Weird high end artifact on mixdown at 44.1/16

[grendelos]

I have a strange mixdown bug that when exporting to 44.1/16, I get a high end artifact added to the track. I can consistently recreate the issue and I am wondering if anyone else can recreate too and help me confirm if this is a system bug or something unique to me?

1. Start with Mai Tai with + init patch.

2. Lay down a basic C chord for a few measures

3. Drop the EQ onto the track and do the following:
a. Set to FFT Curve
b. Set Analyzer Range Min to -120 dB
c. Set the high cut to 2k and 48dB

4. Now do Song → Export Mixdown with these settings:

5. Drop another EQ onto the new mixdown track and set the following:
a. Set to FFT Curve
b. Set Analyzer Range Min to 120 dB

6. Now pin both EQs so you can see them at the same time and play both tracks and hopefully you will see the artifact on the high end:

7. If you want to see it more extreme, then delete the mix down track, export again, but this time mute your master and then look at the EQs side by side.

This artifact does not show itself when I export my mixdown the very same way but choose 48 kHz/32 bit instead of 44.1 kHz/16 bit. So my work around right now is to export 48/32 and convert it to 44.1/16 with ffmpeg. It is an extra step, so I would prefer to export out of Studio One cleanly.

I submitted this as a bug and was hoping to see a fix in the new release, but nothing is in the release notes. So now I am wondering if the problem is unique to me? Can anyone replicate or confirm that they don't see these results? Thanks.

Don
Studio One 7.2.2.107056 Win x64 (Built on Jul 23 2025)
SSL 2+ MKII USB Audio Interface
Windows 11 - Xeon E-2144G 3.6 GHz - 32 GB RAM - SSD

 

[sintil8]

I Followed your instuctions above and ran a check for you and this is what I got...

Judge for yourself, it looks very similar. I may have the C chord playing in a slightly lower register than yours hence the higher bass content
I dropped the Main fader down by 12 Db as the mixdown was clipping. Other than that It looks very similar.
Hope this helps.

Kind regards

 

[SwitchBack]

Can you try export with adjusted loudness for a maximum true peak level of -1dBfs?

16-bit audio has a limited dynamic range, and (clipping) distortion can result in export frequencies above 22kHz. The foldback of those frequencies (and a missing cutoff filter in Studio One) can be the reason for the HF artefacts. Just a thought.

 

[sintil8]

Can you try export with adjusted loudness for a maximum true peak level of -1dBfs?

16-bit audio has a limited dynamic range, and (clipping) distortion can result in export frequencies above 22kHz. The foldback of those frequencies (and a missing cutoff filter in Studio One) can be the reason for the HF artefacts. Just a thought.

My guess is the downscaling from a higher resoluton to 16 bit and the type of dither being used, or not being used, it might be the padding out
of zero's from the rounding down?
I'm not a qualified sound engineer maybe someone on here could comment further?
Best regards

 

[soundofdrums]

My guess is the downscaling from a higher resoluton to 16 bit and the type of dither being used, or not being used, it might be the padding out
of zero's from the rounding down?
I'm not a qualified sound engineer maybe someone on here could comment further?
Best regards

I also wondered if it could be a dithering or lack of dithering issue

 

[grendelos]

The results are pretty much the same when exporting with adjusted loudness.

Since someone mentioned dithering, I went into the audio settings and disabled "Use dithering for playback and audio file export" and I get more of a "even-level" artifact as opposed to the upward slope artifact.

I guess of even more interest is that if I do the experiment and pull the HC filter way back to like 380 Hz, that artifact persists all the way down to the cut off, but if I re-enable the dithering setting, it doesn't go any lower than about 6 k. You can see that in the screen shot below.

So it would appear that enabling dither in general is a good thing, but it keeps the artifact above 6 k and boosts the artifact steadily above 10.9 k.

One last note, I plugged some of this info into Claude.AI and here is what it it suggesting. Looks like none of the possible suggestions could really be implemented and tested and would have to be fixed by the Studio One devs.

 

[soundofdrums]

Have you tried a 3rd party dither to see if that changes anything?

 

[sintil8]

@grendelos
I am curious as to why you are reducing the fidelity of your audio down to 16 bit ?

As the sound is somewhere in the region of -60 to -70 Db and any louder sounds would render it almost, if not, inaudible.
There are also lots of questions around sound masking and frequency build up that are possibly drawn into this discussion, which is why I asked the question in the first place.

Just to put some meat on the bone...
Another way of looking at it might be similar to putting a piece of frosted glass or a fresnel screen in front of a picture. One might give a more focused image the other will probably blur the image.

I don't think I can add any more to this, others may choose to discuss this more.

Kindest regards

Ps. If you have a copy of Izotope Ozone, there is the dither page in it, where you can listen the dither in isolation if you wanted to hear what is
removed when dropping down to a lower bit rate.

 

[grendelos]

Point taken that this may be inaudible in the end. As to why is this of concern to me? Is 44.1/16 bit not the audio CD standard? Seems like at one time that Distrokid had 44.1/16/wav as their submission standard too (though that appears to not be the case now). Even if Distrokid takes higher resolution files, they will down-sample them after the fact, so I would rather be in charge of the final product.

But I guess its the neuro-divergent brain of mine that looks at a down-sampled file using free/open source ffmpeg NOT having the issue, but my paid version of Studio One HAVING the issue that irks me the most. Plus, when I submit content to a radio station, I need the file to look professional and not have a high-end dancing appendage that the programmer might see and think that I am unprofessional and not worthy of considering for airplay.

I can look at a dithering plugin, but I suspect that it won't work because it is ahead of the downsampling process.

 

[sintil8]

Point taken that this may be inaudible in the end. As to why is this of concern to me? Is 44.1/16 bit not the audio CD standard? Seems like at one time that Distrokid had 44.1/16/wav as their submission standard too (though that appears to not be the case now). Even if Distrokid takes higher resolution files, they will down-sample them after the fact, so I would rather be in charge of the final product.

But I guess its the neuro-divergent brain of mine that looks at a down-sampled file using free/open source ffmpeg NOT having the issue, but my paid version of Studio One HAVING the issue that irks me the most. Plus, when I submit content to a radio station, I need the file to look professional and not have a high-end dancing appendage that the programmer might see and think that I am unprofessional and not worthy of considering for airplay.

I can look at a dithering plugin, but I suspect that it won't work because it is ahead of the downsampling process.

OK thank you for explaining. I see your concern. At least now you know it's not something unique to just yourself, as for it being a bug
is debatable, will the Dev's look at as a bug, only time will tell and if it gets addressed as a Bugfix...

It's a very complex subject beyond my skill level so I hope you will excuse me from here on in. Let's hope that someone with experience of Posting stuff on Distrokid can help you.

Kindest regards

 

[ianaeillo]

This is one of the many reasons I never master tracks I mix and why I rarely master *anything* at all. I always suggest to clients that they master somewhere else with someone who worries about those things. There's just so much extra stuff to worry about and I'm always amazed when people pick up on dithering aliasing and stuff in non-classical environments. I'd have never noticed that and doubt I would care if I did, but good job on you for picking up on it.

 

[grendelos]

Thanks sintil8, I greatly appreciate your input on the issue. I have made more progress on analyzing the issue by bouncing it off others than I ever did alone.

And ianaeillo, at least 50% of what I produce would be considered either new age or neo-classical. And I simply don't have the resources to outsource mastering, as nice as that would be.

 

[ianaeillo]

Thanks sintil8, I greatly appreciate your input on the issue. I have made more progress on analyzing the issue by bouncing it off others than I ever did alone.

And ianaeillo, at least 50% of what I produce would be considered either new age or neo-classical. And I simply don't have the resources to outsource mastering, as nice as that would be.

Ah my bad. I was inferring from your stated instrumentation you were using digital instruments only and I mistakingly meant "live" classical instead of just classical. Apologies for the misunderstanding.