After testing few different things, all i have to say is, there’s a better way to spend 89€…
No improvement at all. Feels like 12.0.50
Sad, but true!
Oh for … please see my second post in this thread – nobody is arguing that SL13 is better than anything else in terms of generating karaoke tunes, but to condemn the entire product based on one feature, is simply going too far.
And yes, we all know about MVSep et al.
Any one of the new features such as Voice Declick, Reconstruct or Ambience Heal is worth the update price alone, provided they work as advertised, but of course I haven’t had time to test them in real-world scenarios in the few hours since release, and neither has anyone else unless they’ve been a beta tester under NDA.
I have tested Reconstruct extensively throughout the day on the same job I was doing yesterday with SL12, and it works superbly. It really saves an enourmous amount of time in my workflow and seems to work pretty well.
Ambience Heal I have used less, but it has also worked very well in the fee cases where it was needed today.
Attenuate also works very well, as advertised, but I think it can still improve somewhat (I can explain it later in a different post).
No doubt in time we will all come up with different ideas to improve and refine these new features and SL will continue to be improved as has been the case.
But the upgrade IMO is excellent and I definitely recommend it to anyone in post production.
lalal fails badly on lots of stuff I submit compared to other solutions. I’ve used it off/on for several years. All depends on the source.
@Robin_Lobel
Hi! Robin!
The Fade module is really amazing!
This is exactly what I wanted!
Thank you so much
I found about it just now! I’m so excited!!!
Thank you for mentioning that!
Any improvements on Unmix Noisy Speech in terms of speed or algorithm?
Improvements to Unmix Noisy Speech are also coming with patch 1.
@Robin_Lobel
The New Modules Filter is really helpful! But I’d like to add a module to the existing module list I made.
@Yoshiaki_Watanuki just select your module filter to display it, then check another module in the list, and click “Add a Module Filter” and type the same name as your existing module filter. This will replace it.
@Robin_Lobel It worked for me! Thank you so much
@Robin_Lobel
Sorry for the repeated messages
Is the Reconstruct module AI-driven?
@Yoshiaki_Watanuki Now it’s purely algorithmic. AI could be useful for generating larger missing data, but for short patches it’s better to rely on algorithms which gives the user more control about the reconstruction.
@Robin_Lobel
Wow! That makes total sense!
It’s really amazing that the Reconstruct module now lets us freely adjust Tonal, Noise, and Transient components!
Thank you so much!
Where is Improvements to Unmix Noisy Speech?
https://www.steinberg.net/spectralayers/release-notes/13/
@Yoshiaki_Watanuki I had to delay it, the deadline for patch 1 eventually did not allow a proper improvement here.
@Robin_Lobel Don’t worry,Robin! Please don’t overwork yourself!
I can’t wait for the next update