Reasons for using an LDC

[Aled Hughes]

Apart from lower noise and any 'subtle flattery' they might impart, what is the point of using a large diaphragm condenser? Multi-patterns are inherently compromised compared to either a true figure-8 or an SDC omni, and a cardioid SDC's off-axis response is usually better than an LDC's.

So what gives? SDC self noise is rarely an issue in the studio, and if we discount the aesthetic impact of an LDC, that only leaves the 'subtle flattery' - is that really it?

Aled

[Jack Ruston]

Well it does rather depend what we're talking about - some LDC's sound like SDC's, some offer subtle flattery, and some aren't subtle at all - they flatter. There are of course records where vocals are recorded with all sorts of things, but generally speaking a vocal likes that size that an LDC can bring, especially close up, especially in cardioid (proximity effect). It's not always a subtle difference. But even if it is, we deal in subtleties don't we. We're always, as engineers, going to choose the tool that yields the subjectively most pleasing result, and on most vocals for example, something with a big colour like a 67 is going to be more pleasing than something very clean and flat like a transformerless SDC...the end result is likely to be quite thin, and it's going to have some frequency extension that we neither need, nor want.

J

[Sam Inglis]

Supposedly Motown Studios only had SDCs, namely Neumann KM86s. You probably couldn't criticise their records for being too clean and sterile.

[Jack Ruston]

Well it's not really a 'level' playing field Sam, because you're comparing a very coloured and band-limited signal path/recording medium, to a modern digital environment - It is interesting, as an aside, how that has affected the subjective performance of some classic microphones. For example, in years gone by it might be that eg a C12 was the most desirable microphone for many vocal performances - but if you record a C12 via a relatively transparent signal path straight to digital, it can be rather too toppy! Characteristics that were compensating for the 'tape effect' to yield beautiful detail, can now sound a tad brittle at times...At the same time, microphones like the SM7 are becoming more and more popular for vocals because they don't reveal unwanted HF content. Someone who really understands this is Trevor Coley, and that, for me, is why the Aria is such a GREAT modern mic. It's bright, but it's also controlled a little bit. Going forward I don't see the values of some of the brighter 'big guns' being maintained in the same way as some of the more midrange focused options like the U67. I just don't think they'll wind up being quite as desirable for those who don't still use tape, which is obviously most of us. This trend is also reflected in what might be seen as the 'rise of the ribbon' in the digital era.

J

[Hugh Robjohns]

While Jack's thoughts on the aesthetic values of LDCs are all valid, there is some real science involved too!

Yes, an LDC inherently has much lower self-noise than an SDC because of the increased diaphragm surface area; typically 12-18dB less, in fact. However, that is of limited value in most practical situations as the mic's noise floor rarely determines the noise floor of the recording.

Yes, many LDC's have switchable polar patterns and, although they are typically not as precise or consistent with frequency as a single-diaphragm SDC mic, the convenience of being able to change pattern instantly is attractive for some and in some situations.

However, much more important is the fact that almost all LDCs employ dual-diaphragm capsules, whereas very few SDCs do. To my mind, this is the most significant, especially when it comes to recording close sources, like vocals.

There's a good white paper on the science of dual-diaphragm capsules here:

http://cdn.shure.com/publication/upload ... m_mics.pdf

A short history lesson: when von Braunmühl and Weber first developed (in the 1930s) the cardioid capacitor capsule, they created a time-delay labyrinth at the rear of the diaphragm by drilling offset holes of different sizes in metal discs that were layered together to create the backplate. This labyrinth delayed the passage of sound waves from the rear of the mic to the rear/inside of the diaphragm. The idea was that if the labyrinth delay could be arranged to equal the time taken for sound from a rearward sound source to reach the front/outside of the diaphragm, the net force on the diaphragm will be zero and the capsule will have a cardioid polar pattern. Hurrah! It was a brilliant idea, and it has become the default concept for cardioid capsules.

Incidently, a clever man called Bauer (born Baumzweiger) worked for Shure in the mid- 1930s and came up with a similar labyrinth idea for moving-coil cardioid mics which was called the 'Uniphase Acoustical System' -- later the 'Unidyne' system.

Back to the cardioid capacitor mic... and Braunmühl and Weber were concerned that the static charge needed between backplate and diaphragm in this capacitor capsule would attract dust that might enter and clog the labyrinth... So to prevent that they fitted a second (unpolarised) diaphragm behind the capsule to serve purely as a dust filter.

However, when testing this new mic design they discovered that this second passive diaphragm altered the way the capsule worked, specifically to vary the relative proportions of pressure operation to pressure-gradient operation at low frequencies. This turned out to have a significant and beneficial affect upon the low frequency polar pattern and general bass response of the mic when working with very close sound sources.

In particular, it acts to reduce the proximity effect for close sources, and also reduces the tendency to 'pop' on plosives -- both very welcome properties when recording vocals, for example. And that's why the default vocal mic technology across the planet today is a dual-diaphragm capacitor mic! More airy top-end than a moving coil. More output than a ribbon. And less proximity and blasting than anything else apart from a single-diaphragm omni which isn't usually a very practical choice from the spill/room acoustic rejection point of view.

The white paper I linked above explains the science in more detail than I can give here with a tablet on my lap...

Another side note... Later still, Weber and Braunmühl realised they could polarise this dust-cap rear diaphragm and use it to create a capacitor capsule with two back-to-back cardioid elements, enabling the world's first switchable polar pattern.... And that became the Neumann M7 capsule... which us a whole 'nother story.

So, back to the science, essentially most large diaphragm cardioid capacitor mics -- just because of their dual-diaphragm construction -- perform much better in terms of spectral balance (proximity effect) and plosives-handling with close sources than most small single-diaphragm cardioid mics. And the reverse is true for more distant sources when the mic is working in the far field -- so SDCs are theoretically a better choice for distant stereo arrays and space mics etc.

While there's no doubt that sound recording is an art... It is called audio engineering for a reason!

H

[Tim Gillett]

Also on the theme of audio engineering, tape machines, at least properly maintained pro units, while nowhere near as ruler flat as digital recorders, had little audible rolloff of highs. One badly set up tape machine might sound too bright, and another dull, but a properly set up tape machine sounded substantially neutral. In a serious pro studio such as Abbey Road, using a bright sounding mic to compensate for a dull sounding tape playback would have been regarded as unprofessional, I suspect.

[Hugh Robjohns]

Supposedly Motown Studios only had SDCs, namely Neumann KM86s.

This is interesting: the KM86 did indeed employ small diaphragm capsules -- essentially the same as those employed in the wonderful KM84 -- and they were arranged back-to-back as entirely separate capsules, rather than two acoustically-connected half-capsule as in most LDC multi-pattern mics.

So although it is a dual-diaphragm mic in the sense of allowing user-selection of polar patterns, it doesn't benefit from the close-source bottom end benefits associated with the Braunmühl-Weber capsules.

NB. Text edited because I didn't have my thinking head on when I posted the original

H

[Hugh Robjohns]

I've tried to make it a policy this year of not reacting to TG's posts, but in the interests of accuracy I don't think I can let this one go without offering some important additional clarifying information.

Also on the theme of audio engineering, tape machines, at least properly maintained pro units, while nowhere near as ruler flat as digital recorders, had little audible rolloff of highs.

While the high frequency response of a tape machine is certainly affected by its EQ and bias alignments, the fundamental HF performance is set by the size of the magnetic gap in the replay head, in concert with the tape speed. (Smaller and faster is better!)

The AKG C12 was introduced in 1953 and at that time professional tape machines (like EMI's BTR2) typically had replay heads with gap sizes around 5-7 microns. That meant that even at 30 ips tape speeds they could not physically achieve a bandwidth greater than about 16kHz on a good day. It really wasn't until the '70s that head gaps of 1 micron or less became commonplace and flat-to-20kHz tape recording bandwidths could be achieved reliably (...and at lower tape speeds).

Interestingly, playing back tapes recorded in the 40s, 50s, and 60s on modern 1-micron (or better) replay heads reveal the full audio bandwidth that those vintage recorders' electronics allowed -- rather than the slightly rolled-off replays that the original engineers and artists heard at the time! This is because during the tape recording process the signal is encoded on the trailing edge of the record head's magnetic field and is not constrained by its gap size at all. (In fact, the record head gap is generally pretty wide to ensure the magnetic field saturates the tape to its full depth). So more HF ended up on those tapes than could have been heard at the time, but it can be retrieved with more modern tape machines.

Personally, I'm not entirely convinced of the arguments that many 1950s mic were deliberate voiced to be bright to compensate for restricted tape machine performance at the time. That seems like very fluffy and non-engineering thinking to me.

Moreover, If that really were the case, listening directly off the console while rehearsing would have been quite a fierce and unpleasant experience... and having known personally some of the engineers of that time I just don't think they would have put up with it!

In fact, a much better engineering solution -- and one easily within the technology of the time -- would have been to had some HF lift in the the tape machine's record electronics if they had really wanted to ... but they really didn't!

The reality is that the C12 is a bright-sounding mic because the capsule backplate design (which us quite deliberately different from Neumann's M7 design) gave it that property. It has lots of little internal resonating chambers that boost the extreme HF.

From an patent/IP point of view, AKG had to do something different to Neumann, and from a marketing point of view by having a 'bright-sounding ' muc they could offer a distinctly different sound character from the U47, which was probably considered a useful sales quality.

H

[Matt Houghton]

Personally, I'm not entirely convinced of the arguments that many 1950s mic were deliberate voiced to be bright to compensate for restricted tape machine performance at the time. That seems like very fluffy and non-engineering thinking to me.

I think you're right about that from a design perspective, Hugh. But I also suspect that some recording engineers seeking a brighter sound coming back from tape would have taken to choosing mics which, for whatever design reason, sounded that bit brighter.

[Hugh Robjohns]

I also suspect that some recording engineers seeking a brighter sound coming back from tape would have taken to choosing mics which, for whatever design reason, sounded that bit brighter.

Absolutely... and they might well have chosen mic positions and even applied EQ for the same reasons... but that's a very different thing from mic manufacturer's deliberately designing their mics to counter some notional limitation of unrelated recording equipment. It just doesn't ring true to me -- and especially so at a time when the technology was improving quite noticeably and quickly year on year.

H

[Tim Gillett]

Also on the theme of audio engineering, tape machines, at least properly maintained pro units, while nowhere near as ruler flat as digital recorders, had little audible rolloff of highs.

While the high frequency response of a tape machine is certainly affected by its EQ and bias alignments, the fundamental HF performance is set by the size of the magnetic gap in the replay head, in concert with the tape speed. (Smaller and faster is better!)

Of course.

The AKG C12 was introduced in 1953 and at that time professional tape machines (like EMI's BTR2) typically had replay heads with gap sizes around 5-7 microns. That meant that even at 30 ips tape speeds they could not physically achieve a bandwidth greater than about 16kHz on a good day. It really wasn't until the '70s that head gaps of 1 micron or less became commonplace and flat-to-20kHz tape recording bandwidths could be achieved reliably.

I suspect also that at that time it was believed the upper limit of human hearing was around 15 to 16khz. So an article in 1948 describing the then new Ampex 200a professional tape machine listed this performance spec :

"full coverage of the audible spectrum (+ or - 1db : 30 to 15000cps).

http://museumofmagneticsoundrecording.o ... eview1.jpg

Would a recording engineer at that time seek to boost frequencies that were believed inaudible?

[Sam Inglis]

They would certainly boost 16kHz. The high boost on a Pultec EQP-1A for example can be set to 16k -- and the high cut to 20k.

[Tim Gillett]

I have a soundcard that is essentially flat up to 80khz. Does that prove that humans can hear up to that frequency, or anywhere near it?

[Sam Inglis]

um, no, but I am not sure why you think that is relevant.

My point was that engineers certainly made use of processors that operated above 15kHz, whether or not theory told them it should make an audible difference.

[Jack Ruston]

Yep - my point was not that mics were necessarily designed to be bright because of tape, but that certain models flourished because of it, in a way that they might not do today.

J

[illegal colors]

I have a soundcard that is essentially flat up to 80khz. Does that prove that humans can hear up to that frequency, or anywhere near it?

Very high frequency harmonics or their absence can affect perception of 'more audible' frequencies.

[alexis]

I have a soundcard that is essentially flat up to 80khz. Does that prove that humans can hear up to that frequency, or anywhere near it?

Very high frequency harmonics or their absence can affect perception of 'more audible' frequencies.

Are you referring to hardware-associated artifacts only, or are there real-life effects as well?

[illegal colors]

I have a soundcard that is essentially flat up to 80khz. Does that prove that humans can hear up to that frequency, or anywhere near it?

Very high frequency harmonics or their absence can affect perception of 'more audible' frequencies.

Are you referring to hardware-associated artifacts only, or are there real-life effects as well?

I'm not sure if anyone can answer your question with certainty.

[Sam Spoons]

I think that is 'not proven' just yet though some of the luminaries who post on here are starting to consider it might possibly be true (certainly with digital systems).

[alexis]

I have a soundcard that is essentially flat up to 80khz. Does that prove that humans can hear up to that frequency, or anywhere near it?

Very high frequency harmonics or their absence can affect perception of 'more audible' frequencies.

Are you referring to hardware-associated artifacts only, or are there real-life effects as well?

I'm not sure if anyone can answer your question with certainty.

Sorry, I have no idea why I wrote it that way.

What I meant to ask is: are you referring only to the effects of aliasing of high frequencies back down into the audible range, with subsequent signal distortion ... or are you saying that if our ears are exposed to high frequency (i.e., > 20kHz), it changes our perception of sounds in the audible range?

I would think that even the latter could be tested for ...

[illegal colors]

alexis,
I understood what you meant the first time and I'm still not sure.

[alexis]

alexis,
I understood what you meant the first time and I'm still not sure.

Cool, thanks illegal colors!

[ef37a]

The claim that we "miss" frequencies beyond the accepted range of human hearing, 20kHz for jazz, has been around for as long as I have had an interest in audio, 55 years or so and many companies have used the concept to make equipment, notably microphone amplifiers that extend into the MW radio spectrum.

I have never read any solid science that backs this concept up and remain firmly in the GA Briggs (of speakers fame) camp that "the wider you open the window, the more the muck flies in". In any case, VERY few microphones get anywhere near 20kHz and even those few that do "fall off a cliff" shortly thereafter. I am also very suspicious of DC coupled audio gear.

I have no idea what proportion of engineers were working in broadcasting rather than vinyl back then but do remember that FM chopped off at 15kHz? Hardly anyone seemed to notice that at the time (and when the analogue V digital debate kicks off for the Nth time, the grotty, compared to todays, digital links are conveniently forgotten. 13 1/2 bit or so? )

Dave.

[The Korff]

What I meant to ask is: are you referring only to the effects of aliasing of high frequencies back down into the audible range, with subsequent signal distortion ... or are you saying that if our ears are exposed to high frequency (i.e., > 20kHz), it changes our perception of sounds in the audible range?

I would think that even the latter could be tested for ...

Aliasing isn't the only artifact of band-limited sampling systems — there's also phase shifts in the audible range (which, again, can be negated with higher sample rates and thus higher-frequency anti-aliasing filters). Then there's the pre-ringing associated with those filters (which, according to the people trying to peddle the MQA audio format, is still audible even with very high-sample-rate PCM systems)... My point being, even if you eliminate aliasing (which is surely a sign of a 'broken' system), digital audio as most of us know it still has its (tiny) flaws even when it's working exactly as designed.

As to the question of supersonic sound causing intermodulation in the audible band... As far as I can remember, I've not seen any conclusive evidence one way or the other! Unless anyone has a link to a convincing study?

Cheers!

Chris

[Matt Houghton]

Absolutely... and they might well have chosen mic positions and even applied EQ for the same reasons... but that's a very different thing from mic manufacturer's deliberately designing their mics to counter some notional limitation of unrelated recording equipment. It just doesn't ring true to me -- and especially so at a time when the technology was improving quite noticeably and quickly year on year.

Absolutely agree... hence my sentence that prefaced the section you quoted