No Introprice? + RipX Stem comparison?

hi, im wondering if there is no intro price for Spectralayers 10 Pro?

and did anyone compare the unmix/Stem seperation with RipX?


@Robby29 I’ve set up a comparison page here:


Great comparison, my feeling would be SL and RipX are on a par, with nuances, in terms of separation, but they are also different products, each with a different focus … RipX is almost a “DAW for remixers” and SL is a forensic tool that happens to have excellent source separation as just one of it’s features.

It’s that last bit – the forensic nature of SL – that is for me the most important; it’s non-destructive, which is of course of no interest to remixers. Add to that the forthcoming GPU acceleration, and for me, SL is the winner, hands down. Nevertheless, I use both!

So, as is so often the case, the answer is “it depends” … on what you want to do.

I added another comparison: Restoration Comparison, SpectraLayers vs RX vs Acoustica:



I think, however, we (society) might need to stop and think what is “restoration” and what is “manipulation”.

Whoops! Too late.

Restoration can be classified as restoring audio in a non-destructive way (where for example you’re removing artifacts like tape noise or tape hiss) without reconstructing the audio. So for example if you remove tape noise or tape hiss from a piece of audio and phase invert/phase flip the taped removed/taped hiss audio against the original then you should only be left with tape hiss/tape noise.

Remastering is taking audio (that may have been outdatedly mixed for an outdated sound system, such as the superman movie with Christopher Reeves) and remixing (or sometimes referred to as upmixing) so the audio can sound more modern (sort of like the idea up upscaling film and restoring grainy film from the 1970’s to a 4K HDR format). A lot of audio from the 1970’s have this distinct tone/sound where the mid-to-high frequency is strong dynamically in that range, so rolling off and filtering out the mid-to-high/mid-to-low audio while boosting some of the lows and boosting some of the highs from 1970’s can give make it sound more modern.

Also to add: RipX is okay but I still believe (with some work) Spectralayers is superior. The only advantage that (I see) Ripx has over Spectralayers is real-time note changing and pitch-shifting. If Spectralayers can somehow switch all transformation operations to real-time and you can select any frequency/tone and transform it in real-time that would be a game changer.

Ohhh good lord , i can’t wait for this 1st patch :heart: :heart: :+1:

RipX’s main multi stem model is still Demucs3 looking at the source in the latest update this month. SpectraLayers is all the better for using what seems is Demucs4 maybe, although it used to use Deezer Spleeter I think just as RX10 still does. Even since Demucs4 things have moved on a little bit though in the Music Source Separation community, but solutions for those models are a little hidden in github and google colabs, but to give you a sense MDX23C is seeing vocal SDR of 11.75, Demucs4 is seeing 10.28, Demucs3 is 9.78 & Spleeter is 7.31.
Although there’s still some bugs / limits with MDX23C being ironed out I believe, hopefully we see that in a more public-facing tool soon because it’s astonishingly good overall.