Setting sound levels, theatre and live band mixing

Being relatively new here, I was having a ramble around assorted forums when I came across a post that dwelt on the topic of sound levels in theatre. I noticed that it was locked, so I was unable to add my thoughts there, which, instead, I will do here. If this is a breach of this forum's etiquette, someone might let me know. The start of that other discussion was this summary: "In summary: 1. Your cd player will have a nominal line output, it may be fixed at 774mV as it is not usually a 600 ohm feed, but a 10Kohm feed (domestic). 2. Set the slide fader to 0dB which is the black line at the 3/4 mark (3/4 of 1 volt is 775 mV) 3. Set the main fader to 0dB which is the black line at the 3/4 mark (3/4 of 1 volt is 775 mV) 4. Adjust the channel gain which is usually a rotary knob so you get 0VU on the desk output meter. 5. Adjust you power amp input level, usually a rotary know, for the desired sound level in the theatre. You now have the ideal compromise between best signal to noise (getting rid of the hiss) and not running out of headroom (distorting the sound coming out of the speakers." I had a read of the link that the post also referred to. The basis of my remarks stems from a couple of decades of live and studio sound mixing, so I have reasonable experience and expertise in the area. My thoughts are these: 1 The principle behind the summary above and the Yamaha article is sound; i.e. optimise the gain structure from the source as much as possible. The aim of this principle is to maximise the signal to noise ratio. All elements in a signal path generate some noise, the effect is cumulative, and a maxxed-out power amp at the end of the chain will reveal all this clutter magnificiently. 2 However, there are some difficulties in the practical application of this principle, the main one being access to a power amp's level controls, if it has them (not all do). If the power amp is co-located with mixing desk, there is no problem of access. There is another problem, though. Long speaker cables from back of the hall (or thereabouts) to front of house speakers result in considerable energy and signal loss. The result is that the power amp may have to be driven harder to make up for this loss, therefore increasing the understory of noise. It is no accident that on music stage rigs, the amps are co-located with the speakers, not the desk, to minimise high-energy cable lengths. 3 Furthermore, over the last decade we have seen a huge increase in the use of active speakers, where the power amp is built into the speaker cabinet (which, unsurprisingly, is where the level controls are located). 4 In the past, power amps were bulky, heavy and noisy. They make great boat anchors. Current technology has resulted in smaller, lighter and virtually noise-free power amps (and active speakers). The importance of managing a venue's sound levels by manipulating the amp's controls is not as critical as it may have been once. 5 Within the fraternity of sound engineers with whom I'm acquainted, the common practice is to set the power amp levels to maximum, and to control overall level from the mixing desk's master control (which hovers around 0db). All audio peripherals (e.g. graphic EQs, compressors and so on) are set to unity gain (0 db), and are left alone. 6 Though this is a common practice, it is not universal. There are times when you know that you don't need the amp's full power. For example, I was mixing with my desk plugged into an in-house system that had so much power behind it that I would have been trying to mix with the bottom couple of millimetres of fader movement, which is just plain silly. So I set the power amp levels at a quarter, which was all that was needed for that particular job. There are also some rigs that are past their prime and very noisy . . . and again you have to balance between getting an adequate level without introducing too much noise. 7 I note that the Yamaha article talked about setting signal levels as high as possible to maximise signal to noise, but with the warning to avoid tape saturation or digital distortion. This notion has been in vogue for many years, but is not nearly as relevant these days. Current technology has reduced the intrinsic noise significantly, tape is rarely used, and 24 bit digital recording has increased the inherent dynamic range such that it doesn't really matter how low the signal is. The aim now is not so much to reduce noise, but to make sure that there is plenty of headroom for mixing. 8 If I were to recommend a general practice for setting levels, it would be this: a) Set power amp to max b) Set master fader to unity. c) Set a desk input channel's trim to deliver the best signal without peaking d) Bring the channel fader up to the desired level. e) Only if the channel fader can't give a reasonable working range (i.e. it's too loud when it's nearly at the bottom) would I adjust power amp levels. 9 However, each venue is different, as is each sound system, so any practice has to be tailored to the combination. Finally, one of my greatest thrills is to set up a rig, wind up the amps, return to the desk and listen to the absolute quiet. Then I put on a CD or something . . . and there it is, clean , crisp, undistorted and immensely satisfying.

There is only one way to set gain structure and it is not your way, you may ,on occasions get away with it IF the amplifiers input level happens to match the output level of the mixer or previous device in the system, but luck is not a part of setting up, I have endless problems sorting out noisy systems set up your way, the essential fact is that each element has to be matched to the next, the input of the mixer has to be trimmed to the mic or input source or it will overdrive and distort or underdrive and increase the signal to noise ratio.The same applies to each device in the system including the amplifier. The same rules apply absolutely to powered speakers. Now in Rock and Roll a bit of noise is no problem but in a theatre situation it is crucial to have the lowest noise possible and this can only be done with correct gain structure. Now anyone reading this, do not believe me, or Yamaha, or Rane, or any of the dozens of authors on gain structure, just try it for yourself, in only a very few situations will the correct setting on your amps be full, because most amplifiers leave some extra gain for weaker output devices. I cannot emphasize too much how wrong Mikes advice can be.

I stand partially corrected, but totally unrepentent. I blinked, momentarily taken aback, at your unequivocal and uncompromising statements. Had I been mistaken? It's quite possible; I often am. So I did further investigation, and I do note that a number of audio authorities (e.g. Yamaha and Rane) recommend the method you so cogently outlined, i.e. aiming for unity gain throughout the signal path, with final master volume level being set at the power amp, and I note also there are some advantages in doing so. Armed with this information I went on an internet voyage of investigation, contacting my engineering acquaintances (mainly in Australia and the US) to get their perspective. They too were aware of this idea of final level setting using the power amps. But, interestingly, none have adopted it. I have yet to find a single engineer that does not, as a matter of course, wind the power amps fully up. It's possible they could all be in error, and it is also possible that different practitioners will emerge. I note, though, that both practices emphasize the importance of gain staging, and that the only difference is where the final level is set. In essence there is little difference except for this power amp stage. But I note also that the level control on an amp is just an attenuator; it is not a gain-setting control. In other words, if it was a car, turning the knob clockwise is more like taking the handbrake off rather than pressing the accelerator. It it is, in effect, turning the amp less 'down'. Amps are designed to run unattenuated, and will do so happily and without complaint . . . after all, that's what the power amp in your home stereo does. The volume control adjusts the pre-amp, not the power amp. I note also the increasing use of digital signal processors and speaker management systems that sit between desk and amp, and which do a fine job of keeping the signal under control. I use two twin digital EQs and signal processors. I agree that when mixing rock bands, managing a residual noise level may not be as critical, because of the nature of the music. However, most of my work is not mixing rock bands. Most of it is mixing acoustic acts (for example, folk festivals), and when some doe-eyed singer is singing softly and soulfully about a long lost love, the last thing an audience wants to hear is the cicada-like buzz of a 50 cycle hum affronting their ear drums. I too have had my share of noisy systems, and I agree that getting the source right and following this through the system is how you deal with them. I also agree that luck should not be part of the process. But there is no luck involved in matching a mixer's output to a power amp's input . . . they are designed to match (unless they are old valve-driven relics from a darker and less hospitable age). So, in the end, I acknowledge the value of following the practice recommended by Yamaha et al, and I'm not going to argue with anyone that does so. I might even try it myself one day. I might even be pleasantly surprised. However, I have never experienced poor signal quality or noise with my own rig. I've never blown an amp or a speaker, and even when driven hard, I've enjoyed clear, undistorted sound. There is no imperative for me to change my ways. This bit is not technical, but explains my approach to mixing. I think it is a bit like flower arranging. Each flower presented for arrangement has its individual beauty. The flower arranger's task is to create an integrated, visually pleasing whole from the sum of these individual blooms. Mixing is the same for me; creating an aurally pleasing soundscape for the audience in which all the individual elements of musicianship come together in a cohesive whole that is greater than the sum of the parts. It's problematic trying to do this with a dodgy rig and a poor understanding of how its machinery works. There are many wise pieces of advice about setting up systems. For example, running all power off the one outlet to avoid earth loops, keeping signal and power cables apart and so on. These are all beneficial, and no harm comes from following them. Most of them, though, are unnecessary. I have found that in most venues these days the power is very clean, and you can plug the amps in from an outlet near the stage, and the mixer from somewhere else in the venue, and there is no loop-generated hum. On the occasion when I don't, I happily run a power cable to the desk alongside the multicore, again with no problem. My biggest noise issues stem from light dimmers (specially flourescent light dimmers), but, by using balanced cable througout the system, even thse are minor irritations.

I really shouldn't complain about your technique, it actually makes me money, my clients will contact me because "the hum has come back that I fixed after a band came through".I simply go out to the job and reset the gains fix the noise and take the money. Now you have problems with dimmer noise, surprise, surprise, because your gain structure is wrong, your signal to noise ratio is more of a noise to signal ratio and what is the common source of noise, dimmers! "David, your dimmers are causing noise in my sound, it was fine before you turned on the lights" Well there's nothing practical to reduce the dimmer noise but it's dead easy to stop the PA from picking it up. Yes it is a very common practice to turn to full the amplifiers and yes it is wrong, let us take a simple example, many mixers give you a choice of output levels, according to your ideas it doesn't matter which output level you use as the amps are always set at full, it's totally illogical, saying that processors will overcome bad settings is actually true, but your system cannot be as quiet with the amp gains at full as they are at their proper setting, the basic laws of physics as applied to the signal to noise ratio are inviolable, and if everyone in the world used your method, which they don't, it is still wrong and it is so incredibly easy to demonstrate it is wrong, I cannot understand why people continue the practice.

I think you must have missed what I said. The point is I don't have hum. I hate system noises: hums, hiss and crackles, and my rig is dead quiet. The only time I've had problems with dimmers was when I once had unbalanced cables going to a foldback amp. This was at a venue with a very basic set up. Swapped with balanced cables . . . bye bye dimmer noise. I can understand why people continue the practice . . . because for the most part, it works just fine. Now I am pleased that you are making money out of improper techniques, and I'm convinced there are many patchwork systems out there, and they need the help that you can provide. And nor am I grumbling about the 'correct' method. It's fine. Whoever wants to use it can, and they will benefit from improved signal management. But you won't make money out of people who use my setups.

No I said that under some circumstances when the output of the audio chain happens to match the input of the amplifier it is fine, BUT that occurs only sometimes, maybe 20% of the time, your method will therefore work roughly 20% of the time where the correct method will work 100% of the time, advising people to always set their amps to full is plain wrong, 80% of the time.But as I said to any confused reader, try both methods and find out for yourself.Your system may work perfectly well, but it is luck.

We are not using personal abuse, like some of the bad threads in recent times and having a heated discussion is good in that it raises one of the most common faults around. My main concern is that some student will read Mikes piece then go to school and tell her teacher that she knows how to set up the drama sound system, when she winds up the amp and has to turn down the mixer to get the level down and the hum is rumbling away and she looks stupid, then we may have lost a keen kid to the industry.Lots of students and amateurs read these pieces, so bad information becomes legitimized by being on a website and spreads like a virus.As for a demonstration, sure, anywhere anytime.[subject to schedule]

At Melville theatre, I run the power amps at around 30%. I read the Yamaha way of setting gain, and it made sense to me. Also, if I run them at full belt, I have very little contol on my sliders. In effect, it suits my venue to run the power amps at a low level. It is a converted community hall. I don't have any issues with hiss or hum. www.meltheco.org.au

Whilst I agree with you David on your concerns, I differ in one respect: if a student turns up and insists the teacher listen to them, then perhaps either the teacher needs their brains checked, or the teacher should learn proper use of sound equipment. (Here, my concern is with safety. If a student comes along and tells you not to use a safety harness at heights, is the teacher going to listen? I hope not!) Honestly, how many students will actually do that? More to the point, how many teachers would listen to a student over their own knowledge/lack of? Puppets and patterns at Puppets in Melbourne

I tell the students I teach that they must play with their sound and lighting gear as much as possible, most of them soon know more than their teachers, who seldom have much technical education, like any skill it can only be learnt by trial, and sometimes error,so I always encourage them. The comparison with working at heights is completely invalid, it's very hard to kill yourself on a sound desk. This is no reflection on teachers, the drama teachers generally receive little or no relevant training and they tend towards creative rather than technical mindset.

I have already volunteered me and any gear you want to borrow for the event, subject to schedule,[mainly me, not the gear]

the definitive paper on gain structure is by Chuck McGregor www.gain.pe.kr/spboard/board.cgi?id=audio&action=download&gul=14 anyone with a serious large system would benefit from studying this, the "Yamaha/Rane" method is possibly fair enough for most purposes.

I attended the ANZAC dawn service in Kings Park this morning. There were thousands of people there, and all was silent, except for the very loud hiss from the P.A. www.meltheco.org.au

Here is a quote from the above piece, you can follow the reasoning from the above if you are technically inclined, but it does prove my contention without any doubt, "BACKGROUND YOU NEED TO KNOW Realistically, audio signals at or near the noise floor of a system are not useful because the signal will not be significantly louder than the noise. Therefore, some minimum usable level must be assumed below which the electronic noise is considered objectionable. A signal to noise ratio of 20 dB is considered minimally acceptable for good intelligibility. For a high quality system 30 dB would be a better figure to use. Using this value, the range from this minimum signal level (30 dB above the noise floor) to the clipping level is the usable signal range window for the system (also called the dynamic range in my way of thinking). However, for purposes of this paper, the maximum output to noise floor is used as the dynamic range. Every audio system with more than one electronic component has a "system gain structure". Gain structuring for a system occurs in the signal processing chain between the mixer or another signal source and the power amplifiers. One usual scenario is to set all the signal processors to unity gain and turn the amplifier inputs to maximum. Unfortunately as you will see, given the different maximum outputs and noise levels of typical signal processors, this method will may not come close to the best gain structure." The article is an extremely good read, not understood by a large majority of sound operators.

Do sound techs generally carry around a CRO and actually do this? Should they? I'm trying to follow what is being said, but I'm slowly drowning. I am an electrician, but know not too much about sound. Our theatre has pro type Denon CD players, then a mixing desk (Yamaha) and power amp. The CD has no attenuation controls. Lets assume I have a pink noise CD (which I don't). I turn the power amps to minimum. Set the output of the mixer to 0dB. Set the output slider of the channel (or channels if stereo) to 0dB. Set the input gain control of the channel(s) to 0dB. Adjust the power amp attenuator(s) to give....what? Maximum volume required? 0dB on the output meters? Where does one get hold of a pink noise CD? How do you set the gain on an instrument input? Mic input? My electrician brain tells me that the losses in the speaker cables would be negligible over a 30 - 40 metre run, using even fairly basic speaker cable. Ohms Law must still apply. The current flow can't be too big. (What would be normal?) If current is low, then volt drop is minimal. If you use Monster cable, or similar, is it a waste of money? www.meltheco.org.au

The full complex gain structure is only done once on a system, then left unless a component changes,so one doesn't carry a cro around. A loud cd will normally suffice for a source. I turn the power amps to minimum. Set the output of the mixer to 0dB. Set the output slider of the channel (or channels if stereo) to 0dB. Then adjust the trim so that your output meter is reading 0, on average. Set your amps to the maximum level you may want. One huge misconception is that Full on the amp control is Full output, it is not, your amp can be on 1 but still deliver full output or it can 10 but giving a very low output, if you are feeding in too high a signal you will have your full output at maybe 3 and as you turn the amp up further you simply move it into clipping and distortion and blowing up speaker time. Having set up your mixer output and you want a mic, turn the trim down on the channel, bring the fader up to 0, then bring up the trim to the level you want to work at, you will then know that every channels normal setting is zero, if 6 people walk on stage you just bring up those 6 channels to zero and you have sound, you can then do minor adjustments around the zero mark but you have a good start, and you're optimizing your noise. The speaker cable thing is a red herring, it mainly adds some resistance which has no major detrimental effects except in extreme cases. One might think that I am academically inclined, but I only learned this stuff to get me out of the manure in various sticky situations, my original tech training was 40 years ago and is long forgotten. I don't want to frighten youngsters like Mike off, but his advice could work in some situations but be awful in many others.

Setting amp levels at lower than maximum is absolutely the right thing to do if you would otherwise blow people's ears out or find you have to work with only a few millimetres of fader travel at the bottom of their paths.

There's nothing more annoying that being bedevilled by noise when sensitivity and silence is needed. Sometimes hiss can come from a noisy power amp, but it could equally as likely come from the desk. Budget desks have pre-amps of variable quality, and often, turning up the input gain beyond 2 oclock introduces unacceptable levels of hiss (specially if there is nothing plugged into the channel). It's an easy mistake to make . . . on an unused channel the gain may still be set high from a previous mix, and if that channel is not muted, it sends its inherent noise through the system. It's an easy mistake to make, but just as easy to fix. So I would have expected the engineer to have been able to stop the noise. However, if the hiss was from an inherently noisy power amp, it should have been turned down. But the nature of the venue and its acoustics may have been such that it couldn't be turned down because it was simply not powerful enough in the first place. This is a mismatch of the rig to the venue. In any case, if it was the amp, it is time to get a proper one.

here is an interesting graph on cable loss: http://www.mbaudiosystems.com/index.php?page_id=34&mainmenu_id=27&title=Cable%20Loss&image_grp_id=2 My electrician's brain likewise agrees with you, but audio signals don't behave exactly like AC power

Whilst I understand that differing frequencies will be attenuated unevenly down the same piece of cable, I have always thought that loudspeaker cable salesman were peddling snake oil. Our passive speakers are about 25 metres from the power amp, and connected with figure 8 flex. Works just fine at this distance. I think that the exhorbitent expense of specialist speaker cables would make little or no difference, and certainly not be value for money. Over a long cable run, 6mm^2 electrical flex would be a very cost effective and economical speaker cable. www.meltheco.org.au

I reckon there's a lot of snake oil in specialist audio cables, and I have not had much time for the exhorbitant claims nor the exhorbitant price. I use standard figure 8 cable, or ordinary power cable, just leaving one core unattached. What you are saying is suggesting another train of thought, which I might expand on later once it settles into something coherent, but is encapsulated roughly in the following. There will be loss of signal down long cable lengths. There will be less signal loss down shorter lengths. Does it matter? Probably not. Most likely the system has sufficent capacity to cope easily with any such loss. Despite that, if I am powering up passive speakers, I will still keep the amps as close to the speakers as I can. It is (a) a force of habit, and (b) a technically wise thing to do.

Obviously you haven't been to the dawn service recently Graeme. It is a full on Pro setup with multiple 20 foot high video screens, and a huge sound/video mixing desk with racks upon racks of amps and the like. www.meltheco.org.au

The graph that Mike has linked to shows a loss of 3.5db spl around 200 hz for 100 metres. Litz wire and monster speaker wire etc make use of many smaller size conductors to take advantage of the surface effect where a high frequency signal travels on the outer surface of the conductor and does not use the inner part of the conductor. This effect is very noticeable in very high frequency systems such as fm radio transmitters and in microwave frequencies which is gigahertz, they use hollow conductive waveguides, so a 200 hertz seems a bit unplausable. I would suggest that the loss shown in this graph is insignificent in local theatre for several reasons. 1. When setting up your system, all equalisation should be set to flat, after taking a note of what it was set to as the previous tech/volunterr may have git it tuned correctly. After you set up the system levels, use the main eq or in the case of powered speakers, the speaker eq, to compensate for these fequency dips. The graph shows spl or sound pressure level. You need to convert this to the db you see on your desk meters, it could be dbm which is referred to 600 ohms, or dbr or db reletaive which could be 10,000 ohms for semi pro audio. Also what weighting and response do they use on their sound level meter. The frequency axis of the graph only goes to 1000 hz, This has probably been chosen as if it went to 20khz the frequency dip would appear to be insignificant. A one third ocatave sweep of your theatre venue will probably turn up more peaks and dips due to the acoustics of your venue. You do not need expensive equipment in community theatres to try these techniques as your ears will usually back up your budget instruments response. You can find pink noise files here http://www.dogstar.dantimax.dk/testwavs/ or use Google to find cd's. Unfortunately some young professionals put the minimum effort into some jobs so clients do not get value for money. Some firms are conscious of profit margins or have to work within a clients budget so the outcome is a compromise. In our area of community theatre which is "what this here web site is about", we can strive for excellense which is why our patrons usually end up getting better value for money.

I agree with Greame, about the PA used at the Dawn Services, I have attended. {After holding vigil, overnight, on the monument, for more years, than I care to remember! In the past it has just been as he descibes - normally set up by the R.S.L [at the last minute, if they remember to plug it in?]. From what Jeff mentioned - it has become more Hitech!

The discussion above about signal loss in speaker cables led me down a train of thought, which I will now dwell on a bit. The practice of locating power amps close to speakers is pervasive, and is done to minimise signal loss associated with long cable runs. However, in many situations, the actual amount of loss is minor, if not negligible, and it is not really necessary; the system is more than capable of coping with any small amount of loss. Indeed, many powered mixers, by their nature, require long speaker cable lengths. I want to take slight, but relevant diversion here, and talk about PCs. I moved wholly into digital recording in the mid-nineties. At the time I was running a moderately powerful (for the time) PC; 500 mhz CPU, 256 meg of RAM and 20 gig hard drive. With this I could record and play eight tracks simultaneously, but there was a fragility to the whole process. To be confident of recording succes you needed to optimise your computer for audio work. This include disabling all system polling activities (e.g. screen savers, auotmatic power management systems, virus and spyware checkers), enabling DMA access on hard drives, removing nearly all the start-up programs, and frequent disk defragging. The practice of using two computers was common; one for audio, and one for other stuff (e.g. internet), and never the twain shall meet. These days, I am still using a moderately powerful PC, but now I have dual-core 3.2 mHz CPU, 2 gig of RAM and just under two terabytes of disk storage. I can mix any number of tracks with impunity. The most I've mixed at one time is so far 35 tracks with associated plug-ins, with no sign of system stress. What about optimising this PC? Well, I don't bother. So I have screensaver, virus checkers etc., all going, and I just use the one PC for everything. I could optimise it, and that would result in a performance increase. However, the loss in performance has no impact on audio work, and like signal loss from speaker leads, is safely ignored. (I will concede, though, that optimisation could help with my video work . . . but what's really needed is not optimisation, but far more computing power, optimisation would yield only marginal benefits). Optimising is a risk-management strategy. Risk has two dimensions; the likelihood of an undesirable event occurring, and the severity of impact should it occur. In the case of the first PC, the likelihood of undesirable events was high: dropped samples, clicks and stutters; loss of synchronisation and so on. The impact was high (specialy if the artist just recorded the take of the century and it didn't record properly.). Optimising the PC was very necessary to manage those risks. With the current PC, the impact of these events is still high, but their likelihhod is extremely low. (There is greater risk of operator (i.e. me) error). I could optimise it, which would do no harm, but it would be of little benefit either. In any audio handbook, there are many pieces of advice that are, in fact, risk management strategies: A - keep speaker cables short ( to prevent signal loss), B - keep audio and power cables separate, make them cross at right angle (to prevent electrical interference), C - run system power off the one outlet (to prevent earth loop hums), D - power the system up in order of source to speaker, and power down in the opposite direction (to prevent loud and possibly damaging thumps in the speakers), E - use 'over and under' cable coiling to prevent internal twisting and damage to the cores . . . and so on. I feel now that I have to agree with David in that I would not make a good role model for budding audio students, because I ignore most of these pieces of sage advice. I often ignore B by running the desk power alongside the multicore, and if I do heed it, I ignore C by getting desk power from the nearest available outlet. Every risk management strategy can be examined for its relevance, and I have found that by using balanced cable throughout the system, electrical interference does not happen, and that these days most venues provide stable, clean power, and earth loop hums are rare. Therefore, B and C are strategies devised to deal with a risk that had a high likelihood and high impact, but whose likelihood has diminished markedly over the years. I heed A because there's no point in doing anything else; the multicore provides a mixed signal back to the stage, so if I'm using amps with passive speakers, that's where they'll go. I'll also heed D if I'm using passive speakers, but not if I'm using active speakers. My main rig consists solely of active speakers, and I have developed a habit of not turning them off (or down) when I finish. I just pull out the power leads (indeed, you may well shudder!). This means that when I connect them up at the next job, they power straight up. This is fine, because their internal electronics management prevents that familiar power-up thump. The same applies to the desk and rack. The last thing I do is to patch the multicore into the rack outputs, and I get a very satisfying 'click' through the speakers each time a connection is made, and this tells me that the signal lines have been connected correctly and are working, which I find highly reassuring. I heed E because cables will always deteriorate if mishandled, the likelihood remains high and the impact is high. When I look at Chuck McGregor's article, I note that at its heart it aims to increase the effective dynamic range of a system, thereby maximising headroom and minimising noise. The procedure does not claim to eliminate noise, because even if adopted, some components may just be too inherently noisy. Nor does it claim to deliver the right sound levels, because the amp may just be underpowered for the application. However, like my PC optimisation example earlier, it is the best way of getting the most out of a system, specially if it's one where you need every db you can muster. I have never had cause to be concerned about headroom or noise. So long as I get the source gain sorted (which is part of the procedure), I have not found myself scratching my head pondering about where the volume has gone and why there is so much noise. Maybe it's luck. Maybe it's just having a good set of contemporary, complementary components with ample reserves of power. If I regard it as a risk management strategy, then I concude that, for my rig at least, the risks it addresses are minor, and it's like me not worrying about optimising my current PC. So, kids . . . if you are reading this, I think you are in good hands heeding David's advice as a proxy for Chuck McGregor, and you should ignore me. No harm will befall you if you do this, and if you are working with sub-optimal components, it's the best way of getting the most out of them.

QUOTE from Mike "When I look at Chuck McGregor's article, I note that at its heart it aims to increase the effective dynamic range of a system, thereby maximising headroom and minimising noise. The procedure does not claim to eliminate noise," Of course he doesn't claim to eliminate noise, that is impossible. Proper gain structure does reduce noise to the lowest possible value, in practice that is the difference between painful and acceptable. With the best gear in the world you get value out of gain structure, with the economy gear commonly used in community venues it is essential.