Why I'm hesitating to purchase

SpectraLayers
1

Is it just me or is Spectralayers just hard to use?

Especially within Cubase. I’ve been dipping in and out since it made its first appearance and was always frustrated that playback/looping didn’t seem to work right; it didn’t feel well-integrated. It’s still the same frustrating experience on the v.11 trial I’m running.

It’s especially annoying because some of the new modules are great - just what I’ve been looking for; and I’ve always liked being able to transfer between layers - essential, really.

So I’d love to fork out for this but it’s just such a **** to use! Is v.12 much of an improvement?

Cheers,
C :thinking:

2

It’s actually quite easy to use, powerful and truly ingenious. Great concept, great program,… but: It’s full of bugs in version 12.0.0. As it is now, not recommendable! (Because of all the bugs, just read the forum here and you’ll see…)

2 Likes
3

again, if you are good at reading the spectrogram, then you are ahead of the game

SL is not difficult to use, at all

you just need to learn how to use SL and how to remove whatever audio in which order to best separate sounds (if that is what you are doing)

the unmixers of music tribes will be interested in the latest automated processes and for that, we are looking to the future or elsewhere

4

Which is a goddamn shame, Sunnyman, because I mostly agree with the first bit. But usage-wise:

  1. The Loop-to-Selection option does no such thing. (And never has.)
  2. Hitting Play in Cubase sometimes triggers recording, even though the track is unarmed. [“FIXED”: see below :roll_eyes:]
  3. When I make a Frequency Selection while looping, the cursor resets to the project origin.
  4. When looping from Cubase, the cursors aren’t in sync.

That’ll do for now. I’ll keep an eye on developments.

Cheers for chiming in,
C :+1:

5

I’ve worked out what’s going on in that video. The SL cursor loops back to wherever it was when hitting play in Cubase. If that point is within the loop then it waits for Cubase’s cursor to catch up, then resumes. So it never tracks over the earlier part. If you position the SL cursor at the start of the loop first then it behaves. This shouldn’t be necessary but at least there’s a workaround…

1 Like
6

I know what this is too. It’s pressing I to change to Time Select. If you don’t realise the focus has shifted then you create a record enable point.

7

I’m no help here as my entire philosophy is perhaps different than many…

Spectralayers 12 and earlier have been fantastic for me.

SL (for me) is a tool for surgery. I only use it standalone. The various modules are clear, purposeful, and useful.

Cubendo isn’t even booted up when I’m doing audio surgery within SL etc…and I’m often in there for hours.

Anything I do in SL standalone ends up as an exported file(s) that are then imported into the multitrack after booting up Cubendo.

And since I already pre-know my desired result while in either a surgery-program…or the multitrack, results (for me) are predictable, repeatable, and make for efficient workflows

Separate processes at all times. No ara. No daw-sandwich containing multiple stages of stuff going on simultaneously.

At any rate…that’s how I focus regarding this stuff. I like to de-complicate what others seem to accept as necessary complicated-chains :slight_smile:

Spectralayers 12 is a wonderful, constantly-evolving standalone tool that’s great for certain types of audio surgery. Workflow version to version stays in the realm of what I find useful.

4 Likes
8

Maybe that’s the answer then: go standalone. I agree that the tools are very good

9

me too, I don’t have SL12 yet, tho

10

Yep, that’s where it wins out. Standalone is the way because the integration is not great. I even tried performing the vocal/bv split I’d done in s/a from within Cubase and the results were truly horrible.

So I’m looking forward to getting stuck into my old tape collection. I’ve finally managed, after 40(+! :face_without_mouth:) years to remove the crowd from the first gig I ever did. Now to find out if it was as good as I remember! Another thing that’s been a boon is using the Vocal Denoiser to split main and backing vocal.

Thanks for taking the time to chat. I feel encouraged. :slightly_smiling_face: :+1:

Cheers,
C

1 Like
11

Standalone doesn’t change anything about too many issues SL12 Pro has.

While I’m generally really liking how Spectralayers develops, that can’t cover or excuse the poor quality of this initial release. Especially for unmixing, despite some welcome progress the user experience with the present SL12 is a far cry from satisfying, as long as some basic problems aren’t fixed..

12

For what it’s worth, the stem separation in 12 is so much better than what I’ve experienced in RX. In my first use of it since downloading last week or so, it’s already paid for itself in terms of functional excellence and ease/speed of use. A serious jump forward.

Chewy

1 Like
13

Thanks for the extra contributions. Hmmm. I’ve been looking around the forum and - well, it makes for interesting reading. However, there are enough positive comments to make me think it’s probably worth a punt.

If only there was a trial…

1 Like
14

Know one’s tools…

Made a lame reply and can’t delete the post! Sorry for the noise,

Chewy

1 Like
15

Best advice you could give anyone :wink:

Btw, I’ve taken the punt and so far so good. The song split of an amateurish gig recording has just done a great job :slightly_smiling_face:

1 Like
16

SL12 was a rough release unfortunately. I am still a fan, but I’m also realistic about this release. It fell short of the high standard of release we’ve been seeing lately. Feature-wise it’s great. Bug-wise, not so much.

I am definitely looking forward to the first (and second) update. I really hope the Steinberg team in general (and especially the main dev for SL, Robin, who is a great guy) will PLEASE spend much more time in beta. Cutting edge is only great if it works well. So crucial is testing a wider variety of common hardware. Many people do NOT use Intel and nVidia, for example. Steinberg in general needs to test across more AMD IMO.

Other than that, it’s a great update. Maybe give it a few more months though.

3 Likes
17

100 %

1 Like
18

The problem is AMD neglected programmable GPUs years ago, because it was only a niche field back then, and now it’s so far behind CUDA that pretty much all AI technology runs on CUDA and everything else is just not competitive.

19

So does that mean that us AMD owners can forget about any GPU action for the foreseeable?

20

I get what you are saying, but that’s not really the problem with SL12 in this case. It may be problem-adjacent, but the real problem with SL12 is that it wasn’t tested thoroughly enough on enough hardware before release.

I’m not expecting perfection, that’s not possible, of course. But I am expecting a certain minimal standard from Steinberg at this point, and the SL12 initial release fell short IMO. And keep in mind I love the work Robin is doing overall. I’m in. I’ve been paying for updates for a while now, and grateful for his innovative work. But he and Steinberg in general need to test more, and test more AMD, to be blunt, before release.

And if they find that AMD (or Apple Silicon, or Snapdragon, etc.) is inferior or problematic or limited for this app, and/or they don’t have the resources to achieve parity with Intel and nVidia, then that needs to be clearly explained in the release notes, in the marketing pages, and right there out in the open with the published system requirements BEFORE anyone pays for it. As long as they are up front about it all, then I’ll accept it and then I can decide what I want to do as a customer.

But here’s what is listed as system requirements from as of July 19, 2025:

System Requirements

Windows

Windows 10 (21H2 or higher), Windows 11 (21H2 or higher)
Intel® Core™ (5th Generation or higher), AMD Ryzen™ (or higher), Qualcomm Snapdragon™ X
8 GB RAM (16 GB recommended)
8 GB of free hard disk space (for temporary files)
DirectX 11 compatible graphics card (DirectX 12 with 8GB VRAM or more recommended for AI processing)
1280x720 display resolution
Windows-compatible audio hardware

macOS

macOS 12 (Monterey), macOS 13 (Ventura), macOS 14 (Sonoma), macOS 15 (Sequoia)
Intel® Core™ (mid 2013 or later), Apple silicon
8 GB RAM (16 GB recommended)
8 GB of free hard disk space (for temporary files)
Metal capable graphics adapter
1280x720 display resolution
CoreAudio-compatible audio hardware

There is nothing in that above to indicate any specific requirements or preference for nVidia and Intel.

If Robin wants to make a product specifically for nVidia-based machines with Intel CPUs, then that’s fine by me too – just state that fact up front, and then as a consumer, I’ll decide if I want to buy it.

But that is not what his system requirements says… on the contrary, it has sweeping hardware support. Maybe that’s unrealistic of him or overly optimistic to shoot for the stars like that, but the bottom line is that if Steinberg is going to say those are the system requirements, then they need to reasonably test across those systems and not leave a big gap… OR they need to explicitly detail the limitations of certain hardware combinations up front and center, publicly, right from the beginning, before anyone buys it.

And I’m not trying to give Robin or Steinberg a hard time. I know the ropes with software, been pretty hip to what’s going on in our little corner of the market for years, and it’s amazing what they are giving us. BUT you’d have to be pretty blind to not see SL12 in this case fell a bit short of quality control this time around, at launch. Now I’m NOT ranting about Robin, but rather I’m a reasonable, sane customer and I’m also a fan.

I’m just saying: BUY MORE AMD hardware, TEST MORE AMD hardware, and if it takes a little longer to release, THAT’S OKAY. Take your time! And if you have to charge us a few more bucks or pounds or euros or rupees or pokedollars, THAT’S OKAY TOO.

Nothing personal against anyone. Nothing against the big vision Robin has, which I support. Just PLEASE spend 10% more time and budget testing.

And it goes without saying (but I’ll say it anyway), that the ONLY thing developers get with rough software releases like this is the loss of confidence from their customers. So it’s in Steinberg’s best interest to raise the standard up a notch or two here and listen to their reasonable, long-term, patient and supportive customers.

Cheers to all, and let’s get this baby wrapped up and rocking and rolling so we can all make incredibly cool stuff with it!

2 Likes