Bit depth question

[Billy86]

??? Distrokid is one of the major distributors, and this is from the site: 

What Audio File Formats Can I Upload?

Audio files should be WAV, MP3, M4A, FLAC, AIFF, or Windows Media (WMA).

If you're sending a WAV, 16-bit, 44.1 kHz WAV is typical but pretty much anything works.

https://support.distrokid.com/hc/en-us/articles/360013647753-What-Audio-File-Formats-Can-I-Upload-

[Craig Anderton]

I switched from 44.1 to 48 kHz recently not because I wanted to or felt it would improve the sound, but because Dolby Atmos sessions need to be at 48 or 96 kHz. Rather than switch back and forth, I figured I'd cave to the gods of "Immersive Is Really Going to Get Traction This Time, Really It Will, It Won't Be Like Every Other Surround Fail Since the 70s." We'll see about that.  

So now I do everything at 48 kHz. I guess that means seamless transfers to DATs, LOL.

[John Vere]

Still waiting to hear back from Zoom. But I sort of know what the answer will be already. 

In the meantime I did a little digging and now realize that there's actually no mention of it being a 32 bit device anywhere. 

So now I am totally wondering where I was told this??   It might have been  from another thread here a while ago where someone was wondering why their audio recorded with a Zoom R12?? was only 16 bit when they had been told it was 32 bit. I questioned that myself knowing what I new about the 24 bit max being standard. 

 I remember the thread ended with you needed to select 32 bit in Audio Data settings. This is a setting I had never used in the past  because since the beginning of time I assumed I was recording at 24 bit because that was the little greyed out box in Driver Settings. And of course I was not in the habit of looking at the Transport Module readings.  Big mistake. Lucky for me since I bought the Scarlett 6i6 and later the Motu M4  all my new projects seems to have recorded at 24 bit. But older projects are a mix of 16 and 24?   

I think what can happen is, say I start a project using another interface ( my Tascam us1641)  or even WASAPI mode  then that project can default to 16 bit and you might not notice this.

Bottom line is after all these years I now discover an important setting I was not really aware existed. ( my typical Cakewalk story) 

But, Hey, wait a minute, why is this setting sort of off the radar in an obscure place in preferences and not found with all the other audio settings??  And is this per project or global?  

So when I set up the Zoom I noticed right away that it was showing 48/16 in the transport Module.  With my new found knowledge of the Audio Data settings box I knew what to do. The selection had 32 bit so what the heck, I checked it. 

 Because of this thread and my attempt to get to the bottom of this I just now found that dialog box looks the same no matter what audio device you have connected.  Even my Motu has the 32 bit option.  

So yes the Zoom Live Trak L8 is just another 24 bit Audio interface and now I need to go and remove the videos I made about it and change that setting back to 24.  I will continue to leave dithering off like I most always have done. 

Thanks to everyone for pointing this out. 

Oh and also thanks to those would politely pointed out that there might be nothing wrong with using 44.1.  

 

Edited October 29, 2023 by John Vere

[OutrageProductions]

I'll just insert my 2 cents worth of humor here:

[HOOK]

Is this where I bloviate about what sampling rate folks should use without suggesting they listen to the different sampling rates on their actual device, so they can decide for themselves if one sounds better than the other?

[DeeringAmps]

totally wondering where I was told this?

it was here @John Vere

 

[John Vere]

14 minutes ago, DeeringAmps said:

totally wondering where I was told this?

it was here @John Vere

 

Thanks I couldn't find it. And that explains the mystery as the Zoom UAC 232 is listed as being 32 Bit Float. A rare beast.  From that thread I guess I assumed that ALL Zoom interfaces where 32 bit. Bummer. It wasn't the reason I bought it but it might have been part of the Con list. 

If anyone does take the time to read that thread you will learn that @RexRed does strongly feel 32 bit recording is worth it. He makes some very good points and personal observations which I tend to respect and believe are possibly true.   Sorry if some of you find the topic boring and over done but I'm still learning even after 60 years of music creation. And the Bottom line- What does it sound like! 

@Craig Anderton mentioned DAT machines. That was my first experience with 48 and even I could hear a difference between the masters I made back then as compared to when I was forced to use the 44.1 setting by the  stupid CD replication factories. I mastered to 48 and then had a second machine to make the 44.1 copy which involved using the analog connections. Sony blocked copying your own creations!   So I have my own reasons for liking 48 that go a long way back.   

[whoisp]

What I can tell you working on, pre, post and live broadcasting for ITV as a sound and video engineer. The reality is we want 48/24 for video and music, less issue with dynamic range loss and anti-aliasing can be a real problem, especially with digital platforms. If you need to know more anti-aliasing varying factors with loss and noise read about Nyquist frequency is a good start.

Why we often resample 44 up to 48 due to those above issues, but why use 44 in first place and resample you ask. Well only because some studios use a desk that have SPDIF or ADAT at 44 or gave us CD or Digi Tape at 44 and not 48 due to upgrading studio is costly. Don’t forget, you lose some of that vintage tech classic sounds of the pre amps etc with some old gear. 

i have a desk with AD Conversion: [24 bit Guitar/Bass]  [24 bit phantom mic (Mic)] [24 bit (Line)] [24 bit (Simul) great desk pre amps ] back to DA Conversion: 24 bit with Internal Processing: 24 bit (digital mixer section) however its  limited to SPDIF  44.1 kHz  so i take a line out to 48kHz  s to bypass the SPDIF as a hybrid work a round 

48/24 is not about our ears, we can’t ear the difference from 44/16 but the digital world does due to things like Nyquist frequency, plugins, software algorithms etc, why we get issues like anti-aliasing, drop outs and much more with massive file size when we push our sampling with audio and midi.

If i use the SPDIF,  you have to set the cakewalk project, interface and system drivers all to   44.1 kHz or its will create noise, anti-aliasing, cackles and drop outs. That is important to note when working with any standard make sure everything is the same standard or you will get bugs and issues. Cakewalk project, your interface drivers and system drivers are all set to your preference which should be 48/24 and why i bypass SPIDF as im forever changing my setting if i don't and you want a stable system first.  Some interfaces have a higher internal clock than 24-bit but i say 48/24 is a safe mode standard 

At one time, 44/16 was better due to CPU restrictions of audio interface, hardware and CPU etc.  48/24 can still be pushing CPU restrictions mixing with large recorded sample files size, plugins and loads of tracks (play/record tracking hammers your system) but everyone should be on 48/24 standard really unless you have NASA power and hardware, 
 

Edited October 30, 2023 by whoisp

[Max Arwood]

Someone said there are no 32bit converters. Me I got to google everything lol.  Check out Mytek - True 32 bit integer. It guess this is future proofing your  recording, and emptying your pocket book simultaneously.  Wow 384kHz /32 bit integer. Now some math guy - how many bytes per minute does this thing use ? and how many empty 0’s are in my files per minute?

[Glenn Stanton]

46 minutes ago, Max Arwood said:

Someone said there are no 32bit converters. Me I got to google everything lol.  Check out Mytek - True 32 bit integer. It guess this is future proofing your  recording, and emptying your pocket book simultaneously.  Wow 384kHz /32 bit integer. Now some math guy - how many bytes per minute does this thing use ? and how many empty 0’s are in my files per minute?

next $10M hit = $50K microphone -> 384/32-bit IO (captures even the nose hair resonances), double 64-bit processing with 32-bit storage, output as 384/32 and finally posted on Spotify as 44.1/16 320K MP3 and listened on a shared pair of AirPods (each person gets a left or right channel...) on a subway whilst texting with five other people on a party chat, then switching after 30 seconds to some other song... ? 

Edited October 30, 2023 by Glenn Stanton

[John Vere]

@whoisp I enjoyed reading your post. It had a few points that related to my past experiences with the dawn of consumer digital gear.
As I said earlier I was using 48 to master to a Sony DAT machine which had replaced my 10” reel to reel. I could definitely hear the improvement overall and mostly it was because I now had that nice clean and quiet sound. And a small item to mail in to replication people. 
Shortly after I bought the Yamaha o1v digital mixer and discovered it was set at 44.1 for SPDIF output. Dang. So I would mix both a 44.1 master via SPDIF and a 48 which involves D/A out to A/D in. Guess which one still sounded better.  
 

Later I got my first PC and soon was using Wave Lab and started transferring all my masters into Wave files. The Sound Blaster Audigy II card had optical inputs so I used that to transfer the DATs.
I had already transferred my 10” tape to DAT at 48. 
I assumed they would be 48 but many years later I discovered that they are only 44.1. I have no clue why but no doubt the Sound Blaster is to blame.  
I later yet replaced that with a Interface that had SPDIF and used my O1v as the pre amps so a lot of stuff was recorded at 44.1. 

[Will.]

@John Vere this article might shine extra light.

https://www.soundonsound.com/sound-advice/q-do-really-need-24-bit-recording#:~:text=The bottom line is%2C if,stressful and mixing rather easier.

 

[whoisp]

2 hours ago, John Vere said:

@whoisp I enjoyed reading your post. It had a few points that related to my past experiences with the dawn of consumer digital gear.
As I said earlier I was using 48 to master to a Sony DAT machine which had replaced my 10” reel to reel. I could definitely hear the improvement overall and mostly it was because I now had that nice clean and quiet sound. And a small item to mail in to replication people. 
Shortly after I bought the Yamaha o1v digital mixer and discovered it was set at 44.1 for SPDIF output. Dang. So I would mix both a 44.1 master via SPDIF and a 48 which involves D/A out to A/D in. Guess which one still sounded better.  
 

Later I got my first PC and soon was using Wave Lab and started transferring all my masters into Wave files. The Sound Blaster Audigy II card had optical inputs so I used that to transfer the DATs.
I had already transferred my 10” tape to DAT at 48. 
I assumed they would be 48 but many years later I discovered that they are only 44.1. I have no clue why but no doubt the Sound Blaster is to blame.  
I later yet replaced that with a Interface that had SPDIF and used my O1v as the pre amps so a lot of stuff was recorded at 44.1. 

In my experience SPDIF is always 44.1 however some modern system where being upgraded to what we was sold SPDIF 48khz and some devices did not work,  i am guessing there is something in the PCB chain given its hardware that caused issues.  We always got issues when set at 48, sometimes the internal clock and the optical SPDIF clock would create driver and compatibility problems, the days when you wanted to hit device manger with a hammer. 

There was a time we always got random phantom clicks in recording, tuned out wifi was interfering with the SPDIF hardware and optical cable. We was like great SPDIF no noise from jack plugs etc then we had a phantom click lol We mess about with all the technology to plug a guitar or mic jack plug in anyway lol

[Glenn Stanton]

and for me, it was always 48K which was what i used for a number of years until i discovered many sample libraries used 44.1 and sucked at 48K settings..

from our friends at Wikipedia:

Protocol specifications
S/PDIF is used to transmit digital signals in a number of formats, the most common being the 48 kHz sample rate format (used in Digital Audio Tape) and the 44.1 kHz format, used in CD audio. In order to support both sample rates, as well as others that might be needed, the format has no defined bit rate. Instead, the data is sent using biphase mark code, which has either one or two transitions for every bit, allowing the original word clock to be extracted from the signal itself.

S/PDIF protocol differs from AES3 only in the channel status bits; see AES3 § Protocol for the high-level view. Both protocols group 192 samples into an audio block, and transmit one channel status bit per sample, providing one 192-bit channel status word per channel per audio block. For S/PDIF, the 192-bit status word is identical between the two channels and is divided into 12 words of 16 bits each, with the first 16 bits being a control code.

Edited October 30, 2023 by Glenn Stanton

[JnTuneTech]

16 minutes ago, Glenn Stanton said:

S/PDIF protocol differs from AES3 only in the channel status bits; see AES3 § Protocol for the high-level view.

Back when I was using my now ancient Eleven Rack, I tried both SPDIF and AES, at 48/24 - as my sound interface has inputs for both,   -It was interesting.   My sound interface allowed me to manually detect the clock synchronization status of the SPDIF as well, and the Eleven Rack has a clock source & sync selection. -While using that procedure stopped pops & clicks on the SPDIF connection, it was still a pain, and I just switched to using AES. In AES normal mode, both input methods sounded relatively the same, though AES offers an optional mode that definitely had a more audible high end spectrum balance. Not to my taste, so I left that alone.

I am glad I pretty much gave up with wrestling over that stuff. And I no longer deal in the broadcast standards mess like @whoisp describes, as well thankfully - that can really be taxing, even today with more of the bit level protocols having leveled out somewhat (but now there's MADI, LAN & Wi Fi transports and so on...) - I just can't stay focused on creating fun music worrying so much about all of that. I do still watch out - Listen - for poor digital conversion issues, to be clear, and have a good respect for good performing mic & line preamps - in an audible way, but arguing (and worrying) over the mathematical issues is of little use to me. If I have the opportunity, I listen through as much of the process as I can, and choose what sounds right for me, as I say, when I can, And, it's a constantly moving target, new hardware, new sample libraries, they all have differing qualities and interactions. Never ends.  Just like this topic!

[John Vere]

6 hours ago, Will. said:

@John Vere this article might shine extra light.

https://www.soundonsound.com/sound-advice/q-do-really-need-24-bit-recording#:~:text=The bottom line is%2C if,stressful and mixing rather easier.

 

2008!  probably back when audio interfaces ( sound cards?)  were still mostly 16 bit as well as CD's were still your delivery medium.  

[Will.]

6 minutes ago, John Vere said:

2008!  probably back when audio interfaces ( sound cards?)  were still mostly 16 bit as well as CD's were still your delivery medium.  

You are missing the point with that article. 

[azslow3]

8 hours ago, Max Arwood said:

Someone said there are no 32bit converters. Me I got to google everything lol.  Check out Mytek - True 32 bit integer. It guess this is future proofing your  recording, and emptying your pocket book simultaneously.  Wow 384kHz /32 bit integer. Now some math guy - how many bytes per minute does this thing use ? and how many empty 0’s are in my files per minute?

I have found only one Mytek ADC device, in "discontinued" section. And from its documentation:

Quote

The Brooklyn ADC is capable to record true (not floating point) 32bit files.

Now I have got it, we are using FALSE 32bit files, since they all are filled with floating points... ?

What is not possible to find in the documentation are characteristics of the device, normally given for middle-range audio interfaces and always known for top interfaces. But sure, in HighEnd world "This is absolutely the best analog-to-digital converter ever created by Mytek" covers everything and leaves no questions (those who ask simply can't understand the whole spirit...).

[mettelus]

Something to bear in mind is that even with 24-bit ADCs, there are enough interfaces out there with pre-processing FX that nothing would prevent them from passing 32-bits to the host. Implying that they are 32-bit ADCs is bit much, but pre-processing is essentially doing the same function as the DAW, and no reason they cannot pass that on.

[John Vere]

3 hours ago, Will. said:

You are missing the point with that article. 

I probably read it when it was published. I used to subscribe to SOS. 
There’s not much of a point. Just one personal opinion about pros and cons of using higher bit rate audio which is a topic that has been obviously been on going since the 90’s. 
So we can each make our own choice and for me that is based on my own personal experience over the last 30 years of digital audio being part of my recording process.
I absolutely believe that 48/24 is the perfect choice for me.    

And when it becomes the standard I’ll bump it up to 48/32