[atriolo]

• The noise is now configured as a rate in hits/pixel/s and converted to hits per ROF using the ROF length of each layer. This keeps the noise level independent of the ROF length. The default is 0.2 hits/pixel/s for both TRK and FT3, which corresponds to ~1e-7 hits/pixel/ROF with 18-BC ROFs (default used until now)
• Now it can be passed as configKeyValue as before, but the value to be passed should be 0.2 (or similar values) to simulate a correct amount of noise, instead of the old 1e-7. The definitive noise value is still to be decided, though.
• The ToA of noise digits is now generated uniformly within the ROF instead of being 0 as it was in the first draft

[atriolo]

Hello @shahor02, @sawenzel, could you please review/approve this PR? Thanks!

[shahor02]

@atriolo The code looks fine, but I am wondering if the logic is correct. At least for the Alpide, my understanding is that the noise is defined per ROF, regardless of its duration. Is it different in case of Alice3?

[atriolo]

Hello @shahor02, thanks for your comment. Actually, I was explicitly asked to change the noise definition from "per ROF" to "per second", in order to make it independent of the ROF length, since the designers will be studying the noise values independently of the ROF length. It will help to introduce the noise in the simulation as foreseen/studied by the designers.
Of course, if this changes in the future, or if we notice that this definition does not suit well our needs, I will revert the change. Hope this clarifies, and of course, if you think this should not be done this way, we can discuss a better way

[shahor02]

Chip designers surely know better than me.