What I used to do was to work hard on getting a pair of speakers to give acceptable calibrated response at my sitting position despite a mediocre room and just a carpet, pillows and moving the speakers for acoustic tricks. Then I'd use tones and sweeps to try somewhat match that with headphones/IEMs. Way back then,
@Joe Bloggs made a tuto on aiming for equal loudness contour as an intermediary reference(which was better if you tried to EQ one IEM or headphone into another one by ear). For me, it was laborious(I was so very inconsistent in the subs and upper trebles), but it was also the best approach I could find, so I kept doing it. Later, David Griesinger made a video
explaining something similar with what at the time was a work in progress app(that he generously shared with me after I requested it by email). To be clear, his approach is aimed at binaural recordings, and in his case, recordings he's done with mics in his own ears. So the main issue about having the wrong FR is that it affects the elevation cues, as everything else is already correct for him in the recording, interaural time cues, FR variations with directions. When we listen to standard stereo albums or binaural albums recorded with a dummy head, we have more issues than fixing the FR for frontal elevation.
Still, it's a method to try and match one FR with another so it can be used for other applications. You could try one central speaker and see if that pleases you, or you could do it with usual stereo speaker placement, or even with speakers placed nearly at 90° on each sides instead of 30° to better match the actual panning of headphones. All options have some logic to them, and all are also insufficient and flawed in some ways. So the best we can get is to simply try things and pick what subjectively satisfies us in particular.
After all that I kept looking for a "real" fix to the issues of stereo in headphones, I tried just about all the crossfeed apps, amps, VSTs I could get my hands on, tried many 3D/surround/virtual speaker solutions I could try or afford. And after mostly disappointment, I finally settled on the Realiser A16(where you ideally measure a good pair of speakers in a good room with mics in your own ears at different look angles(you turn your head, measure, turn your head again, measure...). Then with the same mics in the same position you also measure the FR of the headphone you plan to use for the simulation and the machine does the rest and switches to the right impulse responses for convolution based on the head tracker you put on the headphone). The funny thing, and why I even bother you with that, is how at the end of all the fancy technical stuff and impulses and decoding, you still end up having to apply some corrective EQ. The default solution lets you set your equal loudness contour with tones and just assumes you have a fairly standard contour, so it uses some generic curve to match, which won't work great if you don't have a very average ear or if you have some hearing loss). And that's how I came full circle, having to rely on actual speakers and go back and forth with tones and sweeps to EQ my headphone
2) The solution proposed for bass(beside subjectively adjusting it), is to have something like a bass shaker send back some alternative to the tactile bass you feel with speakers or real life sounds. It does tend to work to some extent and give an increased perception of bass "sound".
Otherwise, the trend is to add a little more bass on headphones, and even more on IEMs to try and compensate for the missing tactile feeling. But of course it's not perfect, as if you boost say 60Hz and lower a lot, a portion of the frequencies above that will be masked by our own hearing scheme. So then you might want to increase that area that then masks some higher stuff, etc. And you end up with just a louder gain everywhere.
Of course that has nothing to do with compensating for bass if you happen to have a bd seal, a FR graph for a headphone will usually be done with the best possible seal they could manage to get, you might not have anywhere near that seal level on your own head with hair or maybe glasses or just a shape around the jaw that leaves an opening. That you can compensate as it won't result in too much bass, it will just restore what you're losing through the air leakage.
3) You should be able to use some "virtual cable" like the ones proposed by VB Audio(some adapt to sample rate and stuff on the fly, some are more "bit perfect" oriented but they won't work when you feed them with the wrong settings, up to you to decide how much you can be bothered with all that. There is even an asio conversion thingy). Or if you need even more settings you probably wouldn't understand but exist, the old VAC, virtual audio cable(not free if it's still sold), had all that and more. Whichever you use, they show up on Windows as another audio device so you can just tell whatever app to output the signal to it. Then you still have to get some app or VST host that will contain your EQ VST), you tell the virtual cable to output to it, and you tell it to output to your DAC. It's cumbersome, but I did use all that crap for nearly a decade(with EQ, room reverb app, some "true stereo" crossfeed) before moving on to the A16.
It's simple, you just need time money and experience.
Edit for franglish and ducking autocorrect.
