[DVDdoug]

You are unlikely to get the kind of "loudness" as a pro mastering engineer with as little "damage" as a pro mastering engineer.

But there is something to the psychology, if my music is 1db louder than the next person's, all else being equal people will prefer mine.

Thus, the loudness war!

But there's more... if I write a quiet piece of music, it sounds better quiet, rather than normalizing it so the peak is at -0.1db. I'm exporting some tracks, and looking at a tiny waveform, and thinking it's wrong.

Don't go by the waveform or the peaks. You can use a "loudness" plug-in if you wish, but the normal thing to do would be to make it "sound right" by ear. If you're making an album it should be easy to get the relative loudness for the quiet track(s), but if it's a single you might just want to normalize it (and let the listener adjust the volume) or you can try to match the loudness of other quiet-songs in the same genre.

Recently I've had the idea, listen to some music, decide what volume it sounds good to me through my DAC/Amp/headphones, and then use that volume on the final export stage. Is there something wrong with this method?'

...Should I just find an ideal recording to set my levels to?

Using a good-sounding reference is probably the best approach.

Should I rather increase the final limiter's input level until I have an RMS of at least -10ish? How quiet is too quiet?

If your chosen reference track is at -10dB, that can work. But, an EBU R128 loudness plug-in is better, and your ears are probably better yet. But again, you are unlikely to achieve -10dB RMS with the quality that a mastering engineer can get (and IMO the pro is over-doing it at -10dB too, but that's a matter of taste)

You might also want to read the K-system paper. With the K-System, you calibrate the acoustic loudness (with an SPL meter) and then use your ears to adjust loudness. (That's probably not practical with headphones.) But even with the K-System, you'd probably want to check against a known-good reference, if for no other reason than to "keep your ears calibrated".

Actually, I don't know if it's my new DAC/amp/headphones, but a lot of recordings are sounding bad to me now... maybe I'm just going nuts.

If there's anything wrong with your headphones it would probably be frequency response (too much bass, too little bass, etc.) and it should be identifiable. It's highly unlikely that there are any audible defects in your DAC unless there is noise or unless you are driving it into clipping. You might just be getting bored with those un-dynamic -10dB RMS recordings.



P.S.

if my music is 1db louder than the next person's, all else being equal people will prefer mine.

That might not be true if the loudness comes at the expense of compression. I'll often turn-down highly compressed tracks. Plus, I'm normally using ReplayGain or SoundCheck so all of that loudness-war compression doesn't make it any louder, it just makes the music more boring.

It's my understanding that iTunes Radio and Spotify are applying similar volume-adjustment techniques, so again, loudness-war recording isn't any louder, just more compressed. (Broadcast radio does the opposite, further-compressing everything.)