But if you do it like Zef wrote it up here, then you should definitely listen to everything on some good speakers.
Or better yet, listen to it on all kinds of different speakers, so you get an idea of what it could sound like on different audio systems.

This is a really good point.  I don't have fancy studio monitors, so when I'm working on an EP I listen to the songs on as many speakers as I can.  Computer speakers, in the car, ipod dock, big component stereo system, earbuds, over-the-ear headphones, etc.  Getting an idea of how your mix sounds on lots of different speakers is helpful.

I'm mentioning this, because I noticed that instruments you might hear clearly on headphones won't be as clear on speakers. By listening to your tracks on different speakers you can easily determine which instruments need a volume boost + which don't. Then it often helps to increase the envelope by just 10-20, this should make them more present already.

I try really, really hard to get everything really even and smooth in LSDJ before I record it.

This; definitely.

Quite honestly the DMG gives a really nice clean mix if you just take the time to do all your volume levels right, and if you know how to write so that lines don't conflict. I also record 1 track straight from the jack; it's quicker and cleaner, and your songs sound the same live.

I try really, really hard to get everything really even and smooth in LSDJ before I record it.

Hell yeah!

I'm personally a fan of pure unadulterated Gameboy line-in recordings. (FLAC of course)

The only other thing I do is use the TLs Maximizer plugin to boost and even out the volume.

I try really, really hard to get everything really even and smooth in LSDJ before I record it.

are you basically looking at the track through a spectrum analyzer as it plays and lowering peaking frequencies accordingly?

Not really. I mean I definitely use reference tracks to get a general idea but I definitely use my monitors/ears to make sure it's nice and balanced.

It's not an EQ. I'm talking about multiband compressor. Forgot to clarify that in my last post, my bad! So once again, I just use my ears and get a light squash on the bands that need it.

Mid-Side EQ is a strategy to separate the panorama spectrum into "middle" and "sides" and to EQ them differently. I like to high-pass the sides a bit and raise some air/shimmer slightly and for the middle I just hone in on the kick and cut out all below 30-50hz.

Mastering I really don't do too much other than evening out some peaks with a multiband. I have however been really into using Mid-Side EQ to get your bass all very mono and tight as hell. Definitely an improvement over my previous recordings.

More of a general mastering question here (to which I am also a relative noob), when you're evening out your peaks, are you basically looking at the track through a spectrum analyzer as it plays and lowering peaking frequencies accordingly? If it's a global EQ, how do you get that to not affect other parts of the song where that frequency might not be peaking?

And wat.jpg on "Mid-Side EQ", please explain this magic.

Is the variability of LSDJ's tempo too great to possibly record each track by itself in Live mode mix them in another DAW? Or would that be cheating?

I wouldn't trust it, personally. You can actually sync it with MIDI Clock/ARDUINO or whatever too, but I just Warp the tracks to the click in Ableton Live.

Mastering I really don't do too much other than evening out some peaks with a multiband compressor. I have however been really into using Mid-Side EQ to get your bass all very mono and tight as hell. Definitely an improvement over my previous recordings.

also - This is just me but when I DID multitrack once, I had to delete all the data from the other channels that weren't being recorded. Even on mute, the DMG still processed the data and made some hiccups on my DMG. Recording one lone channel at a time made a far smoother multitracking process, at least for me.

+ yes, it is totally possible to record each track by itself. This way you have so much more possibilites. I was wondering all the time why noone mentioned this before.

So anyway, if you mute the channels you don't want to record, you sometimes get disturbing noises in parts where nothing is played. Especially kit-instruments cause these disturbing noises, but other channels or instruments do, too.

EQ-Protip: You get most of the DMG bassdrums at about 100 Hz.

Edit: I recommend recording single channels with the three other channels muted. The distortion sounds usually don't bother, + if they do, you can cut them out or mute them (in postprocessing). This way it is sure that your DMG will keep the tempo. Even if there are slight variations, they will be on every single channel, so every single channel will be synchron. Variations may occur, but you wouldn't be able to hear them in a full mix, because they would be too slight.