Convert Project Sample Rate?

[javahut]

Is there an easy way to convert a 44.1k project to 96k?

I found an old Sonar post that said if you convert the WAV files in the Audio folder to 96k, the project sample rate would be converted when it opens. So that half works. It seems to import the converted audio file, but around half way through drawing each track's wave form, the wave form changes to a flat line, and when played back... there's 96k audio up until the flat line in the WAV file, and then, as the waveform indicates... nothing. It appears the audio is converted properly, but only for the same number of samples that were originally in the  44.1k WAV file... then no samples at all for the rest of the length of the 96k file. The file time length seems to be correct... but only around the first half of the file's samples are in the track's waveform display.

So not sure why it only "half" works. Tried several things to try to "rig" it to work, but always the same result.

Anyone know of a way to convert a Cakewalk 44.1k Project to a higher sample rate?

Thanks.

 

[scook]

https://www.cakewalk.com/Documentation?product=Cakewalk&language=3&help=AudioPerformance.05.html

[Byron Dickens]

You know there is no benefit to doing this, right?

 

Upsampling it doesn't improve audio quality; you can't add what isn't there in the first place.

Edited August 19, 2021 by bdickens

[javahut]

1 hour ago, bdickens said:

You know there is no benefit to doing this, right?

 

Upsampling it doesn't improve audio quality; you can't add what isn't there in the first place.

I think there is a benefit to it when it comes to effects processing, vitural instruments, and the actual process of mixing multiple tracks together, in general, in some cases. When I've tried mixing 96kHz tracks vs original 44.1k files , the entire mix seems to come together much easier. I personally think the mix process and blend of instruments sounds much smoother and less harsh.

But, yeah, I've been doing this for many, many years, and am totally aware of when increasing audio resolution has a purpose, and when it doesn't.

[javahut]

3 hours ago, scook said:

https://www.cakewalk.com/Documentation?product=Cakewalk&language=3&help=AudioPerformance.05.html

I was aware of being able to do it manually. The projects I want it for have been heavily edited and automated. It'd take up quite a lot of time to do it manually. I kinda was hoping there was a way to convert the project in it's entirety to a higher sample rate automatically. I'm pretty sure in some DAWs it's possible.

The method I mentioned... converting the existing audio files outside of Cakewalk, and having the project open at a new sample rate almost works. Half of the new 96k audio is imported correctly, and all the automation is timed correctly for the new sample rate. If I just knew why the last half of the audio file comes up zeroed and could remedy that... that would be ideal.

Looking like it's not possible though. So I guess if I really want it enough, the only way may be make a manual attempt at it.

Thanks for the info.

[scook]

A slight tweak to the documented process would save automation.

• Export the clips as broadcast wave files.
• Create a copy of the project using "Save As" to a new folder.
• Delete all the audio leaving just the automation.
• Set the new project sample rate.
• Import the broadcast waves.

Alternately, in the existing project upsample all the plug-ins that may benefit from it. Not all plug-ins do.

CbB supports upsampling any plug-in so those that lack the feature may still be upsampled.

I created utility to manage upsample settings for VSTs.

A link to my utilities is in this thread.

 

[javahut]

On 8/19/2021 at 9:36 PM, scook said:

A slight tweak to the documented process would save automation.

• Export the clips as broadcast wave files.
• Create a copy of the project using "Save As" to a new folder.
• Delete all the audio leaving just the automation.
• Set the new project sample rate.
• Import the broadcast waves.

Alternately, in the existing project upsample all the plug-ins that may benefit from it. Not all plug-ins do.

CbB supports upsampling any plug-in so those that lack the feature may still be upsampled.

I created utility to manage upsample settings for VSTs.

A link to my utilities is in this thread.

 

Thanks a lot for this. Really appreciate. Was able to get this to work fairly easily, leaving all the automation intact.

The only thing that didn't really work exactly right was the export as broadcast wave. When I re-imported into new 96k session, having them import at the original time stamps did not move them to the correct locations within each track. It seems there still must be some kinda relationship to the original project sample rate when importing them at or moving them to the original time stamp. It moved them towards the correct location, but not far enough to make it to the correct location, often overlapping with files earlier in the time line.

So anyway... I just abandoned that part of it altogether. I re-exported the entirety of each track as a single WAV file, starting at 0:00 on the time line, so that the entire track would be "auto-aligned" within the single WAV when imported. Then after they were all imported, I just trimmed 'em down to get rid of blank/zero data within the WAV, then bounced those in place to separate individual files within each track... to keep files from continuing to be processed at times there was no actual sound in the file.

Worked great! So for anyone else ever needing to do this... it works!

Thanks again for the info. Really helped!

 

 

[Craig Anderton]

On 8/19/2021 at 6:27 PM, javahut said:

I think there is a benefit to it when it comes to effects processing, vitural instruments, and the actual process of mixing multiple tracks together, in general, in some cases. When I've tried mixing 96kHz tracks vs original 44.1k files , the entire mix seems to come together much easier. I personally think the mix process and blend of instruments sounds much smoother and less harsh.

That harshness may be the results of the virtual instruments and effects process not being oversampled. The simplest solution might be to leave what's 44.1 as 44.1, and upsample virtual instruments and effects, which is something Cakewalk does very well. This gives the performance advantage of higher sample rates in lower sample rate projects, because it's the rendering process at the higher sample rate that gives the cleaner sound. Once it's in the audio range, the sound quality is preserved, even when downsampling back down to 44.1 kHz. 

[javahut]

55 minutes ago, Craig Anderton said:

That harshness may be the results of the virtual instruments and effects process not being oversampled. The simplest solution might be to leave what's 44.1 as 44.1, and upsample virtual instruments and effects, which is something Cakewalk does very well. This gives the performance advantage of higher sample rates in lower sample rate projects, because it's the rendering process at the higher sample rate that gives the cleaner sound. Once it's in the audio range, the sound quality is preserved, even when downsampling back down to 44.1 kHz. 

Yeah, I understand that and use it sometimes. However, not yet sure I'm a fan of all the upsampling and downsampling multiple times at different places within a project. Seems like it would be good to have the option to upsample once going into the FX rack, and then downsample once going back out of the effects chain, instead of what could become multiple up/down SRC several times within a single audio stream, between tracks/effects and bus chains... depending of course on what type of effects/instruments are being used and whether they'd benefit from higher sample rates.

Also, I use Acustica Audio  N4 and Acqua quite a lot, and those don't yet fully support dynamic sample rate changes once they're loaded and saved in a Project. They've just started working on it, but I don't think the kinks will be ironed out of it for quite awhile.

Thanks for the input, though.