A N S W E R S
Many different mic set-ups exist for location stereo recording. Refer to_The New Stereo Soundbook_ by Alton Everest and Ron Streicher, Tab Books 1992, for more.
| Coincident Pair | ||
| Two mics arranged (typically) one above the other, so that sound waves reach both capsules at the same time. The following three techniques are coincident pair techniques: | ||
| XY Coincident | cardioids at 90 degrees | |
| MS Mid/Side | Use a single mic, which may be anything from omni to hypercardioid, facing forward (mid) and a figure-eight facing to the left. Remember that the back lobe of a figure-eight is _out of phase by 180 degrees_ from the front. When you add mid plus side, you get a left-pointing pickup. When you subtract mid from side (mid plus inverted side), you get a right-pointing pickup. If the mid mic is cardioid, the resulting left and right signals are cardioid at 90 degrees. Theoretically the result is the same as XY. | |
| Blumlein Coincident | figure-eight mics at 90 degrees | |
| Semi-coincident | ||
| Two mics angled to encompass the sound stage but also spaced between 6 and about 48 inches apart to add time-of-arrival/phase differences to the amplitude differences caused by the directional pattern. The following techniques are semi-coincident: | ||
| ORTF | Office de Radio-Television Francaise - two cardioids angled 110 degrees, spaced 170 mm. | |
| NOS | Nederlandsche Omroep Stichting - two cardioids angled 90 degrees, spaced 300 mm apart. | |
| AB | Generally two spaced cardioids facing forward to slightly angled apart [dj] | |
MS is nothing but a different mathematical representation of the polar patterns inherent in coincident AB recording. In theory there should be no difference between an MS mic and an AB mic when positioned identically. In practice, the greatest difference is due to the fact that sounds in the centre of the stage hit the M mic bang on axis; such sounds arriving at a coincident AB pair with a 90 degree angle are therefore 45 degrees off axis, and it is a very good mic indeed whose frequency response is anything like as good at 45 degrees as it is at 0 degrees. Therefore centre sounds are likely to be better when MS mics are used. The biggest disadvantage of the technique is that by definition, two non-identical mics are used. At the very least you could use the same type of mic body in each case, equipped with a cardioid capsule for the M, and a figure-eight for the S. Purpose built single body stereo condenser mics are often capable of either AB or MS working, but they are expensive.
Apart from the significance of MS mic techniques, remember that you can convert a stereo signal to MS at any time; altering the relative amounts of S will change the width of the stereo (the obvious reductio ad absurdum is no S at all - you are merely left with the output of the single mono mic fed to both speakers). Likewise you can achieve some increase in apparent width by increasing the relative S level beyond the norm. The reductio ad absurdum here relates to complete removal of the M signal - the output of the S mic appears in anti-phase on each of the speakers and the resultant signal when reduced to mono - zilch! So you have a convenient way of controlling width on a single fader, with the possibility of going slightly "super-wide" but beware; super-wide results in greater than normal out-of-phase content which some listeners find disturbing, and mono compatibility is always compromised. Apart from the super-wide aspect, the control of width available to you is really no different to that which you achieve by closing up the panpots on an AB signal.
MS mic-ing offers no magic solutions. Assuming you are using coincident techniques you are merely using a different model of the same signal. To sum up: central sounds will probably be cleaner as they are on the M mic's axis, but the fact that the two mics are non-identical may bring its own problems. Whether the signals are handled as AB or MS further down the chain is unlikely to make any sonic difference - AB is certainly more convenient. [po]
[dj] refutes [po]'s comments:
>In practice, the greatest difference is due to the fact that sounds in the centre of the stage hit the M mic bang on axis; such sounds arriving at a coincident AB pair with a 90 degree angle are therefore 45 degrees off axis, and it is a very good mic indeed whose frequency response is anything like as good at 45 degrees as it is at 0 degrees.
Not quite. Nearly all cardioids show in fact identical response over a 90 degree cone (+/- 45 from centre). It's the zone from 45 to 90 that's the problem, where the left mic is picking up right signals (by definition, a cardioid is only -6dB at 90 degrees!).
>The biggest disadvantage of the technique is that by definition, two non-identical mics are used. At the very least you could use the same type of mic body in each case, equipped with a cardioid capsule for the M, and a figure-eight for the S.
I don't think so. The two output channels left and right are produced with the same virtual microphone, which is the sum of the two real microphones used. My early MS work was done with a ribbon mic for S (an RCA 44BX, which is flat to 20 Hz, something very few condenser figure-eight’s can claim) and a cardioid condenser for M. This worked just fine. Remember you are measuring very different things: the S mic is sampling particle velocity while the M mic is sampling vector pressure. Each of the outputs L, R contains equal amounts of the two microphones. The big problem with using the same type of mic body is that this assumes, in all currently available interchangeable capsule systems, a 19 to 21 mm maximum diameter. This works OK for a cardioid or omni, but I have yet to hear a figure-eight that small that sounds right, which is the reason I no longer make the KA800 capsule. Schoeps tries to fix this problem with extensive baffling and damping; AKG and Neumann don't attempt it anymore (AKG had a large diaphragm figure-eight capsule in a wire mesh ball which would screw onto the end of a C451 but does no longer); the Sennheiser small figure- eight uses fairly radical electrical EQ in the (non-interchangeable) mic body to make the capsule measure flat. Neumann made a small capsule dual-diaphragm mic that synthesises figure-eight by combining two cardioids back-to-back and out of phase. This is found in their KM56, SM2 stereo mic and the KM88, and is probably the best compromise for a small figure-eight but the frequency response is not wonderful (-6 dB at 50 Hz, +6 at 6 kHz, -6 at 14 kHz).
>Assuming you are using coincident techniques you are merely using a different model of the same signal. To sum up: central sounds will probably be cleaner as they are on the M mic's axis
Given ideal microphones, they are the same signal. With practical microphones central sounds will be very similar between the two techniques, it's the edge sounds that will be most compromised by using XY. Another minor problem with XY is that you can't get the two mics in the same place. With MS, the mics are automatically in the same place and the physical obstruction that one mic makes in the other's sound field is symmetrical for the left and right channels. This is also the case for XY when you put the mics in the same plane, but now you get reflections off one mic's diaphragm grille right into the other mic, which makes for some nasty comb filtering at high frequencies. If the mics are one on top of the other, there is a vertical time of arrival difference of opposite sign between the two channels that can cause odd shifts in the reverberent pickup. There's no free lunch.
>Whether the signals are handled as AB or MS further down the chain is unlikely to make any sonic difference - AB is certainly more convenient.
Yes... which is why many people recording in MS will in fact matrix at the recording site, putting down left and right tracks on tape rather than M and S.
Note I left out of my earlier post on MS, there are two excellent sources for info comparing the various stereo mic-ing techniques. One is the compendium volume "Stereophonic Techniques" which is articles excerpted from the Journal of the AES, and published by the AES. The other is the "New Stereo Soundbook" by Alton Everest and Ron Streicher, published last year by Tab Books. [dj]
There are two principal types of transducers used in mics: dynamic and condenser. Dynamics are often favoured for miking individual instruments because they add a favourable colour to the sound. Condenser mics are generally more accurate than dynamic and are preferable for audience recording. Modern condensers use an "electret" design which enables the mic to operate from a low voltage which can be supplied by an internal battery or by an external power supply or by certain mic preamps. By contrast dynamic mics need no power source. [rg]
This is a type of question which is sure to get many "religious" responses, but I will try to stick solely to the facts. I will also try to keep the discussion in the context of audience recording, since this is probably the most common application amongst DAT-Heads.
| a) Omnidirectional | |
|
This type of microphone will pick up sound evenly from all directions, hence the omni- prefix (they are sometimes called non-directional). For this reason, omnis should be used only when recording very close to the source. It is generally felt that omnis offer the best sound as compared to directional mics, and top omnis are often less expensive than top directional mics. For an audience application, omnis are generally only used for fob (upfront) recordings, since they tend to pick up too much audience noise when used at too great a distance. Another common audience application is to use an omni/shotgun mix from further back in the hall. The shotguns are used to pick up the direct sound from the sound system while the omnis provide ambient information. The reasoning behind this method is to give "presence" to the mix, which would be lacking in a shotgun-only recording. When using a spaced omni setup (several inches to several feet between mics) in quiet venues, distance is less important a factor; quality recordings can be made even far from the source. For binaural recordings (with the omnidirectional microphones mounted near your ears and the recording played back over headphones), closeness to the source is only a minor factor in achieving a quality recording. | |
| b) Cardioid | |
| This type of microphone is slightly more directional than omnis. Much of the sound coming from behind the microphone is not picked up. There is also a small amount of side rejection as well. Cardioids are often used as vocal mics since they will not pick up most of the other noise onstage coming from the amps, monitors, etc. For a Dead-show-type application, proximity to the sound source is important, though not quite as critical as with omnis. Like omnis, you will get the best results when used upfront. Some people use cardioids from further back, but audience noise becomes more of a problem the farther back you go. If you plan to use cardioids from a large distance (i.e. the tapers' section at Grateful Dead shows), the best results are usually obtained when the mics are elevated as far as possible above the audience in order to minimise crowd noise on the recording. Cardioids usually work well in a small club setting and also outdoor amphi- theatres, where the crowd noise tends to be more attenuated than on indoor audience recordings due to the lack of reflective surfaces (i.e. side walls and ceilings). A cardioid which has a 20-20000 Hz frequency response will generally be less expensive than a comparable hypercardioid or shotgun. | |
| c) Hypercardioid | |
| This type of microphone is more directional than standard cardioids but less directional than shotguns. Hypercardioids are *not* shotguns. They can be thought of as "short shotguns." They have more side rejection than cardioids but not as much as shotguns. I find the hypercardioid to be the most flexible of the four polar patterns discussed here (in the context of audience recording). Excellent results can be obtained from far back as well as upfront. They also work quite well in small clubs. Hypercardioids are generally the least common of the polar patterns and you can expect to pay a tidy sum for a good pair. | |
| d) Shotgun | |
| This type of microphone is the most directional of the four. Shotguns have the most side rejection and thus are well suited for recording at a large distance. Shotguns are the microphone of choice in the tapers' section at Grateful Dead shows. Shotguns tend to have a number of drawbacks, however. The less expensive shotguns will not have very good frequency response, especially in the lower octaves. For this reason, lower-end shotguns are often described as "tinny" or "hollow" sounding. Also, the drastic side rejection of a shotgun often results in recordings with a lack of "presence." Expect to pay a fairly large sum to get a shotgun with 20-20000 Hz performance. Because of their large size, shotguns are not usually used upfront. Directionality is useful for increasing the ratio of direct sound (from the stage and P.A. system) to reverberant and ambient sound (from the rest of the room). This becomes more critical as the distance from mic to stage is increased. Generally, one must pay more money for a shotgun in order to get as good sonic characteristics versus a less directional mic.
In contrast to omni's and cardiods, microphone placement is very critical with shotgun capsules, which some believe is more often the reason behind the "tinny, hollow" sound than the technical qualities of the microphones themselves. These mics are very directional, and you really have to consider the fact that they are recording where you point them -- unlike omnis which record the sound where they are located. After some years of experimentation with Nakamichi CM-100 bodies with the CP-4 shotgun capsules, I [sj] have found that the amount of bass in the recording is highly dependent on the position of the mic with respect to the PA. My current alignment results in a very clear bass, almost to the point of considering the use of the "Lo-Cut" switch. It is my opinion that the "traditional Dead Taper" placement does not adequately take these effects into account, hence the resulting "tinny, hollow" recordings that people dislike. |
|
DAT-head recording techniques are generally divided into stealth and non-stealth projects. For the former, generally the technique of choice is to wear a pair of mics more or less near the ears (binaural). This works far better than it has any right to, given that commercial or semi-commercial stereo mic systems can be had for less than $250, or you can also build your own for less than $10 using the Panasonic WM063 capsules that seem to work best.
For non-stealth techniques (including overt on-stage taping when they want you to be there) it's mostly a choice of directional pattern and tone colour. Less directional microphones generally sound better than more directional ones, but often you need to record from a long way away, and the only simple way to do this is with highly directional mics.
Omni mics are absolute pressure transducers and inherently would have response down to DC but for a small air leak built in to allow them to be shipped by air, not explode in tornado’s, etc.
While omni mics, being pressure transducers, almost always produce a cleaner sound, many times the recordist finds that some directionality is required. The practical way to achieve this is with a mic that combines absolute pressure and pressure gradient responses. All of these mics have a directional pattern that follows the equation x + y*(cos of angle) where x+y=1 and the angle begins with 0 being on-axis to the microphone. The standard cardioid is x=y=0.5, hypercardioid is x=0.3, y=0.7.
There are two families of directional mics in common use: phase shift pressure gradient, and interference tube. Phase shift pressure gradient mics are available in a continuum of patterns which can all be expressed with the relation x + y*cos(theta) where theta is the angle from the mic to the sound source. For x=0, the pattern is bi-directional with the implied phase reversal toward the rear and sharp nulls at 90, 270 degrees. For y=0, the pattern is omni. For x=y, the pattern is called "cardioid" (for heart-shaped, not "cardioid", please) and the response at 90 degrees is 0.5 of (or -6dB from) the on-axis response. For x around 1, y around 2, the pattern is called "supercardioid" or "hypercardioid" and there is a small rear lobe out of phase with the front pickup, but sharp nulls at 135, 225 degrees.
Since the pressure gradient is sensed by leaving the mic diaphragm open on both sides, the inherent frequency response is inversely proportional to wavelength... the longer the wavelength, the less response, because there's less pressure gradient at any instant between the front and the back of the mic. This makes a frequency response curve that, left to its own devices, would be very un-flat; rising 6dB per octave. Most unidirectional mics use a lot of damping to flatten out the response at some point, above which it doesn't rise anymore. But below this frequency (typically 100-400 Hz) you have in effect a bass rolloff. Which is another reason that directional mics are often used in concert recording, because all of the low frequency energy is rolled off even before it gets to the first FET in the mic body.
Variable pattern mics are available (AKG C414, Neumann U87, U89, TLM170, Milab, etc.) that are either two cardioid capsules back-to-back (each one of 0.5 + 0.5 cos theta, you can work out the patterns that result when they're added in phase and out of phase with each other), or are (Schoeps MK5) a phase shift pressure gradient type with mechanically variable phase shift parts to make different patterns.
Interference tube mics use a perforated tube in front of a phase shift pressure gradient type of microphone, that gives more directionality at frequencies where the tube is more than a quarter wave long. Most interference tube mics, even expensive ones, have a sort of strange hollow sound which may be more than overcome by the fact that the performer sounds closer than it would otherwise be possible to get, and as you point out there is a lot less crowd noise. [dj]
The general idea is to place microphone selection in perspective. If the microphone is great, but the factors listed below are not considered, the results can be disappointing.
Your actual results will depend on: the spiritual, emotional, and intellectual content of the music; the skill and level of preparation of the composer, musician, and recording engineer; the amount of ambient noise; the musical instruments; the room acoustics; microphone placement; the quality of the recording equipment (including the preamp); the quality of the playback equipment; the environment in which the playback takes place; and the ability of the listener to appreciate music. [pd]
[td] writes:
My conclusions ?
I prefer using the Schoeps, Neumann's or B&K's.
Why? I like the sound. Many other who hear them like the sound as well.
However, they are not without problems.
If the 4011's are used off centre, they sound like shit.
The KMi84's outdoors are a bit boomy (maybe that was just Foxboro)
The Schoeps are a little tinny, etc...
Each of these mics has its use and its place.
That perhaps is some of the most valuable experience a recordist can gain about his or her art. The only way to know is by experimentation and LOTS of listening. However, if you can only afford one set of microphones and you rarely get a chance to compare them with any other, then you'll probably be happy with whatever you have.
Someone once asked me, "...what's the best placement of a microphone to record a piano?" I replied with, "...what's the best place to put a camera to photograph a mountain?" [ss]
The questions: 1 - 49
answers to questions: 7 - 14
answers to questions: 15 - 31
answers to questions: 32 - 49