How I Record Steelpan and Why I Use this Technique


Making a good recording depends on a lot more than putting a couple of microphones near the band and recording what they pick up.  You may be lucky and get a decent or even good recording, but often this technique will lead to disappointment.

The technique I use for almost all my recordings is referred to as "orchestral recording", as opposed to "close miking".   There are a number of orchestral recording techniques, each with advantages and disadvantages.  Most depend, at the core, on using two (one for mono) microphones located so as to pick up the sound of the entire orchestra with proper balance and presence.  Under certain circumstances the recordist may add "spot microphones" to record, often on a multi-track recorder, parts of the orchestra that do not present themselves properly into the main stereo pair of microphones.  There are also variations of some of the stereo orchestral recording techniques which add extra microphones to enhance the recording under some circumstances.

Many of my recordings are of steel orchestras with a few to over 100 musicians.  Occasionally I will record a solo or a duet.  Most of the recordings are of a live performance or a public rehearsal  with an audience present.


Equipment Choice and History:

My current techniques and basic equipment were chosen to fulfill the following criteria:

  • My setup had to be capable of making good quality recordings under conditions I was likely to run into at the venue.
  • The equipment had to be affordable. To a degree you get what you pay for, but I have tried to spend more of my available capital funds on the items most likely to improve the recording.
  • My core set of equipment had to be compact and portable, I wanted to be able to carry it from my location near New York to Trinidad without incurring additional baggage charges.
  • It was highly desirable that the equipment could be battery operable when no power was easily available.
  • The equipment had to be able to be set up and taken down quickly at the venue.


The evolution of my equipment over the years:

The equipment I use has changed drastically over the 20 plus years I have been recording pan.  Part of this is due to changed or more clearly recognized requirements, part of this is due to the rapid advancements in technology that have made it possible to improve the recordings at reduced costs, and finally, part of this is due to equipment becoming obsolete, as newer devices replaced older ones.

Initially, I used a pair of Shure SM81 cardioid microphones connected to a Tascam portable DAT recorder.  I soon added a pair of quality microphone preamplifiers to use in place of the mediocre ones in the Tascam.  The microphones were set up according to an orchestral recording technique called ORTF (for the French broadcasters who invented it.)  Most orchestral techniques require microphones with particular pickup patterns to be located at precise distances apart and at particular angles to each other, based on acoustic physics.  Some of these techniques are listed in my article on mike technique basics.  

I got a number of decent recordings with this setup, including one that is still available commercially in Trinidad.  There were, however, issues that pushed me to move to a different core set of equipment:  The equipment (including moderate length microphone cables) was somewhat bulky and heavy to transport, and with the airlines reducing baggage allowances, I'd be unable to take it to Trinidad.  It required a source of electricity for all but the shortest recordings, and for any recordings if I used the external mic preamps.  There was a definite lack of bass in the recordings, partly due to the response of the SM81 mikes, partly due to the physics of almost all directional mikes. (However, the weak bass could easily be corrected with a little discreet EQ during post-production.)  Finally, the ORTF setup required attention to detail in setting up the microphones at the correct spacing and angle, increasing setup time.


The Mini-Disc Portable Kit:

In late 2002 I decided I needed something more portable to take to Trinidad for the upcoming Steelband Music Festival.  Several manufacturers were selling Mini-Disc recorders, extremely compact and not very expensive.  The downside was that they recorded "compressed bit-rate" recordings in a Sony format called ATRAK, similar to MP3's.  They were applying psycho-acoustics and leaving out "inaudible" sounds in order to achieve reasonable recording times on the small mini-discs.  I was not happy to make this compromise, but listening tests of the highest bit-rate, least compressed recordings compared favorably with CD's.  I purchased a recorder made by Sharp that had good reviews.  One other negative was that there was no way to transfer the complete recording digitally into my computer to edit.  I had to play the disc in real time, recording in analog on the computer connected to the line output of the Minidisc.  Tests showed that there was no audible degradation doing this.

For microphones, I decided to try a pair of DPA 4061 lavalier microphones (the type you see news anchors wearing) that I already owned.  These are omnidirectional mikes (they pick up sound from all directions essentially equally).  The specifications on these mikes are very good, the frequency response is essentially flat from well below to well above the range of human hearing, with a slight "presence bump" in the upper midrange that typically adds a little "sparkle" to recordings.  Their only weaknesses are that they have somewhat high "self noise", hiss that will be totally inaudible in any live recording situation; and the cables are somewhat fragile, more of an issue when worn by performers than when on a fixed stand.

To do orchestral recordings I used a technique used since the earliest days of stereo in the 1930's, called "spaced omnis".  The microphones are placed anywhere from a few inches to some number of feet apart, depending on the width of the "orchestra" you are recording and how wide you want the stereo image to be in the recordings.  The advantages for me were:

  • it allowed me to use the very small and light lavaliers.
  • Omnidirectional mikes typically have very good bass response, better than most Cardioids (mikes that pick up from the front only.)
  • The setup is simple and can be easily varied to suit the needs of the venue. 

    The  disadvantages of spaced omnis are that the recording may not be very "monophonic compatible" (not an issue today unless you plan to broadcast it on AM radio), and that being omnidirectional, the microphones may pick up undesired sound, such as noise in the audience, spill from a PA system, etc.  To me, given my goals, the advantages outweighed the disadvantages.

One remaining issue was powering the microphones.  Electret microphones, such as these lavaliers, require a source of voltage to work.  The Minidisc microphone inputs provided voltage to work with consumer headset microphones.  While it would have worked with the DPA mikes, the voltage was less that that recommended by DPA and would have degraded the sound of the microphones. I could have bought a pair of battery packs from DPA for about $150. each, but I designed my own, built with a few dollars worth of parts from Radio Shack.  I did purchase some extension mike cables from DPA. 

So I was in business.  I tried this setup and was quite pleased with the results.  I used this setup for several years before upgrading.



The Hi-Minidisc and Core Sound Preamp. then the Denon-Marantz Handheld:

A few years after I started using this setup, Sony came out with a "Hi-Minidisc".  It offered the ability to record uncompressed 16 bit audio and the associated software allowed digital transfer of the recorded files to a computer.  I quickly purchased one.  While the uncompressed storage was a feature, I discovered that the microphone preamplifiers built into the Sony recorder were not of a high quality, and I don't think the resulting recordings were as "pure" as the recordings on the Sharp recorder.

That fall, at an Audio Engineering Society convention in New York I saw an exhibit from a company called "Core Sound" of a microphone preamp with built-in analog-to-digital converter and digital optical output.  It was able to operate for several hours off a 9 volt battery or forever from a "wall wart".  Furthermore, it could be configured to properly power the DPA microphones, allowing me to get rid of my "battery box" microphone power supply.  The audio specifications were very good, so I purchased one. I ran performance tests in the lab where I worked, and indeed, the audio performance is superb.

At this point my setup used the DPA microphones connected to the Core Sound preamp and A to D converter.  The digital optical output of the Core Sound preamp was plugged into the digital input of the Sony Hi-Minidisc, which was now simply a digital storage device.  I did add a headphone amplifier for monitoring, also made by Core Sound.  I found this setup capable of making excellent recordings.

Around 2011, Sony stopped manufacturing the Hi-MD discs, forcing me to choose a different recorder.  After some research I purchased a Denon-Marantz (now rebranded Marantz Professional) DM661 handheld recorder.  This recorder uses SD cards as storage media.  I plug the digital output of the Core Sound mic preamp into the digital input of the recorder.  Essentially, the recorder serves as a digital storage device, similar to the HD-Minidics. 

This is still my basic setup today, although I now also own a simple and relatively inexpensive multitrack recorder that allows me to add "spot" microphones (for a vocalist or solo instrument) when used together with a small mixer, such as made by Mackie.


My Current Orchestral Recording Equipment, in the Skiffle Panyard in TrinidadMy Current Orchestral Recording Equipment, in the Skiffle Panyard in TrinidadSummary of my Current Equipment:

This is a list of the equipment which makes up my basic, extremely portable simple stereo recording kit.

  • A pair of DPA 4061 lavalier microphones with available windscreens and extension miniature coaxial microphone cables.
  • A Core Sound Mic 2496 microphone preamp and analog to digital converter set up to directly connect to and power the DPA microphones.  Core Sound now has a newer "V2" model for sale.
  • A Denon-Marantz (now Marantz Professional) DM661 handheld recorder (Again a newer model, currently identified as the PMD661 Mk II is now being sold).  The "S/PDIF" digital output of the mic preamp is connected to the digital input of the recorder.  Both the preamp and the recorder are set for 44.1 KHz sample rate, 24 bit recording.
  • An external headphone amplifier, also made by Core Sound, called the Headline.  An external amplifier is desirable if your setup is close to the band.  The maximum volume from the handheld recorder headphone jack may be inadequate to drive your headphones loud enough to tell what sound is coming from the recorder and what is leaking past the earmuffs on the headphones.  You can in all likelihood find far less expensive headphone amplifiers with enough gain and power.
  • a pair of Senheisser HD280 Pro headphones.  There are other professional headphones available, with varying performance.  Get a pair that you like and are used to.  They must be "closed back" to keep out the direct sound of the band. 

    I strongly recommend against any form of "ear buds", most of them are horrid, and none of them provide much isolation from ambient sound.  I also don't recommend "Beats".  While the built in amplifier may make them plenty loud, the frequency response has a boosted bass, making it impossible to truly determine the balance of your recording when using them.


DPA Microphones Attached to Two Floor Stands, Improvised Platform, Skiffle Steel Orchestra, TrinidadDPA Microphones Attached to Two Floor Stands, Improvised Platform, Skiffle Steel Orchestra, Trinidad Mike Stand Selections:

I have several microphone stand setups:

  • I can use two ordinary tripod stands, preferably with "junior booms" aimed straight up to get the microphones over people's heads. 
  • I can use a single tripod stand and junior boom with one or two extra 5/8 inch 30 inch long stand pipe sections threaded at both ends, together with Atlas Sound 5/8 inch couplings.  The stand sections were salvaged from old stands, and I had a machinist thread the unthreaded end with a 5/8-27 thread to fit the couplings.

    I set this up to look like a "T" with one of the extensions below the junior boom, which is set horizontal.  If the band is small, I tape the microphones to the end of the boom pipe, about 24 inches apart.  If the band is large, I use the other extension as a part of the "T" making it somewhere around 60 inches long.  I tape the microphones somewhere  about 48 inches apart.
  • I have a On Stage SB9600 Tripod Boom Stand, which, when used straight up with a junior boom at the top can get the microphones close to 10 feet in the air if necessary.  Again, I will use the extension stand section to make the length of the "T" longer than 48 inches when recording a large or widely spaced band.


On Stage Large Boom Stand with Spaced Omni Microphones, Brooklyn Christmas Steelband ConcertOn Stage Large Boom Stand with Spaced Omni Microphones, Brooklyn Christmas Steelband ConcertMicrophone Placement:

Correct placement, spacing, and for other than omnidirectional microphones, aiming of the microphones is critical to getting a good recording.  With orchestral recording, the intent is for the microphones to pick up a good blend of all the instruments, similar to what you hear with your ears.  For this to occur, the microphones must be located some distance in front of the orchestra.  If you place the microphones too close to the front of the orchestra, your recording will contain far too much of the instruments nearest to the microphones.  If you place the microphones too far away, you will likely get too much of the room acoustics (reverb, noise).  If there is a live audience present, you may well hear too much of the audience in the recordings.

For recording steelbands I typically try to set the microphones 10 feet or so in front of the orchestra, closer if the acoustics aren't ideal and/or the number of musicians is small and they are close together, further back if possible if there are a large number of musicians, like a Panorama side.  I often end up closer than I would like, due to space constraints or the presence of an audience sitting close to the orchestra.  If patrons are sitting very close, I may ask people not to sit in the immediate area of the microphones, if the people running the event are agreeable.  It is not their physical presence that is the issue, it is that they all too often will start a private discussion in the middle of the recording, or will develop a sudden bad cough!

When making scheduled recordings during what otherwise are rehearsals, such as "panyard" recordings, I will typically block off the area for some distance around and for the width of the band in front of the band, sometimes with caution tape.  One or two people in this space will have a negligible effect on the recording as long as they don't talk, but a crowd will soak up the high frequencies.  Your front line will sound very dead.

For spaced omnis, my usual technique as I explained above, I typically place the two microphones two feet apart for a small sized ensemble, four feet apart for a full sized orchestra.  The trade off here involves distance to the different sections.  If the mikes are close together and the orchestra is "wide" with, as is common, the bases at the far sides, your recording may have weak bass, due to the distance from the bass pans and the microphones.  If you place the mikes too far apart you will get a "hole in the middle", you won't pick up the instruments in the center of the orchestra well.  There are variations to the spaced omni technique to avoid this (beyond the scope of this article), but the simplest solution is to move the microphones further back, making the distance from the microphones to the different parts of the orchestra more equal.  Another solution I use on occasion, particularly at panyard recordings, when there is no space to move the microphones is to suggest to the arranger or bandleader that we move some of the bases (or whatever is weak) closer to the microphones. Since they want a good recording, most are amenable.

The next issue is the height of the microphones.  I try to position the microphones relatively high, around 9 to 10 feet above the musicians if possible.  This is for a very practical reason, not for better sound.  If someone walks or stands directly in front of one of the microphones, the high frequencies disappear, it sounds muffled.  It's like putting pillows next to your ears.  If the microphones are up high, no-one can step in front of them.  If someone steps directly in front of a pan, the sound is muffled, but only from that pan.  Unless that is the only pan playing that part, you are unlikely to hear the difference.  A secondary benefit if there is an audience, the mikes are a little further from the audience and you are less likely to have an issue with private conversations, coughs, etc.

In New York, I have a couple of stand options described above.  In either option, I use a junior boom arm on the top, "T" shaped, crosswise parallel to the ground.  I have some extra stand pipes, threaded on both ends, and some couplings.  So I can use my regular tripod junior boom stand with an extra pipe to add to the height, or the heavy duty tall On Stage boom stand, which will go up over 10 feet.  As explained above, with either stand the mikes are from two to four feet apart

In Trinidad I try to borrow a pair of stands from a friend.  If all else fails, I improvise.  I have used pan racks, a ladder with a 2X4 taped crosswise to the top, the fence around the panyard, whatever is available to hold the mics.

One other item involving the mics, unless it is dead calm, use windscreens (available from DPA)  Even though omnidirectional mics are less subject to wind noise, you will have a problem even in a light breeze.


Recording Levels and Settings:

When recording digitally, it is important to keep the levels below clipping (0dBFS).  Extensive clipping can create very objectionable distortion.  Also, at about the level that clipping takes place, the microphone preamp may also start distorting, and this distortion may well be more objectionable than clean clipping. This is not like cassette recording, where you wanted the meter to bump into the red on the peaks.  If you get a chance to set levels in advance, set the peaks of the loudest part to be about -10dBFS on the meter.  If you are not sure or don't get a chance to set levels, be conservative.  I've gotten excellent recordings where the peaks never got above -25dBFS.  I do use 24 bit recording, which avoids certain kinds of distortion if the levels are low. 

Make sure the recorder is in manual level adjustment, not automatic.  Recording with automatic level control will remove all the dynamics from the music.  I record at the 44.1KHz sample rate, that is the bitrate on CD's, and since my post production processing rarely involves time stretching or pitch adjustment, there is nothing to be gained from a higher bitrate with modern equipment.  One caveat:  If your recording is going to be married to video, as opposed to being burned to a CD, record at 48KHz, that is the bitrate used for all audio married to video.

If despite your best efforts, you get an occasional clip, it probably won't be an issue. There is software available that "unclips" slightly to moderately clipped recordings, provided they don't also have other distortion.  The particular software I use for this is not inexpensive, but I am sure less expensive unclippers are available.


Post Production:  

Following the recording, I copy the files to my computer.  The post production steps which follow are important in making a satisfactory recording.  I use Sound Forge made by Sony and RX4 made by iZotope to edit the files, but you can do an adequate to excellent job with free audio editing software on the web.  The most well known of these products is Audacity.  

The first step is to mark the beginnings and ends of the takes:

At the beginning, if you are burning a CD, you have to leave a little silence or unimportant audio before the beginning of the track.  CD players take some time to "unmute" as a track starts, and if you start the music exactly at the beginning of the track, the CD player  will "upcut" or chop off the first note or so when the listener plays the track.  Different players take different amounts of time to unmute, I have found leaving 0.350 second of silence (or fade in, if there is a lot of room noise) at the beginning of a track works with about any CD player.

If there is applause, I have to decide whether to leave it as a part of the take, or cut it out.  If I leave it, and it is more than about 7 to 10 seconds, I will mark the end at about that length, and later, after all other processing, fade it out over two or three seconds.  If there is no applause, or short applause, I allow a second of "room tone" background sound after the decay of the last note or applause, to apply a quick fade after the track is processed.  You want to be careful to avoid chopping the end decay of the last note off, it sounds unnatural.  Let the note decay completely before ending the track.

I then listen carefully to the take(s) and perform any corrective work.  If it is an indoor recording, I may have to add some gentile EQ to adjust for poor room acoustics.  I sometimes use my restoration software (Izotope RX4) to remove objectionable noises such as cars with noisy mufflers.  Modern editing software will allow you to remove or attenuate all kinds of extraneous noise, you have to decide if it is worth the time and if the result is worse sounding than the original.  I rescued recordings in a catering house in Brooklyn where there was a DJ in the next room with a hyperactive subwoofer, louder than the steelband.  It took many hours, and probably wasn't perfect, but made something listenable out of what was pretty bad. You should be aware that good restoration software, particularly with the spectrum editing I needed to excise the subwoofer's booms, does not come cheap, and here to a good degree you get what you pay for. 

The next step in post is dynamics adjustment and normalization.  You may not know what normalization means.  It is adjusting the levels to the standard desired for release.  Historically, if you were making a CD. you normalized the track so the loudest peak was at 0dBFS, this means it was just at the point of clipping distortion.   If you take your raw recording and normalize it, you will probably find it plays back very softly.  You may think you have to apply a compressor to make it louder.  I rarely use a compressor on my tracks.  I want the full dynamic range of the original track, or at least the audible part of that dynamic range to be available to the listener. 

So how do you make your recording reasonably loud but still have good dynamic range.  If you look at the "volume envelope" of most recordings as displayed by most editors, you will see narrow spikes are the loudest parts.  These momentary spikes are essentially inaudible and can be "shaved off", allowing the track to be normalized usually 6 to 12 dB louder with no audible effect on the music.  To do this, you use a limiter.  The limiter is set up to remove the extraneous peaks, which will allow the normalizer to add gain without clipping.  What is the difference between a limiter and clipping?  A good limiter, properly set up (these days, your limiter will probably be a plug in or setting in your audio editing software) will "shave off" the peaks inaudibly. Clipping will probably cause annoying distortion. 

There is a limit to how much you can limit the recording.  If you shave off too much, you will find the track will loose its dynamics, and you may hear "pumping" where a loud note in one part of the audio spectrum, often the bass, will cause the audio in the other parts of the spectrum, like the midrange to momentarily dip.  If this happens in one of your recordings and you need to limit the audio without pumping, there is a plug-in called a multi-band limiter but its set up is beyond what we are doing here.

Knowing how much to limit is an art, trial and error.  You need to listen carefully to the resulting track.  I do it by eyeball, looking at the volume envelope and deciding how many dB I can shave off without damaging the sound of the track.  If you are not sure, try different amounts of limiting followed by normalization and see what results you like.  Less is better until the track plays back "too soft". 

There is another method recently made available which takes most of the guesswork out of this, but it is only available in more expensive software now.  It is called "loudness control".  You set a desired loudness number, -18 LKFS is a good conservative number, -14 is a little louder, and apply the software to the track.  If it is a good plug-in or software package it will apply a "transparent" limiter as needed and normalize the track to play at exactly the desired loudness and not clip.  Some packages may not include the limiter, rather they will tell you how many dB you have to shave off with the limiter to be able to apply the loudness plug-in without clipping.  I have recently started using a very good loudness plug-in to normalize all my tracks.

The next step is to apply fades at the beginning and end of each track.  At the beginning, I use a graphical fade in, where I completely attenuate the track for almost the entire 0.350 second and then quickly fade in ending at exactly .350 second (see above about the start mute time of CD players to understand this.)  At the end of the track, the fade depends on how the track ends.  If there is long applause, I apply a slow, even fade of two to three seconds at the track end.  I don't like to fade applause faster, as to me it sounds "chopped off".  If there is no applause or short applause, I apply a short quick smooth fade after the last note before the end of the track. 

Finally, you may want to add metadata and either burn CD's of the completed album or make computer files to listen to.  If you are making computer files, the metadata tags are necessary for the player (Windows Media Player, your smartphone software, etc.) to locate the track.  At the minimum you need Track Number (make it track one for a "single" track), Track Name, (Track)Artist, Album Name, Album Artist (may be the same as Track Artist), and probably Genre and Year.  You can add more, like composer, etc.  Most tagging standards don't make provision for Arranger, something many steelband people would like included, but you can usually add custom tags.  Unfortunately most playing software won't show custom tags.  Another way is to use the "comments" tag, with an entry like "Arranger: John Doe". 

Your editing software will probably have a way of making mp3 or m4a files for portables or FLAC files for computer storage playback, or .WAV files to make a CD from if the editing software doesn't include a CD burner package.  One other note:  If you are burning a CD, burn it DAO (Disk at once, all the tracks are burnt at the same time) not TAO (one track at a time.)  Many players will not properly play TAO CD's, and if they do, you'll likely get a loud pop between tracks.


Where and When Orchestral Recording Probably Won't Work:

Much of what I will bring up here repeats the material in the mike techniques basics article.  It bares repeating here.  Every recording technique has its limitations.  This technique will not give good results if the acoustics are bad.  The acoustics are part of the recording.  If you are indoors and the room is reverberant it will likely sound even more reverberant in the recording.  If the reverberation sounds bad, there is not much you can do, record some other place.  If the reverb is pleasing but too strong (like a church) you can try moving the mics closer to the players.  But, then you may not get a good balance between the parts.  If you are in a church and the steelband is playing classics or slow gospel, the extra reverb is probably OK, that music was intended to be performed in a reverberant church.  If they are playing fast Soca, forget it, you will probably end up with mush.

I should comment the steelpan is an acoustic instrument.  If the acoustics are lousy, you may well not get a good recording no matter what technique you use, even close miking.  Close miking (on all the instruments, best multitracked) will give you a chance at an acceptable recording, better than orchestral miking in this case; but if the acoustics are truly bad, even the close mics will pick up some of the bad acoustics.

If there is a PA system, forget it, unless the system is kept low, and the speakers are way away from the mics, not usually the case.  Also, all too often the people doing PA mic pan very poorly and get really lousy sound through the PA system.  I think all of us have been exposed to this.  I have rarely gotten a usable orchestral recording when there was a PA system in use; I usually don't even try any more.

If there is supposed to be PA and you have some control over the event and can get the cooperation of the FOH (Front of House) sound reinforcement mixer and setup person, and you have multitrack equipment available, you can carefully mic each pan (miking whole areas doesn't work with typical live band setups) and either use mic splitters to feed both the FOH SR console and the recording device, or take "prefade" sends from the FOH console to feed the multitrack recorder, and you might get a good recording which you can mix down in "post".  This technique (best with the mic splitter transformers) is used at most pop music concerts when a "live" recording is being made.  What generally won't give good results is to record the output of the FOH mixing console.  The mix which sounds good in the sound system (if it even sounds good) will probably not be a good recording mix.

If there is a crowd directly in front of the band with no space, the only way to possibly get a good recording is to get the mics up high above the crowd or (if you can set up before the event starts) hanging above the band.  Recording through people is like putting the mic in a pillow!  Also, if there is a crowd and little space, you may get far too much crowd noise in the recording.  Sometimes a loud energetic crowd will add "spirit" to a recording, sometimes it just drowns out the music.

Watch out for the audience member close to your mic who decides to have a private conversation during the recording.  Crowd noise is OK in moderation; what were supposed to be private conversations now made public in your recording are really annoying.  This is particularly an issue during soft numbers or parts of numbers, or stops in the music. 

You may be done in by low flying aircraft, motorcycles, loud trucks or buses, dropped folding chairs or beer bottles, barking dogs, you name it.  Environmental sounds are a real problem.  If you have expensive restoration software and copious time, you may be able to clean up some noises, maybe not.

If there is any chance of a breeze, use the windscreen designed for your microphones.  Wind noise is obnoxious.

Rain is an issue.  I will put "baggies" over my mics if it starts to rain, but raindrops hitting the baggies will make loud pops.  I have sometimes rescued recordings made in the rain by using the software designed to remove scratch pops and clicks from recordings of LP's.

Watch out for SLR cameras with large external flashes.  Don't let them be used near your mics; they often make a loud but brief "pop" every time the flash goes off.  That can make a mess of a recording.

Needless to say, cell phones are an issue.  If it is strictly a recording session, get everyone to turn them off, if it's a live event, good luck!  Also, don't put a cell phone near your recording gear.  The signal from the phone, even when not making a call, may get into your wiring and make interference severe enough to ruin a recording.

It goes without saying that the band needs to be well rehearsed. If this is strictly a recording session, and they have to do multiple takes, after a few it all goes downhill, go on to the next song or take a break.  Also, performing for recording is much more critical than a live performance.  If someone hits a wrong note in a live performance, it is gone, and usually forgotten in an instant.  When you record, everyone hears that wrong note every time the recording is played, over and over.  If doing a recording that isn't a live event it is a good idea to do at least two "good" takes, all else being equal, the second take is usually better, and you have a back up if something you didn't hear during the recording is bad in one of the takes.  If you are recording live events to make a composite album, the more events you record, the better choice of takes you have to choose from to make an album.

One final point:  If you are doing a recording without an audience, (i.e. not a "live event") try to have someone like the captain or arranger talk to the musicians and ask them to:

  1. Not drop their sticks in the pan at the end of the take and:
  2. Not to play their own personal little solo at the end of the take.  You need to be able to record the decay of the last note in relative silence. 

Either of these actions sounds very unprofessional, particularly in a recording.