It was good to read such a candid review of their application by the product developers. Unquestionably, there are many challenges presented, and it's always beneficial to learn what goes on 'under the hood'. It gives us a better understanding of why some results are the way they are.
BIAB & RB2025 Win.(Audiophile), Sonar Platinum, Cakewalk by Bandlab, Izotope Prod.Bundle, Roland RD-1000, Synthogy Ivory, Kontakt, Focusrite 18i20, KetronSD2, NS40M Monitors, Pioneer Active Monitors, AKG K271 Studio H'phones
For those that may not have noticed, Dan is quite plugged-in and willing to share what he learns, thanks Dan. Two things jump out to me:
Originally Posted by DrDan
We trained our algorithm on high-resolution audio sources so I'm not entirely surprised that we didn't cleanly separate the bass from the lower quality YouTube audio. This I addressed earlier when I said I believe that PreSonus chose [wisely] to train on "messy", real-world compressed files, the lowest common denominator.
Stem separation is still very much in the early stages of technology development and will continue to get better over time. This is absolutely true.
Kudos to Mr. Schnurrenberger for being open, I certainly wish him and his team the best with their product. Dan, if you think it helpful, feel free to share my SoundCloud bass stem separated with Studio One with him.
https://soundcloud.com/user-646279677 BiaB 2025 Windows For me there’s no better place in the band than to have one leg in the harmony world and the other in the percussive. Thank you Paul Tutmarc and Leo Fender.
I’m glad you picked up on that, since he is saying there is a difference in the results based on the quality of the audio source as I had suspected. Perhaps the next step as I mentioned is to have options to configure these programs for audiophile input versus commercially squashed input. And such an option would cater to two very different uses of this software.
BIAB 2025 Win Audiophile. Software: Studio One 7 Pro, Swam horns, Acoustica-7, Notion 6, Song Master Pro, Win 11 Home. Hardware: Intel i9, 32 Gb; Roland Integra-7, Presonus 192 & Faderport 8, Royer 121, Adam Sub8 & Neumann 120 monitors.
Hmmm, I see different modes of thinking/needs on this (which is why forums can be effective) and 3 possibilities: 1. A prompt for the user to specify whether the audio file is of high or low quality 2. Two entirely separate programs, one for low quality files and another for high 3. No prompts needed because the AI can handle any quality input (I've been foccused on this one)
Options 1 and 2 would require the training of two distinct AI models, but all 3 would have pros and cons.
https://soundcloud.com/user-646279677 BiaB 2025 Windows For me there’s no better place in the band than to have one leg in the harmony world and the other in the percussive. Thank you Paul Tutmarc and Leo Fender.
I just did a fast test of a song I’m working with, comparing stem separation in Studio One Pro (7) and Song Master Pro.
The bass sounds much better - fuller - in Studio One. In Song Master Pro it is more like a ghost of a track.
I have not yet found the way in Studio One to capture this bass track as MIDI, which I can do in Song Master Pro. If it exists, I haven’t found it in the Help.
Matt one way you can do this is select the bass track and choose edit with Melodyne. From melodyne file menu select save as midi file. Then just load the midi file into a track. I have not upgraded to Version 7 yet so I have not had a play around with it yet
BIAB 2025 Ultrapack- Studio One Pro 7 Windows 11, Mac Mini M4 with Logic Pro 11, Melodyne Studio
Thanks Brian. Yes, that would work. I was hoping Studio One Pro 7 might have the ability to export MIDI from stems built-in, as Song Master Pro does.
If you have Toontrack's EZBass, click on the Audio Tracker tab and drop the monophonic bass stem into the box on that screen, and EZBass will create a MIDI file for the audio stem.
ThinkPad i9 32GB RAM 7TB SSD; Win11 Pro; RME Fireface UCX II; BiaB 2025 Ultra Bitwig Studio 5; Studio One Pro 7; Melodyne Studio 5; Acoustica Premium 7 Gig Performer 5; NI S61 MK3; Focal Shape 65; Beyerdynamic DT 880 & 770
I just did a fast test of a song I’m working with, comparing stem separation in Studio One Pro (7) and Song Master Pro.
The bass sounds much better - fuller - in Studio One. In Song Master Pro it is more like a ghost of a track.
I have not yet found the way in Studio One to capture this bass track as MIDI, which I can do in Song Master Pro. If it exists, I haven’t found it in the Help.
Matt one way you can do this is select the bass track and choose edit with Melodyne. From melodyne file menu select save as midi file. Then just load the midi file into a track. I have not upgraded to Version 7 yet so I have not had a play around with it yet
Another way to do this is to edit the audio track with Melodyne (Ctl-M) and then drag and drop the audio track onto an instrument (ie, MIDI) track. It's pretty slick...
DC Ron BiaB Audiophile Presonus Studio One StudioCat DAW dual screen Presonus Faderport 16 Too many guitars (is that a thing?)
Thanks for these great suggestions. I know BIAB pretty well, but thirty-plus years of using a hardware MIDI synth has put me behind many of you who are skilled in software synths and related tools.
BIAB 2025 Win Audiophile. Software: Studio One 7 Pro, Swam horns, Acoustica-7, Notion 6, Song Master Pro, Win 11 Home. Hardware: Intel i9, 32 Gb; Roland Integra-7, Presonus 192 & Faderport 8, Royer 121, Adam Sub8 & Neumann 120 monitors.
I wouldn't base the quality of a stem separator based on the quality of just one instrument on one track.
Wise words...
I now have a 2nd datapoint. I extracted my own bass line from my most recent composition Down Periscope
Several months ago, I originally recorded this bass in Studio One in ver 6.something, and I used a version of my mix that contained no compression. I consider this another tuba test because it contains low frequency submarine rumblings. It did incorrectly extract those rumbles but it also extracted my bass clearly enough to easily percieve the melody, tone and rhythm of my playing; including some timing flubs on my part
Of course anything is possible with AI models including hallucinations. But AI researchers are making big strides in solving the hallucination problem. The proof is always in the pudding and the pudding is saying PreSonus and their partners hit a home run and I'm expecting further improvements yet before ver 8 arrives.
If anyone is able to cause this stem separator to clearly fail, do share your results.
On a separate note, I was quite proud of myself in being able to capture in a 2 page, 16 step process how to create tempo maps in S1 based on a Gregor YouTube . . . another solid piece of PreSonus software engineering in its own right. From what I can glean so far, this lengthy process has been reduced to a few mouse clicks in ver 7.
https://soundcloud.com/user-646279677 BiaB 2025 Windows For me there’s no better place in the band than to have one leg in the harmony world and the other in the percussive. Thank you Paul Tutmarc and Leo Fender.
I wouldn't base the quality of a stem separator based on the quality of just one instrument on one track.
Wise words...
I now have a 2nd datapoint. I extracted my own bass line from my most recent composition Down Periscope
Just to clarify, I'm not saying you're wrong, maybe you're right (and Presonus algorithm is the best one currently availiable for bass), maybe not, I really have no idea. What I want to point out is that drawing such a general conclusion from such a particular case makes little or no sense to me. If someone really wants to compare the different algorithms available on the market to figure out which oneis ‘the best’ for bass, then it would be necessary to compare not one or two examples, but a large number of them, with different audio qualities, different mix qualities, from different eras, different compression ratios, different styles of playing, different bass sounds and techniques, different panning, etc. And of course, if this job is going to be made by different people, the test files must be identical for all of them; posting a youtube link and expecting that everybody else will download the same mp3 with the same quality from that link is probably not a realistic idea.
Bearing in mind that (as far as I know) the vast majority of available algorithms actually use the same common code (derived from the Open Source Spleeter -or, in some cases Demucs-) and differ from each other basically in terms of the amount and quality of material they have been trained with, the result of such an experiment would probably be that there is no ‘best’ algorithm in a broad sense. Some of them would simply work better in some cases than others, and viceversa.
It seems common sense to me that an AI trained mainly with classic jazz material will do a better job separating Paul Chambers' double bass line ion "So What" than another AI trained mainly with 70s funk. I think that the next generation of stem separation programs will allow the user to choose between different trained models, depending on the starting material.
I wouldn't base the quality of a stem separator based on the quality of just one instrument on one track.
Wise words...
I now have a 2nd datapoint. I extracted my own bass line from my most recent composition Down Periscope
Several months ago, I originally recorded this bass in Studio One in ver 6.something, and I used a version of my mix that contained no compression. I consider this another tuba test because it contains low frequency submarine rumblings. It did incorrectly extract those rumbles but it also extracted my bass clearly enough to easily percieve the melody, tone and rhythm of my playing; including some timing flubs on my part
Of course anything is possible with AI models including hallucinations. But AI researchers are making big strides in solving the hallucination problem. The proof is always in the pudding and the pudding is saying PreSonus and their partners hit a home run and I'm expecting further improvements yet before ver 8 arrives.
If anyone is able to cause this stem separator to clearly fail, do share your results.
On a separate note, I was quite proud of myself in being able to capture in a 2 page, 16 step process how to create tempo maps in S1 based on a Gregor YouTube . . . another solid piece of PreSonus software engineering in its own right. From what I can glean so far, this lengthy process has been reduced to a few mouse clicks in ver 7.
Studio One 7 has a brand new tempo map detection that is so much easier than before. You are sure to like it.
BIAB 2025 Ultrapack- Studio One Pro 7 Windows 11, Mac Mini M4 with Logic Pro 11, Melodyne Studio
Just to clarify, I'm not saying you're wrong, maybe you're right (and Presonus algorithm is the best one currently availiable for bass), maybe not, I really have no idea. What I want to point out is that drawing such a general conclusion from such a particular case makes little or no sense to me. If someone really wants to compare the different algorithms available on the market to figure out which oneis ‘the best’ for bass, then it would be necessary to compare not one or two examples, but a large number of them, with different audio qualities, different mix qualities, from different eras, different compression ratios, different styles of playing, different bass sounds and techniques, different panning, etc. And of course, if this job is going to be made by different people, the test files must be identical for all of them; posting a youtube link and expecting that everybody else will download the same mp3 with the same quality from that link is probably not a realistic idea.
Bearing in mind that (as far as I know) the vast majority of available algorithms actually use the same common code (derived from the Open Source Spleeter -or, in some cases Demucs-) and differ from each other basically in terms of the amount and quality of material they have been trained with, the result of such an experiment would probably be that there is no ‘best’ algorithm in a broad sense. Some of them would simply work better in some cases than others, and viceversa.
It seems common sense to me that an AI trained mainly with classic jazz material will do a better job separating Paul Chambers' double bass line ion "So What" than another AI trained mainly with 70s funk. I think that the next generation of stem separation programs will allow the user to choose between different trained models, depending on the starting material.
Cerio, I have no time or desire to do the exhaustive investigation work that you propose. The two conclusions that I am drawing are as follows:
1. Studio One Pro ver 7 has produced a state-of-the-art stem separation capability for bass that is second to none. I now have a wonderful new tool in my toolbox that meets my needs in this regard.
2. No one on this forum has demonstrated a program that has a bass stem separation capability superior to that of Studio One Pro ver 7.
Because of these conclusions I'm moving on to other things.
However, I encourage you to personally perform the very comparison work that you propose. I’ve given you a head start on this with a list of programs at the top of this thread. I’ve even given you a good “tuba test case” that you can use for part of your comparison work. I also encourage you to share your results and conclusions here on the forum. And if at the end of your study you find that your conclusions differ from mine, great. Learning something new is good.
My instincts rarely fail me and they are telling me that you will decline to do this work. My hope is that you can prove my instincts wrong.
PS> A spreadsheet may be a good way to organize your results.
https://soundcloud.com/user-646279677 BiaB 2025 Windows For me there’s no better place in the band than to have one leg in the harmony world and the other in the percussive. Thank you Paul Tutmarc and Leo Fender.
[quote=Cerio] However, I encourage you to personally perform the very comparison work that you propose. I’ve given you a head start on this with a list of programs at the top of this thread. I’ve even given you a good “tuba test case” that you can use for part of your comparison work. I also encourage you to share your results and conclusions here on the forum. And if at the end of your study you find that your conclusions differ from mine, great. Learning something new is good.
My instincts rarely fail me and they are telling me that you will decline to do this work. My hope is that you can prove my instincts wrong.
Yes you're right, I won't do that job, mainly because I don't really have time / need to, but also because my instincts (which rarely fail me either) are telling me the result of such an experiment wouldn't be too far from what I've just described above, for the reasons given above.
Having said that, I've also made some tests in the past comparing these kind of tools, here's just an example:
IMO, you can hear some differences between those three programs in that particular case, but as I said before, a test based in a single use case, while interesting and useful, is simply not enough to draw any serious conclusion about the quality in general of these tools.
It’s way too early in this game to draw conclusions. Let’s revisit this after a few months of working with these programs.
I disagree.
Think of conclusions like weather forecasts. Just as meteorologists make predictions based on current data, our conclusions are based on the information available at the time. However, as new data comes in, the forecast can change. Similarly, as we gather more information about software tools or as circumstances evolve, our conclusions may need to be updated; and often are.
At one point in my musical journey I concluded that Audacity would be my final DAW; then I concluded Reaper would be, then a friend introduced me to Studio One and my conclusion changed once again. It's a process of learning and growth.
This also appears in the animal kingdom. Ever watch a lioness or cheetah stalk a gazelle? She's updating her conclusions continually as she processes her inputs. The process is fascinating to watch
It's never to early to draw conclusions, and often never too late to change them.
https://soundcloud.com/user-646279677 BiaB 2025 Windows For me there’s no better place in the band than to have one leg in the harmony world and the other in the percussive. Thank you Paul Tutmarc and Leo Fender.
So if a conclusion was reached now, what would it be?
(This is not in any way meant to be challenging, I really don't know, as I have not used any of the programs, but if a conclusion needed to be reached now, I would be keen to know what it would be. This thread has enjoyed significant input and much interest.)
BIAB & RB2025 Win.(Audiophile), Sonar Platinum, Cakewalk by Bandlab, Izotope Prod.Bundle, Roland RD-1000, Synthogy Ivory, Kontakt, Focusrite 18i20, KetronSD2, NS40M Monitors, Pioneer Active Monitors, AKG K271 Studio H'phones
Ok, using that line of thinking, I have made a preliminary conclusion that I cannot yet conclude one program is better than the other when comparing stem separation of Studio One Pro (7) and Song Master Pro. Was that helpful?
BIAB 2025 Win Audiophile. Software: Studio One 7 Pro, Swam horns, Acoustica-7, Notion 6, Song Master Pro, Win 11 Home. Hardware: Intel i9, 32 Gb; Roland Integra-7, Presonus 192 & Faderport 8, Royer 121, Adam Sub8 & Neumann 120 monitors.
Update your Band-in-a-Box® 2025 for Windows® Today!
If you’ve already purchased Band-in-a-Box® 2025 for Windows®, great news—a new update is now available! This update introduces a handy new feature: a vertical cursor in the Tracks window that shows the current location across all tracks, and more.
Video: Band-in-a-Box® 2025 for Windows®: Boot Camp: The AI Lyrics Generator
With Band-in-a-Box 2025® for Windows®, we've introduced an exciting new feature: the AI Lyrics Generator! In this video, Tobin guides you step-by-step on how to make the most of this new tool.
Band-in-a-Box® 2025 for Windows®: Boot Camp: The AI Lyrics Generator video.
Video: Band-in-a-Box® 2025 for Windows®: Using VST3 Plugins
Band-in-a-Box 2025® for Windows® now includes support for VST3 plugins, bringing even more creative possibilities to your music production. Join Simon as he guides you through the process in this easy-to-follow demonstration!
Video: Band-in-a-Box® 2025 for Windows®: Using VST3 Plugins
Video: Band-in-a-Box 2025 for Windows: Using The BB Stem Splitter!
In this video, Tobin provides a crash course on using the new BB Stem Splitter feature included in Band-in-a-Box 2025® for Windows®. During this process he also uses the Audio Chord Wizard (ACW) and the new Equalize Tempo feature.
Video: Band-in-a-Box® 2025 for Windows®: Using the BB Stem Splitter
Check out the forum post for some optional Tips & Tricks!
Congrats to Misha (Rustyspoon)…downloaded/installed a full Audiophile 2025!
Breaking News!
We’re thrilled to announce that Rustyspoon has made PG history as the very first person to successfully complete the download and install of the full Band-in-a-Box 2025 Windows Audiophile Edition (with FLAC files)—a whopping 610GB of data!
A big shoutout to Rustyspoon for stepping up to be our test "elf!"
With the launch of Band-in-a-Box® 2025 for Windows, we're adding new videos to our YouTube channel. We'll also share them here once they are published so you can easily find all the Band-in-a-Box® 2025 and new Add-on videos in one place!
Whether it's a summary of the new features, demonstrations of the 202 new RealTracks, new XPro Styles PAK 8, or Xtra Styles PAKs 18, information on the 2025 49-PAK, or detailed tutorials for other Band-in-a-Box® 2025 features, we have you covered!
Band-in-a-Box® 2025 for Windows is here, packed with major new features and an incredible collection of available new content! This includes 202 RealTracks (in Sets 449-467), plus 20 bonus Unreleased RealTracks in the 2025 49-PAK. There are new RealStyles, MIDI SuperTracks, Instrumental Studies, “Songs with Vocals” Artist Performance Sets, Playable RealTracks Set 4, two new sets of “RealDrums Stems,” XPro Styles PAK 8, Xtra Styles PAK 19, and more!
Special Offers
Upgrade to Band-in-a-Box® 2025 with savings of up to 50% on most upgrade packages during our special—available until December 31, 2024! Visit our Band-in-a-Box® packages page for all the purchase options available.
2025 Free Bonus PAK & 49-PAK Add-ons
We've packed our Free Bonus PAK & 49-PAK with some incredible Add-ons! The Free Bonus PAK is automatically included with most Band-in-a-Box® for Windows 2025 packages, but for even more Add-ons (including 20 Unreleased RealTracks!) upgrade to the 2025 49-PAK for only $49. You can see the full lists of items in each package, and listen to demos here.
If you have any questions, feel free to connect with us directly—we’re here to help!
One of our representatives will be happy to help you over the phone. Our hours of operation are from
6:00AM to 6:00PM PST (GMT -8) Monday thru Friday, and 8:00AM to 4:00PM PST Saturday. We are closed Sunday. You can also send us your questions via email.
One of our representatives will be happy to help you on our Live Chat or by email. Our hours of operation are from
6:00AM to 6:00PM PST (GMT -8) Monday thru Friday; 8:00AM to 4:00PM PST (GMT -8) Saturday; Closed Sunday.
Dansk
Deutsch - Windows
Español
Français - Windows
Italiano - Windows
Nederlands
Polski
Português
Svenskt
русский
Japanese
Simplified Chinese
Traditional Chinese
Korean