mics some years ago because of the silkness and warmth .. just to balance the B&K coldness !
Pierre

 

[Rossi]

Given the fact that there are mics out there that are almost impossible to overload and have decent output, P48 seems to be sufficient. Take for instance a Neumann TLM103. [...].

This is a perfect example of the issue, with numbers that look really good and a microphone that's far from great sounding. Technical requirements are one thing, sound is another it seems. Unfortunately we sometimes have to balance the two.

Martin

Well to be fair, we don't know what the K103 capsule sounds like with a different circuit as it is only used in the TLM103. It is a K67-derived capsule and maybe it should have some HF correction instead of a flat amplifier. But that's just an aside.

Basically, we have to think about what we really want. For a long time the direction was clear: People wanted transparent microphones. Ideally, the recording should sound like what you hear with your naked ears. The microphone was seen as no more than a mediator, something that came between you and the actual experience. Reducing distortion (and smaller capsules) seemed like a good idea to reduce the mediation loss and thus bring the recorded experience closer to the actual experience.

That kind of thinking is still present in classical recording (But even there things are gradually changing).

In pop music, things are very different. Nobody cares how a something actually sounded in the room. What people want is a sound that is interesting, engaging. Also, we create an experience that never was. If you listen to a pop record, you listen to something that never actually happened. Sounds were recorded individually and then placed within the mix. The sense of room, dimension etc. is mostly an illusion. In order to create sounds rather than record an experience, you need different mics, or preferably a great number of different mics. You need mics that make a voice sound engaging and large and you need mics that make the snare sound snappy etc. Close miking requires different mics than distant miking. Yes, there are some one-size-fits-all mics that will do for most sources and situations, but if you strive for the ultimate experience (and when you think about it, that's what pop music is all about) you want to use different mics that will help you to place sounds in the spectrum and from front to back.

So what we want in vocal mics is this engaging quality. And some tube mics do have that. The U47 certainly does. What we don't need is über-cleanliness. Besides, a large diaphragm capsule produces quite an amount of harmonic distortion, some slight distortion on part of the amplifier doesn't really matter too much.

What I think manufacturers underestimate is the fact that people don't use one mic for everything anymore. Which means, it's not necessary for a mic to  handle both kick drum and vocals. Of course it's possible, a TLM103 can handle very high SPL and its self noise is low enough even for the quietest sources. But who actually uses one mic for such diverging tasks?

So, IMHO one way to come up with a better sounding mic is to design for a more specific set of tasks. A truly great vocal mic will likely be a so-so kick drum mic. And that's fine. Just as there are great kick drum mics that, for good reason, won't get picked for vocals or voice over.