Can I Upgrade Old Sonar Audio Tracks

[aleo]

I recently opened an old cwp file that I created back when I was using Sonar.  it contains eight audio tracks that I was in the final stages of mixing.  The tracks' volume and pan envelopes still need some work and I would like to finish this project, but I noticed that everything was recorded 44.1kHz and 16 bit.  From the standpoint of preserving the fidelity of the recorded sound, would there be any advantage in exporting these tracks as 48kHz, 24 bit wave files (using pow r 3 dithering) and them importing them to  a new CbB project set to work at 48KHz and 24 bit and finish my final mix in the new CbB file?  if I do this should I remove the old volume and pan envelopes prior to exporting the tracks?

 

[Jimbo 88]

Since you recorded at 44.1 , 16 bit there will be no improvement in sound by resampling at a higher rate.  Added plugins might perform better. Things like reverb, pitch correct etc... But the difference will not be very perceivable.

I recommend sticking with your current rates, they are fine.  I'm guessing you will end up mixing down to MP3 in the end anyway.

 

[aleo]

Thanks for responding Jimbo.  I may want to use some plug-ins during the mixdown.  The tracks will be mixed and exported as wave files then the final track will be burned on a CD.

[John Vere]

Because most of the songs I record end up on. CD or as mp3 I’m totally fine with 44.1. Ain’t nobody going to tell the difference.
And 16 bit is nothing you can improve on as changing it  to 24 will only add more zeros to the data. 
What’s more important is the quality of the content. 

[msmcleod]

It really depends on how much more mixing you've got to do.

Changing from 16 bit to 24 bit won't add quality as such, but it give you far more dynamic range, which will give you better quality at lower volumes, allowing you a bit more headroom.

[scook]

No need to force the import to 24bit. Leave it at the original bit depth. The engine runs at 32 or 64bit based on the Double Precision Engine setting. The other setting to check is Render Bit Depth used for bounce, freeze and export. By default, it is 32bit. Any new recording uses the Record Bit Depth setting on the same page as Render Bit Depth.

[aleo]

Many thanks MS, Scook, and John!  I'm learning a lot about these settings.  In all of my current projects I'm working with 48kHz. and 24 bits .  In Preferences, 64 bit Double Precision Engine is always checked; Record Bit Depth is 24, Render Bit depth is 32.

[Clovis Ramsay]

On 3/19/2021 at 10:04 AM, aleo said:

Many thanks MS, Scook, and John!  I'm learning a lot about these settings.  In all of my current projects I'm working with 48kHz. and 24 bits .  In Preferences, 64 bit Double Precision Engine is always checked; Record Bit Depth is 24, Render Bit depth is 32.

Someone smarter than me should follow up and confirm my hypothesis regarding 64 bit Dbl Precision Engine....I do believe one must be working with 64-bit audio files for one to need 64 bit DPE engaged.  

That is how I interpreted it's purpose.  ...-waiting for forum user tech person smarter than I to confirm and/or correct me--  until then, be excellent! 

[Will.]

4 hours ago, Clovis Ramsay said:

Someone smarter than me should follow up and confirm my hypothesis regarding 64 bit Dbl Precision Engine....I do believe one must be working with 64-bit audio files for one to need 64 bit DPE engaged.  

That is how I interpreted it's purpose.  ...-waiting for forum user tech person smarter than I to confirm and/or correct me--  until then, be excellent! 

Not really. What all this mean is . . . 

In a 32 bit OS you could only access to 4 GB of RAM for each process, and in 64bit that = 192 GB max (if I remember correctly.) 

So, a 24 bit audio is exactly the same as a 32bit or  64 bit file, these numbers are for the memory address mapping of the OS, the audio is not affected in any form (other than adding headroom.) Internally, most modern DAW's work at 32 bit floating point.

Most well known producers will tell you: that they still prefer a 16/48khz audio file for all its grittiness it brings to the audio - over a 24bit file. 

We can't even hear the difference between 32bit and 64bit. FWIW: This goes for "dithering" as well.  

This subject is slowly attracting the same attention the "LOUDNESS WAR" did back in 08/09. 

At the end of the day - its all about taste / and what works best for you at your disposal. 

[msmcleod]

Enabling the 64 bit engine means that all internal audio processing is done using 64 double precision floating point arithmetic, rather than 32 bit floating point.

This has nothing to do with the bit depth of the audio files.

What it does mean though, is that there is a huge amount of dynamic range when the engine is internally mixing tracks.  Internally, the audio can go to 1000's of dB without ever clipping, before it scales the mixed volume back down based on the track count. Only then does it convert it back to 16 bit / 24 bit integers.

If you were just using 16 bit integer arithmetic (or even 24 bit integer arithmetic for that matter), you'd lose quality very quickly as your track count increases.

[Lord Tim]

Beat me to it, Mark! And just to untangle a little further, the bit depth of the OS has nothing to do whatsoever with the sound of the audio files. It's two entirely separate things.

[aleo]

Good discussion guys.  Thanks again MS!

[Will.]

2 hours ago, Lord Tim said:

Beat me to it, Mark! And just to untangle a little further, the bit depth of the OS has nothing to do whatsoever with the sound of the audio files. It's two entirely separate things.

Oh - but it does. you need to read what I've said again, cause it expains itself. 

Using Double Precision in a 32bit OS does not give you nearly as much of headroom as in a 64bit OS. Read carefully what I've said. This does not affect "Quality" but only headroom between the ceiling of the wave file (Dynamically.) You don't get nearly as much of headroom in 32bit as what you get in 64bit. Like Mark said it db thing. 750db (considered as infinite) over 1500db (also infinite) 

 

Edited March 25, 2021 by Will_Kaydo

[Lord Tim]

Come on Will, tell us how a 32 bit operating system gives you more headroom on a wave file. I'll wait.

32 bit and 64 bit BIT DEPTH certainly can affect it, but it has NOTHING to do with the operating system. You can have 32 bit and 64 bit double precision stuff happening on either a 32 bit or 64 bit operating system and the wave files will be absolutely identical sonically. It is entirely unrelated.

[msmcleod]

1 hour ago, Will_Kaydo said:

Oh - but it does. you need to read what I've said again, cause it expains itself. 

Using Double Precision in a 32bit OS does not give you nearly as much of headroom as in a 64bit OS. Read carefully what I've said. This does not affect "Quality" but only headroom between the ceiling of the wave file (Dynamically.) You don't get nearly as much of headroom in 32bit as what you get in 64bit. Like Mark said it db thing. 750db (considered as infinite) over 1500db (also infinite) 

 

This is incorrect.

64 bit double precision floating point numbers are exactly the same in both 32 bit and 64 bit operating systems - they're both 64 bit structures.

This is not the same as 32 bit vs 64 bit address pointers which is what you were referring to in your earlier post.

[Will.]

32 minutes ago, msmcleod said:

This is incorrect.

64 bit double precision floating point numbers are exactly the same in both 32 bit and 64 bit operating systems - they're both 64 bit structures.

This is not the same as 32 bit vs 64 bit address pointers which is what you were referring to in your earlier post.

Okay. Maybe I'm wrong OR maybe i'm being misunderstood. Either way, i see the same answer between what I said and your answer. 

Edited March 25, 2021 by Will_Kaydo

[Jimbo 88]

Here's the thing,  If you work with the same audio track for hours on end, as I have been forced to do in my career,  you have a chance to hear the difference between 48K and 44.1, 16 bit vs 64.  After hours and hours you might hear the difference, but a week later you won't be able to pick out which one is which.  I have tested this out myself not because I really wanted to, or was trying.

If you start your project at say 96K,  64 bit you will notice your recordings sounding better and your plugins responding better. That will translate.

In the old days 48K stayed in sync better with picture, but that's not an issue anymore.

 

 

[Glenn Stanton]

and then it's all crushed down into a 128K MP3 to be listened to on tiny ear buds ? 

Edited March 26, 2021 by Glenn Stanton