I have been using Wavelab Pro for about a year and things have been going well. I usually record at 32bit/384KHz PCM using my RME ADI-2/4 Pro SE Converter. A few days ago I updated to 12.0.30 and things were still going great. But early this morning I made a recording and while the file was 32bit/384KHz the bitmeter showed it only had 24bits of audio. I have been pulling my hair out over this all day. I have gone over my RME setup carefully, removed and reinstalled the drivers just in case. After that I did a full reset of the RME. I also reinstalled Wavelab after removing the settings from app data in my profile.
Any suggestions that might help me resolve this issue?
Apologies if I misunderstood: as far as I know, there are no audio converters on the market today that are capable of recording at more than 24-bit depth (which corresponds to roughly 144 dB of dynamic range). Sessions can be set to 32-bit or higher these days to maintain audio quality above 0 dB.
It seems your bit depth meter is giving you an accurate reading. Even if the session were set to 64-bit, the meter would still show 24-bit from the recording, as there is no additional information above that.
This is simply wrong.
Stagetec TrueMatch (27bit)
Sound Devices (32bit float),
Tascam Portacapture (32bit float)
Steinberg UR C series (the analog front end isnât capable of the 32bit, so itâs more a marketing thing)
Zoom recorders
Neumann MT48
there are more I guess.
That means calculation precision. But you can save the renderings in 32bit too.
Iâm not sure, but I think the RME drivers are not able to transport more than 24bit.
On the other hand, these converters produce better recordings than the 32bit capable UR-C series from Steinberg.
But why? I would understand 32bit float.
There is no advantage over 24bit in the actual situation. The best 32bit converters do not max out the performance of the 24 bit transmission. Neumann specifies a dynamic range of 136dB for their 32 bit converter. And they can transmit standard compliant 24 bit AES streams.
Ok. Thanks for the reply. Initially I pushed back on that suggestion based on my read through of the RME user guide for the ADI-2/4 PRO SE. But after some further testing with SoundForge Pro 17 and Audacity I get pretty much the same result. So I will leave some room for the possibility that you may be correct.
What I canât explain is why numerous recordings I made previous to this issue appearing to have 32bits of audio as shown below. The example shows 32bit FP but I have also done 32bit INT.
What seems clear enough to me is that the RME converter is the issue, whether by design or due to some misconfiguration that I canât seem to track down. So I am taking this issue to RME Support because I no longer believe this is an issue with Wavelab Pro.
I can understand and appreciate the skepticism but I really have no desire to explain or justify myself. My only concern here is what the hardware and software can deliver.
You may be right about the RME driver. I have put that question to RME Support.
But I canât see the benefit from a bad recorded 32bit file with real dynamic range around 120dB over a real good conversion with 132 dB dynamic range in 24bit.
Donât get me wrong, I think the RME is a fantastic A/D converter.
I get that. I really do. But itâs beside the point and a topic for another discussion (though I am fairly certain it has been discussed at length on numerous forums over the years).
I heard back from RME and they said the ADI-2/4 PRO SE can send up to 24bit/768KHz PCM or up to DSD256 over USB. So that is the answer. What I was seeing in my unprocessed recordings with 24bits of audio is correct.
What it doesnât explain is why I have a lot of recordings that have audio out to 32bits. But I will make an educated guess that there is a conversion that is occurring somewhere in my workflow that I was unaware of and havenât accounted for. I will probably spend some time tracking that down just for my own information, but I consider this matter closed.
Havenât had enough coffee yet, but does a bit meter always show how many bits are available or does it show how many bits are âusedâ? If itâs how many are âusedâ wouldnât there be the possibility that a non-âscaledâ 32-bit float signal would appear to only have 24 bits used?
Know what I mean?
I suppose it could be tested by simply slapping the meter onto a track and running a signal with/without scaling.
Itâs a long shot but it may have actually depended upon a change of sample rate or some other subtle change. For example, if you had rendered to 32-bit while simultaneously having changed the sample rate this would cause the bitmeter to read 32-bits in the resulting file (somebody correct me if Iâm wrong). Or it could be some similar workflow or change which caused the audio data to be expressed in a greater number of bits.
Hi @MattiasNYC yes, but I guess the response may also depend heavily upon the source signal itself. A real-world signal or a sine wave will likely show a more âconsistentâ response than a square wave. Results would also vary depending upon the settings of the bit meter - Wavelabâs bit meter can be set to intuitive or true mode. Changing the level of a square wave while viewing the bit meter in intuitive mode gives pretty unintuitive results!
Yeah I donât disagree with any of that, just pointing it out to the OP that there are ways to think about it other than âit says itâs 32-bit float so I should see 32 bits in the meterâ.
With all due respect I think you guys are making this much more complicated than it really is.
The guidance provided in the Wavelab documentation is very straightforward and simple enough to effectively use, as someone else has already pointed out.
It happened that I came across this thread of discussion. I switched from 24 bits to 32 bits in ripping vinyls since a year ago and I went through a lot of issues which I would like to share with you. I have been using Sound Forge for more than 10 years and everything was good in recording and processing in 24 bits. I bought the RME ADI-2/4 a year ago when it came out to the market and did my first 32 bit 384kHz vinyl rip using Sound Forge. Everything appeared to be good until a played the file using my home DAC which is Accuphase DC-37. The DAC displayed the file as 24 bits! I used a free hex editor to examine the file. I just scrolled to the middle of the file to look at the samples and found out that every sample was indeed 32 bits but the most significant byte (8 bits) were always zeros. So DC-37 is not stupid. I logged a support ticket to Magix and so far they havenât taken any action. The interesting thing is that when I recorded the file as 32 bits 384kHz and saved the file as 32 bits 192 kHz, DC-37 displayed the file as 32 bits. I thought I had worked around the problem as I think 32 bits 192kHz is close enough to analog sound as far as my ear could comprehend. As Sound Forge records directly to harddisk and I often came across significant amount of missed samples. I think it is because disk access cannot catch up with the speed of 32 bits 384kHz.
I decided to switch to another DAW and so I started playing around with trial copies of many DAWs which I think (not 100% sure) including WaveLab. The DAWs either do not support 192kHz or 32 bits integer or have the same problem as Sound Forge. The exception is NCHâs WavePad. Then I ripped a vinyl in 32 bits 192kHz using WavePad. The recorded file sounds much superior to those 192kHz 32 bits files recorded by Sound Forge. What a big surprise! I suspected that Sound Forge used 32 bit floating point arithmetic for its internal processing and so when I saved the file as 32 bits integer, it just upsampled from 32 bits floating point (i.e. 24 bits integer) to 32 bits integer. Since then I have switched to WavePad for ripping vinyls.
