The install automatically switched to SL Go…
Is there no option to test the full version?
Check that you have activated your SL13 license in the Steinberg Activation Manager.
It doesn´t show up there and the site still shows SL 12 Pro.
Might be too early and we´ll have to wait a moment…
Website shows 13 now
Not the demo link.
Hi guys,
Does anyone know how to get a trial version for SpectraLayers 13? I’m on SpectraLayers 12 right now and want to try the new version before upgrading.
I already installed SpectraLayers 13 via the Steinberg Download Assistant, but it keeps demanding a license. I know Steinberg usually offers trials because that’s how I tested version 12 before buying it, but I can’t find the link for the v13 trial.
Thanks in advance!
Trials always come around a month later after release (for all Steinberg products).
Soooooo is it using the mac gpu now???
Ah, forgot that. Updated - now let´s see 
Unmix song for a ten minute track - estimated about 40 min. on an M4 imac - so maybe not…
Voice DeClick — it’s all up to you whether I upgrade.
@Robin_Lobel thanks for listening to the community!
I look forward to many of the new additions and modifications
Even if the unmix module was blazing fast, it’s not very good. My advice, especially if you’re trying to unmix a song, is to use mvsep. com or lalal.ai. The models are fast, cost pennies to use, and produce incredible results. Far better than SL can do.
The new reconstruct module seems to have replaced the “Heal” function
We can now “heal up” or “heal down” to the surrounding levels (the controls I once asked for, go robin!), and the new noise, tonal and transient modes make for VERY seamless heals.
I’m still getting a hold of it, but it seems spectacular.
Congratulations @Robin_Lobel !!
So let me see … you registered on the forum a couple of hours ago, just after the product was released, to post this anonymous comment? Nice first post – I can’t wait for your follow-up in-depth review and analysis, complete with examples, use cases and comparisons with other products.
Has it occurred to you that maybe someone can be an active user of prior versions but not a regular forum user? I was in this thread because I too was waiting to see when version 13 was released.
My comment was directed at an earlier poster that was concerned about lack of GPU acceleration. The point was, stem separation really isn’t a strength of SL, so most people would be better off not worrying about the lack of GPU acceleration and use either a service or local models that do a much, much better job than what SL is capable of (because SL doesn’t ship a Python/Pytorch environment, which is necessary to use things like the latest Roformer-based models). Don’t take my word for it, compare something as simple as vocal+instrument unmix in SL against mvsep’s BS PolarFormer model.
SL is a great tool, and in fact very necessary to clean up the results of even the best stem separation. But the model it’s using for stem separation is completely outclassed by even the free models you can use in UVR5.
I agree completely.
SL stem separation is fine for saving you in a pinch, but if users need it as a main tool with top performance, it’s clear that they should use specialized tools for that.
I use a combination of models from Huggingface in the “local version” of Mvsep, called ZFTurbo MSST, and in Ultimate Vocal Remover to make instrumental versions of my favorite music. The quality is exceptional.
SpectraLayers definitely cannot do this.
So true, but people don’t listen. This happens often, if nothing better is known. Things of medium quality are taken for granted and called “superb”. Sad? Yes, but that’s how the world works.
SL is first and for all a spectral editor - and by far the very best, no question about that. Stem seperation is rather a goody, a nice to have, but not an essential part of an editor, imho.