[taileon]

First off, hello! 
 
I'm relatively new to the audiophile scene, having purchased my first "proper" headphones a few days ago (the ATH-M50). They sound absolutely brilliant!
 
But there's something that I've been trying to figure out! I've seen many people say that the M50s project recessed mids (compared to the lows). I kind of know what "highs" are supposed to be, and I definitely know what the "lows" are supposed to be, but what part to "mids" play in a song? Is it the vocals? Certain instruments?
 
Help is very much appreciated! (I apologize if I posted this in the wrong forum!)

 

[TMRaven]

I'd say it's the area spanning from 250hz to 5000hz, give or take. It's the section of music where most instruments and vocals are situated. Sometimes people give vocals as the dominant example of mids, but that's hardly the case imo. The stronger the mids are, the stronger both vocals and instruments will be. Audio Technica M50 had good lower mids if I remember correctly, but a bit tamed upper mid section. It's been a while since I had them. Upper mids affects things like female vocals, electric guitars and some brass-- among other things.

 

[obobskivich]

Here's a nice graphic from an EQ manual I have (observe that the scale is not 20-20k, but 16-32k).

 
The manual defines the mids as between 400hz and around 2500hz. I roughly agree with that. Basically it's the range you'd expect to see handled by the mid-range driver or mid-woofer on a modern multi-way loudspeaker. It's not quite the entire human vocal range, but contains both most of the vocals and a lot of instruments. Sucked out mids basically mean that the lower frequencies (like 125hz and down) and the treble (like 6k and up) are going to be more prominent in the presentation. That's actually rooted in an ISO standard (226:2003) based on the equal-loudness contour (our hearing isn't ruler flat, it's V-shaped), but sometimes it's taken too far and you get a very "recessed" sounding headphone. I'm not saying following ISO produces the best headphone ever mind you, because it's based on an AVERAGE of human hearing response, with smoothing, not an exact model of your specific hearing (which is likely not even ideal to the average due to a variety of factors including individual variation). 
 
A very simple example, DTMF tones all exist within the "midrange."