Quick EQing question

[XZit]

Hello everyone. I have some basic simple questions regarding basic EQ 'protocol'.
I've researched tons on the topic, yet have come up with various answers sort of all over the board, and not really a nice clear-cut answer.
So basically as I'm about to purchase some studio monitors I've been wondering about proper EQing. Not referring to EQing in the daw, but more of actual physical EQing on my input mixer. Mind you I would be doing compression via software/via the daw so I don't have a physical/analog compressor. If it makes a difference my mixer is firewire thus my interface.


*From what I understand the best strategy is to create the track with the gain down, the treble, mid, and lo end all set at medium/default. After a track is made and mixed downed then bring the gain up slightly until I hit that "sweet spot"-then test the track using various listening devices (headphones, other monitors, in the car, cd player, iPhone etc.). Any input regarding this?
*In relation to this --on my physical mixer there is the individual input/tracks volume, main mix volume, and control room/headphone volume. Should the main mix volume be on central for its default-and then bring up the input/track levels volume accordingly? Mind you there are separate outputs "main mix" and "ctrl room/hdphs". I'm assuming the studio monitors are to be connected to "main mix" output and "ctrl room/hdphs" output can be for headphones/other studio monitors for sound comparison?

*Last quick question: Is volume separate then EQ? I know sounds silly , but for example when I turn up the volume obviously it will increase overall track volume, yet I can't help but wonder if it seems the lo end/bass is being boosted due to volume increase. Perhaps this is due to my sub-par speaker setup I currently have.
So it shouldn't matter what volume I mix the track at, as long as I can hear the full spectrum of sound from the monitors?



Thanks everyone! Any response greatly appreciated. All of you guys are quite knowledgable

[madtheory]

There's no clear cut answer because learning about this is an iterative process.

1. How bassy/ trebly it sounds also depends on how loud it is, because in our ears loudness and frequency are interdependent. Look up "equal loudness contour". You've used the terms "volume" and "eq" which are often used interchangeably with "loudness" and "tone" or "timbre". Using the terms in this way is incorrect in sound engineering. There are important differences, physically.

2. No idea where you got your mix strategy from, but it's a chinese whisper- a mixed up bunch of concepts that have been watered down, twisted and turned into nonsense.

3. For recording, you set the gain so it's high enough to be above the hiss, but not so high that it will clip. That's assuming you want to capture the sound accurately, without changing it or distorting it. If it's a "live" source, where you're capturing a performance, you should leave some headroom below 0dBFS (the clipping point) standard is -20dBFS on your side of the pond. This is to allow for the performer getting excited (which is a good thing).

4. For listening, or in this case monitoring the sound, there are ways to calibrate the speakers to a standard "loudness" by measuring the sound pressure level. For you, put on some dynamic music you are familiar with, and set the monitor level. Mark that on your dial and use it as your reference.

As you will find when you look up equal loudness contour, the level you monitor at will affect the balance and tone. Then there's the whole problem of dynamic range and how tracks are made to sound "loud" without going over 0dBFS.

5. For eq, adjust it til you like the sound. But if you're programming your synths well, you won't need much. Just an occasional touch of top end or bass. Best done in the mix, not while recording through your desk. Adding bass or treble will increase the level ("gain" as you called it) because you're adding more energy.

I'll stop there because I've opened a can of worms, but that info is enough to get you started. Soon, you will have even more questions

[ninja6485]

I don't want to go against mad theory's well thought out strategy to not bombard you with too much information first, but here is one thing that can help you optimize your source/mixer/ monitoring setup, especially if you ever get to the point where you want to start putting some gear in between things at various stages.

Generate, download, or record a simple test tone: I use a sustained digital sine wave that plays at -12db.

What you do then, after turning the volume controls down on your monitors, is play the sine wave through a channel on your mixer.

Since you know the sine wave is -12db, with the volume fader set to zero (as in not boosting, not cutting), adjust the gain control so that the signal level on the mixer reads -12db.

Some mixers have a separate meter for track levels, and one specifically for the master levels. Either way, you want the test tone to read -12db through the whole signal path. The master fader should also be set so that it is not not boosting or cutting the signal, and all of the EQ controls (the trebble bass, mids, etc) should be not boosting, not cutting.

Check that the panning controls are set to center, that the mic/line level settings are appropriate to the what you are playing (Synths are generallly line level...). In other words, you want the sound to pass through the mixer unaltered at every stage, so that when you put a single test tone of -12db in, you get -12db out.

So lets say you hooked up your test tone, and everything is reading -12db. Great! Now, slowly turn up the volume on your monitors to a comfortable level. It will probably help if you set it so that -12db might even sound a little loud.

For each new piece of gear that you add in between the source of your sound going into the mixer, and the sound coming out of the outputs, make sure, when the gear is set so that when the signal is not being altered, the -12db signal reads -12db at every stage, and is outputting at -12db.

Believe it or not, calibrating your signal path in this way will make a huge difference! This is is your blank slate, and you your monitors are set so that when the mix is averaging -12db, you hear it as being comfortable to a little bit loud on the monitors. DONT adjust the level on your monitors to compensate for how loud your mix sounds! Fix the problem in the mix.
--
So lets say you've calibrated your settings, and you want to begin mixing. Right now we are just concerned with getting two sound sources to play together at the appropriate level. With your calibrated setup, you know that when you hook one up to the mixer, you can see that if it reads -12db, that it will come out comfortable to a little loud. If you take two signals that are both at -12db and mix them together, they will come out of your speakers sounding really loud!

What you can do, is hook up the first, and do what you did with the test tone - have it go through at -12 at every stage. Now, keeping the gain control (usually at the top) where it is, adjust the individual fader for that channel so that it is -inf, or all the way down (attenuating the signal so you can't hear it). Now, do this with sound source number 2.

At this point, both signals are going in at -12db, but nothing is coming out because your faders are set all the way down. THIS is where you bring up the two faders, and adjust the volume levels of the two sound sources so that they play together the way you are intending to combine them.

Again, right now we're only mixing two sound sources! So now when you bring the faders up, and you want to get a good mix, you know that if the two signals combine are coming out of your monitors at -12db (like the test tone), and the master outputs is reading around -12db, you are in pretty good shape. No matter how many tracks you mix together, you want your overall mix to average around this setting. If your signal path is calibrated so that -12db sounds comfortable to a little bit loud on your monitors, you will be able to tell much more accurately when your mix is too loud, or two soft, since you know when you start to combine the sounds that when they all come together and you're getting above -12db, it will start to sound really loud!


There are other signal levels you could use, some people prefer -14, or even -20 - I probably wouldn't go above -12, obviously with more dynamic sound sources your signal is going to peak over that, and dip below that. These things are more advanced for now.

[madtheory]

OK ninja, you've let out some of the worms, let's let out some more!

Using a sine wave as described is all perfectly fine- except when it comes to setting the "loudness" of your monitors. The loudness depends on the chosen frequency of the sine wave, as well as the level you've set. Again, the reason for this will become clear when you look up "Equal Loudness Contour" (the wiki article is excellent). So for monitor calibration, it's better to use pink noise.

An added complication is your room. It resonates at some frequencies and not others, depending on the dimensions and materials in the walls. So a frequency that sounds loud in your room might sound quiet in another, differently sized, room. This is the main reason why people check the mix on different speaker systems (car, club etc.). The solution to this problem is acoustic treatment.

Level setting for recording live sources is different to level setting for mixing what is recorded. A further complication is that "-12dB" (that's a capital B for Alexander Graham Bell, who was a really great guy so respect him) is meaningless for loudness. There are many different flavours of dB (go on, ask why, or Google it!). In this case it's dBu which is a measure of amplitude in the mixer. It's used to help the user set levels that are above the noise floor, but below clipping. But in your DAW the levels are measured in dBFS. In there, your final mix should peak very close to 0dBFS. No headroom, because you're mixing sources that are coming off the hard drive or MIDI so are completely controllable.

[madtheory]

A couple of reasons why Mr. Bell was a great guy:

In 1880, Bell won the Volta Prize from France for his invention of the telephone, and utilized the winnings to set up the Volta Bureau, a library holding information on deafness. This later became the amazing Bell labs (they came up with stuf like the transistor, the laser, Information Theory which makes the internet and digital audio function). Ten years later, in 1890, Bell set up the American Association to Promote the Teaching of Speech to the Deaf, with the objective of promoting oral communication (which later morphed into the Alexander Graham Bell Association for the Deaf and Hard of Hearing).

So, respect. dB.

[Stab Frenzy]

But in your DAW the levels are measured in dBFS. In there, your final mix should peak very close to 0dBFS. No headroom, because you're mixing sources that are coming off the hard drive or MIDI so are completely controllable.

That's a good way to get your mixes sent back by any mastering engineer who ends up working on them. Premaster mixes shouldn't have any peaks over -3dBFS as a rule.

Agreed with you on the unsuitability of using a sine tone for setting your monitoring levels, pink noise is the tool for that job. 1kHz sine tone is what you'd use to calibrate your meters.

Regarding the whole 'how to EQ/how to mix' question that this thread is about; it's too much to go into in one thread. It's a science and it's an artform and it takes years to master. I guess start off with the goal not to clip anything and then go from there.

[ninja6485]

MT & Stab, you are both right of course! 1kHz sine calibrates the meters, pink noise calibrates the monitors. Some great history on AGB here also!

My nomenclature is indeed wrong, and I am talking about dBu in a hardware mixer. In both a DAW and on a hardware mixer, I personally try to keep the average around -12 dBu or dBFS, and the peaks not more than -3 dBu or dBFS, or sometimes even -6. The pre-master (Is that the right term?) is not where I handle the loudness of the song anyway, so the most important things are that the tracks sound they way I want them to sound in relation to one another, and that I'm doing exactly what MT prescribed: keeping the mix well above the noise floor, but safely away from clipping.

[madtheory]

That's a good way to get your mixes sent back by any mastering engineer who ends up working on them. Premaster mixes shouldn't have any peaks over -3dBFS as a rule.

That's not necessary. If the engineer sends it back then he doesn't know what he's doing. There's a lot of that about. If there's a problem with inter sample peaks, then it means you've over cooked the limiter, so you've basically clipped it. In actual fact the thing you don't do is clip it. Leaving 3dB is not necessarily going to prevent that. If you've clipped it, then a remix is a reasonable thing for the mastering engineer to ask for. So don't use a limiter. Leave that to mastering. If the mastering uses an analogue chain, and the engineer doesn't know how to set it up for headroom, that's not your fault. I would use a different engineer, one who knows what he's doing.

It's been a long time since I sent something for mastering (I do my own now), but we always mixed to 16 bit DAT at close to 0dBFS because that was doable with a Yamaha 02R. Never had a DAT returned.

[XZit]

Wow! Amazing replies I must say. I'm sort of overwhelmed! lol. A lot of great info to comb through and apply.
Regarding the gain--I dunno where I heard/read about the gain being on low setting then being turned up to find a "sweet spot", but recently I've read several sources that indicate its better to say start with the gain knob at 12'o clock and then bring it up or down appropriately.
I found a nice link with basics http://www.alesis.com/tipsdec08 that helps me to review some fundamental concepts.
Obviously EQing goes much deeper then the concepts discussed so far--i.e. room sound treatment.
Regarding the EQ knobs on my mixer---should I ideally try and keep the lo, mid, and treble in neutral positions and do most eq adjustment in my DAW? I use Reason if anyones curious, I love the interface and options, and it has a great compressor, equalizers, maximizers etc. Latest version has a spectral analyzer as well. So with all the daws options us doing eqing on the mixer itself very necessary?

I don't want to get to far off topic--yet regarding very basic, core sound treatment for a room; I read somewhere that a good first step would be bass traps in r/l corners facing you/corners behind speakers, then treat behind the speakers, then behind the user, side walls, ceiling, diffusers, then expanding with foam distribution/diffusers etc. Sound treatment seems like quite difficult large topic as well, with lots of personal taste, choices, and product options, but once again don't want to get to far off topic

[madtheory]

Sound treatment, or acoustics, is a science. It's precise, there are measurements. Opinions are to be ignored if they're not backed up by evidence.

Quick and dirty room treatment:
1. pick a room in the house that is not a cube.
2. Stack two rolls of rockwool or fibregass insulation in each corner.
3. Make rockwool panel absorbers. Put them at the mirror points, and one overhead (that's three panels).
4. Position your speakers using the 38% rule. equilateral triangle with where your head will be most of the time. Read the manufacturer's instructions to set the roll off filters for proximity effect.

Done! Now you're 60% -80% better off.

http://ethanwiner.com/acoustics.html

[XZit]

Sound treatment, or acoustics, is a science. It's precise, there are measurements. Opinions are to be ignored if they're not backed up by evidence.

Quick and dirty room treatment:
1. pick a room in the house that is not a cube.
2. Stack two rolls of rockwool or fibregass insulation in each corner.
3. Make rockwool panel absorbers. Put them at the mirror points, and one overhead (that's three panels).
4. Position your speakers using the 38% rule. equilateral triangle with where your head will be most of the time. Read the manufacturer's instructions to set the roll off filters for proximity effect.

Done! Now you're 60% -80% better off.

http://ethanwiner.com/acoustics.html

I've seen a few videos on making diy rockwool/fiberglass panels. Do you have any personal recommendations for videos to watch?
I think the diy rockwool/fiberglass panels is a great idea, and very cost saving. Regarding studio foam, I have read that there are various types of foam that are more effective then others (i.e. using a 'eggshell' mattress will obviously not provide good results vs. some Auralex panels). I do see some pretty good deals for acoustic foam on craigslist occasionally. I guess I would have to weigh the overall costs between diy panels, and a good deal on acoustic foam. For example with the diy panels I need materials (wood framing, rockwool, cutting saw, fabric) and then theres the time aspect.

Regarding my room; it's not a total cube, has a slight dropped ceiling, isolated from rest of house, about 12x15ft, no windows, and dedicated solely to becoming a nice music studio.

EDIT: I did find this great link on building your own acoustic panels http://acousticsfreq.com/blog/?p=62

[madtheory]

The Ethan Winer site above has plans for panels. I don't dig tutorial videos, personally.

[Stab Frenzy]

That's a good way to get your mixes sent back by any mastering engineer who ends up working on them. Premaster mixes shouldn't have any peaks over -3dBFS as a rule.

That's not necessary. If the engineer sends it back then he doesn't know what he's doing. There's a lot of that about. If there's a problem with inter sample peaks, then it means you've over cooked the limiter, so you've basically clipped it. In actual fact the thing you don't do is clip it. Leaving 3dB is not necessarily going to prevent that. If you've clipped it, then a remix is a reasonable thing for the mastering engineer to ask for. So don't use a limiter. Leave that to mastering. If the mastering uses an analogue chain, and the engineer doesn't know how to set it up for headroom, that's not your fault. I would use a different engineer, one who knows what he's doing.

It's been a long time since I sent something for mastering (I do my own now), but we always mixed to 16 bit DAT at close to 0dBFS because that was doable with a Yamaha 02R. Never had a DAT returned.

Welcome to the future! We record at 24bit now, so you can leave headroom in your mixes and mastering engineers expect 24bit premasters that don't peak over -3dBFS.

Regarding the gain--I dunno where I heard/read about the gain being on low setting then being turned up to find a "sweet spot", but recently I've read several sources that indicate its better to say start with the gain knob at 12'o clock and then bring it up or down appropriately.

Start with the gain on the minimum setting and then turn it up to where it needs to be. There was nothing in that link that said anything about starting at 12 O'Clock (which would be a different amount of gain on every device) and turning it up or down, where did you come up with that idea?

Regarding the EQ knobs on my mixer---should I ideally try and keep the lo, mid, and treble in neutral positions and do most eq adjustment in my DAW? I use Reason if anyones curious, I love the interface and options, and it has a great compressor, equalizers, maximizers etc. Latest version has a spectral analyzer as well. So with all the daws options us doing eqing on the mixer itself very necessary?

You want to be recording as cleanly as possible, so leave the EQ flat, and if possible switch the EQ out of the circuit completely.

On the other hand, if you have a nice sounding EQ you can EQ on the way in, but just remember that you can't undo it that way, and if you screw it up you've screwed up the take.

[XZit]

Hey thanks again guys for all the awesome responses. I have tons of reading and research to do.
First step is to get the monitors I'm looking at, pick up some acustic foam cheap off the craigslist, and potentiolally build some nice diy bass traps and then branch off from there as far as sound room treatment goes.

As far as getting everything sounding good and eqed I'll be referring to this post quite often! lol
Thanks for the note on keeping the EQ on the mixer ideally flat. I don't really have the need to use the mixers eq knobs, only time I may do that is recording with a xlr input/mic or whatever-or maybe a hand-drum which needs some low end make up or something like that. The daw is more then capable of EQing in a variety of ways.



take care everyone

[madtheory]

Don't bother with foam until you've built the traps using rockwool (see Ethan Winer link). Those low frequencies will be your biggest problem. Foam can't deal with LF. In fact I think it's likely you won't need any foam, just the traps. When you put panels at the points I described, they will take care of a very wide range of reflected sound.

Welcome to the future! We record at 24bit now, so you can leave headroom in your mixes and mastering engineers expect 24bit premasters that don't peak over -3dBFS.

The point of the example was that I've never had a mix returned because it was peaking at 0dBFS. Bit depth is pretty much irelevant. That said, I guess I should've specified that, like everyone else in the world, I record 24 bit. I made an assumption.

But more importantly, you've raised another myth. You can leave headroom in 16 bit too. In fact you should, when recording. Not while mixing. EBU standard for 16 or 24 bit is -18dBFS= 0 VU. SMPTE is -20dBFS= 0VU.

So again, if the engineer is any good, a mix peaking at 0dBFS (that is not actually clipping) will not be a problem. It's much better to understand what clipping is, than to mistakenly assume that 3dB of headroom guarantees no clipping.