Slater
Headphoneus Supremus
Just for the sake of curiosity, how do you modders know what to mod? Is there a set rule to the process (like, doing A will have effect B), or do you go around plugging random holes and such and hope for the best?
Depends on if you're the pioneering type.
There aren't any set rules per se, but as you get more comfortable doing it, you'll find that many of the principles apply (and these same principles apply to full size and car audio as well). Stiffer cabinet materials, enclosure filling, dampening with absorption materials, port tuning, etc.
Personally, I do 1 of 3 things:
1. Wait for more experienced members whom I either trust or I know have similar sound signature tastes as me to develop/test out mod first. Many of these people are in the audio business and often have fancy testing equipment to validate the results of the mod testing.
2. If I get something that I am not happy with the sound signature, or it was cheap enough, then I try some mods to change the sound signature to something more to my liking.
3. Wait until multiple people have tried a mod and there is a general consensus as to the change.
Many mods are easy to try and only take seconds - taping off 1 or more ports, pad rolling, tip rolling, removing felt covers (which can be taped back if you don't like the change), quarter mods to foam pads, and changing to a better cable (on units with removable cables). And most full size headphones are only held together by a few screws, making them easy to try out a little fiberfill, or peel and stick on a layer of Dynamat or acoustic foam mat. Also, not all mods are acoustic - I always tweak the clamping force on full size headphones to make them more comfortable and increase the length of listening sessions. Often I adjust the amount of headband padding to correct weird pressure points and improve comfort.
I'm much less apt to perform more advanced/permanent mods - cutting holes in housings to make them open, swapping drivers, or adding resistors to increase impedance - unless there's a really good reason to or it's a result of #1 or #3 above.
The general argument is often 'omg everything is carefully designed by engineers and how can you know more than them?' Well, everything is designed with compromises - cost compromises, form over function (good looks but uncomfortable), model lineup compromises (we don't want model x to be too much competition with model y), material compromises (thin plastic or cheap crappy cables for example). And products are tuned to a particular sound or particular target audience, which may or not match my preference. So I can use mods to more closely match what I like. For example, I like an airy sound with big soundstage (who doesn't), but tight, warm, and powerful bass. But those features aren't always present together - open headphones generally lack bass, and closed headphones generally sound compressed and closed in. But even in those cases you can do things to tweak the sound to be more your liking.
There's no free lunch though - as you change one thing, you always change something else. Making changes to bass may affect mid range. Reducing sibilance may lose clarity and sparkle. You may get that monster bass but it may turn muddy. I've undone mods because I wasn't willing to accept the compromise of whatever it changed, so in those cases I may relegate that particular IEM/headphone to a particular musical genre or even movies or gaming only.
You'd also be shocked at the detuning manufacturers do because of product lineup compromises. A perfect example is the Sennheiser HD558 and HD598. It's literally the same headphone, same drivers and shells. The only difference is a few $0.50 strips of foam inside the housings, yet one is 50% higher in cost. Sennheiser can have more offerings at multiple price points, and lower models won't step on the toes of their higher model (and higher retail) siblings. Auto manufacturers are experts at this, and often the only thing between 2 marques at very different price points is $20 in badges - the rest is all marketing.