I forgot to mention that I actually have a mic very similar to the MXL 2200... but it's a "no name" mic that I bought from someone for like $30. I don't know what the brand is. I used it for a M/S recording for a choir once and everyone was happy with the results... so I guess it works OK. I think I might prefer something that I KNOW myself is good.

i'm not a professional, so take this for what it's worth.

i went to sam ash one day to buy a couple mics to do exactly what you want to do. i out of them, i tried out the Rode NT2000 and the NT2-a. Honestly, i didn't a difference. The only thing different between the NT2000 and the NT20-a is that the NT2000 has the variable polar patterns, bass roll offs, and pad. If i'm not mistaken, they actually share the same capsule, but don't quote me on that. I ended up going with the NT2-a since it's $200 cheaper. Also, try playing the used market to see what you can come up with; i liked my rode so much that i just bought another on ebay for $260.

as far as the shure ksm mics, i have the ksm27 and love it...got that on ebay too. i don't have the ksm32, but when my band recorded our ep, our producer Lee Dyess actually used one to cut the vocals over a AT4060 (a badass mic, btw) and a Neumann M149...to me, that says something.

You could also check out the audio-technica stuff. the at4033 and at4040 are definitely worth checking out. do a search on here and you'll find a lot of fans of both.

-dave

Don't spend any real cash until you've at least checked out the AT 4000 series.

The 4040 and 4050 are two of their best, and I use them a lot on a variety of things, including choral and instrumental work. The 4040 is simply a good work-horse single element LD cardioid mic, with a -10 pad and bass rolloff switch for under $500. The pricier 4050 has those features as well, but is a dual-element capsule, which means it also has a figure of eight setting for bluemline use, as well as an omni setting for use as an omni on its own, or used with another 4050 in MS mode. very nice options to have in a mic that's under $600. They look great onstage (disappear on most TV shoots since they're jet-black) and they're quite rugged.

Another good one to check out is the 4033, in the same price range as the 4040. Their lower end mics aren't bad either; the 3035 is wonderful too, regardless of price. (Usually about $299 each.)

+1 on Joe's post.

Search for Audio Technica on this here site.
Many of us have given a rundown or offered an anecdote or two about their mics.

Just curious - can the AT4050 be set between the pattern positions, to give a wide-cardioid pattern for example?

Just curious - can the AT4050 be set between the pattern positions, to give a wide-cardioid pattern for example?

no it can't...just the 3 patterns.

-dave

JoeH wrote: The pricier 4050 has those features as well, but is a dual-element capsule SNIP!

Interesting...

I'll be using a pair of 4050s to record some 'lhamo' (Tibetan Opera) in the nicely reverberant hall at the Tibetan Institute of Performing Arts (TIPA) in 1.5 hours from now. I'll be using them in MS Blumlein cofiguration (MS with a bidirectional M capsule, facing forward), because I am not sure if they are a matched pair or not; so putting them in MS spreads any tonal differences between the mics across the stereo image, rather than one side only. Also, the soundstage is not huge - one main vocalist with two others standing behind him, forming a triangle. So most of the direct sound will be in the M channel, with the S channel mostly capturing the hall reverberation. There'll also be some reverberation picked up by the rear of the M microphone, of course, due to its bidirectional response. Coming from the rear, rather than the sides, means the reverberation in the M channel will be nicely decorrelated from the direct sound (assuming the mic is a good distance from the rear wall!) and ought to sound fine when summed to mono for radio broadcast (a lot of broadcasters in this part of the world are still using mono, especially national shortwave stations).

While I generally don't advise LDCs for choral, orchestral or concert band (although concert bands fare far better from LDCs than the others...) I can whole heartedly recommend the Shure KSM line as well as the AT 4040/4050. They're both great mics for the price and work well in a myriad of situations.

Also consider the Blue Bluebird. This one has a bit more bite than any of the mics I've mentioned so far, but it's a good sounding mic with great off-axis response (so critical in this type of work).

It's damn cheap too for what you get. If I had heard and used this mic first, then someone told me it cost $900, I'd still have dropped the cash.

Cucco wrote: While I generally don't advise LDCs for choral, orchestral or concert band (although concert bands fare far better from LDCs than the others...)

Hmmm... I'm curious as to why you say this. Please elaborate. I guess I was under the impression that LDC's are the best to use (as opposed to medium or small DC).

I've taken into consideration the AT 4000 line of mics. They are reasonably priced (mostly in the used market... though I hate buying used mics). The 3035 looks good too.

Thanks!

SDC's by nature exhibit a faster transient response, as well as better stereo imaging and capturing of low frequencies.

LDC's are slower to respond, due to the added mass of the capsule, comparatively speaking, and have a tendency to color sound more so than an SDC.

In addition to what Bent stated -
Because of the larger surface area of LDCs, there is a greater potential for non-linearity (both caused by tolerances in material and breakup due to high or conflicting SPLs) and the smaller the diaphragm, generally the better off-axis response.

This is why you will rarely if ever see any major orchestra recorded with LDCs (there are exceptions, but fewer than you might think. I think the Neumann Solution D was probably the most commonly used LDC over the past few years due in part to Neumann's aggressive marketing of the technology.) The M50, which looks like an LDC, is definitely not an LDC but is without a doubt, one of history's (and present, but not as much currently) most popular mics for this task.

Currently, the most common mics for this task are SDC mics from:
Sennheiser (MKH line)
Schoeps (Collette)
DPA (40xx)
Gefell
Earthworks

The only real disadvantage to a SDC is a potentially higher noise floor (though WELL under the capabilities of most playback systems) due to the laws of physics and the smashing together of air molecules.

Ardroth wrote: I guess I was under the impression that LDC's are the best to use (as opposed to medium or small DC).

When choosing microphones for this kind of work, you need to find the right balance between two virtues: 1) low self-noise, and 2) good off-axis response.

As the diaphragm gets larger it is capable of collecting more sound energy and therefore has lower self-noise - i.e. the microphone is quieter. But the larger diaphragm (and the housing around it) creates a larger obstacle to sounds arriving off-axis, resulting in poorer off-axis response. The result is a coloured sound. LDCs are fine for close-miking, but not so good when you're miking from a distance and need to capture all sounds cleanly.

As the diaphragm gets smaller it becomes less of an obstacle and so the off-axis response improves, resulting in a more natural sound when miking from a distance. The downside is that there is less surface area to collect sound energy, resulting in higher self-noise.

SDCs from 10mm to 15mm or so generally provide the best compromise between good off-axis response and low self-noise. Hence they are very popular with people who record acoustic music.

To give two examples from DPA's stable: their 4060 series of omnis use 5.4mm diaphragms and have a textbook omnidirectional response, but their self-noise is 23dBA or thereabouts. At the other extreme, their 4041 omni has a 24mm diaphragm and offers a self-noise of only 7dBA, but it's off-axis response deviates considerably from omnidirectional, especially at higher frequencies.

Horses for courses. If low noise is a priority, there are few mics that can come close to DPA's 4041. But if excellent off-axis response is a priority, it's hard to beat their 4060 series.

Most DPA owners have the models with the 16mm diaphragms, such as the 4006 or similar, which offer a good compromise between off-axis response and self-noise.

I hope that's helpful...
 

Simmosonic wrote: This edit was added after discovering that Jeremy had posted something similar while I was writing this. Sorry for doubling up...

Yes, but you did it so much more eloquently.

The staph made me do it...

[see recent post under 'Recording Orchestras' for clarification...]

Sweet. Thanks for the info... I did not realize this about the diaphragms.

Unfortunately, those mics recommended are FAR too expensive... I'm no professional yet! I just have a project studio that's probably only worth 2 of those mics. For recording live acoustic music (small choirs, chamber, solo, etc.), would a M/S (figure 8 ribbon paired with a LDC or SMD) work well?

Perhaps there are cheaper alternatives to SDC's which has a good balance between low self noise and off axis coloration? I guess low self noise IS indeed a big priority for me. Mostly because I'm using cheaper pre-amps (Presonus) and cables (Peavey w/Neutriks)... also, the hall I generally record in contains a lot of noise from the air system (HVAC I think it's called??!). I'm guessing that the only way to get a good balanced mic is to spend no less than $1000... Then it's either noisy mics, or less accurate mics... right?

Thanks for the help guys.

I would not focus on the low self noise at this point. Even the mics with self noise of 20+ dB would suit most every recordist and would still be lower in noise than the traditional concert hall of most mid-level ensembles.

Not to mention, self-noise is only have the equation. Sensitivity is the other. SDC can be made to be quite sensitive (although, usually beaten out by their larger siblings). A mic with a high sensitivity and a medium to low self-noise will seem quieter on the recording in most cases than a quiet mic with low sensitivity.

This is how Earthworks gets by with their "Quiet" series mics. Check the specs, they're no quieter than their non-QTC series mics, they're just far more sensitive.

In an affordable range of SDCs, you might want to consider the Shure KSM series, AT 4041/4051, Rode NT55 and similar.

Cheers-
J.

Cucco wrote: A mic with a high sensitivity and a medium to low self-noise will seem quieter on the recording in most cases than a quiet mic with low sensitivity.

Yes, I used to think so, too, but that's not quite the way it is... Self-noise is not related to sensitivity, it's measured in SPL.

Other way round: Sensitivity is measured at 94 dB SPL. Regardless of the electrical level that 94 dB SPL will translate into (aka sensitivity), the mic's S/N ratio (usually quoted in relation to 1Pa/94 SPL as well) at e.g. 20 dB of self noise will be 74 dB, and at 6 dB it will be 88dB. Higher sensitivity will not change that. (Hope I'm not stating the all-too-obvious...)

I've found an Earthworks SR78 (which does not claim to be particularly quiet) to be way too noisy to be even used as a harpsichord spot mic...

As for good low-cost SDCs, I'd like to mention the Beyerdynamic MC930.
But I've also use a pair of AKG 414 B-TL (not TL-II) as AB main mic quite successfully...

Daniel

d_fu wrote: Yes, I used to think so, too, but that's not quite the way it is... Self-noise is not related to sensitivity, it's measured in SPL.

I know what you're saying and agree - however, the quoted spec of a fixed dB noise rating for each mic is directly related to (or more accurately to say - "impacted by" or "affected by")the amount of gain applied to that mic using your mic preamp. Where you have to use less gain, you will have lower (relative) noise levels. However, there's no escaping the S/N ratio - it's a matter of maximizing the headroom and gain of the mic at hand.

Ardroth wrote: Unfortunately, those mics recommended are FAR too expensive... I'm no professional yet! I just have a project studio that's probably only worth 2 of those mics. For recording live acoustic music (small choirs, chamber, solo, etc.), would a M/S (figure 8 ribbon paired with a LDC or SMD) work well?

There are many suggestions being made here...

I'd forget about MS for now, mostly because you'll need a means to decode it, and also because it is not the most versatile technique around for the money.

I'd put in a recommendation for a pair of Rode NT5s, which are SDC cardioids and are sold in matched pairs. A pair of those will allow a range of near-coincident (ORTF, etc.) and coincident (XY) methods, making it equally useful for large or small ensembles.

If you can stretch the budget a bit further, a pair of NT55s with cardioid and omni capsules is also a good choice. IMHO, the NT55 with omni capsule performs well above its price point. That set-up won't cost you much and adds spaced omnis (AB) to the near-coincident and coincident methods offered by the cardioids.

They're also very versatile for close miking, which sometimes is the best thing to do...

Decode it?
It does not take much effort at all to copy one track and phase reverse it.

Yes, but decoding it live while recording can be a bit more tricky. It's tough to monitor an undecoded MS signal without thinking you've been licking some acid earlier in the day.

Makes sense.

BTW, tones of really great info in this thread, so thank you every one for expanding my understanding of microphones.

Cucco wrote: It's tough to monitor an undecoded MS signal without thinking you've been licking some acid earlier in the day.

Prior to getting my Nagra V (which has MS decoding on board), I made half a dozen or so MS recordings without decoding during monitoring. By paying careful attention to the individual M and S signals, and judging what I thought *ought* to be in the centre and what *ought* to be in the sides, I figured I could do a pretty job of it.

Every single one of those recordings was a spectacular failure. I would've been better off dropping that acid when the curtain went up and spending the rest of the concert cowering in the corner of a bathroom, afraid of the soap dispenser...

Cucco wrote: [quote=d_fu]Yes, I used to think so, too, but that's not quite the way it is... Self-noise is not related to sensitivity, it's measured in SPL.

I know what you're saying and agree - however, the quoted spec of a fixed dB noise rating for each mic is directly related to (or more accurately to say - "impacted by" or "affected by")the amount of gain applied to that mic using your mic preamp. Where you have to use less gain, you will have lower (relative) noise levels. However, there's no escaping the S/N ratio - it's a matter of maximizing the headroom and gain of the mic at hand.

This is not correct Jeremy. The self noise figure is completely independent of sensitivity, by design. Its measured in equivalent sound pressure level, thus removing sensivity, ie mv/Pa from the equation.

Well, the sensitivity and the self noise are related and it boils down to Signal to Noise Ratio. If the SNR is the same between two mics, the only variable at this point is the preamp and its noise imparted at given gains. So a low noise, low sensivity mic should, in theory, be no noisier than a higher noise higher sensitivity mic if the noise and sensitivity go up by equal levels and the SNR is the same.

Let me just put it in different terms.

Take 2 mics. One of which has a 7 dB self noise and a low sensitivity (11dB/PA).

The second has a high sensitivity (40dB/PA) and a 20 dB self noise.

To get the same level of signal out of the first mic, you'd have to amplify it by let's say 13dB (I haven't done the math so I'm probably way off - let's just work on hypothetical.)

Now, the equivalent noise from mic a is 20dB since you've boosted the existing noise by 13dB. (Last time I checked, you boost everything that comes down the line, including noise.)

So, if everything else were equal between these two mics, you should get a pretty darn similar sound and noise figure.

If this is incorrect, please state how as this has been my understanding as long as I can remember.

Let's put it in a less simple term -

SNR is determined by the delta between the fixed *input* level of 94dB at 1kHz (1 Pa) and its self-noise figure which is determined as the electrical *ouput* of the mic with no load and no gain.

Since this delta always results in a figure of less than 94dB, is it not safe to assume that NO mic is suitable for 24bit recording?

My point is, determining SNR is a good factor for demonstrating a mic's self noise, but little else of relevance since it would be more accurate to measure the output of the microphone with a specified input and its output with no input.

Since the self noise output does not vary based on input yet output signal does, how does the existing SNR have any relevance to actual performance of the mic?

I remember we had a debate about this very early on. Could find the thread but nevermind, he is the explanation.

Take two mics, A and B, with the same self noise of 15dBA, but mic A has a sensitivity of 10mV/Pa and mic B has 40mV/Pa.

Put them up in front of a lute which produces 25 dBA at 1m. Mic A preamp will be on a setting which is 20 log 40/10 more gain than mic B's preamp setting, but the recording will have the same SNR for both ie 10dBA, ASSUMING the mic preamps noise floor is insignificant.

This last assumption is my only qualificiation and is true for most modern pres, they have insiginificant noise floors compared to mics.

Whats depressing is that I am unable to fully understand what you just said.
I think it is time for me to do my home work.

Cucco wrote: Since this delta always results in a figure of less than 94dB, is it not safe to assume that NO mic is suitable for 24bit recording?

If you were to find music (and acoustical surroundings) that could actually provide a dynamic range beyond 100 dB, the microphones' self noise might become a limiting factor... But this very unlikely to happen, isn't it? :D
Also consider that practically all 24-bit converters in the market have an SNR that does not even reach the theoretical SNR for 20 bit (120 dB)...

24-bit recording will help keep converter noise away when recording levels are set too carefully, e.g. with peaks at -20 dBFS or less.

And while a noisy mic's self noise (e.g. 26 dB) might become noticeable in quiet surroundings (the DPA 4090 being a case in point, totally useless as e.g. AB main mic), a quieter mic's self noise of 6-15 dB is likely to be inaudible.

My point is, determining SNR is a good factor for demonstrating a mic's self noise, but little else of relevance since it would be more accurate to measure the output of the microphone with a specified input

That's what sensitivity measures (mV/Pa).

and its output with no input.

That's the self noise...

Since the self noise output does not vary based on input yet output signal does, how does the existing SNR have any relevance to actual performance of the mic?

SNR and self noise are directly related, and both will tell you whether a mic is noisy or not. In relative or absolute terms...

Mind that SNR (as in the above examples) is a standardized measurement, allowing direct comparison between different microphones. A mic's SNR is not equal to its dynamic range. The latter will take into account a mic's max SPL handling. (Max. SPL - Self noise = dynamic range). This figure does not allow direct comparison, since some mics handle greater SPL than others. Hence the SNR figure, based on a fixed SPL value.

Cucco wrote: Take 2 mics. One of which has a 7 dB self noise and a low sensitivity (11dB/PA).The second has a high sensitivity (40dB/PA) and a 20 dB self noise. To get the same level of signal out of the first mic, you'd have to amplify it by let's say 13dB (I haven't done the math so I'm probably way off - let's just work on hypothetical.)

The problem with the math here is that sensitivity is measured in mV/Pa, not dB/Pa.

Daniel

My bad - I mistyped "db/Pa" vs. "mv/Pa"

All else that's stated is stuff that I'm aware of and do not dispute. (Actually, where you quote me as stating "and its output with no input." is equal to self-noise - I had made that statement in the previous paragraph, so we're on the same wavelength.)

David's information is helpful. Given his examples, I'd like to do some testing of my own to validate what's been said.