You are not logged in. Please login or register.
RME User Forum → FireWire & USB series → Recording in 32-bit float
Logic Pro has a new version out (10.8) that allows recording in 32-bit float (up from 24 bit). There is a big warning that the audio interface needs to support this. In a first test with my UCXII everything seems to work fine.
My question - is this officially supported?
The tech specs for the UCX II say:
Resolution AD: 24 bit
But of course the recording file can be 32bit anyway, having unused bits and wasting storage…
The data format of the audio sent from the interface to the DAW is integer.
Besides, any modern DAW processes audio in floating-point format.
In order to benefit from higher DR that 32 bit FP audio format can provide, the recording interface has to have such a wide DR. This is usually achieved by utilizing several ADCs in a staging configuration. Most studio interfaces do not use this arrangement, because studio environment is much more controlled and one can do very much without using such complex ADC architectures just by carefully setting input levels. 32-bit FP recording is thus much more useful for field recording applications, and is primarily found in such devices.
And in any case, this technique does not solve the issue of analog clipping, which is as important as having wide enough AD conversion DR.
Is there a RME interface that supports this?
That "warning" comes from where ? And what says this warning ?
The Logic upgrade is from today, you don't really expect the device industry to upgrade their product catalag within a day ?
Is there a RME interface that supports this?
What is your use case and which specific problem do you need to solve by recording in 32-bit FP format?
That "warning" comes from where ? And what says this warning ?
The Logic upgrade is from today, you don't really expect the device industry to upgrade their product catalag within a day ?
Well, you get the warning from Logic if you select 32-bit float bit-depth. I have no expectations from the "device industry" here, I just wanna know if it's supported - some audio interfaces already do. RME will do their own thing anyway and not blindly follow a trend.
What is your use case and which specific problem do you need to solve by recording in 32-bit FP format?
None yet. I think it's just a "nice to have" so you don't have to worry about digital clipping. 24 bit works just fine in a controlled setting where you know your levels. The most interesting use case might indeed be field recording via DURec.
25% more bits would also lead to an increase in bandwidth during transmission between the recording interface and the PC.
Ultimately the PC would also have a higher load, as you can already see with other things like increasing the sample rate.
Regarding Dynamic:
24 bits already allow a dynamic range of about 144 dB, if I googled that correctly, which is already 24 dB above our hearing threshold of about 120 dB.
120 dB is already unbearably loud and already represents something like the pain threshold for our hearing and is very damaging to our hearing even with brief exposure.
Values above 120 dB according to Google: 130 dB riveting hammer, 140 dB airplane take-off. 150 dB forging hammer. 160 dB gun bang.
So I think for AD/DA conversion and transport from/to the PC, 24 bit should actually be completely sufficient, shouldn't it?
When it comes to further processing in the DAW or volume control, things may look different again.
Breaking the whole thing down in an understandable way might also be an interesting topic for an RME YouTube video.
The shown alert from Logic pro X v10.8 is only meant as a hint to the fact, the recorded audio is not 32bit (floating point) when '24bit only' audio hardware is used. Nothing more than that. As I wrote above, the files would be much larger but without any benefit. More and more field recorders and now also audio interfaces come with the 32bit feature. So of course Apple now made Logic pro compatible to avoid users switching to other audio software / DAWs for editing the files.
I'm 100% with ramses, for studio work there is no need for 32bit recordings.
In my opinion it's much more important to avoid 16bit recordings because there easily can be quantisation noise on quieter passages.
I thought the purpose of the recording 32 bit was because the daw's operate that way internally and if you do something like a render in place of a virtual instrument or plugins, you won't lose anything in that process.
32 bit is a lifesaver for people who only get one chance to record the source. News gathering, fi.
In the studio it shouldn't be necessary. It wouldn't harm either.
The current 32 bit interfaces use two 24 bit ADCs to get to 32 bit. One ADC digitises the lower level, one handles the higher level. Both data streams get stitched together to output a 32 bit floating signal.
It's not even new. There is/was a company in italy that had/has a 32 bit float interface 20 years ago. Expensive, tho. And I can't remember the name of the company, sorry.
RME User Forum → FireWire & USB series → Recording in 32-bit float
Powered by PunBB, supported by Informer Technologies, Inc.