Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ster4610]

I have Windows 10 Pro 64 bit. My dac is an Auralic Vega which streams up to 32 bits. Is there any advantage to using MC 23 64 bit?

[kr4]

Yes.  The advantage is for better processing, not necessarily more resolution.

[ster4610]

Thanks for the reply. So the Auralic driver will work equally well with either version of MC 23?

[Awesome Donkey]

The computer's architecture (32-bit or 64-bit) has nothing to do with the bit-depth (16-bit, 24-bit, 32-bit, etc.) of media files (or what the DAC supports).

https://en.m.wikipedia.org/wiki/Instruction_set_architecture

https://en.m.wikipedia.org/wiki/Audio_bit_depth

Performance should be equal between the 32-bit and 64-bit builds of MC23. Though the 64-bit build should be generally a little faster than the 32-bit build.

[ster4610]

Thank you very much!!!!

[kr4]

The computer's architecture (32-bit or 64-bit) has nothing to do with the bit-depth (16-bit, 24-bit, 32-bit, etc.) of media files (or what the DAC supports).

Of course, it depends on how the math is done.  At AES last week, Michal Jurewicz of Mytek presented a paper about high resolution recording work process and advocated the redesign of DAW software, which is currently mostly 32bit, into new 64bit structure.  One issue is that with 32bit software, he says, one can expect, at best, accurate retention of only 23bits and that comparing 24bit source material processed in 32bit and 64bit with the same algorithms reveals the superiority of the latter.  With higher resolution source material, the difference increases.

[Hendrik]

That sounds largely bogus to me. Both 32-bit and 64-bit applications can use the exact same precision in processing, a 32-bit application would just be slower at it, but achieve the same perfectly accurate result.

Since he is citing 23-bits of accuracy, he is most likely talking about floating point, where "single precision" floating point (which in total is 32 bits) has 23 bits of precision for the fraction, and 8 bits for the exponent (and one sign bit) - while a double precision floating point value (64 bits) would have 52-bit of precision for the fraction, plus 11 for the exponent. Media Center uses double precision floating points all the way through in its DSP engine, both in 32-bit and 64-bit builds.

If anything, he might have been talking about getting DAWs to use double precision internally, but not about the architecture it is built with - because that wouldn't make sense. Certainly Media Center behaves identical in both.