DSP Robotics Support

Have you considered adding a 'hold' element to help smooth ripples?

I've added a hold section into your test schematic so you can see what it does.

Through some experimentation, I have found '2 pass' of a short peak hold time better at reducing ripples than a single longer hold time.

Edit: there is something wrong with my Peak Hold code that can screw up the audio output even if is only connected to a meter. It shows the principle, but Don't use it in your schematics.

Hi,


Thanks. Using the HOLD function gives interesting results.
To my mind a 'dual pass' mechanism will give the same result as a 'single pass' mechanism with doublt the 'HOLD' time, if hold is long enough to flatten the sine wave shape.

Rocko wrote:Hi,


Thanks. Using the HOLD function gives interesting results.
To my mind a 'dual pass' mechanism will give the same result as a 'single pass' mechanism with doublt the 'HOLD' time, if hold is long enough to flatten the sine wave shape.

I thought that too, but 2 pass seemed better when I tested. But now I can't replicate the behaviour

In any case, you don't need a particularly long hold time to flatten envelopes in the entire audible range.