Fletcher-Munson Equal Loudness Contours

L

little guy

New member
I'm using SONY Soundforge 8.0d. i wanted to compare my work to some professionally done projects. I grabbed a couple songs by the Crystal Method. I scanned the RMS using the Fletcher-Munson Equal Loudness option, and i also scanned the RMS without using the Equal Loudness contour. I found that The Crystal Methods songs a little variation between the two scans, maybe 1 - 2 db difference.

So, i then took my project. did both types of scans. quite different. i had atleast 6 db in difference. Using the Equal Loudness option my mix was about 6 db quieter than without using it.

I understand what the Fletcher-Munson Equal Loudness Contours are. How the human ear is more tuned to hear freqs at about 2000 hz than freqs above and below this. very low and high freqs don't seem as loud as freqs in the 2000 hz, even though they may register the same db.

i guess my questions are:

1. how important is using the fletcher-munson contours when getting your volume up?

2. how many of you use the fletcher-munson contours?

3. i know it's all about the ears and how it sounds, but geesh, how do i narrow up that gap between using the contour and not using it? (stupid question)

4. i'm looking for a good overall balance of the frequency spectrum as percieved by the human ear. if i minimize the gap when scanning RMS between not using the contour and using it, will this help me acheive this?

thanks to anyone who can enlighten me.

p.s. i know it's all about how it sounds and not how it looks on the screen. i'm just trying to learn a bit.
 
little guy said:
1. how important is using the fletcher-munson contours when getting your volume up?

2. how many of you use the fletcher-munson contours?
Click to expand...
I personally have never used them even once. This a red herring issue, IMHO. And if you think that's a bad answer, just wait... ;)

little guy said:
3. i know it's all about the ears and how it sounds, but geesh, how do i narrow up that gap between using the contour and not using it?

4. i'm looking for a good overall balance of the frequency spectrum as percieved by the human ear. if i minimize the gap when scanning RMS between not using the contour and using it, will this help me acheive this?
Click to expand...
If you don't look at or even think about the contours and mix to the song itself, the "gap" will take care of itself.

Seriously, checking a Fletcher/Munson contour when mixing a song is rather like measuring how many centimeters one has the gas pedal depressed when driving a car; while it can be an accurate way of measuring how much gas one is feeding into the engine, it's not the technique one should use to determine whether they are giving the car enough gas. It's derivitive information only.

Yes, it is all about the ears. But really that's even a bit derivitive. Fundamentally, it's really all about the song and it's content. What does the song WANT to sound like? While you may want it to sound like a perfect equalization curve, it may have a different idea. Listen to it; let it tell you what it wants.

First, if you do want to give it some balance (which is, BTW a good idea), when trying to find the balance, find it not in the curves, but rather in the soundstage. Find the right balance and choreography in L/R pan, in frequency, in depth, and in time. When one has acheived good sonic balance in all four of those dimensions, the ears will be pleased. The resultig curves may or may fit some theoretical curves, but it really doesn;t matter at that point, does it? if it "sounds" balanced, it is for all meaningful intents and purposes, balanced. At that point the curves become deceitful if they tell you otherwise.

On a more practical level, a true F-M fit can only be had if one is mixing or EQing pink noise. In the real world, where we are mixing drums, guitars, keyboards, vocals, etc., we are at the mercy of the curves of the instruments and the color of tracking devices themselves. It is entirely possible to have a mix that is as balanced as it can get and still signifigantly deviate from a "perfect curve" because the instruments themselves do not give out equally in all frequencies.

Plus, just to complicate it even more, some of the best engineers purposely alter or modulate that balance as needed to fit the "feel" or "emotion" of the song. They have listened to what the song wants and then engineered to bring that desire out. Some songs are purposely mixed with an unbalance in frequency spread, due to a combination of the types of instruments and arrangement used, and the selection of specific "colors" or "emotions" that deviate widely from "balanced"; e.g. they may have provided an over-heavy bass line to indicate impending action or a serious tone, or a harsh midrange on the vocals to generate tension or alarm, or crisp clear air on the high end of the spectrum to reflect smoothness and calm, etc. These "pro" commercial recordings would be happy to kick both Fletcher and Munson out the door and on their asses :) (See some of the classic recordings by the likes of Tom Waits, Radiohead, Elvis Costello, David Bowie, Moby, Nick Cave, Graham Parker, Jimmy Cliff, and Led Zepplin, among dozens of others, for perfect examples of this.)

G.
 
Last edited:
wow, thanks for the reply southside. i appreciate it. the mix i'm working on does have a lot of low end, and that's what i'm looking for. i just kept feeling like something was missing in the mix. the mid range was there, the mid-low was there, low was definately there. the mid-high to high needed a little pushing. so as usual i did my best just listening as made some EQ adjustments. got it to where it sounded balanced throughout the mix, with just a bit more on the low end. i was happy. so for giggles i opened up the mix in Soundforge. Scanned both the RMS with and without the contour. to my surprise it closed the gap.

I'm gonna work on this concept a little more tomorrow, it's late now.

BTW, i wasn't able to find what was missing by using a paragraphic EQ. i used a 20 band graphic EQ and messed with each band until i found the sound i was looking for.
 
I'd read an article saying that "get those frequency response as flat as possible". Rather than trusting my ear i follow that "advice", i tried my best to get that perfect flat response using spectrum analyser.... which make me crying :). After long hours of "crying"... i decided to listen to reference songs, listen, analyse.

Suddenly, I realise my mix too much of "muddy". IMHO, comparing my mix with the ref song is a great way to learn from the pro..
 
I think in the end, you don't really have control over it anyway. If people set their radio on a low volume, you got a bass and treble problem right there. If they pump it up it's gonna scream. You can ofcourse try to minimize this by making your mixes fairly flat, but even in a perfect situation (which is hardly ever the case) it's gonna sound different at different volumes.
 
in my reading i've found that 85 db is approximately where one should be mixing at, for that's where the ear perceives the frequencies to be the closest to being flat.

1. how do i know what db i'm mixing at? is there a device that will tell me?

2. if so is it expensive?
 
cool, thanks. how important is the quality of these. like is there a huge difference between a $50 and a $100 meter?
 
little guy said:
cool, thanks. how important is the quality of these. like is there a huge difference between a $50 and a $100 meter?
Click to expand...
No, not huge. There is a big difference between the $50-$100 meters and the $1500 meters :). But nothing that folks on our level need to worry about.

As far as setting your monitoring level, you can do it well w/o metering. Take a good quality reference CD with good tonal balance across the spectrum and play it back through your monitoring system. Start wil the volume turned down, listening to the bass response. The bass will sound weak at lower volumes. Turn up the volume slowly until the bass catches up with the rest of the mix. That's where you want to be. On the other side of the volume coint, if you turn it up so far that the high end starts feeling fatiguing, you've gone too far.

G.
 
speaking of fatiguing. at work i've been listening to music with my headphones (office enviroment). lately my ears have felt like they need a break. plus i could be mixing at a volume that's tooooo loud.
 
I think the curve is a good thing to know, but in the same way you know how to balance on a bike. It just comes instinictually (or should).

I can't remember any situation where any engineer that I've worked with or myself has had to bring up the chart as part of production work.

The only valuble information it will give you is a broad consensus of what average ears react to. I think they settled on 83db as being the "optimal" level for the best mixing.


Any higher and your extreme highs and extreme lows may be lacking, since our ears are sensitive to the mid range frequencies.

Significantly lower than 83db, and you might have boosts in highs and lows that you didn't anticipate.

But that's just a text book explination. It's just as Glen said, you mix according to your situation.

Sometimes you have to mix loud and sometimes you have to mix at very soooooooooft levels.


As they tell me, "you be the judge".


Just make sure you can tell when your ears are in need of a break. You might need those in your older years.
 
You must log in or register to reply here.
Back
Top