At Classical Music recording sessions or concerts, the object is to make sure the wide dynamic peaks avoid not hitting 0dBfs on the digital recording medium. Ideally the recorder’s metering should have the neg.12dBfs point, half way across it meter scale. Then it is easy to ensure that the peaks neatly arrive into the area above the neg.12dBfs area, by minimalist gain riding with linear faders. Back up recorders can be set up a few dBs lower.

With future editing in mind, excessive gain riding at sessions can make things difficult at specific edit points, apart from the fact that the performers also produce level changes themselves! So after the editing with session recordings, or just a concert recording, you are presented with a range of peaks, some sensibly close to 0dBfs.

Apart from these peak levels, there will often be long sections, may be 30dB or more lower. So, what should happen in the Post Recording/Editing stage? Some ‘purists’ do say that it should be ‘left as real’, but there is a way of ending up with a sensible overall raised level, avoid long very quiet areas, and yet definitely retain the musical dynamics of the peaks.

In the ‘pop music’ field, over a number of years, there has now been a total ‘musical destruction’ as the recording are ‘bashing zero’ in the search of LOUDNESS. Apparently this loudness aspect is now more important in the judgement of the ‘music’ than its composition, content and performance!

If you view these ‘pop music’ recordings on an editing computer, you find the whole track is just a rectangular display, all dynamics removed. In Classical Music, and for that matter, Jazz, Big Bands, Brass Bands etc, the retaining the dynamics is musically essential.

Current rumour has it that some Classical Music CDs, are now being issued, where this overall loudness factor is now being over used. This Paper outlines a sensible way of dealing with the dynamics in Classical Music, for playback in our usual listening environments, without getting near anywhere the current situation with that other genre.

DYNAMIC RETENTION

The first thing is to view the whole final edit of a session recording, or that obtained at a concert, on a computer editing program’s screen display. Lets assume that there are one or two really high peaks, a dB or so from full scale, many other areas five or six dB from the top, and long quiet sections, like slow movements, where they could be below neg.30dBfs.

After the editing, transferring ‘un-mastered’ to CD will not be very acceptable to the majority of listeners. Mainly the quietness of the long sections mentioned, but also with the sudden loud peaks. If the start is very quiet, it is very likely they would have turned up the initial playback volume a lot.

The writer knows of published reviews that have complained about this, and also of a composer, who’s 30dB or more leap in level, some three minutes in from the start, did reduce his loudspeaker’s tweeters to shreds, as he had raised the volume a lot, to properly hear the beginning.

Overall then, the way to proceed with the Mastering is to trim the peaks across the board, so that none dominate the rest, and deal with some of the long quieter sections by raising their levels appropriately. Then it will be possible to raise the overall level of the CD, maybe by even up to 6dB resulting with most of the peaks reaching a neg.1dBfs peak.

TACKLING PEAKS

The writer’s ‘Fast Edit’ computer editing program goes back the early 1990s. Manually, via the mouse, you can use 1dB reductions in the level of a ‘high-lighted’ area on the waveform display. You can’t hear these 1dB reductions. So, this method is used to tackle the ‘excessive’ peaks, by increasingly spacing out high lighted areas, each side of the peak and do multiple 1dB level reductions.

The first thing to do is to deal with the one or two excessive level diversions. You can initially zoom in and do a 1dB reduction right on the peak, then move outwards, either side of the peak slightly and have another, and then a little further outwards, a third. Depending how much headroom the peaks originally had you already have the chance to raise the overall level of the CD by 3dB or more.

But you can go further. If you move away from the immediate peak region, you can spread out some more of the 1dB reductions, say 30, 20 and 10 seconds from the already treated area. This vitally allows the dynamic to be related to the content before and after its position in the waveform.

Of course, you can immediately playback, at the click on the mouse, to hear the run up and beyond of the peak area, after each reduction is carried out. Because you have shared the direct ‘attack’ on the peak with other 1dB reductions in the wider area, you still hear it as a musical dynamic.

QUIET MOVEMENTS

With the chance of having some 6dB or more available for the overall increase in the final levels, the long quieter movements, do perhaps now only need to be boosted by a smaller amount? This can be done by another spread out set of highlighted areas with the 1dB level rises. Again it’s these distances apart, over which these ‘micro rises’ are done, that makes them un-noticeable.

After this and when you view the whole waveform you can see any final deviations you need to deal with, and listen to the transition from the louder parts to the quieter pieces, you have just dealt with. It should be a more enjoyable musical experience than from the session’s edited stage, or what a ‘fixed level’ live concert recording gives you.

You will also see the final amount of overall level rise you can make use of. The writer believes in only going as far as the neg.1dBfs digital point, although a particular duplicator of CDs once insisted that it would be ‘significantly louder’ to go the full distance – but he was mainly involved with ‘Bashing Zero’ Pop releases!

A USEFULL EARLIER WAY OF MASTERING

There is a ‘real time’ playback way of Mastering, which the writer still does use in some specific cases. It involves using the digital output from the editing Computer’s Waveform display, via manual faders on a digital mixer, such as the Audio & Design DMM-1 Digital Mini Mixer, as it was known. The output then carries on in digital form to a recorder like the Edirol R44 or the R4Pro.

The essential thing you are doing, in real time playback, is to be able to see where the peaks are, how far off dBfs they are, and use the DMM-1 faders to control the transfer levels, to suit the need to approach the peaks gently, yet get the peaks raised to neg.1dBfs, if that is where they are wanted, in musical terms.

Obviously the levels of the quieter passages can be raised gently as you go along, and brought down inaudibly, prior to arrival at the display of a louder part of the edited Master. Overall you are hearing the improvements as you go along. Should a ‘gaff’ overload occur, you merely go back a bit via the computer mouse, and carry on, but needing to later edit out the repeated section caused by the level ‘gaff’ removal.

Some other advantages to the manual approach is that the DMM-1 allows you, with the use of two faders to pan-in the left & right feeds to narrow the stereo. Also you can also subtly alter the left right stereo balance as you go.

Additionally, with the use of digitally coupled external kit, such as the TC Electronics Finalizer for increasing or reducing the Stereo Width and their M2000 Digital Reverb unit for adding more Reverb, you have further Real Time facilities available to use. The Finalizer is used in series with the digital signal flow, the M2000 as a reverb return via the DMM-1 ganged fader 3.

The writer now realizes that current Editing programs offer a host of facilities, apparently covering all that has been reported on here, all on one screen, but the chances of the writer changing his ways are as remote as anyone ever getting ‘absolute perfection’ in this Black Art World of Audio Recording and Playback.