SMPS tips and discussion

As you may know I have a very keen interest in music. For next week my intention was to host a workshop for mid2smps and using it with FL Studio, complete with my whole work process and converting a song all the way through via livestream, but due to unforeseen circumstances that is now impossible as my main machine is unusable with a PSU failure. Whilst I wait for a new PSU, I thought it might be a good idea to still help people struggling with the musical side of their hack, to just get a bit more out of it and generally improve us all.

You may have noticed that with mid2smps out it's a lot more accessible to have custom music in your ROM hack, and in addition with the MD Midi driver by ValleyBell, SMPSPlay also by ValleyBell and MegaPCM by vladikcomper, you have a very powerful arsenal to make a really good sounding piece of music with SMPS with relative ease.

The problem is rarely does anyone actually do this. I'm still hearing the same issues as before and no one really knows how to prepare their music to make use of what they've got. Even the SFactor, which has music I loved a lot two years ago, bugs me now because what's possible with SMPS is more accessible than it's ever been and it's not too difficult to make a full, rich experience with the limitations. Granted, this is mainly due to advances in our knowledge of the SMPS engine and it sounding dated to stuff compared to recent projects, I'm going to give you a few guidelines to help you.

First of all, I think the most important thing is you really need to have a proper keen interest in music. If you're new to this whole thing, avoid working with SMPS, there's far better entry level things to use (I'd heavily recommend FamiTracker, for a start!) because when I started, it was the exact mistake I made. I didn't know FamiTracker existed, infact, I didn't know *any* alternatives existed. I had no interest in music and I had to develop it and this made my task extremely difficult. Through persistence I gradually improved and began to understand SMPS and music in general. But the point is SMPS is definitely not an entry level music related THING to use.

Secondly, you'll need to understand that you're going to face criticism for whatever you do. It's how we all improve. If someone says your music is bad and points out why, don't lash out at them and defend it like a holy grail. It's a phase we all go through, but once you accept they're trying to help you it's a lot easier to take in. We're all trying to improve.

Third, and perhaps one of the most important things, appreciate what you're working with! You have five (six if you're not using a DAC) wonderful, luscious FM channels that can provide a massive timbre of sounds. You have three to four programmable sound generators capable of 50% square waves and two noise modes. You have a programmable sample channel. Use these. It's a difficult road to start with, but there's a lot of ways to ease into SMPS and the Mega Drive itself. Working with a tracker like VGM Music Maker or Deflemask (although I detest Deflemask, you might not) or making fakebit mockups with VOPM and FL Studio will introduce you to work with the FM channels and not rush in blindly.

The main complaints I have with SMPS music today is no one really knows how to use volumes or solve their phasing issues. Some people have music that's too unbalanced with quiet and loud volumes everywhere, or they're too quiet, or worst scenario, they're too loud. Let your music breathe! Use your dynamics. It's important to keep it to a reasonably higher volume, but not an excessive one. Keep everything together, volume and panning is your mix's glue, if it's bad it'll fall apart very easily. Similarly keep an eye on your pitches and make sure you don't have something too high pitched or too low pitched. Phasing, the other issue, comes up very often with PSG channels and echoes. It's very usually solved by lowering the volume (I usually would lower it anywhere between 70%-50% of the original volume) and detuning the echo. It doesn't seem logical to have something out of tune, but this is extremely handy. Phasing is what happens when two identical waveforms play over each other and they cancel each other out. Sometimes this is a desired effect, but commonly it is extremely undesirable and should be eliminated. For FM channels, I detune my instruments down or up anywhere between 10-7 cents, and PSG channels 3-4 cents. Similarly, with FamiTracker, many people would recommend using the Pxx command to detune the echo so it doesn't phase with the main channel. With this in mind it's very easy to make your port or song sound a lot better.

Lastly the biggest thing I can say to help you that comes to mind is don't expect 5 minutes of effort to sound as good as something someone has spent hours into. I work extremely hard to make sure everything I make is as good as possible and fill out my arrangements to make the full use of the hardware. Obviously, some people don't have as much musical skill and only want something a bit simpler for their hack, but even still you should put some effort into making it sound as good as you can. Don't be afraid to ask for help, either! There's a lot of talented guys around that I'm sure wouldn't mind helping you at all.

There's a lot more I could go over, like the process of converting in detail, but I'll leave that to another day. If there's any questions or areas you'd like to know a bit more, I don't mind answering, and if anyone else experienced with SMPS or audio in general would like to step in feel free! It's really important to spread the knowledge so we can all improve.

Lastly, as a note to Aquaslash, I'm not bashing your hack at all!!! I love it and still enjoy your music, but I was using it as a reference as for something I once seen as the benchmark sounds a bit dated by today's standards. No hard feelings.

 

I like the cut of your jib.

Interestingly, I was wondering if there would be a way of "calculating" (forgive the odd terminology) the overall volume of the music, and working out how much to raise or drop the volumes in order to normalise it correctly.

 

What I'm thinking is you'd check how many decibels there are on average using a sound level meter, or (the option I'd try) what you could try is routing up the Mega Drive output to the input of a DAW/mixer and graphically see how loud something is. With SMPS I've usually just gone through trial and error until everything sounds 'normal', though.