Enjoying my new monitors I thought I’d take a break from the heavy involvement in an album project I’m working on and go back to revisit a track I never got right in the first place. Rewind back to 2003 and we have a song called Vulcanella Self Destruct! Made as an entry to a Triple J remix competition, I took Adalita’s (from Magic Dirt) vocals plus the songs chords and built a lumbering industrial wasteland groove around it. I didn’t win the competition, but I liked the cathartic dirge the song ended up becoming - the only trouble was that I never got the mix right with the limitations I had at the time. Now that I have an accurate mixing/mastering setup I can finally get the beast to behave how my ears always wanted it to. So as a point of comparison I’d firstly like to share the original 2003 mix: listen at your own peril! - get it here. What follows is the new definitive version…
I feel it necessary to discuss some technical issues here, so if you find this stuff boring read on to the links and grab the songs. To start things off I upped the entire mix from 32bit 44.1khz sampling to 32bit 96khz sampling, which is what all my projects have been in for the last year or so. The higher sample rate brings new life and vitality to some rather ‘cheap’ sounds and effects that I used in the original, even though it’s a pretty noisy song. This allowed me to really give everything a good polish and take my time with the monitors to get the placement and tone of everything just right. Then again at 96khz I made up a mastering chain of good plugins, the best stuff I got at the moment. So my 96khz mix and master was complete, and I was happy enough with it. However, the trouble with using 96khz is that you have to at the end of the process get the song back down to 16bit 44.1khz sampling for mp3 and CD playback. I’ve been doing this step in the work using a pure digital process: i.e. mathematically using software to upsample then downsample the audio to get the sample rate remotely representing the original high quality mix, and a process called dithering to move down from 32bit to 16bit. Despite the quality of such digital processes the mathematics isn’t perfect and the sound is slightly compromised as a result. It’s good enough for what I’ve been doing so far, but I’ve wondered how I can get around this downsampling/dithering problem.
This is where my new reel to reel tape deck comes into the picture! I set up the Tascam BR-20T with some high quality tape to record my 32bit 96khz master audio file. The soundcard I’m using is the M-Audio Delta 1010 (which isn’t jitter free, see more on jitter for an entirely other technical issues) connected with balanced cables to the tape deck. The BR-20T (pictured above) is a last generation mastering tape deck and the specs are pretty fantastic (e.g. low noise) - I used the IEC EQ profile which is an industry standard for tape. After fiddling a bit to get everything calibrated I recorded a clean take of my high quality mix and checked it on headphones. After rewinding I opened up a new blank file on my computer set to 16bit 44.1khz sampling. I hit record on the computer and played the recording off the tape - so I was recording the analog tape sound back into digital sampling. This was so I didn’t have to do any editing on the digital file (aside from top and tail edits) and I didn’t need to do anything to it that would muck with the sound digitally after the recording. Playing back the end result was quite magical. As some of you may know, tape introduces all sorts of magic to audio - it changes the dynamics in nice ways; it smooths harsh sounds; and it makes anything in the mix that has an organic quality to it (like vocals for example) sound warmer and real. On the other hand the EQ profile and limits of the tape format plainly showed that I had lost some of the crispness in the high end frequencies, and had lost some of the sub frequencies. So if I felt I really needed those parts of the sound I’d have to be mindful that the tape looses those aspects a bit at the benefit of producing a really lovely organic sound. Already I’m thinking that tape would be appropriate for some genres and styles of sound, and for others not so much.
I decided that I’d let people decide for themselves what they think of the difference. So I have made up not one, but two versions of the new mix/master of Vulcanella Self Destruct. The first is the ‘regular digital method’ of using the software to do the downsampling and dithering; the second is the tape mastering version. For web, I’ve made them into 192kbps mp3s, which you can download here:
I’m interested to see what you think of the comparison. Can you pick the difference? Which do you prefer? I’m also interested to see what you think of the song. It’s certainly nothing I would ‘compose’ these days - definitely represents a different era of headspace. But I do think now the idea is properly ‘represented’ for what its worth. Give us a shout if you like what you hear.
Incidentally, Mick Rippon has been testing some new web software for my mastering website that allows for streaming A-B comparrisons. We thought we’d test it out on the two versions of this song. You can try it out here!