The Ballad of 80’s Cheesecake.

I’m surprised that cassettes are still ‘a thing’. The meaning they had back when there were few alternatives is different to the meaning they have now, where they are used to refute alternative delivery formats. We just used to see them as ‘making do’. What hasn’t changed is that cassettes are, if not disposable, a place for sketches and ideas that can always be erased if they don’t work out.Terse 17 Eighties Cheesecake

The longevity of Eighties Cheesecake is a bit of an issue. First released in 1982, it was one of two cassettes that had a large overlap of material expressed differently. Snappy Carrion was the more pop of the two. Both had long stretches of minimal sounds and a clutch of drum machine driven tunes. Around 1984-5 some of the pop tunes were put on the B-sides of vinyl. The expected life cycle of the music was a couple of years maximum.snappy carrion

The vinyl sparked interest in the tapes, which started to appear on eBay in 200x for too much money. I was able to make CDs that cost far less and make sure the music stayed cheap. Good in some ways, bad in others. A CD had to lose most of the sound experiments – fair seeing as they were ‘cassette ideals’. Also it required some hard work on improving the sound so that the hiss wasn’t the main feature.

And now in 2014, vinyl. It seems the fates didn’t like our ‘cassette ideals’.

After a bit of listening and soul searching the process is now under way. You may already know that we have a remarkable digital library of nearly all our recordings due to a SONY boss man visit in 1985 (an anecdote I’ll tell again later). So I have various transfers of the tracks off open reel and cassette made back at the time. When making the CD I used a little noise reduction and a little EQ and that was it, CD is quite forgiving.

Vinyl brings more concerns: about 22 minutes a side, so the running order has to be fussed. Poor stereo at the bass end, distortion at the treble, a cloudiness which people think is ‘warmth’.

If you’re going to remove tracks then it’s worth sacking every one of them and making them reapply for their jobs. That’s happening. It’s worth asking if the 3 minutes of dogs barking could be expressed more succinctly. That’s happening. It’s worth checking if there was anything that really ought be there for the first time and yes, we have some candidates.

Working with old tape sound.

First step is to load the track into the editor and find any damage. If it’s a digital drop out then it’s usually easy to find a few milliseconds from some other part of the recording and paste it into the spot. Tape drop outs are harder, it could be that you leave them alone as a artefact of life, but in a few places they can hand painted to e.g. raise the level a bit.

Now you might try reduce the noise level, but it’s very likely going to screw with the overall sound, making it more metallic and expanded. You can make things better if you gently repeat a little treatment, carefully judging the result each time. For the vinyl I’m banking that surface noise is going to be worse anyway, and I’m leaving the background noise alone. Removing rumble below about 30Hz makes sense as the tape never had signal there anyway, and surprisingly make the bass seem deeper.

Unless there’s a strikingly obvious EQ problem (like Dolby B), the next step is multi-band compression, particularly at the bass. The kick drum has a kick and a boom. The boom makes the drum sound flabby when it’s slurred by the tape recording. Compressing at the frequency of the kick with an attack keeps the impact, tightens the boom and makes the bass melody distinct. Same goes for the synthetic snares which I used to make way too loud in 1982. Compression makes them seem just as loud but everything else is heard clearly as well.

Up top you might need to raise the volume of the high hats which have been lost in analogue dubbing. But then again compression to avoid the vinyl splattering the tail end of the sounds. The playback should seem ‘fresher’ and ‘less woofy’ at this stage, and other problems should be revealed that need a little detailing.

In some cases you’re still hearing mud between the kick and the bass melody. That’s very likely solved by reducing the stereo separation below a frequency where the kick crosses with the bass. Make the kick mono, let the upper end of the bass be a bit wider than that, let the mids be wide.

Next big step is exciting! With an exciter! What this does is synthesise extra harmonics from the existing material, quite a different result to EQ which raises a band of frequencies. It may seem artificial but it’s what tube amplifiers do – the tubes bring back harmonics lost in dubbing and mixing. The exciter is pretty much that, but with adjustments for the specific frequencies that you’re trying to address. For example a voice that’s muffled can be brought out by some attention at 1-2kHz.

Old Tracks for New.

So how is that some tracks may end up on the vinyl* that have never been heard before? Throughout the archive are abandoned recordings. Some have bad drop outs, really bad mixes, didn’t get finished, or had some really bad idea permanently burned in. The latter is the problem here – in 1982 I started a couple of interesting recordings and then added a really bad element that ruined them.

Fixing this is hard and it wasn’t possible up to this point. You have to cut out the bad bits, remake the ‘bed’ from the good bits, rediscover the parts then were ‘under’ the bad bit and re-layer them while cross checking the match with the original. For 80s Cheesecake I had to go through all the cassettes from which I had taken samples, locate the right phrase, place it again. I could only do that after digitising all the cassettes I had in 1982 – which I did a few years ago.

The tracks are entirely made up of the original sounds, which is good because they’re based on equipment and ideas that are no longer available to me. One potential track called Kai Kai is based on Juno 6 synthesizer being controlled by a micro-composer, both long gone. It sounds just the same – with out the horrible vocal backing I’d added at the last moment!

Eighties Cheesecake is to be re-issued on vinyl in 2014. Stay tuned.

* Currently the track listing has not been finalised.

3 thoughts on “The Ballad of 80’s Cheesecake.”

  1. External comment and reply:

    very interesting read as i’m currently doing just about the same thing, running into very similar problems and coming up with more or less similar solutions…my biggest issue is total absence of low end, and finding a balance between hiss, high treble to open up the recordings a little and hihats that are recorded at just that frequency that hurts your ears

    Well if there’s no bass there’s no bass, maybe that’s the truth of the recording. But there’s a plan:

    Kill everything below the 30Hz flabby area, ain’t much there particularly if it’s tape.
    Find the spot where kick and bass are overlapping, use phase to try tease them apart (mono the kick). Phase is terrible for low frequencies.
    Squish down the flab with compressor gate.

    OK, so if you raise the bass level, nothing there?
    Going to have to synthesise it with (or something cheaper).
    Or actually remake it. Yeah I know, but when well done the result is so satisfying. An 808 kick is still an 808 kick and the frequencies it makes below a certain point can be reproduced and painted back in.
    Example = A fixed up live track where the bass wasn’t there.

    As for treble, the annoying frequencies are around 1.5-2.5KHz, while air starts at 4KHz up. Maybe set a de-esser at the high hats and then dial in a gentle high band EQ rolling down from 4KHz.

    1. The kick has a boom that lasts a while. It may last long enough to subtract from other instruments using the same frequency range by phase cancellation. A good kick does its job, then goes away. Best done by compression at that frequency range, with an attack so that the first part of the sound is left alone.

Comments are closed.