I'm currently studying Sound on Sound's Synth Secrets series. Specifically: Part 25 Synthesizing Brass Instruments.

The amplitude response seems simple: Faster attack with higher amplitudes. Attack stage gets shorter as the velocity increases.

I'm implementing this in SuperCollider as attack = (1 / velocity) - 0.99. (This implementation gives very long attack times at low velocity, but I need to know the attack ranges before scaling it -unless someone can suggest a better implementation, which I'm sure there is-)

The problem is that I don't know the attack time ranges of a typical brass instrument. What's a common/typical fast attack time in brass instruments? What's a common/typical slow attack time in brass instruments?

All these assuming no crescendo, or other modulations, are being applied. Just the amplitude envelope trying to emulate a note being played on a brass instrument without crescendo.


As there was no answer and I couldn't find information about this elsewhere, I performed a small analysis.

Using the samples available here, I found the following attack time ranges (all times in seconds):

Trombone pianissimo: [0.070, 0.234]
Trombone fortissimo: [0.041, 0.050]

Tuba pianissimo: [0.033, 0.072]
Tuba fortissimo: [0.012, 0.025]

Trumpet pianissimo: [0.040, 0.060]
Trumpet fortissimo: [0.025, 0.031]

This was not an extensive research by any means, with only 5 samples for each instrument and intensity being analyzed. As an approximation, an attack time range of [0.010, 0.240] might work when synthesizing a general brass sound.

| improve this answer | |

Typically, Attack ranges on instrument are subjective as to the stylisation of how they are played. My recommendation for you would be to record the brass sounds in 192 KHz in 32 or 64 bit. High Sampling frequency would give you accurate time representation and high bit depth would give you accurate amplitude representation from which you can derive ADSR structure.

however to achieve effective attack times , you would have to record the middle C in the octave ranges of all the selected instruments.and then a quick analysis would give the attack times.

| improve this answer | |
  • I disagree. Attack ranges on instruments are not subjective, they can be measured, analyzed, and the dynamics and concepts involved are well defined. As I specified in the question, there is no modulation involved: just a flow of air at constant intensity. I'm already doing some basic research on this and will post the results soon. Also, 16 bit 44.1 kHz are enough to accurately perform these measurements. – Tom Cat May 23 '14 at 13:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.