Magazine Archive

Home -> Magazines -> Issues -> Articles in this issue -> View

Digital Mixing Magic

With Sampling Keyboards

Vocal line out of tune? Lead guitarist fluffed his solo again? Craig Anderton reveals some ways of using a sampler to fix these, and other problems, in the mix.

You thought you heard something strange on the tape you've been given to mix, and on replay, there it is: an audible click on a saxophone part that nobody noticed during the recording process. As you keep listening, you find it gets worse: the singer didn't quite make that high C, there's a wicked pop at the beginning of some words, a snare drum is hit way too late, and the background singers who did such a great job on the second chorus didn't fare as well when they reached the third one...

In the days prior to sampling, you would have cursed the original recording engineer, and settled in for an evening of intimate communion with a razor blade, splicing block, and splicing tape. Nowadays, though, you can simply boot up your sampling keyboard and fix these and other problems in minutes instead of hours.

Although most people think of samplers as musical instruments, you can also think of them as digital recorders, controlled by a series of switches that just happen to come in the shape of a [keyboard. In fact they] are very good digital recorders indeed, often boasting 16-bit analogue-to-digital (A/D) conversion, a frequency response up to 15kHz or higher, enough memory to store several seconds (and sometimes a minute or more) of sound, hard disk compatibility for virtually instant recall of long samples, extensive synthesis-related options (enveloping, modulation, filtering, etc) that even the most sophisticated digital recorders don't have. And to top things off, many samplers are compatible with at least one of the numerous computer-based visual sample editors on the market; this can provide editing options that digital tape recorders can only dream about. If this sounds like just the tool for fixing a mix, then read on.

Most of the following techniques involve sampling a flawed part on tape, editing it within the sampler, and 'flying' it back to tape (ie. re-recording the sample over the flawed part). However, we'll also go beyond that to describe effects that don't just fix, but enhance, the mix.


Sometimes even the best singers will pop a 'P' or 'B', ruining an otherwise good take (usually, according to Murphy's Law, the one that was perfect in every other respect). Punching-in one word can be very difficult, and sometimes the problem only becomes annoying during mixdown, when the singer might not be around to re-record the part.

The solution is simple: sample the offending vocal phrase (it should start with the sound containing the pop). Next, use the truncation function on your sampler or a visual editor to snip away at the sample start point; remove as much of the pop as needed, making sure that the sample starts on a zero-crossing (you don't want to get rid of a vocal-induced pop only to replace it with a sampler-induced click). Spot-erase the original phrase on tape, then roll the tape and fly in the sample. If the timing is so tight that you find it difficult to play the sample at the right moment, and you have a sequencer synchronised to tape, then you can enter the trigger point in the sequencer and edit it as necessary.

If the timing is really tight, or you are punching something like a single word, here's another trick: transpose the sample down an octave, slow the tape down to half speed, and then fly in the sample.

Sometimes a singer will have to redo a performance just because of a pitch problem in one small section of a take. To fix this problem, again sample the offending word or phrase, erase the original, and fly the sample back to tape but use the sampler's pitch bend wheel to correct any pitch problems. Sequencing the sampled part may allow for more precision in pitch bend editing if the pitch really wavers.

Speaking of wavering, I once worked on a tune where the singer held a long, sustained note at the end of a passage. The original idea had been to add reverb on just that note to trail into a solo, but I felt that was too cliched. Instead, I sampled the phrase and, while flying it back to tape, added vibrato with the modulation wheel. Although this effect is possible with most digital delay lines, I often find that a sampler's modulation qualities are more regular and 'musical' than those found in a delay line.

Flying in vocals is a pretty well known technique. This involves sampling a specific vocal phrase that repeats throughout a song and overdubbing it at various places in the song, thus obviating the need for the singer to record multiple overdubs of repetitive material. Frankly, I greatly prefer the minute differences between performances imparted by a human; save your sampler for doing things that humans can't do.

For example, sometimes a vocalist might be able to sing a part perfectly except for one or two notes, thus necessitating a key change. Or does it? Have the vocalist sing the out-of-range note in a comfortable range, sample it, then transpose it on the sampler's keyboard as you fly it in. This does require a little effort; for example, transposing upwards alters the timbre (tone) of the sound, so the singer may have to affect a 'Darth Vader' voice when singing in order to obtain the proper timbre upon transposition. Also, remember that transposing upwards will shorten the sample length, and transposing downwards will lengthen it. Still, with a little practice, any singer can have an extended vocal range.

And now for something completely different: it's possible to change the timbre of a voice by sampling it and using a visual editor's editing capabilities. For example, the Macintosh program Alchemy lets you edit a sample's harmonic spectrum. This can completely change a voice's character, to the point of allowing a singer to accompany his or herself yet sound like a different singer. (Also note that Alchemy contains a function that lets you change the sample pitch without changing the sample length — just the feature you need when transposing a note beyond a singer's range.)

Visual sample editors have other uses as well. You can sample phrases and cut out breath noises and other mouth sounds (licking lips, swallowing, etc). Careful, though; don't take away too much, or the voice will lose some of its human qualities (then again, maybe that's the effect you are after). After making the changes, you guessed it — erase the original phrase on tape and fly in the new sampled one.


If a drum hit is ahead of or behind the beat by an objectionable amount, sample it, erase the original, then overdub the sampled version. This process can also be used if you want to place, say, a snare drum a little behind or ahead of the beat. If you decide that a drum hit should have been a flam, no problem; sample it, and record the overdub on a separate track.

Samplers are excellent for replacing weak drum sounds. Let's assume you are mixing a track where the snare has no real punch; either load in a disk with a great snare sample, or sample the weak sound and beef it up using the sampler's onboard signal processing options (ie. mix in other samples, chorus, EQ, etc). Then sync a sequencer to tape, play the new part in time with the track (or trigger it directly from the audio track, if your sampler can do this), edit the sequence if necessary, then record the new part on tape or treat it as a 'virtual track' during mixdown.

On one project I was called in to salvage, an oboe player had accidentally hit the microphone while playing, producing a nasty pop. Having no visual editor available, it would have been difficult to sample the part and remove the pop. The solution was simple: fortunately, later in the song the player played an identical melody line containing no pop. So I sampled that, and flew it in over the flawed part. (I might add that doing things like this can really impress your studio clients, who assume that the only option is to re-record the part, at considerable extra expense.)

As a guitarist, I've used samplers not so much to fix parts, but to create effects that would be difficult to do otherwise. One of my favourites is polyrhythmic echo. It's easy to set up an echo unit to give simple quarter-note echoes and such, but what if you're playing a piece in 5/4 and want to emphasise the first and fourth echoes? Samplers to the rescue. Just sample the note to be echoed, and use the sampler's velocity response to play in an echo part with each echo having the desired dynamic qualities. You can also do tricks like create arhythmic echoes, or echoes that change pitch (just transpose using the sampler's keyboard). Or increase the modulation at the tail of an echo... get the idea?

This technique doesn't work well with long passages — all that sampling can be tedious — but if you just need to echo a few notes, this technique will let you do so creatively. What's more, using a sampler for echo frees up your digital delay lines for other purposes.

Then there was the time that I wanted to shift a guitar power chord down a fifth and back up again, but have all the strings remain in tune (like using the Steinberger TransTrem). Since I didn't have a TransTrem on my guitar, I sampled the chord, looped it (not an easy task, but I suppose sometimes one must suffer for one's art!), and used the sampler's pitch bend wheel to create the bending.


Once I needed to cut a five minute song down to about three minutes. The song was very complex, and I was sure that splicing the 2-track master tape would result in an obvious splicing glitch. The only solution was to splice the 24-track master so that during mixdown, reverb and other ambience generators would help cover up the splice.

One can get pretty nervous about splicing a 24-track master, but again a sampler can help. Most samplers let you splice two samples together to create a longer sample, which is perfect for testing tape splice points. Record the section just before and just after the splice into two samples, and test the splice point in the sampler. Once you've found a good splice point, get out the razor blade and slice up the multitrack.

I've also found it handy to record an A=440Hz tone somewhere on a sample disk. It won't take up much space, and can be recorded at the start of a tape as a tuning reference. This is helpful if you transfer the tape to another studio where the recorder speed is not quite identical, and you need to adjust the varispeed to a standard. The A=440Hz can also serve as a quick tuning reference or test tone generator for tweaking up your mixing desk levels. Most samplers have a 'non-transpose' mode, so that hitting any key will play the reference tone.

Our final sampler application will save wear and tear on your tape recorder's heads, as well as time. Picture a situation where the guitarist needs to work out a new guitar solo; normally this would involve tying up the multitrack and running tape over and over while the guitarist works out the part. You could transfer a quick mix over to something like a cassette deck, to prevent wear on the master tape and heads, but this still requires shuttling tape around (unless the deck has a 'block repeat' function).

A better alternative is to sample a rough mix of the basic tracks and to loop the sample. The guitarist can then practice along, and since the multitrack won't be tied up, the engineer can be doing something else (like recording vocals, getting preliminary levels for a mix, etc).

Once the guitarist has the solo perfected, only then do you need to record the results on the multitrack. Of course, a long solo would require a lot of sample time, but you may be able to lower the sampling rate and trade off sample bandwidth for more sampling time.


Hopefully by now the point is clear: a sampler can do a lot more than play back string, brass, and percussion sounds. As with so many other pieces of musical gear, a sampler can do a lot more than just sample. Next time you need to fix something in the mix, or create effects that others say are impossible, boot up your sampler. It just may provide the simplest, fastest, and most cost-effective solution to your particular recording needs.

Craig Anderton is a musician, author, lecturer, and industry consultant. His latest book is Power Sequencing (With Master Tracks Pro/Pro 4), his latest recording is Forward Motion, on the Sona Gaia label (a division of MCA Records).

More from these topics

Browse by Topic:



Previous Article in this issue

Fostex 454 Mixer

Next article in this issue

Quinsoft Trax

Sound On Sound - Copyright: SOS Publications Ltd.
The contents of this magazine are re-published here with the kind permission of SOS Publications Ltd.


Sound On Sound - Aug 1990

Donated & scanned by: Mike Gorman

Feature by Craig Anderton

Previous article in this issue:

> Fostex 454 Mixer

Next article in this issue:

> Quinsoft Trax

Help Support The Things You Love

mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.

If you value this resource, you can support this project - it really helps!

Donations for June 2022
Issues donated this month: 0

New issues that have been donated or scanned for us this month.

Funds donated this month: £49.00

All donations and support are gratefully appreciated - thank you.

Magazines Needed - Can You Help?

Do you have any of these magazine issues?

> See all issues we need

If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!

Please Contribute to mu:zines by supplying magazines, scanning or donating funds. Thanks!

Monetary donations go towards site running costs, and the occasional coffee for me if there's anything left over!

Small Print

Terms of usePrivacy