The sample quality myth: Part 4

The human factor

Some composers prefer their virtual orchestrations highly realistic in every thinkable way. Others don’t mind playing loose with reality as long as the overall feel and expressiveness of the music paints the right picture. Even if you’re in the latter camp, there are certain basal realism aspects that we all need to acknowledge and pay careful attention to. First one is:


As mentioned in Mixing and processing a virtual orchestra, orchestras have an immense dynamic range and can go from whisper-soft to thunderclap-loud in a heartbeat — and often do. This is probably something most people are aware of, but what isn’t as obvious is that the dynamics in an orchestral piece doesn’t just change radically from one moment to the next. It’s happening all the time, in more subtle ways.

For example, you are not likely to find an orchestral instrument capable of playing sustained notes that does not change the volume and timbre of the note over time. All of them do. Notes will start with a soft attack and swell in loudness, or start loud and recede to a lower volume. Sometimes both, like the ebb and flow of the tide. Long notes rarely just go abruptly quiet, like flipping a switch, there’s always a bit of taper at the end.

The most common way of controlling gradual dynamics in a MIDI scenario is by using Expression (CC#11) or whatever function your sample library uses for this purpose. Don’t be afraid to use it!

Tempo and timing

Working with a DAW and MIDI there’s really no escaping that grid-based, fixed-tempo paradigm that makes a lot of modern music sound like it’s performed by (and for) robots rather than humans. Sure, you could just ignore whatever BPM your DAW is set to and record your music on the fly, just like you would a rock band that doesn’t play with a click track. But as this means you’ll have some major and possibly showstopping editing hurdles coming up sooner or later, I wouldn’t recommend it.

What you can do to alleviate a lot of that roboticness is make sure to always perform parts live using a MIDI keyboard or other controller. You don’t have to be Vangelis to do it, all it takes is a decent sense of timing and knowing where the notes are on the keyboard. As touched upon here you will mostly be recording parts one monophonic line at a time anyway so virtuosic keyboard skills are hardly a requirement. Moreover, MIDI is a pretty forgiving medium for recording music.

• You can always lower the tempo when recording fast or otherwise complicated parts.
• A few wrong or sloppy notes in an overall good take can be edited by hand after the fact.
• A performance that feels just a little too loose can be tightened up with 50-75% quantization.

This isn’t “cheating”, mind you. It’s the way everyone who values their time does it, and there’s no reason why you shouldn’t also.

Natural limitations

All acoustic instruments have natural, physical limitations. Things such as range, pitch stability, note length, loudness and others. Sometimes these limitations aren’t present in the sampled versions of the instruments, enabling you to play things that aren’t physically possible. So that is something to watch out for.

Brass and wind instruments for example, which are powered by the player’s lungs, simply can not sustain forever. The player will need to stop and breathe at regular intervals. A sampled instrument with a looped sustain phase doesn’t have this limitation, so it’s your job to be aware of it and play the instrument like it’s normally played, i.e. with ample rests and no crazy long sustained notes.

Range is a limitation that skilled players sometimes can stretch a bit above and below the instruments normal tonal boundaries. This is something best used sparingly though, for solo passages. For most players the extremes of an instrument’s range will be just an out-of-tune squeal or powerless hiss, so having entire sections playing notes that are extremely high or low will sound very odd. If the sampled instrument is mapped across the entire keyboard, edit it so that it matches the exact range of the actual physical instrument, just to be on the safe side.

Some instruments are also more difficult and demanding to play than others, most notably contrabasses and other instruments in the bass register which due to their size require more physical force or air to produce sound than their higher-pitched counterparts. And even instruments that are normally quite easy to play might have certain techniques that quickly take their toll on the player. Your sampler might let you play tremolo in the strings for minutes straight, but in reality you would have a string section full of cramped up players in just a fraction of that time.

Going into all the limitations of orchestral instruments is beyond the scope of this article. Suffice to say that this is something that you need to be familiar with. Treat your sampled instruments like they’re actual human players and compose your music with that in mind.

Keep it simple, stupid

People in general have a way of overcomplicating everything. We overthink stuff, we approach problems from the wrong angle and seek out solutions in roundabout ways. Reasons may vary — sometimes we’re trying to be more clever than what the problem at hand really requires, especially if we feel that we have something to prove, and sometimes we’re so close to something that we’re not seeing the wood for the trees. This is part of the human condition and musicians and other artistically inclined people are no exception. Maybe we’re even worse than the rest?

Therefore it’s good to take a step back once in a while, look at the bigger picture and ask yourself things like:

• Does this convey the image in my mind?
• If not, what is the problem?

Sometimes it’s just a question of finding the right instruments for the part. Or, if the part is too weak/strong sounding, adding or removing stuff to create the desired effect. Many other times though, at least if you’re being honest with yourself, you will realize that the problematic passage isn’t intreresting enough. It fails to capture your imagination because it simply isn’t doing anything special. At that point it’s time for a follow-up question:

• Have I really done my best with this part?

Yup, this is a trick question. If the part is uninteresting and doesn’t capture your imagination, the answer can never be yes. Go back, rework it, kill your darlings, think outside the box and just do something completely different. And whatever you do, don’t give it the “more is better” treatment!

Simply throwing more parts at a musical passage that isn’t interesting on its own isn’t really going to help. At best it will sound just as dull as before, at worst the added instrumentation will only make the uninteresting-ness even more obvious. This problem is very apparent in many Hollywood soundtracks these days, where the music is very bland and and the sheer power of the orchestra performing it only serves to make the blandness impossible to ignore.

In fact, you should always be wary about adding tons of stuff just because. Even if you have a musical passage that you’re happy with, it’s deceptively easy taking the arrangement a step too far. Same way as a convoluted style of writing will make a text tiring to read, a convoluted musical performance will often get in the way of the musical idea at its core.

So, try to keep it simple. Approach every new part of your composition as if it’s going to be the best music you’ve ever written. Decide on a basic, solid idea first of all, then experiment with different arrangements and instrumentations to find what best brings out the character and feeling of the music at hand.

Is this easy? No. But if you feel that you’re going through the motions, you probably are, and going back and revising things at an early stage — before you’ve gone Hollywood on the poor music’s ass — will save you a lot of time and spare you lots of frustration.

Did you enjoy this article? If so, please consider donating (monthly, if possible). Your donations will go towards making my free time less limited, enabling me to write more material like this. Thanks.

One Response to The sample quality myth: Part 4

  1. Pingback: The sample quality myth Pt 4: The human factor |

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.