What is soundstage?
If you've never checked out the head-fi.org forum, you need to join that community if you're serious about learning the ins and outs of audio.
It was on this site that I found the perfect and most sesinked answer to the question of "what is soundstage?" and "do headphones control soundstage at all?"
So instead of claiming this knowledge as mine, I'd like to give credit to the user fjrabon, who is absolutely brilliant in his knowledge of all things audio! This answer to the question of what is soundstage dives really deep into the details of headphone audio, so if you're not comfortable with audio jargon you might want to have google open as you read.
I've seen a lot of posts where people state they don't really understand the concept of soundstage "depth". Here I will relate how the concept of soundstage depth works in general, why headphones have a problem with it, and their "imperfect" solutions to the problem.
The vast majority of my knowledge about these areas comes from Gary Davis and Ralph Jones' Sound Reinforcement Handbook, Second Edition, as produced for Yamaha. It's a FANTASTIC book for learning about how audio works. It's aimed at live sound reproduction, but its so exhaustive that you can learn a TON about sound reproduction in general. I don't know of a comparable book for home audio, though I am sure there are some out there.
The fundamental stumbling block to headphones, with regards to all aspects of sound, but especially soundstage depth, really doesn't have to do with their ability or inability to reproduce sound in a quality manner, it comes from the fact that mastering is done for speakers, and headphones aren't speakers. This is sort of a background fact to keep in mind when I talk about why things are the way they are. I will constantly refer to the "producer's intent", what I mean there is what the producer/engineers/etc intended for the record to sound like, I mean proucer, just in terms of the party in charge of producing the record, which isn't necessarily the credited Producer with a capital "P". In fact a lot of people hate how producers want things to sound, and that's an entirely different topic, but for the purposes of this discussion, the goal will be to get as close to the producer's intent as possible.
So, since producers generally master for speakers, lets first talk about how speakers create soundstage depth:
Soundstage depth, in general, is created by the fact that high frequency sound waves are absorbed and dissipated more easily than low frequency waves are. This is why you can clearly and easily hear thunder from several miles away, but you couldn't hear a high pitched siren very clearly from even a quarter mile away. Or why you can hear your neighbor's bass, but not any of the words to the song. So, what your brain does is calculate the differences in amplitude of the fundamental frequency and the frequency's harmonics. Any tone is made up of a fundamental, which is the main tone, and harmonics, which are numerical multiples in frequency of the fundamental. Because the higher frequencies are absorbed by the atmosphere faster than the lower ones, if there are a lot high harmonics compared to the fundamental, your brain thinks that the sound is close by. These higher harmonics are often what is referred to as the presence frequencies. More presence frequencies = sounds closer to you.
So, speakers create soundstage depth from two areas. 1) is what is in the recording. This has to do with microphone positioning (far away mic'ing sounds farther away) and EQ'ing (pulling out or boosting the presence frequency of an instrument can make it sound closer or further away). Typically mic positioning is preferred to EQ'ing for creating soundstage depth. it's easier and doesn't have nearly as many complications. For instance an instrument with a very large range like a concert grand piano is difficult to "move" with EQ, because the presence frequencies of its lowest registers are different from its highest. But if you mic it with several different mics at different distances, you can control the soundstage depth by mixing the relative levels of closer or further away microphones. 2) is the simple distance from the listener and the speaker. This is what gives speakers the feeling of not having sound pumped into your brain, that you are watching a performance. The distance from you and the speaker will knock some of the presence frequencies off just from the air.
Now you can probably already see what the fundamental issue with headphones is going to be here. Headphones, by definition, have the driver extremely close to your ear. There isn't very much air to "knock down" those high frequencies. Meaning that a totally flat measuring headphone would sound insanely high pitched. Like an icepick being shoved into your eardrum. While with speakers the goal is a totally flat frequency response, this simply is not possible with headphones, it wouldn't sound like what we are accustomed to hearing as flat. The problem is we are trying to use a source that is mastered to be heard from a distance of several feet through headphones that are less than an inch from our ears.
Now this probably sounds grim, but headphones solve this problem by rolling off high frequencies. Even a headphone that is considered to be VERY titled towards the high frequencies, the AKG K701, has all frequencies above 8kHz MASSIVELY rolled off.
However, this is still an imperfect solution, because it's hard, if not impossible to perfectly replicate the natural, atmospheric rolloff that occurs in air with frequency response on a headphone. It's imperfect in the same way that EQ'ing an instrument in the studio is an imperfect way to create soundstage depth in the studio. Different instruments have different presence rages, and thus its impossible to accurately reproduce the correct adjustments with a single instrument that has the same sort of "EQ setting" all the time. Some do a better job than others, some are quite outstainding to the point of ALMOST sounding "speakerish" in this regard.
So now onto how some headphones create unrealistic soundstage depth. Some headphones aren't just content to try and approximate headphone depth, they want to try and beat it in some way. They do this by boosting and pulling back certain presence frequencies. A notable example is the AD700. When flat, it pulls the vocal presence back a touch, allowing vocals to less forward, which creates a sense of depth. The AKG K701 goes in the opposite direction.
It pushes the female vocal presence frequencies forward, and drum and bass presence backwards. This creates a deep sounding soundstage for female vocal performances, because the vocals sound immediate, the drums and bass sound like they are in the back and most other stuff sounds in between. However, it can become discombobulating for a bass solo, because the bass solo refuses to move upfront. Or on a male/female vocal duet, because the female vocal sounds much more forward than the male. it kind of sounds like romeo and juliet singing a duet during the balcony scene where you are standing next to juliet. This is what people mean when they say that the K701 has an artificially deep soundstage. It's not that the soundstage is "too deep", just that its artificially, and at times oddly, deep.
It's my opinion that headphones just CANT accurately reproduce soundstage depth. The best a headphone can do is strive for a flat SOUND (as opposed to a flat reading on a FR chart) and then whatever that makes the soundstage depth is the best they can do.
Now, please don't focus on the couple of headphones I mentioned here. They were only used because they were example of headphones that I personally believe exhibit a certain quality, and not because I like or dislike them. Hopefully this helps some understand what soundstage depth is and how it works.