ASIO-32 Source 64 Question

[Phutureprimitive]

Regarding the ASIO2-32 Source 64 Driver:

After reading the manual and many posts on Planet Z, I've determined this driver is capable of 24 bit and 32 bit resolution (If this is incorrect please explain.) So here is my question. What exactly determines if it outputs at 24 or 32 bit?

Ex: If I'm sequencing with 32 bit audio files in Cubase does this driver automatically output 32 bits of data? More importantly, if I'm using 24 bit audio files, but Cubase is calculating FX, etc in 32 bit floating point, then what does this driver output?

I'd like to know so that, if I decide to leave the box via Adat my audio doesn't get truncated, should for any reason this driver be outputting at 32 bits. (I'm surprised now one has developed a bit meter for scope yet.... that I know of).

Thank you in advance!!

Cheers,
~Rain

[FrancisHarmany]

I think ALL signals inside scope are 32bit ? (someone hit me if I am mistaken)

also, the ASIO2-32 will always output 32bit I think (another spanking please, if I am wrong)

Maybe you are confused with the ASIO2 DEST. module, which outputs depending on what SOURCE driver is used (I think)

[garyb]

yeah, you lose the bottom 8 bits when connecting to a 24bit environment. there is a 32bit float module as well(Asio2 flt), but i don't think that the reality of connectivity and filesize is worth all the fuss. if you can't do good work in 24 bit to truncate or dither to 16bit for the cd(or worse, mp3), then it's not the bit depth.

it's important to remember that your system, which is likely to be the BEST place to hear your music, as far as hearing what you heard when you mixed it, is ultimately the least important arena for the music to be heard. it'll be in formats that are popular and played on a variety of systems, very few which will be part of a critical listening experience that can detect the difference between something that was recorded and mixed 24bit or 32bit. in my studio, i can hear the difference, so i don't doubt that in your studio you can hear the difference too, but if you make a cd and mp3 and take them around town and play them on many different systems, i'll bet you'll have a pretty tough time telling which is which.

sorry to bore, but i would just use 24bit 44.1khz, and get on with the fun part, don't even waste any time worrying about it.

yes, mostly 32bit, but even higher within many mixers.....scope and headroom, aaahhhhh.....

yes, FrancisHarmany, that's correct about the asio2 dest module. likely that's where the confusion is. the asio2 dest module outputs whatever bitdepth the source module is.

[Phutureprimitive]

also, the ASIO2-32 will always output 32bit I think (another spanking please, if I am wrong)

Thanks for the input Francis.

Can anyone else confirm or deny this with certainty?

If this is correct, than anytime you use this particular driver and leave the box, your audio is getting truncated to 24 bits, unless you have the forethought to dither it first.

[Phutureprimitive]

sorry to bore, but i would just use 24bit 44.1khz, and get on with the fun part, don't even waste any time worrying about it.

Thanks for the info Gary. And I agree with you on many points.

I'm simply trying to avoid truncation wherever possible......Would rather dither if I must. I would just use the 24 bit driver instead, but it doesn't provide enough outputs (limited to 32). So just so I'm clear then, as you understand, when using ASIO2-32 Source 64, it is always outputting at 32 bits? If so, then connecting a source channel for this driver to an Adat Dest and leaving the box would automatically result in truncation (unless you dithered first). (This is also assuming I'm using Cubase which would be outputting 32 bits).

[Music Manic]

Why don't you record to VDat?

[nightscope]

bits

Gary,

This topic has been floating around my brain for a bit.

I have some tracks from way back which are @ 16 bit. I have always assumed that using the ASIO32 sources with 16 bit recorded material is fine. Sounds OK to me.

Is it? Anything I should take heed of?

ns

[voidar]

With Cubase and most hosts, just use the FLT drivers. Your hardware will then handle the float2integer conversion.
Cubase will then output at 32-bit float. Don't use any dithering.

[garyb]

yeah, truncation isn't a problem usually when going from 32 to 24 bit, when a 24 bit stream is what is coming from the converters. the additional 8 bits can be lopped of without problems in my experience. the same with feeding a 16bit stream through 24 or 32 bit drivers, the only thing that happens is headroom...

[nightscope]

truncation isn't a problem

Thanks for that confirmation. I shall continue cheerfully lopping off as is my wont.

ns

PS I have just disabled 2 stray IRQ shares as per your instructions elsewhere. A USB port and something called SMBus controller. Not quite sure what the SMBus is for but it doesn't seem to be that necessary for much. Appears to be some sort of laptop battery, fan gizmo. Despite disabling it fans are still whirring away, no batteries. Scope projects now load MUCH faster.

[valis]

SMbus is an old protocol that defines system level messaging for communicating between devices on the motherboard. It shouldn't really impact the amount of time it takes to load scope unless you're using a power profile that skewed your cpu usage to being very conservative. Are you running in ACPI or 'standard pc' mode, and what is your cpu/motherboard?

[nightscope]

hi valis,

I'm in ACPI mode on an Intel DG965RY & E6600. I wasn't figuring that knocking off the SMBus was helping things though it might have appeared so from my post above. More the disabling of the conflicting USB port. Thanks for the info on the SMBus, I suspected it was of little importance but better to know than wonder.

AND A RENEWED THANKS FOR THE FAMOUS VALIS SAVIOUR OF ALL MANKIND BATCH FILE WITHOUT WHICH LIFE IS NOT WORTH LIVING!!!!

ns

[garyb]

yeah, SMbus won't make problems. that usb controller will though...i don't think that there's a problem disabling SMbus, though.