SpectraLayers 7.0 - First looks and, "Oh Wow!"

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

JonUrban

Forum Curmudgeon
Staff member
Admin
Moderator
Since 2002/2003
Joined
Mar 2, 2002
Messages
17,681
Location
Connecticut
There's been a lot of talk around here about stems and upmixing lately, and that's all good. The SurroundByUs guys started it all and their SPEC program has been a trail blazer in the field, and is still today one of the top choices for folks who like to upmix stuff to quad, 5.1 and beyond. There are also many new comers like Penteo (which isn't really that new, it's founder, John Wheeler was/is a member here from way back when), spleeter, Acoustica 7, and the newest version of Acid Pro are other programs that can extract stems.

SpectraLayers has been around for many years, first as a Sony product, then a MAGIX program, and now it's owned by Steinberg. All along it's been a powerful spectral editor but it has a very steep learning curve and unless you do this kind of thing for real, like a job, most hobby guys don't have the time or patience to really get into it and learn it. Those that do, well, more power to them.

This week I got an offer to upgrade my copy of the software to Version 7 for $79, so I checked it out and saw that it too now had a stem separation feature, so I decided to go for it. I had used the program in the past to tweak up some audio on some conversions I'd done so I knew the deal. It's fairly sophisticated and I know I will never be an expert or even good at it, but I like to play around with this stuff.

The first thing I noticed is that it opens a 5.1 file, for real! It also supported playback in 5.1 on my PC, so that was very cool to see and hear. You can see the setup here:

Spectral Layers 7 - 001.jpg



Once you have a 5.1 file in the editor, you can solo or mute individual channels with the "Channels" control panel, as circled below.


Spectral Layers 7 - 002.jpg




So, anyway, back to the stem extraction. It works very much like Acoustica 7 or other Spleeter implementations, although I do not know if this is based on Spleeter or not. I would guess that it is not. You select LAYER from the main menu at the top of the window, and then UNMIX STEMS. You then get the pop up window shown below:

Spectral Layers 7 - 003.jpg


Once you select which stems you want to extract, you just press OK and the PC/Software does the rest. My desktop tower is fairly powerful, and it did not take but a minute or two to extract the stems from the Al Stewart song "Song on the Radio", as shown here:

Spectral Layers 7 - 004 Stems Created.jpg



What's cool is, as you can see from the jpg above, is that each stem is shown in the spectral picture of the wav file in a different color, as indicated in the LAYERS box in the lower left of the above jpg. You can use the MUTE or SOLO buttons to add or remove what's displayed in the spectragraph.
 
Last edited:
I found it very interesting to see the audio split out this way. It's cool to see it visually. Now, if I knew what I was doing, I could go into each layer and edit it to make it sound even better.

Spectral Layers 7 - 010 All Stems.jpg


So, what's it sound like "as is?" Well, here are two examples, less than 1 minute each, of the resulting audio from the files above, which is the Al Stewart song "Song on the Radio"

The first clip is the start of the song, the second is from the middle of the song. I did no tweaking or futzing with levels or EQ at all. This is what I ended up with using everything as default.


Here's a link to a folder with the two clips. If the Sax is not in the rears then you are not hearing the 5.1.
Your best bet would be to download them and play them if that's the case.

https://mega.nz/folder/zMpQXA7Z#2sma16q1Ptr_58PLEWT_KA
 
I LOVE this program.....Yes it is a bitch to learn but any program this powerful is gonna be tough at 1st....Thanks for the heads up Jon about the update....I need to update my version now!

Mike,

For $79 it's a steal! I am promising myself to spend some time with the tutorials and the YouTube stuff to try and actually learn a bit about this program. While it does seem intimidating, it looks like you can really get into the nuts and bolts of a file with it.
 
Wow, that's an impressive program

It really is. I was watching a few of the tutorials on youtube, and in one of them, they claim that if you unmix a song and do not adjust the levels of any of the stems, just rearrange them and put them back together, there will be no artifacts at all.

I am going to try this out some more tomorrow afternoon. This could be pretty cool. If I come up with something that knocks me out I will post a less-than-a-minute sample here in the forum for y'all to hear.
 
Thank you Jon! Very interesting. I bought the very first Sony version of SpectraLayers Pro 1.0 and, as you described, found it required more of a learning time commitment then I felt like piutting out. I lost me computer and the back up recently and went to relaod as I still have the key. That's when I found Migix to reload ALL my other Vegas and Sound Forge products but no support for SpectraLayers. I just gave up but with these encouraging words on the new version, I'm trying get a reload and possible upgrade. My old key was not recognized by the Steinberg site so I sent a support request. Fingers crossed. Keep up the great detective work so I can slide into using an updated version after all YOUR hard work :)
 
I would be interested in knowing:
a) how simple/intuitive/impossible it would be to recalibrate the stem filter settings for the different timbres of instruments used
b) if there are further stem filter settings for other instrumentation, such as a cappella choir, string quartet or jazz combo
c) what sort of processing power would be optimal for, say a nine-minute track, or a 35-minute track
d) if you could, let's say, gate the drum sound tighter on the "drum" stem filter, so the reverberations and other qualities spread out over the soundstage
e) if you've done any simpler adjustments with it, such as goosing the rears (pardon my french) in the "Sgt. Peppers" 5.1 mix (or, "Lotus" quad)?
 
I would be interested in knowing:
a) how simple/intuitive/impossible it would be to recalibrate the stem filter settings for the different timbres of instruments used
b) if there are further stem filter settings for other instrumentation, such as a cappella choir, string quartet or jazz combo
c) what sort of processing power would be optimal for, say a nine-minute track, or a 35-minute track
d) if you could, let's say, gate the drum sound tighter on the "drum" stem filter, so the reverberations and other qualities spread out over the soundstage
e) if you've done any simpler adjustments with it, such as goosing the rears (pardon my french) in the "Sgt. Peppers" 5.1 mix (or, "Lotus" quad)?

Hey DillyDipper!

I don't have a lot of time for PC today (family stuff), but I did mess around a bit with it this morning. There are two extraction modes, you can extract the stems, but you can also extract 'components'. There is a tutorial on this where a guy dissects the percussion stem into different bits. I have not tried that.

I doubt that you can pull out a soloist from a choir recording without serious artifacts, but I would guess you'd have to be really familiar with the software to try. Me? I am a total clueless newbie at this program. It's always intimidated me and because they have added more AI and more tools, a rookie like me can do more than I could before.

You can definitely "goose the rears" of a 5.1 file, but you could do that with any wav editor. As for modifying the stem filter settings, I don't know.

Like I said, I am just fumbling with this thing right now. If I continue with the tutorials I may be able to answer more questions. I would say that if you are interested in this things, have $350 and the determination and time to learn it, it might be worth it for you to check it out.

I will say that the upgrade costs are fairly reasonable from Steinberg. Like I said, the upgrade from 6 to 7 was $79, and I noticed that the upgrade cost from versions 1-5 was $199. But again, you do need the eLicenser.

NOTE: I just noticed that the upgrade price for 1-5 was moved up to $199.
 
Thank you Jon! Very interesting. I bought the very first Sony version of SpectraLayers Pro 1.0 and, as you described, found it required more of a learning time commitment then I felt like piutting out. I lost me computer and the back up recently and went to relaod as I still have the key. That's when I found Migix to reload ALL my other Vegas and Sound Forge products but no support for SpectraLayers. I just gave up but with these encouraging words on the new version, I'm trying get a reload and possible upgrade. My old key was not recognized by the Steinberg site so I sent a support request. Fingers crossed. Keep up the great detective work so I can slide into using an updated version after all YOUR hard work :)

I think you can get your key. The upgrade page for version 7 actually lists version 1 as a valid upgrade source.

Uprade Spectralayers.jpg
 
Here is a look at each stem in the spectragraph:

VOCALS
View attachment 55484

PIANO
View attachment 55485

DRUMS
View attachment 55486

BASS
View attachment 55487

OTHER
View attachment 55488


When you have the stems extracted, you can save them as individual wav files by selecting EXPORT ACTIVE AUDIO LAYER
Studying these pics tonight Jon, I was wondering what the "Blue - Other" stem contained, it does seem a little less focused and more continuous as the others? And did you listen to that Other stem by itself to hear any cross artifacts? Thanks.
 
Back
Top