So, you've tweaked and re-tweaked, listened over and over, loud and soft, and you think, "This is it! I'm done! Let's save out that final version and make a million!" Well, hold on just one more minute! It's easy to get excited about your new masterpiece, and in the excitement it's easy to overlook a few things that could make or break your long-term satisfaction with it, and ultimately the success of your published song. You don't want to look back in a week or two and say, "I just wish I'd..."
Here's a checklist of items to consider at this point. (I assume the reader is familiar with Reason and digital audio in general.)
Content
1. The Dreaded Master Tune
2. Making Sure You Hear the Same Thing Each Time
3. Auditioning Tracks Individually
4. Unused Devices
5. Final EQ
6. Clipping
7. The Car Stereo Test
8. Missing Samples
9. The End and Loop Markers
10. The Rack Layout
11. The Song Info
12. The Final Tests
13. Summary
The Dreaded Master Tune
Unfortunately, the Master Tune setting (in the Preferences menu under Audio) is not saved with a song. What this means in practice is that the relationship between the timing of your samples, and the events triggered by the sequencer, is dependent on something completely out of your control when saving a song. This is because Reason retunes samples by resampling, which does not preserve the tempo of the sample. If you've tweaked your song carefully, but your Master Tune setting is different from the person who opens your song, the song he or she hears may be quite different (and awful) indeed.
The solution? Set your Master Tune to zero cents and keep it there. Also, hope that your listeners do the same.
Making Sure You Hear the Same Thing Each Time
When you create a song in Reason, it's simulating the use of a sequencer and a bunch of audio gear. What you are not doing is recording the actual audio into separate tracks and mixing them. Audio gear like synthesizers and effect boxes don't necessarily sound exactly the same when you play the same note or pass the same audio through them. Usually these subtle differences don't make a huge difference in the overall effect, but sometimes they do.
The Portamento
First a brief description of Portamento: When the Subtractor or NN-19 plays a note in a patch with Portamento, the starting pitch is determined by the last note played, not the current note being played. The pitch then ramps up or down to the current pitch, at a rate determined by the Portamento setting. The higher the value, the slower the climb or fall to the next pitch.
How can this a problem? Well, what pitch does the first note played sweep from, after the device is created? The answer is, zero Hz. Therefore, you can be listening to your track all day and your sweepy portamento patch sounds great, but when someone else loads it up for the first time, the first note or even group of notes sweeps all the way up from zero, which is not at all what you intended. In fact, even if it is what you intended, you won't get the same results every time (what if the person plays the song to the end, then starts at the beginning again? This time it won't sweep from zero). If you do want exactly the same sound each time, you need to take steps to ensure it.
The "standard" way to fix this is to have the instrument volume start at zero, insert a "lead" note, with a desired starting pitch, at some point before the first note in the sequence, and bring up the instrument volume to normal right before the first audible note hits. For a high portamento setting, this may require a longer "lead time." If you don't want the first note to sweep at all, you can use the same pitch for the "lead" note. Another simpler method in this case is to start the portamento at zero right before the first note hits and quickly snap up to the desired value during the time the first note is playing.
The LFOs
LFO stands for Low Frequency Oscillator, and at one time was simply another oscillator in a synthesizer which could be mapped to affect various synth parameters such as pitch, amplitude, filter frequency and so on. The LFO itself is always "on," even when it's not affecting anything, and in Reason, even when the sequencer is stopped (ever wondered why your CPU usage indicator stays high even when you're not playing anything? This is partly the cause). This can be a problem if, for example, you have a very slow LFO rate which is radically affecting a synthesizer's pitch, and you want the pitch to hit a certain point at a precise moment in the song. There's no way to guarantee it; it's completely dependent on the state of the LFO at the moment at which the song is started. The same problem might happen for LFOs in the Phaser or Chorus/Flanger. The slower the LFO, and the more pronounced the effect, the more the sonic differences can be perceived.
The solution to this one is a lot tougher than for portamento. There is no way (as of this writing, Reason 1.01) to "hard sync" an LFO, so the phase LFO wave is always at an arbitrary position when the song starts. One method that could improve things is to replace whatever your LFO is modulating with a Matrix curve output. Use a bipolar curve for this purpose (the switch is on the back of the Matrix next to the CV connections). The Matrix is always locked down to the sequencer timing, so the output at a given point of the song is guaranteed. However, this method has drawbacks. You are forced to manually draw the curve, which can be time consuming and imprecise, and control over the rate is limited to the number of steps you choose to use, and the Matrix's Resolution knob setting. Also, since the matrix "curve" is really a sequence of steps, rather than a true curve, the end result may sound quite different than what your LFO was providing, and can even introduce pops and clicks into the sound. One possible advantage of this approach, however, is that by selecting an appropriate number of steps in the pattern, the "virtual LFO" can be synchronized to the beat of the song.
Another method is to draw one LFO cycle manually for the parameter value that you want to affect, directly into a sequencer track assigned to the device. You could then carefully repeat this section by copying and pasting the length of your song. This is possibly more time consuming that using a matrix, but may give better results if the curve is drawn carefully (with no step-like snapping).
Note that unlike a true LFO, both these "fixes" are affected by the sequencer's tempo setting.
This issue makes it clear that a "LFO sync gate CV" input on the Subtractor, NN-19, and effect boxes would be very useful. In this way, you could use a note CV signal (any pitch) from the Matrix, or the main sequencer, to synchronize the LFO output at that point. Another thing on the reason wish list would be a few interpolation modes (e.g. step, linear, spline) on the Matrix curve output.
Auditioning Tracks Individually
Your song is like a nice rich creamy Indian curry. Each ingredient needs to be perfect for the final result to be perfect. Listen to each one of your tracks, one by one, for mistakes, unwanted noise, double notes, parts that are too loud, too soft, anything that makes you go, "hmmm." Whether you do this by soloing sequencer tracks, or mixer tracks, is up to you, but be aware that chaining mixer inputs always loop through the output. You may also want to first solo your ReDrum, then solo each sound on your ReDrum one by one. If you tweak anything, make sure to go back and listen to the section again in full context (i.e. all sounds on). The full context, like the whole curry, is all that ultimately matters.
Unused Devices
Scan through your rack, and remove anything that is permanently switched to "bypass." Check to confirm the signal routing remained correct. This also holds for synths and samplers with unconnected outputs, or effect boxes with either no input connections, or output connections, or both.
For PEQ-2 effects, make sure the "B" switch is off if the B gain knob is set to zero. If both gain knobs are set at zero, the EQ can be removed altogether. The same goes for Compressors set with a ratio setting of zero, etc.
Final EQ
Take a break. Walk around the room for a minute listening to nothing (or, if you live in Los Angeles like I do, nothing but sweet police sirens, obnoxious motor bikers, and gunshots). Then listen to your favourite CD, preferably something close in genre to what you are making in Reason, through the same speakers or headphones you are mixing with. (Make sure to turn off all EQ and effects specific to your CD player in both your CD playing software and your audio control panel.) Then go back to Reason and listen to your song again. Does it sound like it came from the same world? Is it too bright, too boomy, or does it sound too much like your aunt's portable AM radio?
If so, consider a final PEQ-2 parametric EQ between your final mix and the hardware interface. Here, you'll usually want a "Q" value at or near minimum, and very subtle adjustments to the gain knob(s). Don't always think in terms of boosting frequencies; consider cutting back on those that are too pronounced instead. Don't go overboard; it's easy to overcompensate with Reason's +/-18dB EQ gain. Listen again. And again.
Clipping
If you don't care about digital clipping or think it sounds "phat," or "warm," or something otherwise unlike complete crap, skip this section. In fact, skip this entire tutorial.
Start at the start, hit play, and stare at the "Audio Out Clipping" LED. If it ever lights up, you know there's a problem. Clipping means that Reason is generating a signal that cannot be properly expressed in a linear, clamped PCM system (which is what your sound card, and also a .wav file, is). By linear, I mean that the string of numbers generated correspond one-to-one with the desired displacement of your speaker at that instant in time. By clamped, I mean that there is a limited range to these numbers.
To borrow a term from the movie, "This is Spinal Tap," digital clipping happens when Reason tells the audio hardware, "play this sound at 11." The audio hardware shrugs and plays it at 10 instead. Thus, what you get is not what you put in, and the result often sounds like wind breaking. The only way to fix it is to bring the signal Reason is generating back within the expressible range. In other words, you need to turn down the volume.
If you notice the clipping is happening at specific parts in the song, say when a particular instrument hits, you may be able to fix the problem entirely by bringing down that particular instrument in the mix. A compressor on that instrument may also be the solution.
If it's happening throughout the song, first try reducing the overall levels in your mixer(s). Turn your monitors or headphones up to compensate for the drop in volume, and then listen again. If you found you've lost some "punch" that the distortion from the clipping was providing, try putting a compressor as a last step before the hardware interface. Start with a midrange ratio value, between 2:1 and 4:1, a high threshold value, and a very low attack and release level. Experiment, and observe how the compressor's gain LEDs react as the audio is being played through. Look out for the distortion that a compressor can create, especially with very low (0-3) attack and release settings. You also have to be careful of introducing undesired "pumping" in the track, with attack and/or release settings that are too high. Also, just because one part of a song sounds perfect with your current compression setting, doesn't mean the entire song does. If this is the case, consider recording changes to the compressor in the main sequencer. Avoid recording changes to the bypass switch, as this can create an audible "tick" when it's switched. Ramp up and down from a zero ratio setting instead.
Be aware that compression reduces the dynamic range (i.e. the difference between the loudest and softest level) of an instrument or mix. Some genres, especially very electronic-based genres are typically heavily compressed, while others much less. Again, experimentation and comparison to professionally recorded songs will help a lot.
The Car Stereo Test
As several people have mentioned on the ReasonStation boards, it is worth writing a .wav of your song to disk, burning it to CD, and listening to the CD on as many other systems as possible. Don't use MP3 or any other lossy compression format to transport your music if you can avoid it; all such schemes change the sound in subtle ways. Unless you are planning to really get serious and process the .wav in some audio software before burning it to CD, export to 16-bit, 44.1 kHz. This is the (only) format for CD digital audio, and if you output to anything else, you are depending on your CD burner software to perform the conversion. Resampling digital audio is a tricky job, and your burner software might not be doing it right. You may want to export out a few different versions, with slightly different compression and EQ settings, taking careful notes regarding which settings are used with each version. This will give you a few immediate listening options and might save you some guesswork later.
When first listening to your CD, again, first set up the system perfectly to listen to the genre you are aiming at, listen to a couple of your favourite tracks, then listen to your own. Listen for the parts that sound good and those that don't. Listen for places where the bass, treble or midrange completely drown out the rest of the song, and take notes.
Car stereos are a good place to start. They are sometimes of very high quality (many people I know have better systems in their cars than in their living rooms). Also, they are acoustically coupled (you're effectively sitting in the subwoofer), and they force specific positions of the speakers relative to the listener's head. Finally, they can often be turned up reasonably loud without bothering too many neighbours. Find a friend who has such a system and ask him to spend a bit of time listening to your CD. Similarly, friends with good quality home stereos can be a big help here (plus you get to impress them with your music making skills!).
Missing Samples
Do not hit "save" or "publish song" on your final version before doing the following:
File menu, "Song Self-Contain Settings...", Check All, OK. Missing samples means wasting your listeners' time.
Check the .rns file's size. If you're planning to upload it to ReasonStation, the maximum file size as of this writing is five megabytes, so if it exceeds that, look for a way to cut it down. Remove unused samples or edit your samples to remove unused sections. If a sample is used only in mono but saved in stereo, convert it to mono in an audio editor. If you're still over the limit, convert your samples to lower sampling rates, starting with the ones that don't have much high-frequency content, such as filtered basses, or ones that are less predominant in your song.
The End and Loop Markers
The end marker ("E") in the sequencer, as far as I can tell, serves only two purposes in Reason. First, it affects the look and feel of the sequencer (the background to the left is illuminated, and you can't scroll past into material to the right of the marker). It does not cause the sequencer to stop at that point, so make sure that you haven't forgotten anything to the right of it, or your listener might end up hearing 300 bars of drum patterns you were working on in the early stages of the song, thinking that it's an intended part of your final product. You can easily check this by zooming the sequencer out, clicking on the end marker and dragging to the right.
More importantly, the end marker determines the range of the "Export Song as Audio File" function. If someone loves your song enough to want to export out a copy to listen to everywhere, they will get everything the sequencer plays from bar one to the end marker. Play your song, and make sure there is nothing sounding past that marker, or else the .wav or .aif file will be clipped at the end. This is especially true for slow fade-outs at the end of songs, or sounds that decay slowly, like string pads. It's good practice to set the end marker a few seconds past the last audible sound, since if the next song the listener has burned to CD or queued up in his MP3 player starts right away, there will be a couple of seconds of silence to separate the tracks.
Is your song going to play forever? Personally, I don't like this, since it forces me to stop the song at the point when I realize that I'm sick of hearing it because I've heard the same section four times. If you insist on saving the song with the loop setting on, set the left marker to the very beginning, and the right marker at the end marker. Make sure it's clear from the song that it's over, and starting again, and not just one long continuous loop. My preference is for loop turned off altogether, or a short loop of silence past the end marker of the song (just to prevent to sequencer from going any further).
(A quick note, the "Films (Gary Numan cover)" song that I posted to ReasonStation does have material to the right of the end marker, which you can hear if you turn off looping, but in this case it's an intentional inside joke (cookie) for Gary Numan fans.)
The Rack Layout
Are your devices arranged, collapsed and expanded in the way you want to present to the world? A quick look at the aesthetics of your rack might be in order. Some people like to simply collapse everything; you can do this very quickly by Alt/Command-clicking on the triangle on the rack left of any expanded device. I personally like to minimally keep the main mixer(s) open, so the listener gets a bit of interesting visual feedback when playing the song. In addition, labels on devices help a listener study your song. Alternately, you can use the device labels and sequencer track names to be creative and expressive, and some artists on ReasonStation have done. Secret messages, hidden cookies... There is a lot of potential for fun tidbits in Reason.
The Song Info
Who are you? What is this song about? Fill in the song info field. Is this song yours? A cover? Copyrighted? Unfortunately, you only have 255 characters, so keep it simple. Check all the fields to make sure they say what you want. If you really want to be cool, include an original splash screen image. The "Show Splash Screen on Startup" checkbox can be on or off at your whim. Some may find it a bit annoying; others (like me) think it's a great way to "get into" a song by seeing a cool picture or some informative text pop up before hitting "play."
The Final Tests
Are you going to share your song as an .rns, or .rps? This is a personal choice. Read the Reason documentation on what .rns and .rps allow someone to do with your song. Note that this has changed slightly from Reason 1.0 to Reason 1.01. Whatever you decide, make sure to keep an .rns copy! The protection in an .rps file can't be undone by anyone, not even you, the author. Whatever you decide, note that manually changing the filename extension not only doesn't provide any protection, but may cause Reason to report a "bad format" error when trying to open the file. (I've downloaded song files that had a "bad format," and changing the file extension corrected the problem.)
Give the file a decent filename! Something like "song.rps" is not a good example. Be creative! Some people on ReasonStation like it if your RS login name shows up at the start of your filename; this is also a nice way to have all your songs easily found on your listeners' hard disk. Consider it.
Exit Reason, start Reason again, and load up your newly saved file, as if you were a listener opening it up for the first time. Does everything seem OK? Hit play. Is everything in place?
Everything's fine? Then post the song to ReasonStation! Pick an appropriate genre, edit the sub-genre if you want, and above all, talk about the song in the song info. What are we about to hear? Try to stir up some excitement by being specific; don't just say "my new song." And please, please avoid ALL CAPS in any text you post, including the song title.
Finally, just to make sure the upload went properly, download your song from ReasonStation and load it into Reason.
Summary
Now that you've read the details, here's a brief summary that can be used as a quick checklist:
Check that your Master Tune setting is at zero cents.
Make sure the song sounds the same every time it's played.
Listen to each track individually for problems.
Make sure all devices are actually used in the song.
Consider using a final EQ to even out frequencies in the mix.
Check for and avoid clipping.
Burn a CD of your work in progress and listen to it elsewhere.
Ensure all samples are included in the song file.
Set the End and Loop markers appropriately.
Make the rack layout and device labels look nice. Or not.
Edit the song info and splash screen.
Decide on .rns or .rps, publish it, and download to test.
Any errors, omissions, or other comments about this tutorial can be sent to me, BlackWorm, either as a private message on ReasonStation or at the e-mail address indicated in any of my published songs. Questions can also further be posted to the "Tips and Tricks" message board.