Fan Rant and a question

SpectraLayers
1

I’m so glad to see you picked up by Steinberg. I was an early user from back in the days of Divide Frame. I’ve purchased every version since Spectral Layers was first released. Nothing else does what this program does.

My primary use for the program is in audio enhancement and clarification. Which brings me to the question, and I’ll ask it the long way around.

I still use SLP v2 for my enhancement work. I can create layers for noise removal, attenuation, and gain using polarity inversion only. It is so fast.

Choose a layer, select something, preview it, add to the selection or subtract it with Undo, then commit and it is copied to the layer. The selection is based on the aggregate audio on the screen. You can always see the current stage of your work.

Best of all (especially for legal work), nothing is truly deleted. Only the polarity is reversed so any layer can be soloed to verify that only unwanted noise was reduced or removed. The source layer is never touched, so it can be soloed as a reference at any time.

The selection tool does double duty. It limits the area that is being worked on by the brush or frequency selection tools. It can also be used to copy and paste blocks of audio to a layer.

Every version since then has been a step toward a Photoshop-like approach where audio is selected from the source layer then copied or cut and pasted to another layer. The selection tool does not work as a boundary any more. It’s just another tool.

Even when the audio is shown as a composite, it doesn’t mean that a selection can be made from an empty layer. The user has to always go back to the source layer, which still contains that loud noise that was inverted in another layer, or still contains that road noise that was attenuated in another layer. This leads to the temptation (especially for attenuating noises) to perform destructive editing on the source layer, which necessitates making a copy of the source layer in advance as a reference.

I know that all of the changes and versions you have made are as a result of user feedback and probably the first couple of versions were a little geeky for the average user. But if you ever decide that you might want to make a separate version of the program for the kind of non-destructive use that v2 uses, you can count me all the way in.

2

Thanks for your support :slight_smile:

Yes, moving from polarity inversion paradigm to selection based cut/paste paradigm was a choice starting with SLP3 to make it more accessible to new users. However, with SLP5/SLP6 you can actually somehow get back the polarity inversion workflow from SLP1/2 as well, thanks to the Composite view.

The process would then be to activate the Composite view in the Display panel, then select something in the source layer, copy, paste into a new layer and invert its phase (it’s now a phase icon in SLP6, still next to the volume control of each layer). Because the Composite view sum all layers, you’ll then see what you removed by phase inversion, non-destructively.

SLP6 also introduces “Copy to New Layer” (Ctrl+Shift+C) if you want to directly copy your selection to a new layer without having to copy it to the clipboard first+paste.

3

Thanks for the reply. I can see how the composite view allows me to see the effects of polarity reversal, but I still have to select based on the source layer.

As an example, the harmonic tool is affected by the presence of high frequency noise. If I select a partial of speech in an area with a lot of high frequency noise, it will sound very unnatural due to the relatively loud upper partials detected by the harmonic tool.

With SLP2 I could select this area of noise and place it on an attenuation layer. Then I could highlight a partial of speech with the harmonic tool and get a closer idea of the speech without the unnatural sounding upper harmonics since it was working from the combined sum of the audio from all of the layers. This is not possible from SLP3 going forward. The only solution is destructive editing of the source layer, which I wish to avoid.

Another example is of a short impulse noise that occurs over speech. With SLP2 I could select the noise and send it to a 100 percent reversed layer. Any speech that occurred just before or just after could be detected and salvaged.

Now with composite view, I can see the result of the phase reversal, but I can’t salvage the speech as easily since the noise still exists on the source layer. Again, the only solution is destructive editing of the source layer.