Setting sound levels, theatre and live band mixing

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.