♦ Recording & Production
Created with the new Korg KRONOS workstation.
This lengthy, detailed description may only be of interest to musicians and recording geeks, but here it is!
The original song was written in 1986 two days after the event (see The Story of the Song), and my partner David Alvey and I recorded a demo of it then at Dave’s home studio - it was recorded on his Otari MX-5050 mkIII-8 - an 8 track tape recorder utilizing 1/2” tape and a fairly high-quality unit for the time. Given this was 25 years ago, I used a Yamaha DX7-II for most of the keyboard parts with the exception of an overdub from my Yamaha C-7 grand piano (recorded to tape at my house and flown in to the 8 track over at Dave’s), an Oberheim DMX drum machine, a LinnDrum, and Dave’s Synclavier for some very cool sound effects that sound like deep sweeps of “space noise” (for lack of a better comparison). Dave played lead and rhythm guitars, we both sang background vocals, and I sang the lead. Since we only had 8 tracks to work with, there were many comps and bounces of various tracks into 2-track stereo submixes.
Getting the original analog tape transferred to digital format
So first, in thinking about redoing the song, I wanted to get the original demo tracks from the analog tape as a starting point, transfer them to digital format, and maybe use some of them in the new version I would be producing on my Mac, using Digital Performer. Dave still owned the Otari (pretty amazing in itself after 25 years, given all the gear we’ve bought and sold over the years), but Dave’s entire home studio was in storage, given he was in the middle of a massive remodeling of his house.
But first up, we had to locate the master tape. We could find no masters labeled “Challenger” between us - and we both are packrats who have diligently saved every possible tape and master from everything we’ve ever done - so that was the first worry. But Dave had one reel of 1/2” tape labelled “Technicolor” (the name under which we were working on our other project years ago) - and that was the only thing that seemed promising. But it had no track sheets, no notes - nothing.
Then, he had to dig the Otari out of storage (it’s not small) and bring it over to my studio for us to be able to check the tape. The unit still seemed to work (after being in storage for about 10 years), and after a thorough head-cleaning and demagnetization process, I started playing the “Technicolor” tape. It turned out to be a work tape of some other songs we were working on in the studio at that time, just experiments, and after 7 songs it was seeming kind of unlikely that the song was on the tape. But wouldn’t you know it - right at the end I located the Challenger song.
The problem was, the tape was playing terribly - slowing down, speeding up, making weird noises...at first I thought the Otari maybe needed servicing, but further research revealed that the problem was likely something known as SSS - “sticky shed syndrome” - a fairly common occurrence with analog magnetic tapes that are that old, where the tape actually starts decomposing and becomes sticky with pieces flaking off - and cannot be played unless it is scientifically baked in an oven for a certain period of time.
Searching on the Internet for information about tape-baking, I came across a great article by a guy named Steve Puntolillo - which led me to Sonicraft A2DX Lab - his company that specializes in transferring old analog tapes to digital format - baking them if necessary, doing whatever it takes to get them transferred (see image of Sonicraft scientific tape-baking oven at right). I quickly realized that this was the answer to my problems - he had the capability to transfer just about every tape format known to man including 1/2” 8-track to 48k 24-bit digital audio - and better yet, he was in NJ about an hour south of me. I quickly set the job up with him, Dave drove the tape down there (didn’t want to chance losing it in transit), they baked it, and I soon had the audio tracks back, perfectly transferred (DBX noise reduction and all), and set up in Digital Performer on the Mac.
What to keep and what to do over?
Next, I had to decide what, if anything, I wanted to use from the original demo. Being only 8 tracks, nearly everything was bounced down into stereo 2-track submixes (i.e. drums and bass together, all keyboards together with background vocals, all guitars mixed together, etc.) I decided to use the original lead vocal I sang 25 years ago, given that I was younger then and my voice was in top shape from constantly performing, which I no longer do much as a music software designer. Everything else would be re-sequenced and re-recorded, using the original tracks as a reference.
I also specifically wanted to use the “space whooshes” from the Synclavier, (see image at right) which is an unbelievably cool sound created by taking cymbal crashes sampled at 192 kHz and tuning them down several octaves, along with some pitch envelopes. Because of the super-high sampling rate, even if you tune them down two octaves, you’ve still got effectively a 48k frequency range, and there is all of this cool airy high-end noise that you simply don't get taking a 48k cymbal crash on a regular workstation and dropping two octaves - at which point you’ve only got 12k of high frequencies - believe me, I tried that.
The problem was, the space whooshes were mixed in with the drum sounds in a stereo submix on the original tracks - no way to isolate the sound. And Dave no longer had the Synclavier. I remembered that I sampled them into my Fairlight Series III many years ago (which I still have, see image of myself to right from 1987), but it hadn’t been powered on in about 10 years. I did that, and it fired up like a champ - but the sounds were not on the hard drive, and it turned out that the tape streamer (which is how you stored all the Fairlight sounds in those days) was eating tapes. No way to get it quickly repaired and get the sounds that way.
I then remembered that we used those same sound effects on a 24-track 2” master tape we were working on in a studio around the same year, in a different song - and they were isolated on separate tracks on that tape. Which prompted another tape being driven down to Steve at Sonicraft, more tape-baking and cleaning, before I had the tracks from that tape back in digital format and could extract those very cool sounds and add them back into the mix on audio tracks.
Getting old MIDI tracks from an Apple II+
Meanwhile, I decided to see if there were any existing MIDI tracks that I could import, rather than redo the playing of the keyboard parts - not that I couldn’t recreate them, but more from the idea of trying to get the original vibe and performance. However, it was 25 years ago that they were sequenced. At that time, I was using an Apple II+ as a computer, and software from Music Digital named “Studio II”.
I dug through my attic, and uncovered my 30-year old Apple II+ (which I had kept all these years specifically because it was used for all of our Technicolor projects in 1985/1986, and all of the 5.25” floppy diskettes storing the data and programs). I remembered throwing all of that stuff, and the manuals for various programs into a box and deciding I would keep it. Unbelievably, after bringing the computer, 2 disk drives and CRT down to the studio (the computer actually had dead leaves inside it and around it, I don’t know where it was stored for some of that time, probably my garage), and downloading from the internet Apple II manuals so that I had an idea of how it was supposed to work - it powered up immediately and loaded the program from the disk drive! 25 years later and it still worked flawlessly - kudos, Apple!
But the computer keyboard seemed non-functional - only a few keys worked. I used a can of compressed air both on the keyboard itself, and inside the computer on the underside of the keyboard assembly, and more keys started to work. And as it turned out, with repeated workings of the keys, eventually they all started to function again. Pretty unbelievable!
I found a data disk labelled “Challenger”, and from that I was able to load a song file into the Studio II program - an adventure in itself, re-learning 25 year old computer software (driven by an upper-case only keyboard) that bears no resemblance to any modern graphical mouse-driven software of the last 20 years or so. The only existing file I could locate was the bass part for the song (I swear there were others, but I never did find them). Unbelievably again, the ancient Passport MIDI card inside the computer functioned perfectly - I was able to set it up to play the track, send the MIDI data out, send MIDI clock, I could synchronize Digital Performer to it, and I recorded the bass line into the project. And, with minor tweaks, the bass part exists as it did 25 years ago. That’s pretty cool.
Sequencing and recording the new parts
As part of my job, I work with the Japanese keyboard company Korg as a developer, having licensed them my KARMA algorithmic music technology, and I happened to have a prototype of a brand new unreleased keyboard workstation named the KRONOS here at my studio. This is an amazingly powerful keyboard with nine different synthesizer engines, including one named MOD-7 that would allow me to emulate some of the original DX7 sounds that I used on the original song. I wanted to keep some of the flavor of the original, while adding more modern sounds and production techniques. I decided that I would try to create all of the musical tracks of the project with the KRONOS alone (other than real guitars and vocals).
So I began about a month of sequencing all of the parts (except the bass line) from scratch, using the original tracks as reference guides. All of the sequencing was initially done in Digital Performer on a MacPro Quad, since I’m addicted to using a computer DAW with a 30” monitor. I selected a new song in the KRONOS sequencer mode to be my 16-track song setup, where I would select the various programs the MIDI would play, and assign various effects to the parts.
Click to see full-size version of Digital Performer project
I began with the body of the song, leaving the intro for later, as I had some ideas for extending it from what the original was. I was able to greatly improve on the original drums (which were done on 8-bit drum machines) with the KRONOS’s HD-1 stereo drum kits and percussion programs, giving me many more options than I had 25 years ago - I particularly like the addition of the “orchestral snare” phrase that shows up every now and then, giving it a slightly military touch, along with the timbale, congas and other percussion parts.
Click to see the MIDI Prog/Mixer and Mod-7 Patch Panel pages of the KRONOS
Using the KRONOS’s MOD-7 engine, I was able to import some of my original DX7-II programs and layer them with new sounds, staying fairly true to the original DX7-II piano layer that was the main keyboard phrase, and the icy pad sound layer, which I modernized slightly by chopping it with KARMA. Other keyboard parts were provided by the AL-1, Polysix, and STR-1 engines of the KRONOS - and there’s a subtle strumming acoustic guitar part generated by KARMA throughout the song. Speaking of KARMA, I added a panning delayed mallet-sweep glissando right before the third chorus that works well with the visuals I later created for the video, which also appears twice in the end section.
Recording Guitars and Backing Vocals
Once most of the song was fleshed out, it was time to record the guitars. We spent several days recording Dave’s clean and muted guitar parts, all of which were doubled left and right, sometimes with octaves or different chord inversions, following a sort of “question and answer” approach between the guitars, keyboards and vocal that was a trademark of the sound we were trying to develop back in the 80’s. We took a direct feed out of Dave’s Mesa Boogie amp, with a resistor strapped across the speaker terminals to fool the amp into thinking it was driving an 8 ohm load so as not to blow the output transformer. For the rhythm guitars we dialed in a slightly overdriven crunchy clean sound and Dave used his Strat. Ultimately, we recorded about 10 tracks of clean and muted guitars, using a technique in the choruses of overlapping the phrases on different tracks so the notes could ring out over each other in a way that is impossible to actually play, to get more of a bell-like effect.
Then it was on to the distorted lead guitars. Dave added a Danelectro FAB Metal stomp box to the input of the amp to provide that warm, crunchy metal distorted sound (well, it became warm after a lot of the high frequencies were rolled off), and we recorded the solo section and end sections using Dave’s Ibanez RG350. I had an idea to do sort of a Brian Mays/Queen orchestrated guitar harmony part on the final phrase of the solo section, and we recorded 5 tracks of harmony there, spread somewhat across the stereo spectrum, which then fades out in a lovely suspension when the vocals come back in.
After that, we attacked the background vocals. I wanted to get a much huger sound than we had originally done (we had probably 3 background vocal parts in the original, doubled left and right), so I recorded, with an AKG C414 B-ULS microphone, about 20 tracks of myself singing the mid and high parts (10 left/10 right doubled and quadrupled) and at least 12 more tracks of Dave singing the mid and low parts, again doubled/quadrupled left/right - and it sounds fairly massive at that!
Creating the Ending
While working on the parts, I was sending Dave rough mixes for comments and suggestions, and Dave had a very cool idea for the ending: instead of just fading the song as we originally did, we should fade out into some “space sounds” and telemetry beeping, possibly some audio of someone trying to contact the Challenger...and with that in mind, I went online and located some space static recorded by the Cassini Project, some cool recordings of “Earth Auroral Kilometric Radiation”, a recording of a shuttle rocket launch and other audio elements that ultimately were fused into the end sequence. There was no audio existing of someone trying to contact the Challenger, and I didn’t want to make something up, so I came up with the idea of using the last communication of Mission Control and the Challenger, and repeating it as it fades out - because the idea of the end sequence and the visuals of the video are that the Challenger goes on, still moving upwards, onwards and outwards into the Universe, still exploring. There’s also a coded message in there, for anyone clever enough to find it.
Back to the Intro
I had left the intro for last - the original intro began with simply the ascending pipe organ chords, and then right into the song - but I wanted to extend the intro somewhat, having the idea (after working on the ending) to incorporate the actual NASA Mission Control recording of the liftoff sequence into it, along with related visuals. So in the intro, you’ll hear some wave-sequencing from the KRONOS HD-1 engine, some panning left and right metallic sounds generated by KARMA, the pipe organ climb with low synthesized bass notes under it, shuttle rocket engines, a KARMA-generated harp glissando as the pipe organ climbs, and then an abrupt chop as it segues into the main body of the song.
Getting it all into the KRONOS
With all of the tracks completed, I decided I wanted to put everything, all of the completed MIDI and audio tracks into the KRONOS, mix and master it in the KRONOS using the KRONOS’s effects, and bounce it to a finished 24-bit WAVE file on the KRONOS as well - so that ultimately, you could load the song in the KRONOS, push the play button, and have the whole song right there. In doing so, I had hoped Korg might consider using it as a demo sequence, but I think they felt that the subject matter wasn’t appropriate for that use - and maybe they are correct.
Also, it’s worth pointing out that this song was not, at all, written as a “KRONOS Demo” that showcases the various programs, synth engines, and “amazing sounds from each engine” - not at all. I picked sounds for use that I felt were appropriate for the creation of a real song, in the style I desired, to capture the message I wanted to deliver. But ultimately, it’s pretty amazing that all of this is playing out of one keyboard workstation, when you press the play button. Perhaps I’ll make the files available later when the KRONOS is actually released.
So even though I used Digital Performer to record the guitars and vocals (there were approximately 30 tracks of background vocals and 20 tracks of clean and distorted guitars, although it may not sound like it in the end mix), and recorded some of the keyboard parts to audio tracks as well since I had more than 16 MIDI tracks, all of the 24-bit 48k audio tracks were bounced down to multiple two-track stereo WAV file submixes: clean guitars L/R, muted guitars L/R, lead guitars L/R, E.Piano L/R, backing vocals L/R and so on. And 16 tracks of MIDI data to be played by the KRONOS were exported in Standard MIDI File format to a MID file.
Click to see the Audio Track/Mixer and Track Edit pages of the KRONOS
Then, it was simply a matter of transferring the WAV files and MID file to the KRONOS hard disk, loading the MID file into my song setup, loading the WAV files into regions with the Track Edit > Region Edit dialog, and using Audio Event Edit to insert the regions into the 16 audio tracks of the KRONOS, finding the proper start points.
After that came the time-consuming job of assigning various appropriate effects to the audio tracks, along with the effects on the MIDI tracks (KRONOS has 12 Insert Effects that can be applied to any of the MIDI and audio tracks), the two Master Effects and two Total effects, and mixing the song.
Click to see the effects pages of the KRONOS
There’s a 3-band EQ with sweepable mid-range band on each track, which comes in very handy during the mixing process. Once I was relatively happy with the mix (you can never be certain it can’t be better), I used the KRONOS “Bounce All Tracks To Disk” feature to record it to a 24-bit 48k WAV file - and the audio was done!
Click to see the eq and audio (bounce to disk) pages of the KRONOS
Producing the Video
During the work on the sequencing of the song, I was also spending long hours every evening searching the Internet for images and video related to the Challenger disaster, and other Challenger missions, and the space program, and space images, and starting to formulate ideas for a video screen play. You can find links to some of the NASA databases and other sources of images I used on the Resources & Links page.
In addition to sourcing video and images on the Internet, I purchased several DVD sets of official NASA footage of the mission and astronauts, trying to locate the best quality footage (so much of what you can find on the internet has already been compressed, and re-compressed to the point of being horrible quality). Ever image and video clip was documented in a spreadsheet with where I found it, so that I could go back if needed and find it again, or search for more at the same places. But ultimately, I ended up with far more than I could possibly use, and so much fantastic (and humbling) material that taught me so much more about the mission than I already knew or remembered. I also decided not to limit myself to only images of the Challenger or STS-51L, but to use other images and (better quality) video from other shuttle missions and space missions to enhance the experience, with the idea that all astronauts are “Challengers”.
I was able to collect enough imagery of each of the seven astronauts that my video screenplay eventually began to coalesce around the idea of giving tribute to each member in a separate section, which ended up working very well with the lyrics in the verses about “What makes a man go against the odds…” and “What makes a woman abandon fear…” and so on.
The video was produced with Apple’s Final Cut Pro Studio (v6). The project was set up as HD 720p (1280 x 720). The first step was to import and layout the stereo 24-bit 48k final mix of the audio on the timeline, then import all of the video and image assets and start laying them out above the audio.
Click to see full-size version of Final Cut Pro project
In the end, about 3 weeks went into the editing of the video, trying to find the best matches of the imagery to the music and lyrics. I wrestled with whether or not to show the actual explosion (you can read more about that here), finally deciding that it needed to be shown, and matching it up with the climax at the end of the third chorus, where it could be shown briefly and then fade to black for a “moment of silence”.
The final movie was output using Apple’s Compressor software, with settings of QuickTime H.264 video with AAC 44.1 320 kbs const bit rate, highest quality on everything (takes several hours to render a 6 minute video) - but recommended for getting the highest resulting quality when uploaded to YouTube and they compress it all - and it does look pretty darn good on YouTube, but nowhere near the quality of the final 1 GB .mov file!
If you read all of that, congratulations! You are a certified keyboardist, musician or recording geek - and I mean that in the best way possible!
Many people have said the song sounds “80’s” - well, it was written in 1986 two days after the event, and I stayed fairly true to the original sounds and production ideas we used in the original demo, so I suppose that makes sense. There were good 80’s and bad 80’s - I hope that ultimately it resembles the good 80’s.
- Stephen Kay, January 2011