Danny de Graan

I came across some strange behaviour from the [coll] object in max 4.6X. When pasting a float value with more than 3 digits behind the dot results in rounding differences. So if the paste this:

I'm building a microtonal instrument so it is vital it produces the correct value.

I came across some strange behaviour from the [coll] object in max 4.6X. When pasting a float value with more than 3 digits behind the dot results in rounding differences. So if the paste this: 

coll-rounding-issue

Much debated topic. 32-bit floating point numbers come in a finite number in your computer. That's why. 134.521 just doesn't exist. 134.520996 does. Check the archived discussions.

Much debated topic. 32-bit floating point numbers come in a finite number in your computer. That's why. 134.521 just doesn't exist. 134.520996 does. Check the archived discussions.
J-F.


And it's not a coll issue, of course. Just try a [flonum] object: type the number, hit Enter.

Thanx for your reply. I didn't know this yet. I will take a look in the archives for a solution.

Instead of a solution you will find out there is no problem...

coll rounding issue