mixing and mastering from very old tapes

This was a myspace post that we put up after the album 'Thirty Three' was finished. It's a tech-fest supreme with a generous side dish of lightly fried analogue tape. Read carefully; we'll be testing you on it later......!

Thirty Three was a mission of a project to re-mix. I went straight from this to mixing Comus' first live festival appearance in 35 years for a DVD release. Different musics, different recording technologies, but both referring inexorably back to the late '60s / early '70s .....

The Red Square material dates back to 1974 and the majority of the tracks were either recorded on a Akai 4000DS reel to reel tape deck using 7" Scotch tape, or on a pro Nakamitchi field recorder using chrome cassettes.

The live Circuitry tracks were all recorded in stereo on a two track Revox. Both of the reel to reel machines ran at 7 1/2 in/s. The Nakamichi ran at 1 7/8 in/s.

Tape hiss was inevitable. Resulting problem 30 years on? Well; tape hiss just isn't as sexy as vinyl crackle.........

Generally, when we rehearsed, we would plonk a couple of mics down in the room where we were playing, and press record. Sometimes one of the mics might be on a chair, or on the mantlepiece. If one of us was close to one of the mics, so be it. We tended not to have enough room to set up for a textbook stereo recording, and I'm not even sure that we knew what one of those was at the time!

Resulting problem 30 years later? Some weird and wonderful panning to deal with.........

However, before I even got down to the re-mixing, we discovered that the Scotch tapes were suffering from the feared tape 'squeal' (you'd know it if you heard it!). The problem, which affects a lot of tape stock from the mid to late 70s, is caused by hydrolisation. The binder compound used at that time to 'stick' the magnetic particles to the tape's plastic backing absorbs water from the atmosphere over time, which eventually distorts the tape.

So. We took the tapes to the Copyroom in London, and they set about the painstaking task of slow baking the tapes in a convection oven for hours and hours, before eventually transferring all of the audio over to digital media.

Apparently, after about three weeks, the tapes begin re-absorbing moisture all over again.........!

When it came to re-mixing the tracks, I worked on the stereo pairs as individual mono left and right tracks to allow for re-balancing where necessary. All the material was 'normalised' to achieve maximum output level within the system, before any compression or limiting was added.

Panning was set to around '10 past 10', (70: 30). This was one of the means of combating the extreme left / right separation between instruments on some of the tracks. I also used a stereo-width enhancer plug-in on the master output, again to help combat the original panning extremes.

I wrote custom eq's for each 'group' of tracks. Much of the eq work had to do with reducing tape hiss whilst maintaining presence, and also with reducing low end rumble.

The hiss / presence question inevitably involves a compromise.

I worked on the basis that presence is infinitely preferable to dullness (ie a too ferociously applied low pass filter!). On the Circuitry tracks, for example, I cut the top end, added 5dB of mid-high lift to add air and presence and cut below 70 Hz to clear out rumble. In addition, each track had individually contoured and automated sweep eq settings where appropriate, used mostly to reduce hiss in areas where the music was at a low dynamic level.

All of the eq work on the tracks was done with track and / or main output compressors in place. This ensured that the hiss 'noise floor' was dealt with after any compressor make-up gain had been applied.

I added compression with care, wanting to preserve as wide a dynamic range within the material as possible. Most of the thresholds were set in the -23 to -17 dB range, most of the ratios in the range of 4 to 2:1. Attack was never below 10ms in order to preserve onset information. I used a limiter set to 0dB across the final output to pick off any transients.

Reverb was used very sparingly, and certainly not on all tracks. Generally I used it to help glue together the left and right sides of tracks exhibiting extreme panning anomalies. None was added to the live Circuitry tracks.

There were various instances of 'extraneous' noises that cropped up in some of the tracks. If I quite liked them, or if they were too deeply embedded in the audio (Roger's notoriously squeaky drum stool and Ian's buzzing amp and distortion pedal artifacts spring to mind!), I let them pass. On balance, they add to the live feel of the takes - and these tracks were all improvised live takes.

Jon