tell me how to fix this mix

[stupidfatnugly]

I know it's amateur because I'm amateur

critique the poop out of it please

http://myspace.com/stupidfatandugly

[Johnny Rojo]

As I Listened on my JBL Nearfield monitors, I notice the piano line is very boomy in the lower register, almost to the point of distortion. If you are using a tube preamp, you might want to roll the tube gain down a bit, and if not, you may wanna roll some 125 khz out of the EQ.In general, the piano line could use a little brightening up. The synth sounds great, cuts through nicely, but not piercing.

Cool song though. Overall mix not bad, just some tweaking on the piano line.

[stupidfatnugly]

the piano is just straight from my keyboard into the interface.
should I re record it through an amp?
is it ok to play your keyboard through a guitar amp?

you're right I've got to fix that piano

know of any piano samples that I could use in REASON?

thanks Johnny Rojo

[KineticSound]

I agree with Rojo... too much on the bottom of the piano. Either it was tracked too hot or it's mixed too hot. I don't think it has to do with your sample, just the level setting and eq.

As he indicated, roll off the bottom of the piano with eq to see if that removes the low end distortion. Solo the track: if there's distortion in the recording, you'll need to make another pass at the piano recording with much less gain.

I do like that your guitar & pan flute parts take up very different parts of the frequency spectrum. Well done there.

[stupidfatnugly]

Originally Posted by KineticSound

I do like that your guitar & pan flute parts take up very different parts of the frequency spectrum.

how do you know this?

[stupidfatnugly]

i made a duplicate track of the piano part and panned one left and the other right all the way.

is this causing the distortion b/c I didn't hit any red lights when I recorded?

rolling off the low frequencies isn't helping

I lowered the fader volumes on both piano tracks. should I lower them more or get rid of the duplicate?

[KineticSound]

Originally Posted by stupidfatnugly

how do you know this?

By that I meant that they occupy different frequency ranges. The guitar and pan-flute (or whatever patch it was) didn't have a lot of bottom end, which was a good thing. They weren't battling for position against the piano.

There's more to setting the input level when recording than "not hitting any red lights". You want a signal with plenty of headroom.

[Farview]

Redo the piano part. This time, take the left and right output of the keyboard and record both on different tracks. Pan those tracks hard left and right.

Set the levels so when you hit a not, the sustain of the note sits about half way up the meters. Hopefully, that will take care of the distortion problem.

What kind of keyboard is it?


Another thought: This seems pretty loud for the type of song it is, what was the level of the mix? Did you do anything to make the mix louder once it was done?

BTW, you generally don't plug a keyboard into an amp if you are just trying to record the sound of the keyboard as it is. You would use an amp if you were trying to distort the keyboard.

[pehjott]

brian eno recorded synths,keys through amps which makes them sound warmer...everything affects the sound of your recording...mic,air,sweat,amp.when going out of a line out,maybe without DI, it will sound good, but one can improve and get more brightness with an amp

[stupidfatnugly]

Originally Posted by Farview

What kind of keyboard is it?


Another thought: This seems pretty loud for the type of song it is, what was the level of the mix? Did you do anything to make the mix louder once it was done?

BTW, you generally don't plug a keyboard into an amp if you are just trying to record the sound of the keyboard as it is. You would use an amp if you were trying to distort the keyboard.

roland xp-80

I limited the master fader at the end


I'm mostly wondering if I'll damage my guitar or bass amp by playing the keyboard through it.

[stupidfatnugly]

Originally Posted by pehjott

brian eno recorded synths,keys through amps which makes them sound warmer...everything affects the sound of your recording...mic,air,sweat,amp.when going out of a line out,maybe without DI, it will sound good, but one can improve and get more brightness with an amp

it does need more brightness

thanks

[stupidfatnugly]

Originally Posted by KineticSound

By that I meant that they occupy different frequency ranges. The guitar and pan-flute (or whatever patch it was) didn't have a lot of bottom end, which was a good thing. They weren't battling for position against the piano.

There's more to setting the input level when recording than "not hitting any red lights". You want a signal with plenty of headroom.

you can just hear it then?

what does headroom mean exactly?

[Farview]

Originally Posted by stupidfatnugly

I limited the master fader at the end

You probably went too far, that caused the distortion.

Originally Posted by stupidfatnugly

I'm mostly wondering if I'll damage my guitar or bass amp by playing the keyboard through it.

No, but it will dirty the sound up. You are having problems just cleanly plugging the keyboard into the interface, it would be better if you get that together first before you start worrying about doing something artsy.

[Farview]

Originally Posted by pehjott

brian eno recorded synths,keys through amps which makes them sound warmer...everything affects the sound of your recording...mic,air,sweat,amp.when going out of a line out,maybe without DI, it will sound good, but one can improve and get more brightness with an amp

What sort of amp? Preamp? Guitar amp? Bass amp?

There are a lot of things people do to mess with their sounds. I've been known to run drums through a Marshall. I wouldn't try to pass that off as common practice or good advice for someone just starting out.

[Farview]

Originally Posted by stupidfatnugly

it does need more brightness

thanks

I'm sure the keyboard was plenty bright. You need to figure out what happened to the sound that came out of the keyboard. Somewhere it went south. You either cooked the recording levels, cooked the mix levels, EQ'd it, beat it to death with a limiter or a combination of these.

[KineticSound]

Correct. It's about letting different sounds reside in their own little part of the frequency spectrum. Think of it this way... If you have two bass signers singing similar parts at the same time, it's difficult to distinguish the two, and the individual sounds get lost. But put a bass singer and a soprano singer singing similar parts together at roughly the same volume, and separation becomes quite easy. Mixing instrumentals isn't that different...

But I digress. Your immediate challenge is leanring to set proper levels. Farview was pointing you to this, "Set the levels so when you hit a note, the sustain of the note sits about half way up the meters. Hopefully, that will take care of the distortion problem." He's right, and he's leaning you toward headroom. Basically, the headroom is how much dynamic (volume) space you have between the signal and the point where your system says "here comes the distortion, buddy". Learn to use your dB / dBfs meter... a quick search of the forums on "setting proper input levels" will provide you with some tips.

[stupidfatnugly]

I've done a little searching and found

-18dbFS

db is for decibel correct?

what's dbFS?

basically I need to have less amplitude on my sound wave when I'm recording, correct?

[Farview]

db doesn't mean anything by itself. You need to know which db scale you are talking about.

dbfs is the scale that's used with digital equipment and is the scale that is used in your DAW.

dbVU is the scale used on analog mixers.

[KineticSound]

Originally Posted by stupidfatnugly

-18dbFS

basically I need to have less amplitude on my sound wave when I'm recording, correct?

Yes. Set your input so that you're hitting -18dbFS. That gives you plenty of room before you're in danger of clipping or distorting.

[stupidfatnugly]

-18 is considerably less than what I was doing

I mean that seems really quiet, like I wouldn't even get peaks. it would be just a straight line nearly.

are you serious?

I put another version on there and I sampled another piano and made a new track that I am playing along with the old one but at different frequencies.
I think it still sounds too hot but it has brightened up a bit.

what do you think of it now?

In those previous posts they say "tracking" what does that mean exactly?

is -18db what I think it is?: anything in the positive range clips so you have no choice but to record at less than zero
most the stuff I have done before has been maybe up to -1 db so this seems odd.

in wavelab I have a decibel meter but not in protools, I think. Should I check that it stays under -18dbfs in wavelab?

[KineticSound]

Yes, we're absolutely serious. I don't quite follow what you mean by a "straight line". I'm not suggesting that you limit or compress everything to maintain exactly -18dbFS all the way through. I'm suggesting that this is the target for input metering.

I'm going to point you to an article on SouthSIDE Glen's page. He explains it far better and more clearly than I am able...

Metering and Gain Structure

[born]

Originally Posted by KineticSound

Yes. Set your input so that you're hitting -18dbFS. That gives you plenty of room before you're in danger of clipping or distorting.

The -18dBFS thing is a semi-standard for calibrating VU meters to peak meters. To keep this really simple and leave out the complexities of metering standards, different rise/fall times, digital/analogue meters and what not, peak meters show you the peak voltage/level of a signal instantaneously - as it happens at that very moment, and VU meters, on the other hand, show you the average level of a signal over a period of time (still a fraction of a second though). Because of this, transients can slip through VU meters almost undetected whilst simultaneously going over 0dB on a peak meter - hence the need to calibrate. As an example, if you were to feed a continuous sin wave to the two meters and they are both at -3dB, then add a cowbell into the mix at a decent level, you'll barely move the VU meters, but will probably go over 0dB on the peak meters.

Now, say we were mixing on an analogue desk (with VU meters) into a DAT machine... in order to make the VU meters useful for metering and to help maintain correct gain/levels through the board we have to account for the transients which won't show on the VU meters, but will overload the DAT. So what we do, before we begin to mix, is feed a sin wave into a channel, send it to the master (which is sending to the DAT) so it's hitting obD on the master channel VU meters, then calibrate the levels on the peak meters so that the same sin wave coming from the console is sitting at -18dB (or whatever other level you may end up using, which can vary for many reasons) on the peak meters to allow headroom for the transients that will slip through the VU meters whilst otherwise overloading the DAT.

-- stupidfatnugly,

This doesn't apply to you, but may clarify why I'm going to contradict the advice that has been given to you thus far, and may explain why I believe it is a little confused.

In your case, as long as you stay under 0dB you won't be distorting the analogue to digital conversion stage... this depends on what you are using to meter, and how accurate and/or fast it is, so a few dB of headroom won't hurt you. You are quite correct with being a little shocked by the -18dB advice. Leaving headroom is good, especially when going to digital, but enough headroom that you can have a rock band turn up and play while you're recording a vocal is not good. I'll listen to the mix tomorrow.

For now, just stay under 0dB and then you can almost be sure that the problem is not overloading your interface.

[Farview]

Originally Posted by born

In your case, as long as you stay under 0dB you won't be distorting the analogue to digital conversion stage... this depends on what you are using to meter, and how accurate and/or fast it is, so a few dB of headroom won't hurt you. You are quite correct with being a little shocked by the -18dB advice. Leaving headroom is good, especially when going to digital, but enough headroom that you can have a rock band turn up and play while you're recording a vocal is not good. I'll listen to the mix tomorrow.

For now, just stay under 0dB and then you can almost be sure that the problem is not overloading your interface.

Here is the problem with that advice: While it is true that you won't overload the converters, you will still have to run the preamps feeding the converters well over the level they were designed to run at. Also, you will invariably end up turning everything down when you start mixing, otherwise you will overload the mix buss.

Peak levels are pretty useless when setting gain structure. The peaks don't matter as long as they are under 0dbfs. The average level is what matters.

Now, for instruments with a large peak to average ratio, like drums, having peaks up towards 0dbfs is perfectly fine. For instruments with low peak to average ratios, like sine waves, violins, etc... you would need to run your preamps at +15dbVU in order to get the level just shy of 0dbfs.
15db over line level! Does that really sound like a good idea?

Add to that the fact that most people on this board are using cheap or bang-for-the-buck preamps and outboard and you have a recipie for thin, scratchy, and generally terrible audio.

This is why there are so many of us running around trying to educate people about proper gain staging and trying to stop the assult of the outdated 'track as hot as you can without clipping' advice.

[stupidfatnugly]

Kinetic,

after first glance of that article, I'm thinking that in protools I must have a VU meter for my faders.

I must be thinking -18 db on the VU meter which I think you would agree is crazy low.

do I even have a bdFS meter anywhere in protools 7.3 or wavelab?

[SouthSIDE Glen]

Originally Posted by stupidfatnugly

Kinetic,

after first glance of that article, I'm thinking that in protools I must have a VU meter for my faders.

I must be thinking -18 db on the VU meter which I think you would agree is crazy low.

do I even have a bdFS meter anywhere in protools 7.3 or wavelab?

Any standard peak level meter on the digital side of the analog/digital conversion is a dBFS meter (unless it's a special plug-in specifically meant to mimic the action of an analog VU meter.) If it's a digital meter that ends in 0 at the very top, it's a dBFS meter.

There are two things to remember about VU meters on the analog side. One is that they are slow-responding - that is that they tend to show more-or-less average levels and that many of the transient, fast-acting peaks never register on your typical VU meter. Two is that they are calibrated so that 0VU equals the line level voltage that the inputs and outputs are designed to accept and deliver - and that the internal circuitry is designed to operate "best" at ("best" is in quotes because this ignores those exceptions where you purposely may want to introduce some analog distortion effects.)

So, taken together, this means that if one has their VU meter on the analog side so that the needle (or LED) seems to be averaging somewhere around 0VU (sometimes a bit higher, sometimes a bit lower), that one is running their signal at about that level at which their equipment was designed to be run. Which is usually a good thing.

The next thing to consider is the conversion calibration for the A/D converter - i.e. what dBFS level the converter is designed to convert 0VU analog into on the digital side. The unfair and unfortunate truth is there is no standard for this. While much gear is designed to operate at 0VU = -18dBFS, this is only an "average" amongst gear. It can range anywhere from -14 on some older European models to -15 on DATs and ADATs on the low end, to -20 to -22 dBFS on some late model Japanese and American DAWs and digital mixers.

If one *really* wants to understand their system's gain structure and get their system's metering and operation calibrated correctly, they should find out this conversion factor and calibrate their metering accordingly. there are instructions for this in the online app that Kinetic linked to.

Once calibrated, then one can know what they are really looking at and set their level accordingly (setting them on the analog preamp side of the converter, not in the DAW software.)

Just keeping peaks below 0dBFS,while it can get you by, is only part of the story and is not really following optimum gain structure. Just for example, if one actually has a conversion factor of 0VU = -22dBFS (which is more and more common every day), and their actual signal only has a crest factor of 10dB (the distance between average and peak level), then pushing that signal to ride just below 0dBFS actually means adding 12dB of gain to the preamps on the ADC, which can potentially push one into unwanted distortions.

And if you boost it 12dB on the digital side instead of the analog side, all you're doing is raising the converted analog noise floor and decreasing digital headroom by 12dB. Besides, when you start summing tracks together, you'll just have to turn it back down anyway.

In summary, in a perfect world everybody would stop just looking for the shortcut answers, whether it be the "below 0 peak" answer or the "-18dBFS average" answer, because while both of them are somewhat serviceable, neither of them is quite accurate or optimal. Find out what your own signal chain actually is and how your own conversion factor is calibrated, and you'll be cooking with gas. And give that answer to every one else, instead of perpetuating the "easy, but not quite right" answers, and before you know it, it will be second nature for everyone.

Unfortunately it's not a perfect world, it's a world where evryone is lazy and wants to take shortcuts - even if the long way is really only 5 minutes longer - and everyone winds up having to repeat these same debates and threads over and over and over and over and...



G.