What are you listening for?

[Andrew B.]

Having said that, the esteemed Dave Rat (FOH for Red Hot Chili Peppers and others) uses a third method, which is to set the faders to the relative positions he thinks they should be in the mix and adjust the gains accordingly. So if the vocals should have an output level a few dB above the keys he will put the faders there and adjust the input gain so that it sounds right with the faders like that. There is an explanation on this page, somewhere:

http://www.ratsound.com/cblog/

 

Andrew

[indyld]

Some interesting thoughts so far, thanks.

 

I tend to fall into the "good sound is the absence of bad sound" camp, which more often than not comes down to understanding what the source sounds like in the first place and then not wrecking it with the myriad of choices when it comes to transducers and processing. I find the keenest of beginners or those with a little knowledge need help in distinguishing between sound design and electrical fiddling - the middle stage in the way we commonly reinforce sound.

 

Of course, when anything sounds "wrong" the first stage is to understand which component part of the sound is "bad",then why and then to know how to deal with it.

 

From the coherent (and manageable) answers from this thread so far, I wonder if anyone would care to define speech intelligibility, again in easy to understand but sonic language. I don't mean "well, if you bump n by +3 with a medium Q...." or "it's the consonants, guv", again it's "What are you listening for?"

 

BTW, I've totally nicked that Dwarves thing too, it fits in well with a session I do. Thanks for flagging that up.

[dbuckley]

There was a period, many years ago, when I was fussy about keeping faders at unity and adjusting (where necessary) by gains. My reasoning was that if I had the mix "right" I shouldn't need to adjust realtive levels of instruments much, as the band would self adjust sufficiently, and I used EQ to adjust colouring to emphasise things as needed.

 

Then I changed my prevalent mixing style from "make what I had louder" to "live production", and got a lot less fussy about adjusting gains, accepting the faders may well be above or below unity and thats OK.

 

However, I still try to start with the faders near unity. So initial level adjust is set fader to about unit, wind gain until volume is about right. As long as the meter says I have headroom then I'm happy, and the gain structure will be adequate. Probably not perfect, but adequate. Many other engineers set the gain by the meter, with the fader at zero, and then set the fader to taste.

[Andrew B.]

There was a period, many years ago, when I was fussy about keeping faders at unity and adjusting (where necessary) by gains. My reasoning was that if I had the mix "right" I shouldn't need to adjust realtive levels of instruments much, as the band would self adjust sufficiently, and I used EQ to adjust colouring to emphasise things as needed.

 

Then I changed my prevalent mixing style from "make what I had louder" to "live production", and got a lot less fussy about adjusting gains, accepting the faders may well be above or below unity and thats OK.

 

However, I still try to start with the faders near unity. So initial level adjust is set fader to about unit, wind gain until volume is about right. As long as the meter says I have headroom then I'm happy, and the gain structure will be adequate. Probably not perfect, but adequate. Many other engineers set the gain by the meter, with the fader at zero, and then set the fader to taste.

 

I think that this approach is wrong in theory as well as in (most) practice, depending on what you are mixing. Certainly if you are mixing monitors from front of house then you can't mess with the gains after the soundcheck without disturbing the monitor mix. Yes you could fiddle with the gains and then change the aux sends but that's fiddly and an unnecessary complication. Also if you have any dynamics inserted in the channel then you are going to be changing the way they act by fiddling with the gains since the threshold will be "biting" earlier or later than before. And of course in terms of gain structure you want to maximise the signal coming down the channel strip. If you need to lose volume in the house then the gains are not the place to do it and you need to consider the amps/speakers first - it is a system issue not just a mixing desk one.

 

Here's a quotation from the Allen & Heath GL3800 user manual, that I happened to be reading a few days ago:

 

Mixing with faders or gain controls? There is a technique used by some operators where they set all the faders to ‘0’ position and balance the mix using the channel gain controls. We do not advise this method as the signal to noise ratio and control resolution can be severely degraded. In addition, it is impossible to mix monitors from FOH in this way as changes to the gain settings affect the monitors too. The correct method is to use GAIN to match the source to the operating level of the channel for optimum dynamic range, and then use the FADERS to balance each source into the mix. With correct system gain structure, prime sources such as vocals would have their faders operated around ‘0’ while sources low in the mix such as backing vocals and acoustically loud brass and drums would display their true contribution with their lower fader positions. This is a much more visual and accurate way of mixing.

 

Andrew

[J Pearce]

My position on this is that if we were meant to mix on gains they'd be at the bottom of the desk on nice ergonomic slidey things....

 

It ruins any possibility of pre and post fade sends working properly, it mucks up inserts, it's harder to do, and doesn't give you optimum gain structure.

[Andrew C]

Mixing on gain pots is madness, but I can see that there might be varied opinion as to where to set them in the first place. Suppose you set all your faders to 0 and then set the gain to what you want to hear. Then any variation during the gig is on the fader. Plus or minus a little bit. If you fade a mic out, you know when it comes back in later, it comes in at near 0.

 

Different to the above comment on having a graphic representation of what is contributing how much to your mix. And as I understand it, not the best use of gain structure, but is it an option in real life?

[peteshepherd]

surely gain structure is about allowing an optimum amount of gain in to the channel (and as a consequence the insert chain) to allow it to work efficiently, generally at or near 0dB on your meters. the fader being the fine tune of the above structure that you present to the output and thus your customers.

 

mixing on gain pots ? what nonsense is this of which you speak

[Shez]

I'm sure we've been here before. And I suspect this should be split off in to a separate topic...

 

I've always been of the school of thought that initially sets the gains at a position that allows the faders to all be at or close to the 0dB mark. Once the gains are set like that, leave them alone and mix on the faders. Working like this is good practice for a number of reasons: you get the most fine control of the faders at that position; the gain structure is actually optimised like that (I'll come back to this point) and you know that throwing the faders back to their 0 marks will always be a reasonable mix. This last one is particularly useful in theatre - rather than having to remember that jack's mic needs to be at +4 and Jill's needs to be -3 every time they're put up, you can just put them to 0 and know that they're in the right place.

 

On the subject of gain structure, setting each gain so that the PFL level is around 0 makes sense on the surface but quickly breaks down. Anything that's acoustically loud on stage will need its fader to be much lower than 0 to be at a sensible level in the mix. You've just added loads of gain with the gain pot and taken it away again with the fader - not good gain structure and you get rather course control right at the bottom of the fader's travel.

 

Adding inserts does cloud the waters a little but they have adjustable thresholds for a reason.

[S&L]

setting each gain so that the PFL level is around 0 makes sense on the surface but quickly breaks down. Anything that's acoustically loud on stage will need its fader to be much lower than 0 to be at a sensible level in the mix.

 

For the more dense and inexperienced amongst us (i.e. me) could expand on the above, by way of an example?

[Bobbsy]

I know where Shez is coming from.

 

One example I can think of is when you put mics on a drum kit in a relatively small venue. Most of the drum sound is still going to be live, with the mics used just to add some "fullness" where required. If you set all the drum mics for PFL 0 then if you push the fader up to the zero point you'll have a very overpowering drum sound when the "live" and the "PA" sounds are combined.

 

That said, I normally don't mind having faders down relatively low, as long as they're not SO low as to be in that range where a small movement makes a huge jump in level.

 

There's no "hard and fast" rule but, in the theatre environments I work in, I tend to set the gain trims so that zero on the faders represents a good, nomal level for a solo line, be it a singer or a musical instrument. This means that, for example, people in the chorus are normally set well below zero on the fader but I have leeway to bring them up when necessary. As I said above, I quite like the visual cue I get when I can see the two soloist's faders at zero and the chorus at -20 or whatever.

 

Of course nowadays, this plan is modified slightly with the use of VCAs for the mix but that's another thread....

 

Bob

[paulears]

We've floated away from the main question here, which was to do with good or bad sound and how we even talk about sound without using 'pink fluffy' adjectives that mean certain things to some and wildy different things to others.

 

I've got no real answers myself, but tend to think in terms of realism not so much quality. Good hi-fi systems sound different from cheap ones - but sometimes the equipment is pretty similar, at least to look at. Why exactly do big, flown line-arrays sound different from disco ones made from cheap chipboard and pressed frame drivers? Is there some real physics in play, and if so what is happening. Using an infinite baffle as the simplest form of allowing a loudspeaker to function, I wonder if anyone has actually tried swapping drivers of all designs and price ranges? With proper enclosures then there are many individual components to the final sound. Without a 'clever' (or even simple) cabinet the changes to the sound must be the driver. People do shootouts with lighting and the results seem quite quantifiable without clever and fancy adjectives. Is sound really so subjective that it defies accuracy of comment? I suspect we just haven't developed proper language that is truly universally understood. We have colour temperature to describe bias in the lighting frequency response - the best we get in audio is 'smiley face', but never 'smiley face centred on X Hz, with the smile extending down to XHz and up to XKHz'. Sound people always talk in almost meaningless technobabble. We use the dB as if it is something fixed. Hardly ever do people mention the thing it's referenced to - and if you use dB without the reference, it's meaningless. We have faders on mixers labelled with dB, but wouldn't a simple 0-10 would actually make more sense? For years everyone, including the experts, referred to the polarity switch as the phase switch. Now we take exception to this term. Nothing changed. If it is wrong, then it was always wrong - but sound people just accepted that while technically incorrect, it was acceptable to use the term because everyone understood. Nowadays we can shift sound in time, very accurately. When this wasn't possible, nobody objected to using phase for polarity. Sound still has elements of magic. Mystery and intrigue. Lighting has become able to be quantified to a very, very fine degree. We can measure it in ways that make sense. Sound still can't do this. We can measure volume, but we argue about which scale to use - so there's fudge room. We still have a real issue with volume. We constantly (here on the forum and elsewhere)have trouble with equating volume and Watts. We explain that Power isn't the arbiter of volume, and it's driver efficiency and distribution and spectral content that makes 'volume' - but it's clear that not even the manufacturers have ever come up with a better way to describe loudness. To be fair, this is now causing the lighting people the same problems as now they have lost the ability to equate Watts to brightness.

 

Over the years, the best advice, and perhaps the most common advice is to use your ears before spending money. However, it's clear that for many purchasers, the spec drives the sale - and because the specs are missing a measurement for quality, many people are disappointed with their purchasers, despite the specs.

[indyld]

... Is sound really so subjective that it defies accuracy of comment? I suspect we just haven't developed proper language that is truly universally understood. ... Sound people always talk in almost meaningless technobabble.

 

I think this is the key question I have for sound experts. Surely in order to develop, communicate, to teach or learn about something so aural and universal, it can't defy comment using simple language? I can understand how theoretical physics could only be defined buy complex and obscure mathematics, but "good" sound (or, the lack of bad sound)is pretty much understood by every audience member and yet the professionals go straight to Hz or the highly suspect dB etc. There must be a process in between the ears and application of learned technological knowledge, surely? Some of this process has been well described in previous posts, with clear ideas about what makes up the poster's preferred mix in simple language. There must be more, however, lurking in the brains of those that work with sound reproduction every day.

 

I makes me smile when an interesting aside about mixing and the methods of setting input gain turns into the bulk of this thread. While this is no doubt up for discussion and useful information, it does seem that the world of soundies finds much comfort in gain structure and the electrickery of common reinforcement techniques, Volts and Hertz - and comment flows once again.

 

Sound is aural and exists quite happily without having to turn it into electricity, thanks very much. So, perhaps another way of looking at it would be "What is good/bad about THAT sound" when it is entirely natural and is never going to be forced down a wire. Obviously, to make changes involves a lot more physical work such as acoustic treatment, reflecting boards or whatever. But that takes transducers and log pots right out of the equation...

 

I was having a conversation with one of our specialist sound lecturers the other day about the pros and cons of introducing dynamic signal processing to learners using things like plugs-ins for DAW software such as Reaper/Pro Tools etc. The consensus was that as sound is aural, the visual nature of the plug-in interface is helpful to understand things in terms of gain reduction graphs (good for visual learners, I guess) but actually makes it harder to get people to make aural decisions. What does a ratio of 4:1 sound like? It's all very well telling someone to set a threshold to x, with y ratio, attack to z, release..... but how do you give them the tools to be able to just crank the pot until they hear "that thing" that is "good"?

 

That brings me to the advice and learning that is available when reading live sound books or online etc. It seems as if words such as boxy are offered, with a handy frequency reference and some tables that suggest a cut or boost in this area. These are sound engineering books, so they describe things from that point of view, but often don't attempt to describe sound itself. What does an early reflection sound like?

 

I know that our industry is progressing, technology improving and our understanding of how our available technology relates to acoustic theory is as good as ever. I wonder at how this happens when even the experts can't talk about the subject without resorting to Volts and Crest Factors and avoiding describing "good sound" all together.

 

To be fair, this is now causing the lighting people the same problems as now they have lost the ability to equate Watts to brightness.

 

True, and frequent questions about how many equivalent "Watts" a particular LED PAR shoves out seems like trying to claw back that known scale. Funny thing is, lighting has all kinds of mathematics that deals with brightness and reflection, both absolute and relative, and I've pretty much never used any of them in the field.

[Andrew C]

Perhaps the way forward for the books is to include a CD with a sample of "boxy" and "warm" sounds. Include any artefacts that are reproducible in the medium, mic choice maybe compare a C1000 with a Neuman.

 

Maybe a simple song that everyone knows, with simple instrumentation to make it easy. Not Happy Birthday because that is copyright, maybe a trad folk song? Call it "A Young Persons Guide to the Recording Studio". And it might be just as useful to us old farts too!

[Shez]

We use the dB as if it is something fixed. Hardly ever do people mention the thing it's referenced to - and if you use dB without the reference, it's meaningless. We have faders on mixers labelled with dB, but wouldn't a simple 0-10 would actually make more sense?

In my mind it makes perfect sense on a fader. That deals with relative levels, rather than absolutes, so there's no need for any letters to indicate a reference. One should always be mixing with one's ears anyway, so I guess you could argue that having markings on the fader that relate directly to its attenuation shouldn't be necessary. In that instance having a 0-10 might actually be more useful in terms of writing down a value to later bring the fader back to.

 

Back on topic, in this book:

http://images-eu.amazon.com/images/P/1592006914.01.MZZZZZZZ.jpg

the last chapter is concerned with the engineer as an artist, and moves away from the more technical considerations. It has a couple of pages of frequency ranges and the terms that are commonly used to describe them - boxy, clarity, bite etc. I'm sure I've seen a more comprehensive set of translations in another book but I've yet to locate it.

[Andrew B.]

Sorry for being partly responsible for dragging the thread off track - although at least I put my head over the parapet to give the first real answer to the original question! I would like to say just one last thing on the subject of gain structure (which really IS important to sound QUALITY too), which is that it is far better to maximise gain at the top of the channel (mic preamps) and cut it at the faders, than vice versa. That will certainly improve your signal to noise ratios and becomes very important when you have a lot of channels in use. This is also true on a digital desk where you need to feed a decent signal to the A/D. If you find yourself with all the faders low including the master then you know that you've got the rest of the system (amps/ speakers) wrong for the venue, in my view.

 

Andrew