PG Music Home Forums General Discussion Recording, Mixing, Performance and Production Noise Reduction in Studio One

Thought I'd share a success/enlightenment story that may be useful to others.

I do a fair amount of bass guitar, keyboard and drum pad recordings and on my last one a significant amount of noise ended up on my bass track. Studio One has a handful of built-in noise reduction tools and Strip-Silence is one I go to frequently. But someone mentioned the Stem Separator. At first I dismissed this idea since (in my mind) the Stem Separator is intended to separate instruments, not noise. But, the curious guy I am drove me to give it a try.

Here was my test:
1. I dragged in a complete song and separated out the bass (no problem as I do this quite frquently).
2. I then created a track of white noise using the Tone Generator.
3. I then exported the bass line and white noise tracks as a mixdown.
4. Then I dragged the mixdown back into Studio One to verify the bassline was heavily contaminated with the noise (and it was).
5. Then I Stem-Separated the contaminated track to extract the bass, and it worked like a charm. The bass was noise-free.

I have to say Kudos to the Presonus engineers for designing such a robust stem separator, it even works as a noise reduction tool.

As a side note, I'm beginning to look at Studio One as a general purpose audio-editing programming environment rather than just a "musicians DAW".
I know other DAWs are capable but the more I explore Studio One the more I realize how powerful it is.

Thanks, this is very interesting and good to know.

I am a bit qurious about the actual stem separation itself.
Will the separated bass stem be more or less exactly at the same quality as the recorded played bass itself or is it degraded in some way?

This is a very good question and I don't have (but would also like) an answer.

From a practical perspective, meaning to my ears, an S1 separated bass stem sounds clean, crisp, full and fat enough for my purposes. But my ears are not perfect, no one's ears are.

But the engineer in me says that there must be some (perhaps tiny) amount of distortion that results from every separation process. Especially if there are other low frequency content in the neighborhood of what the bass guitar is putting out. I remember a test case last year on this forum involving SongMaster, but as I recall a deep acoustic analysis was never done.

Perhaps someone here has a tool that can do the following subtraction: Original Bass Track - Separated Bass Track = The Difference
If "The Difference" isn't exactly zero in evey bar and at every millisecond, then there would be some level of distortion or artifacts at play.

Are there any digital audio experts in the house?

Use any DAW. Load the original bass on one track and the separated bass on another, making sure they're both aligned. Using a plugin or tool of some sort, invert the polarity of ONE track. Then play back both tracks together and you'll hear the difference. If you then render the output you can then see the difference.

Well, I have an answer and my engineering instincts once again did not fail me.

As suggest by Simon I inverted the polarity and can clearly hear a muffled but audible difference when non-inverted and inverted are played at the same time.
Definitions:
"Bass Track A" is the separated bass from a popular song from the 70s. The source audio is an MP3 file.
"Bass Track B" is "Bass Track A" with significant white noise added.
"Bass Track C" is the bass separated from Bass Track B (which contains no audible noise).
"Bass Track D" is the inversion of Bass Track C (and when played sounds just like Bass Track C)

Test Results:
1. The first test I ran was to ensure that the inversion took place properly. So I played C and D at the same time and got no audio out. This makes sense as they cancel one another out.
2. Then I played A and D and the result is an audible, distorted and muffled bassline that I could easily identify as the bassline to this song. I interpret what I hear when these tracks are played as error (albeit small error) produced by the separation algorithm.

Conclusion
1. Based on this one "simple" test, there is in fact distortion and artifacts generated from the Studio One stem separation process. This makes sense because an AI trained algorithm is generalizing how to extract bass from a finite number training examples. We can't expect it to be mathematically perfect.
2. Bass Track C contains distortion and/or artifacts even though someone with good hearing can't hear them.

But from a practical standpoint the separations are so good, especially when mixed with other tracks that I (and I'm sure others) can't tell the difference.
The other point worth mentioning is that this was somewhat of a best case (low stress) case because white noise is sonically much different from bass guitar and therefore should be relatively easy to separate from bass. On the other hand, songs with bass guitar, low-end keyboard, etc would be more difficult for the algorithm. But even in such cases it does a good job for my ears.

Here is an interesting spin-off question. At what point is the error (observed and described above) deemed unacceptable to the developers of the algorithm? I'd think this would be an important question for those releasing product containing stem separation capability. Presonus obviously got it right.

Simon, thanks for mentioning inversion, I wouldn't have known Studio One could do this without your idea.

Happy to help! I'm not terribly surprised that the extraction isn't perfect, however if it sounds good then I'd be happy.

Although the stem separation in Studio One, and I would say all stem separators, can only be approximations of "unbaking the cake", nonetheless Studio One's implementation is so good that the unavoidable artifacts are below my threshold of hearing. And I have good hearing, so yes, I'm very happy with it.

This is my recap of what I learned thru this exploration:
1. Studio One's built-in stem separator is so strong that it can actually remove noise from my bass recordings
2. Studio One continues to impress me with its collection of availabe plugins, my most recent discoveries are the Tone Generator (for white noise, pink noise and pure tones) and the Mix Tool (for inverting audio waveforms)
3. There is so much more to discover in this DAW

I'm guessing that sound engineers of yesteryear would have given their right arms for capability that is readily available today.

And don't forget Simon's trick to duplicate track output then invert the track polarity of one of the tracks to compare if the waveforms are the same. Added bonus, you found out Studio One has the capability to invert the polarity of a track.

If there’s a DAW that can’t, I’ve not heard of it.

Mike, very likely true.

I based my comment on Bass Thumper's last comment in +++ THIS +++ post.

Some of us users have no or little experience with using a mixer, console, observing or working in a control booth, recording in a studio or investigated all the features available in a DAW.

When you're self learning it's easy to have wide gaps in what many consider basic knowledge because you don't know what you don't know until you encounter it or need it and go hunting for it. Even if you're a big studio veteran recording pro you may not have much knowledge about what goes on in the control booth if you're always on the recording side of the studio window.

Very well put. The reason I went down the inverter path was to help answer shlind's question about stem degradation. And in that process I ended up learning something.

If you're not growing/learning you're dying . . . and I'm not yet ready to die

A very good philosophy. I'm planning to hang around for a while longer also
(I better keep learning...)

I enjoyed reading this thread. The experiment was logical, well-documented, and beneficial.

I might only suggest the post title be modified to reflect that it was an experiment and used stem separation, because I chose not to read it on first glance thinking it was about a routine function.

I'd be more concerned about the source of the problem.

I use ReaFir in Reaper to sample and filter noise I can't deal with at source, (in my circumstances that's RF/EMF interference peak at around 20Khz that sneaks in on ALL tracks recorded in my space. The only solution that would work at source would be turning the room into a Faraday cage.

The BIAB stem splitter works rather well.