I love SpectraLayers and use it a lot. At the time being mostly for cleaning up old recordings and remaster them to modern standards.
When doing that I often use the Unmix feature for stem separation and then I process the layers separately.
Very often I find a lot of “audio detritus” there that needs to be erased to make it sound clearer, i.e. between the harmonics of, say the bass tones, there’s a lot of granular content that mostly represents unwanted distortion. So I try to tidy that up, either via Unmix levels or manually with a square Eraser. The results are usually very good, the bass will sound distinct!
To identify what to keep and what to erase it would have been nice if the program somehow identified and marked up what to keep. I know I can use the Harmonics selection tool, and that’s fine, but it would take very long time to use with this purpose (whole songs…).
Since the program obviously knows how to Identify this tonal content I want to keep, it would have been nice - and sped up the workflow - if one could chose a “harmonics colours layer view” for this. I mean, we have the Composite view and we have the regular Layer view where each layer has a different color. What if there was a view variant where the harmonics were showed in a different color than the non tonal/granular content of that same layer?
Yes, I can go there by Unmix components and harmonically select the tonal component and then change to the noise component layer and invert the selection and erase there but it’s a long and tedious way to go.
It would certainly be easier - and a great timesaver - if there was a color coded harmonics view when doing this kind of manual tidying.
Would it be doable?
Can you do a mockup of what it is that your suggesting. I believe I understand what it is that you’re trying to suggest but (by your post alone) I’m confused and having trouble trying to comprehend.
I’m so sorry! I read this at least 50 times and I’m still not quite sure what it is exactly you’re suggesting. Can you please try to explain one more time (3rd time is the charm).
I’m simply suggesting a view based on harmonics.
The regular view shows a layer in only one colour.
In Composite view I believe the different colours are based on the amplitudes of the frequencies shown. The hotter the amplitude the hotter the colour. Like a temperature scale.
In this suggested view two different colours for a layer would be fine; one showing the harmonics/related tones and one showing the other content. Like when harmonic selection is used, but for the whole layer and without any selection (could maybe be optional).
It would have been nice to have this content presented automatically just by chosing this view for a specific layer. That way it would be easier to manually erase unwanted material (non-harmonics) with the eraser.
In a layer with complex audio this wouldn’t be of much use but for a separated stem layer like vocal or bass it would be a great complement to make it easier to clean it.
The problem would of course be for the program to know what tonals to look for and show.
Oh well, one is allowed to dream, right?
Ahhhh okay! Gotcha!
I already suggested this, but for selections not for the gui(the view).
A harmonic view would be extremely useful.
I’m sure you have your own workarounds. For vocals + backing/harmony I’ll sometimes run it through the amazing Bitwig Harmonic Split device that gives me Harmonic A & B + Nons stems, then bring those into SL for editing.
I agree but the best way to implement this is not for harmonics but for selections. It is much more beneficial to implement a color selections feature(like the idea of choosing which color to apply to which layer) then to implement a whole new color code system based on harmonics.
It wouldn’t make sense to color a harmonic based off the frequency because you can already color code layers. You can already select a harmonic (any harmonic) and cut to new layer and change the color of that layer to any color you choose.
The way I’m suggesting is a solution that works for everyone. The best to implement something like this is to implement a color code selections system (based off the idea of the saved selections system) where you can save different colors for different selections. That way (down the pipeline) it would be easier (for the developer) to implement a split selections to each own layer based off of color feature. That way you can have each harmonic that is selected with it’s own color applied to its own layer.
For example (further explanation) let’s say you have two different singers singing in a unison harmony together and you would like to color code the 2 different singers within the spectrum (so that the viewer can visually see the spectrum analysis of 2 different singers). It would be a lot more simpler to select any harmonic or frequency or use the magic wand tool to make a selection, change the color of that selection(the color of the selection and not the actual color of the layer), apply that selection to it’s own layer, apply that selection with its own color (or the same color as the layer), create a group of the selected harmonics together and edit the selection as one merged group (for example, singer A and singer B). That way you can have singer A show with green harmonics in that group while singer B shows with yellow harmonics while still having the ability to edit both singers harmonics independently.
That way down the pipeline, the developer can sinply implement a locking/freezing editing system (based on the color that is selected), where only the color of the selection you choose can only be affected during the editing process. For example if you choose the green color selections (and there are many other selections that have been applied with other colors) then only the green selections can be affected. Meaning if you have a green selections selected and lock all edits to the green selections while having many red selections and you choose to use the retangle selections tool to delete a selection (where maybe you have a scenario where the green and red selections intersects with each other) then only the green selections would be deleted whereas the red selections is left unharmed.
The idea was to have more than one colour for one layer, like in the composite view.
I think Marshall & I are coming from the point of view it could be constructive to edit in an Harmonic intensity way rather than an Amplitude intensity only. e.g. this is a quick and dirty spectral Harmonic Intensity view of a vocals fundamental (top), the harmonics (middle) and non-harmonics (bottom). That colour-coded view in the main SpectraLayers window could be useful for selecting harmonics both with the tool and manually.
Ahhhhh okay, I get it now
Yeah, Sam_Hocking, you’re better with words than I.