In what ways would fix point be better? Protools has 48bit fixed point wich is proven to be superior to 32bit floating point, how does 48bit fixed compare to 64bit floating? What are the differences of fixed and floating?

your title says 64 fixed versus 64 floating... do you mean 48 fixed versus 64 floating?


IIRC, fixed point is better in general than floating because it utilizes a "double sampling technology" wheras floating does not..

more bits is better ;-) .. so 64bits floating point is better than 48 bits fixed. fixed-point is
often used when the range of values can be bounded .. so they can reserve less bits
for the exponent part and more for the mantissa part.


see: IEEE floating point

i think for audio, i'm not sure if there would be a significant advantage of fixed vs. floating
at 64bit representations. you've got 52 bits for the mantissa .. that's a huge range for
audio.

jeff

because it utilizes a "double sampling technology" wheras floating does not..

what does double sampling have to do with the way the number is represented?

ORIGINAL: raggletaggle

your title says 64 fixed versus 64 floating... do you mean 48 fixed versus 64 floating?


IIRC, fixed point is better in general than floating because it utilizes a "double sampling technology" wheras floating does not..

trying to learn difference of both 48bit and 64bit fixed over 64bit floating,there is no application that i know of that has 64bit fix so protools is only example i can give with it's 48bit fixed,so is 48bit considered a SIGNIFICANT difference over 64bit floating?If it is not would 64bit fixed be considered a significant leap in quality of sound and processing compared to floating?I am curious about all-64bit float/64bit fixed/48bit fixed.

IMHO, it doesn't really matter. The human ear can't hear distinctions in levels
so fine. Granted, performing many mixing and effect calculations can introduce
errors into the soup. But I doubt anyone could hear the difference.

We would need a massive ear and cognitive upgrades for our hardware to be able
to discern the difference.

ORIGINAL: jmarkham

more bits is better ;-) .. so 64bits floating point is better than 48 bits fixed. fixed-point is
often used when the range of values can be bounded .. so they can reserve less bits
for the exponent part and more for the mantissa part.


see: IEEE floating point

i think for audio, i'm not sure if there would be a significant advantage of fixed vs. floating
at 64bit representations. you've got 52 bits for the mantissa .. that's a huge range for
audio.

jeff

Thanks for info! One reason i ask is because cakewalk seemed like they where scared to say wich is better 64bit floating or 48bit fixed (in a video i watched on this site) but they did admit both (64bit float and 48bit fixed)where better than 32bit float.This brings me to my original question again, how does 64bit float compare to 64bit fixed and is the difference significant?

I don't think any system uses 64-bit fixed today.

In theory 64-bit has the ability to represent more dynamic range with greater precision than 48-bit fixed point. In practice maybe there is some esoteric DSP that is more stable/accurate using 64-bit float vs. 48-bit fixed.

ORIGINAL: Ron Kuper [Cakewalk]

I don't think any system uses 64-bit fixed today.

In theory 64-bit has the ability to represent more dynamic range with greater precision than 48-bit fixed point. In practice maybe there is some esoteric DSP that is more stable/accurate using 64-bit float vs. 48-bit fixed.

So Ron your saying that 64bit fixed in practice would not be as stable or precise as 64bit floating?and that 64bit float is better than 48fixed aswell?

how does 64bit float compare to 64bit fixed and is the difference significant?

depends on where they fix the exponent ;-) .. i'm pretty sure ron's right .. never seen a
64bit fixed point audio system.

jeff