The Artist’s Statement of Which Heart Rate Zone You Are In
You are playing an NES game with a friend or family member. The game is encouraging you to keep and maintain a target heart rate zone. You might be doing this on the Power Pad. This is the last all-out intensity push on the last level of the game. Don’t give up and get your heart rate in the zone!
That’s right! DJ Tent Mode’s* Zoneful new track drops right now. Drop everything, give yourself 20 pushups because you’re doing it for yourself not for me! Get in the Zone!
You can download an MP3 of “Get In the Zone” here.
You can download an NSF of “Get In the Zone” here. The NSF is a ROM that will execute on NES hardware, in an NES emulator, or in an NES music player. I like to use GaMBi on my phone.
*No, I still haven’t picked a better name to brand my music projects.
Another NES Technical Primer
The NES supports 5 simultaneous sound channels, 2 Squares, 1 Triangle, a Noise, and a PCM. PCM is “Pulse Code Modulation”. It’s a way of digitally describing an arbitrary waveform. “.WAV” files are PCM. Being specific, the NES can do both PCM and 1-bit DPCM. PCM is more computationally expensive and often required the NES to stop doing much of anything except play back the waveform. Differential PCM is a short-hand way of encoding PCM data which is lossier in compression, but much easier to unpack and play back. Being completely honest, I’m not sure if DeffleMask is using DPCM or PCM to encode its samples, but that’s probably not important.
There were two common applications of the PCM “Back In The Day”. One was for digital speech, like the title screen of Blades of Steel or Double Dribble. The other was for adding texture to music often supplementing the noise channel. Konami is famous for this and you can hear poppy drums in games like Teenage Mutant Ninja Turtles or The Adventures of Bayou Billy. Super Mario Bros. 3’s steel drums and bongos are PCM. Many of your favorite NES soundtracks utilize the PCM channel in some way. Many of them - like the MegaMan series - ignore its existence entirely.
In terms of how DefleMask deals with it, you more or less have arbitrary total sample space, but you have only have 1 bank loaded at a time. You have to send a specific tracker command to switch banks if you want to change sample banks.
Each sample bank can contain 12 samples and are assigned to a different key on the piano. A chromatic scale has 12 distinct tones so each bank corresponds to one scale - which octave you play in doesn’t matter. So if you play a C in my sample channel, you get a “Get in the zone”, C# is “Zone”, D is “One”, etc.
The “Get in the Zone” NSF ROM is about 104 KB. When compared to MegaMan 2’s 257 KB size is still a pretty exorbitant at 40% of the space for a late 80s NES game, but not technically impossible.
This Month’s Prompt
Like others, I immediately thought “Sonic the Hedgehog music” and decided against it. DeffleMask does support the Genesis, but it’s a very different system with substantially different strengths and weaknesses and I’m not practiced with it yet. I also considered doing a Master System Sonic zone which is much more similar to the NES’ sound architecture, but I don’t have the familiarity with that system or the 8-bit Sonic games, so I went in a different direction. I do a ton of road biking and knowing what my heart zone is at is something I can talk about so the song became an exercise anthem.
This Month’s Process
This month, I tried to integrate PCM samples into my NES songs. It’s a powerful tool, and it’s also difficult to use in 2022. Now that digital space constraints are non-existent by 1980s standards, sampled sound quickly sounds anachronistic. Additionally, this “Behind the Code” series posits that there was a bug in some developers’ audio toolchain for NES that flipped DPCM samples on a per-byte basis so they all sounded way worse than they could of if they were properly implemented. But since they are famous samples, its what we associate with the capabilities of the NES even though it’s probably inaccurate.
I originally got some free use drum kit sound effects and imported them into my DefleMask track and realized that they sounded way too clean. I was going to mirror my noise drums and it stopped sounding like an NES song. It sounded like someone playing a drum kit on top of an NES song. It just wasn’t the aesthetic I was going for.
So I scrapped that idea and headed towards creating my own samples. I was going to have these mixed in with the PCM drum line but this is the only PCM portion that is left in the song. I recorded them with my PlayStation headphones (please don’t tell Doug Bowser or whoever runs Sony) into Audacity and had a similar problem - they sounded too good in DefleMask. So I cranked the sampling rate down from 44KHz to 16Kh, and crushed it down to 8-bit PCM on export. If you’re not a digital signal nerd, sample rate is how frequently the audio information is recorded, and how many bits is how many different values you get to use to describe the volume.
After re-doing the samples I had forgotten to reassign my sample patterns and I got surprised when the drums were replaced by my own voice. Please click here to listen to what happens when sample tracks go awry.
Much like how the Triangle channel does not have volume control, the PCM channel does not have volume control either. The samples are imported at whatever volume they are recorded at and you need to now normalize the entire song against two static volumes. This month’s piece doesn’t have as much dynamic range as others I’ve worked on as a consequence. It’s just a pair of trumpets blasting in your ears at volume B (out of 0-F). I think it ended up working OK for the kind of marching band anthem the song is.
Another mixing trick I read about years and years ago and think about a lot is from an interview I read citing Butch Vig. Vig was producer on Nevermind, Gish, and other huge 90s alt-rock albums as well as band member and producer of Garbage. He said if you mix the vocals down a bit with respect to the instrumentation, this causes the listener to instinctively turn up the volume to hear the lyrics. This ultimately makes the song sound louder and thus fuller because you get more dynamic range in the piece. I struggled with whether or not to re-sample the “heart” and “beat” samples to bring them up in volume and ultimately left them under the instrumentation because I think they worked better as vocal texture. It certainly caused me to turn up my volume when making the track.
Musically, I’m pretty happy with how the piece is structured and the harmonies. The sustained 2nd interval (the one that makes your ears prick up in the opening anthem) was challenging to resolve musically for me. I was writing the 2nd part and kept on getting chords that didn’t sound quite right. I ended up with that and its a pretty heavy suspension, but I really like how attention grabbing it is.
It’s also the first song I’ve done that executes a key change. The final verse of the piece is the part of the track that I’m the happiest with. The callbacks that are made in both rhythm and pitch to the opening anthem and the previous verses tie the whole thing together well.
The coda didn’t come together until the very end and this is the first track I’ve written that isn’t intended to loop. The final “Get In The Zone” solo sample is a fun punchy mic drop.
Final Thoughts
I don’t know if I can say that I completed my goal of having NES-style PCM drums, but I learned a lot about samples themselves and the process. Also, huge props to any vocalist that can “commit to the bit”. It’s not something that I practice and I’m not sure if the flutter in my enunciation of “zone” sounds very confident. This is probably something a vocal coach or an audio director would have had me re-record, but it’s certainly how I performed that piece!
