• All remastered music should transcribe the originally intended instruments and overall style, whilst updating them to today's musical standards. Obviously style can be enhanced by adding some more detailed notes, good instrument sets, stereo panning and velocity changes, as well as overdubs or complete drum kits and loops. The tempo and notes themselves should remain mostly untouched, but of course, new notes are also welcomed to avoid shallow sounds. In order to capture the ambience that Sonic Team originally intended, we prefer remastered music than remixed music. Here is an example from one of our top music contributors. The original genesis tune plays first, followed by its remastered counterpart: Click here to download!
To further clarify exactly what the project management is looking for, here's three key guidelines.
Is the submission faithful?
-the piece is easily recognizable as representing a given zone or theme from the original game
Does the submission go above simply recreating the original?
-the piece makes use with modern instruments/samples to create a new work fitting the theme and atmosphere of each area.
Does it inspire?
-not only does the piece show a faithful and reimagined take on the original, but leaves its listeners impressed that this is one of the big ways Sonic 2 HD is living up to its promise.
So what does all that mean? Here's a few examples for you.
Ransom Rath's Metropolis
Mesmonium's Oil Ocean
Blast Processing's Death Egg
• All remastered music MUST be submitted in the OGG Vorbis format at 320 kbps. This is largely because the MP3 format has some inherent limitations when it comes to looping music, which makes it unuseable for the project. When uploading music submissions to the database, it is preferred that they be included inside a .ZIP file.
• You can download all the original music from Sonic 2 (including individual tracks from each tune) from our database, as a reference. All of the community's current music submissions are also available to download from the database.
--- scubaSteve's abbreviated guide to keeping your pieces from making people say "It's too MIDI!" ---
• First off, if your sampler patches don't have velocity mapped to note volume, you'll need to do this yourself. (As far as I know, any sampler in existence can do this - it's a very basic feature) As I said, you MUST use note velocities! This is always the best first step towards creating a more natural sound. Really take some time to imagine how each part should be expressed, how they should flow into each other, etc., and adjust your note velocities accordingly.
• Also, if your patches do not have multiple velocity layers - I.e., different samples for different ranges of velocities - you should consider on instruments with large ranges of timbres, such as brass, mapping velocity to a filter which will open up as velocity increases. Generally you'll want no greater than a 12dB/oct lowpass, as anything higher usually sounds unnatural. If your sampler can't do this (though most should), you'll unfortunately have to automate the filter by hand if you want the effect.
• Regarding the attack lengths, if your sampler is capable of changing the sample start time, this should your first choice for adding variation to attack times. For many instruments and applications, mapping the sample start along with volume to note velocity will do the job just fine; however, if you're looking for more control - ex. if you need a softer note with a short attack - you'll need to map the sample start to an external control (mod wheel, expression, aftertouch, etc.) and automate that.
• If the attack lengths seem fine but the strong notes just don't have enough punch, this is where a compressor becomes useful. Set it up with a short attack time (10-100 ms) and a long release (>200 ms) so that the attack of each note comes through before the compressor squashes the rest of the note down. You might also want to adjust the threshold so that only the louder notes are compressed.
• Lastly, on parts with a strong rhythmic presence (so definitely drums), it often helps to very slightly randomize the positions of the notes. This is especially important on instruments where you might have multiple notes triggering simultaneously - the toms on a drum kit are a particularly good example. Every DAW that I've used has a randomization function for this purpose, so I'd assume that most others have it as well.
I think that we also need to clarify just what it is we mean when we say "It sounds too MIDI!", as there seems to have been some confusion over this lately. Being the bunch that we are, I'm sure we're all well aware that even the oldest MIDI compositions can sound great today - it really comes down to the quality of the composition itself (Rob Hubbard, anyone?). However, I believe it was the initial attempts at using sampled instruments - especially in games - that gave MIDI a bad name, and is often what we're thinking of when we use the term informally. So I think, in a nutshell, what this usage of "MIDI" has mutated to mean today is simply "unnatural". Therefore, when we say something "sounds MIDI", it doesn't necessarily have to do with the quality of the samples - it's how you use them!
More general advice:
Generally you should EQ instruments individually as you add them, and shouldn't do a full-mix EQ until you're close to being done. Furthermore, as a rule of thumb, on your final EQ you should never need to apply more than 2 or 3 dBs anywhere - any more and it's likely that you have a single instrument (or two) that's causing problems.
For any other questions regarding music submission to the project, send a PM to scubaSteve or Canned Karma