What is 'attack' in hearing aids' configuration?

The Starkey Handbook on “compression” has a good explanation of the effect of attack and relaxation constants of compression on hearing aid function.

The attack and relaxation of compression is a very different phenomenon than hearing aids changing their amplification volume or changing their degree of droning noise suppression or their directionality when transitioning between quiet and noisy environments with or without speech present in an environment. Music program settings are special because they need to use longer attack and relaxation constants than programs designed for listening to speech.

Since I’m a bit rusty on hearing aid knowledge, fooling around with other stuff lately, I’ll leave detailed explanations to other folks that are truly knowledgeable on the topics I’ve touched on. The post referenced below has a link to the Starkey Compression Handbook in it. The Handbook does a good job of explaining how hearing aids work, at least as far as compression goes.

https://forum.hearingtracker.com/t/difficulty-hearing-the-beginnings-of-words-in-soft-speech-what-to-change/40898/3?u=jim_lewis (member archive)

See also the following post for the attack/relaxation settings that ReSound recommends its HCP’s use in setting ReSound aids.

https://forum.hearingtracker.com/t/difficulty-hearing-the-beginnings-of-words-in-soft-speech-what-to-change/40898/8?u=jim_lewis (member archive)

@tom12010 said:

the concept of ‘attack’ wrt hearing aid

“Attack” is when your HAs jump off the nightstand in your sleep and…

No, “Attack” and “Release” usually is about Compression.

Program-switching timing probably has other names.

Compression: we can have a gain-knob and turn-down manually when too-loud, turn back up when the loud is passed. This is still done in radio and music production, but humans are not that good at it.

Automatic compressors were developed. Skip the electronic details. Imagine you enslave a little demon. You order it to watch this loudness-meter and turn this knob down/up to keep the loudness out of the red zone. How fast should the demon react?

On sudden impulse sounds, Humans react and turn-down too slow. Leaves a burst of distortion or pain. If our demon were infinitely fast, that’s avoided, but the sudden change of gain makes a POP or CLIK. And short bursts of distortion go unnoticed. The first job of the timing decision is to find “fast enough but not too fast”. For clean music I’ve used 50mS to 5mS. Speech has a lot of short loud sounds and it is sometimes clearer to let these splatt a little rather than have compressor turn-down the following speech sound so much. So even out to 200mS. (When cutting records, transients as short as 0.05mS make trouble, so the compressor is very fast.)

You don’t want the first big sound of the day to turn-down sound the rest of the day. The simplest way: when not too-loud, sneak the gain up over some period of time. If too fast the natural sound decay is cancelled, sound is not natural. (However this is how musical drum sounds are processed in the studio.) If too slow, one loud bang will punch-down a lot of following sound. Release times from 200mS to 2,000mS are traditional choices.

There are “smart” timing schemes which look for patterns and react in a way that may be better. Dual-constant limiters have fast release after a single transient but slow release after multiple transients. That’s 80 years old now, and modern computers can do much more complicated pattern matching. I would not be surprised if some HAs have 1mS attack and 5mS release for certain speech sounds. My little Marvel has a fairly long release after extended loud-sound.