Demix Music Track Yourself Using This Incredible A.I. Stem Separator
Вставка
- Опубліковано 11 вер 2024
- In this video we share a way you can separate individual stems from both mono and stereo Beatles tracks by using a remarkable online A.I. tool. We also look at a superb new book for Beatles collectors about their UK EPs.
Click here to go to LALAL.AI: www.lalal.ai/?...
Audicity free audio software: www.audacityte...
Popsike (lists ended eBay auctions): www.popsike.com/
Book order information:
Extended Pay In The UK: www.apcor.net/...
Made in The UK: www.apcor.net/...
If you enjoy what we're doing, please consider supporting the channel in any of the following ways:
1. Channel Membership gives you early, ad-free access to our latest videos and exclusive members only videos: / @parlogram
2. Join us on Patreon for updates, exclusive content and behind the scenes stories: / parlogram
2. Click on the 'Thanks' icon below the video to donating an amount of your choice.
3. Browse our new merch in our Spring store: parlogram-auct...
Check out some great sounding Beatles and 60s vinyl on our website: www.parlograma...
If you would like to get in touch with us, you can do so in the following ways:
Email: andrew@parlogramauctions.com
Facebook: / parlogram
X: / parlogram
Instagram: / parlogram67
or via our website: www.parlograma...
Thank you!
Many thanks for the feature on my book Andrew. I'm glad you like it. Features like this are a great boost and make the work worthwhile. It's greatly appreciated to be able to reach a wider audience. I hope your subscribers are as enthusiatic as you are about it.
Seems like a very interesting book. Wouldn't have heard about it otherwise.
You're welcome, Richard! Well done again.
@@user-ht5ed2ij5p Thanks for your order. A friend of mine has just collecting Beatles' records again and blamed my book!
@@user-ht5ed2ij5p Thanks for buying my book. A friend of mine has just started collecting The Beatles again, and blamed my book!
@SteveRestorer we'll see.......
Remember when CDs came out, and record companies fretted about the fact that in terms of quality they were basically selling the masters to everybody? Now, with AI, every Beatles fan can be like George Martin, and make their own original masters. We've come a long way, baby.
You don't have the raw recordings, just the processed, compressed versions. It's better to use the Rock Band tracks but they're lower quality. I like the AI because you can tweak older recordings where the levels might be off.
I remember people saying CDs weren't as good as vinyl, but vinyl deteriorates, the more times you play it.
@@sharpvidtube there's CD rot affecting older releases but I don't know if it's happening to the newer types of discs.
Could do with this to get rid of yoko onos screaming that ruins John Lennon's Baby please don't go and anything else that's irritating such as that also...
Yep, so they started trying to make the CDs lower quality than the masters (eg. brickwalling). And then they tried to phase out the format entirely so you couldn't own a digital copy of an album in physical form. Luckily it looks like they've failed, at least in the case of the latter point.
I remember being blown away when I encountered the Rockband stems 10 to 12 years ago, but being disappointed at the inability to cleanly separate the drum and bass parts on the earlier recordings. Incredible how technology has progressed in that short period.
I was just thinking about those Rockband MOGGS! Scary indeed: Now you can separate anything and clone voices.
That software is genuinely amazing. I’m particularly impressed by the vocal separation - I would expect it to be capable of recognising vocals and separating them from instruments - but the fact that it can also retain the *reverb* and keep that isolated from everything else in the mix is fairly mind blowing.
I put the Classics IV song "Spooky" through the extractor, and was impressed that the backing vocal (wordless singing) was not affected. Really amazing software.
As a professional sound engineer myself, I too was amazed at being able to hear the reverb intact with the vocal track:).
I studied audio engineering back in 1991 in Salford. I DJed and thanks to the Beatles I've always buggered around with mixing and recording stuff as a hobby. So when I got the Rock Band stems years ago and bunged them through my digital recorder I could mess around bettter than before. And now as AI has improved with demixing, I was working on doing the earlier Beatles records myself before I learned of the red and blue official demixes. I've been religiously listening to the red album now.
I've used lalalai and a couple of other online tools, but I ended up buying Izotope as it has a few other audio cleaning and mastering tools I wanted. It's fabulous and I couldn't be happier.
Along with people on UA-cam uploading their own stems now, I've been buggering around with all sorts of music. But Beatles stuff will always remain my go-to.
i used the app to separate tracks on a cd that my band recorded in 1997 and, for cost reasons, we chose to not keep the 2 inch tape, which was an instant regret. this is the first time since then that i’ve been able to hear the vocals by themself. thank you for sharing this ! it’s awesome!
You're welcome!
Oh my gosh you're right I got a couple things I need to do thank you
I’ve played with it a bit and it works amazingly well! Opens the doors to remixes and more “parallel universe” Beatles albums like Everyday Chemistry.
Prepare for a 100 years of original mono recordings to suddenly become available in "true" stereo and multi-channel surround mixes for the first time ever.
I found the AI website a few weeks ago after watching a different video on youtube and I think the best results are if you don't use MP3s. If you own any of your favorite albums from HD Tracks in the highest bit-rate, I noticed that there's less flanging and the separation is a lot clearer but it's still far from perfect. This was a different and groovy video.
Thanks for the tip!
I used Lalalai in the past when I still had the free minutes. It truly is the best stem-separation tool I've used in terms of the quality of the separation.
Unfortunately it still suffers from pumping with some tracks, and at the moment can't handle fully electronic tracks. However, it did do a reasonable job of processing and producing separate stems for Ultravox's The Voice when I tried it.
DMix seems better. I use La to remove vox
@@darrell6800 searching google for “dmix separation” and similar terms doesn’t provide any results, are you sure you didn’t misspell it
Is it better than FL Studio's built-in stem separation feature?
@@Fredrik-iz4ou I don't use FL, so I wouldn't know.
Wow! This is SO cool, Andrew! From listening to the Beatles on 45’s in 1964 to separating parts of the songs on a computer. I would never have imagined these advances. I’m going to check out this amazing software!
Unglaublich! Ich kann mich noch erinnern als ich ein Teenager war und die Beatles den ganzen Tag hörte auf Schallplatten.😃
Brilliantly informative video as usual. The possibilities are endless. Many thanks Andrew.
Indeed Steve!
I just used this programme to fix the hard panning on some of their early stereo tracks, and to bring out the bass and drums a bit. Very impressive technology, even at this early stage.
I think it would be pretty cool to see what you can achieve with it over the coming months, Andrew, as a bit of fun!
People have been using this type of spectral demix tech for at least 5 years to remix the Beatles. I've remixed my favorite Beatles songs putting the drums cleanly in the middle of the stereo. Just doing that makes an incredible improvement with the older stereo mixes. There's also tools to reduce reverb. LALAL is an excellent tool.
I'm not a big supporter of AI but I could see this helping me with my own music somehow
@TheAlliesTheBand Same here. 😺
I've tested the A.I. website you talked about and I was just flabbergasted by it. I've tried it with many artists and I was sometimes just so overwhelmed by the beauty of just hearing the voice that's isolated, it often brought tears to my eyes. Thank you so much for making me discover this technology Andrew. Cheers from Quebec Canada.
Glad you enjoyed it!
Andrew, this is great stuff. I wish you would have used one of the songs, from the star club tapes for a test! Thank you for all your hard work.
You continue to amaze me with every video. Keep up the Fantastic work!!
Thanks Michael!
I've had Audacity for a long time, never thought of doing this. Thanks Andrew for the valuable tip.
I've used Audacity since 2002. It has its strengths and weaknesses when it comes to mixing multi-tracks. I can remember mixing with a 4-track in the 80s... a single track could be panned "live" back-and-forth between the left and right channels when making the stereo mix master by simply rotating its pan-pot manually back-and-forth - if that's what you wanted to do. I've yet to figure out a way to easily do that with Audacity.
Great shout out for the ai software. I can now remix my own master as the multitracks have long since disappeared!
I was thinking the same. I recorded a lot of music with a friend in 1985 using his 4-track cassette deck and have only had the final stereo mixes from the 4-track masters which only he could use. Now I can demix my own stuff from 40 years ago and fix some things that were previously burned-in for all these decades.
As a musician, bass player mostly, its a great tool to isolate bass lines to help learn them for parts you cant hear clearly ..I was using la lalai the other day.
Well now…a different, yet fun, episode!! Yes the AI can be fun in this instance I totally agree! After years of working with my software for removing vocals and/or making my own “underdubbed” versions of tunes for myself (like Abbey Road and MMT), I’ve been enjoying some of these type things on UA-cam.
If this tool you featured today can do that without phasing effects or miserable piercing hi-end then I’m in!
The books look great! And I have the US equivalents (Bruce Spizer) that are just beautiful.
Thanks for the inspiring episode! I’ll dust off my ears and give it a go!
Glad you enjoyed it, Brian. Do give it a go, it's great fun!
For myself, I might use something like this if I just wanted to "deep dive" into one aspect of a Beatles song...vocals, guitars, bass, drums...and REALLY hear what's going on. But as far as making my "own" re-mixes, I think I wouldn't want to do that. My mind would "wrestle" with the integrity of how it was originally conceived and recorded by The Beatles. But again, to isolate one stem and by itself would be fun.
Ah, “Beware the Weighty Tome!”. Thanks for this one, Andrew! Although I’ve built actual studios over the last half-century for a number of concerns (Radio, Television, Recording and even myself)… some of which are in operation right this second, this AI software sounds like a ton of fun. I’ll be trying it out as soon as I have a chance, although it won’t quite be the same as mixing on my large 8-Bus Trident 32-8, designed by the great Malcolm Toft. It (and its ancestors & descendants) mostly made their debuts at Trident Studio, in London’s West End. The Beatles and The Rolling Stones are each known to have recorded at Trident, when they couldn’t get the room they wanted at EMI Abbey Road (Mr. Toft sold his studio in the late 1980’s and went into the console business full-time). I prefer working on big mixing consoles to mixing “in the box” (in a computer) but hey, I’m CERTIFIED Old School!
Given the evolution of AI thus far, this will probably be free before summer.
It is free. It's called DEMUCS
It is free, there's a website called MVSep that does it great for free.
@@CarlosPCmxthat one isn’t very good
@@TheBeatlesTracks Eh, you get what you pay for
Already is. There's dozens. Just Google "stem creator".
I was on a work assignment in London in 1992. One evening in the hotel room, I watched a BBC TV show called "The Making Of Sgt. Pepper", featuring George Martin at a studio mixer playing around with, and commenting on, the multitrack recording of that album - isolating, muting, etc. different parts throughout the album from the 4 (or was it 8?) tape tracks. Oh my! I still remember thinking "_I_ want to be able to do that myself!" As some others have commented here, I'd done some primitive stuff like that with individual stereo channels, and mixing them out of phase and copying those interesting "mixes" to tape (usually muting out the center-panned vocals and enhancing the instruments). In the intervening years, many decent-quality multichannel stem files have appeared on the web, derived in various ways, including from the Rock Band or Guitar Hero games. Those were available as mogg files, easily played and remixed with Audacity. We've also had some music (Beatles and others) available in 5.1 mixes to play with from DTS, SACD, DVD-Audio, and Blu-ray discs. [Those had digital copy protection, but they could still be copied in the analog domain.] [And also Dolby Atmos, which I haven't gotten into myself.] Now this latest step with LALAL.AI and other software, at least similar to what Peter Jackson used for the Get Back movie - and they'll surely get better, more sophisticated, and cheaper quickly. Great ways to hear many parts of the music that often "get lost in the mix" on the original commercial releases, which is what I like about it. Thanks Andrew!
make your comment a bit longer
@@JoopZweetsok I could have, but I trimmed it down for brevity. 😁
It was ITV.
I actually use this tool already for my own songs. The other day recordet the voice for our new song with our vocalist at her garden house on the cell phone, without backing track or anything, playing along with an acoustic guitar. Later on I could isolate her voice 100% and use it for the proper demo.
In freeze-frame, I just learned everything I wanted to know about the elusive Wall's Ice Cream record. Thank you.
One of the great joys of listening to The Beatles was muting one speaker and listening to the other by itself. Produced some odd effects, like the "oooooh-ooh-wah-wah-wahs" of "You Won't See Me," or "Revolution" without half the guitars, or "Lady Madonna" without the piano. I'm not likely to do it myself---I really haven't got the gear to support it---but, now that I know where it's coming from, I'm more likely to catch these as they appear on UA-cam.
(Thanks, too, Andrew, for bringing me an experience of The Beatles that I had never even thought of.)
i hate the use of ai in art but this is actually a good use of it! not meant to replace or create art, but let people play with it in entirely new ways. great video as always!
Great tool. I've used it to separate old tracks of my own where the tapes have been lost to time... although I couldn't do much with the tapes if I had them as the old machines are way beyond my budget!
I have used LALAL for months now. It is fantastic! You can turn mono tracks into stereo, remix tracks to your personal taste, and even clean up some bootleg recordings.
I did have some trouble trying to remove Linda McCartney's daughter Heather's voice in 'Told You Before' and 'Dig It.' The limitation appears when the track is buried in an overpowered primary sound. George Harrison's voice/vocals are typically buried in the tapes as well.
I was able to pull what I think is George Harrison's guitar work to the front on 'Now and Then.'
I wonder how close LALAL compares with MAL?
Now to try and pull the screaming out of some concerts!
Excellent topic Andrew! Thank You!
Glad you enjoyed it!
@@Parlogram It's incredible software. I've given it a try by taking The Doors' American Prayer posthumous album and separating Jim's poetry and the surviving Doors' backing tracks which is something I wanted to be able to do for thirty years and it worked flawlessly.
Aside from being able to remix the tracks, this AI separation allows you to hear otherwise buried backing or harmony vocals. For example, I was amazed to hear that it is in fact Lennon singing his own high harmony on the "Shine, the weather's fine" parts on 'Rain'. My money was on McCartney there, as you would expect. But no. Fascinating.
I pre-ordered the book a while back and it arrived promptly on printing last week. As with other books and items from Apcor it is wonderful. An astonishing level of detail and well produced.
Would love a video of you sharing some of the mixing decisions you played around with and liked!
Good idea Joseph, but it's difficult to get more than a few seconds of music in a video before it is blocked.
That AI software is incredible, Andrew! The book on the Beatles UK EPs also looks vey interesting so thanks for bringing these two items to our attention!
Fascinating technology and cannot wait to try it. Well done, Andrew. The Noller book (on one of my personal favorite Beatles subjects) looks excellent. Will be ordering that! Thank you for showing it. 👍
Hope you enjoy it, Anthony!
@@Parlogram Thank you! PS: I may try to improve a few of those Giles mixes 😉
Interesting video with a great tool for mixing. Just purchased the EP book….. I have the previous Apcor books. Thanks so much Andrew!
It's a great book, Bill. Hope you enjoy it!
Lennon said in his 1980 Playboy interview that he would rearrange/change many of The Beatles songs, only he could have said that and had he still been with us I’m pretty sure he would have been all over LaLaAI - even though he was never one from going rewind, always forward. It’s an amazing tool always thought Rain mixed with She Said She Said would have been a blinding mix.
Ringo’s Rain separated drum roll sounds so much like the Sgt Pepper drum roll, never realised it before. Great update Andrew😎👌
Thanks Steve 😎
Hi Andrew, great video as always. I ordered the book as soon as I heard about it and should be here any day. If it’s half as good as the singles one it will be worth every penny.
Many thanks as always for the great videos. All the best, Mick
Cheers Mick. Glad you enjoy them!
Hi Andrew,
certainly a wonderful tool, the only risk is that if you don’t have the taste and the expertise, let’s say, of a Giles Martin, you could easily end up producing a mess. Interesting video as usual.
Thanks, Andrew! I've seen a lot of videos on UA-cam that used this tool, and I'm looking forward to trying it out!
I'm sure you'll love it!
Ok, i was sceptical and started muttering about reverse phasing one channel and mono'ing left & right. However this is impressive.
Me deja paralizado tanta tecnología. ..será por la nostalgia y el miedo a lo desconocido pero me siento cada día más a gusto escuchando los viejos vinilos monoaurales en un tocadiscos valvular.
Amazing technology! This is what I'd call a good use of AI - not generating SEO clickbait or putting artists out of a job.
Thank you so very much for sharing the information about that AI program that separates vocals from the background music. Apart from the Beatles I'm a big fan of Elvis & loads of other music too, including Hindi Film music ftom the '60's '70's. I have been playing around with this program & the results are truly mind blowing 😮 👌
Glad it was helpful. Amazing isn't it!?
Its amazing how it preserves the reverb and ambience so well.
If I had the time Id love to have a go at doing a LOVE type project.
This technology is just amazing. I always loved Ringo's vocals on "Don't Pass Me By" but I am not a big fan of the arrangement with the fiddle etc. I was able to replace the music with an excellent cover version from the Georgia Satellites while keeping Ringo's original vocal tracks. Now Ringo is truly rocking this tune :) - BTW - Thanks for all the excellent content you keep posting on your channel !
I’ve had the app for a while, it’s really fun to take songs you love and just mess around being specific aspects
Great stuff, just bought the second book... Keep up the good work!
Thanks, will do!
Alright! Looks like I'm getting my own underdubbed mixes of other Paul's records too! Thanks Andrew, I became a member just so I could get your videos a day early.
Thanks! Much appreciated!
Once again many thanks for a fascinating video Andrew. I had my reservations about Ai at first, but after the Revolver, the Red and Blue and the Now and Then releases I have changed my mind to some degree. With the advances probably made with this tech since those releases came out, I can only say that the future of the recording industry is assured, we can look forward to releases from all artists of the highest possible sound quality, which to my mind is good news to all music fans. I'm also a fan of classical music, and to hear what could be done with those recordings is mouth watering. I shall look out for this new book, again this must've took a lot of painstaking, but rewarding work for Richard Moller, so congrats to him. Looking forward to next weeks vid, take care and have a good week.
Glad you enjoyed it, Terry!
Yes, LalaAI is a window into all kinds of wonders. Its very transformative in interpreting the whole process, from song writing to mixing. And if you have your own song demos lying about , you can sort of rediscover how clever you are… or not.
I would love to hear some more of what you can do with the mixes. . Truely amazing how the separations sounded . It just made my imagination buzz lol.
Hi Andrew, a fascinating video! Very interesting book also.Cheers 👍🏼
My pleasure, Fuzzy!
it even added the double tracking..impressive.
The AI website is such a fun tool! Thank you for sharing!
You’re welcome!
Now if the powers that be would release the mastertape digital files. Sure could have some fun! Very cool Andrew thanks. In ten years as AI technology exponetially grows, there will be so many fan remixes (great and terrible) that the originals will, for better and worse, become the rarity.
All of this will end of course because the AI robots will end us before our personal Beatles remixes are complete :(
;)
Thanks again Andrew!
Fascinating as usual Andrew. Using AI I think it’s quite possible, someone could, using any of the Beatles backing tracks or anyones, together with their own vocal mixed together to a record company as their demo. Justin Biebers grandmother basically did that and look what happened to him.
Playing with Lalalai now and loving it! Back in the day it wasn't just cutting out LEFT or RIGHT; we used to cut the vocals by cross-circuiting our speakers. This removes anything out of phase, cutting the center channel which was usually vocals. I made a little box to do this so I could record as well. I still have it somewhere, but now you can do it digitally of course.
I did a similar thing by reversing one channel from a record deck's cartridge using a switch to reverse the polarity and then mono'd it. And hey ho, no vocal!
@@DaveInYateley That's very cool! I never heard of that method. I did rig a cassette player to play tapes backwards, back in the 70s. Basically, the A and B side of the tape are running by the play head at the same time, but the B side is running backwards. I just removed the play head and glued it back on upside down, so it played the wrong side of the tape, and voila!
Fantastic as always! As a "2nd Gen" (b. 1976) superfan now pushing 50, I've always wanted to dive deeper into collecting original vinyl pressings. When I was in middle school, I happened upon an original DJ copy of an early EP (pre-Ed Sullivan I believe) in near mint condition (its A-side showing more wear of course!) at a flea market. Of course not knowing better, I played (and scratched !!!) the heck out of it but it's still a prized possession regardless of any skips and pops. I was unaware of this book until your video so hopefully I can snag a copy from here in the states!
Thanks for this video Andrew. I paid for the 500 minutes and I have been playing all morning from DEVO to Pete Townshend.
It's great isn't it? I can't stop using it.
@@Parlogram It's highly addictive! I'm splitting Motown tracks now haha
Did you get the discount? I got the 500 minutes for half price ($35 USD). It’s still coming up, so I wonder if I can buy another 500 minutes at $35?
As well as muting a channel, you could also extract vocals etc very successfully using out-of-phase stereo before AI came along. I use RipX for stem separation, which is insanely good but also ridiculously expensive. The benefit is that it's yours for life, and all processing happens on your laptop.
You say you have only been using Lalalai for a few weeks. I have been using it for over a year. I still have well over 5000 minutes left. I have been using it to convert a load of my monos to stereos. It is a bit late at night for me to explain how i have been doing it. But I can give info. Because I am a regular customer, they have been giving me loads of discounts. I used it to convert all Beatles' versions of Love Me Do into stereo, and got much better sound quality than Giles Martin who does some pretty awful remixes. I also used it to convert all Beatles Binaural stereos on their Please Please Me and Rubber Soul albums into voice centered stereo. To add to the coincidence, I also use Audacity as my mixing desk.
Bravo! Andrew you have provided us with theultimate Beatles tool - I can't wait to try it!🇬🇧
Enjoy!
Andrew, I used to attempt remixes of Beatles and other songs using L/R channel separation and volume alterations by equalization/frequency ranges, which of course offered limited options and benefits. The AI separatons are high enough quality to overcome those limitations, though at times they would still be unable (I think) to separate out premixed Lead and Background Vocals from each other unless/until the AI learns to differentiate, say, the McCartney, Lennon, and Harrison vocal timbres and then a full remix could be accomplished. People wonder whether this sort of diddling around with the original mixes is legitimate or not--beyond being loads of fun for people like you and me--and I think that because the technological improvements general BENEFIT the accuracy of the original sounds they are legit. Also, Martin, Emerick et al. in the 1960s were working under their own deadlines and put out the best they could manage in the available time. Well, the future being more 'available time' means current remixers can tweak and improve more at liesure and in more ways than the original Abbey Road production team could. If we could have time-transported the 1960s Beatles into a current era recording studio the contemporary production team could have taken their performances to even greater heights, though I think it's always wise to REFER back to the mixed sounds which George Martin was going for on the original recording issues.
One remixing scenario which intrigues me is converting Abbey Road stuidos Beatles to live-soundstage Beatles ambience (Paul and George standing to the left, but we'll give George his own mike/track separate from Paul now; John to the right, and Ringer centered in back, all in a nice and slightly reverberant concert hall stage soundscape). One song of John's in particular which I tried to remix myself by additing Background Vocals, a Bass part, Drums, and Keyboard was his Guitar/Voice demo of "Bad to Me." I used Finale music notation software(!) to do my remix; did MIDI-controller tempo changes to match the drums to John's varying tempo; added in the other instruments as notation, etc. and it too, like "Free As a Bird" and "Real Love" comes alive even better by those additions provided that John's Lead Vocal is kept prominent enough in the new mix. I do hope someone at Abbey Road re-separates and enhances John's Lead vocals from FAAB and RL at some point to better match up with the quality of his Lead Vocal on "Now and Then."
Maybe fifteen years ago I anticipated the development of computer-assisted separation of mixes into their component instruments and voices, but there still remains another partially unachieved (and more artistically questionable) objective for this technology: Sampling an artist's existing Vocal Reccordings to extract their Vocal Timbre and Vocalization/Performance characteristics. The iterations of this technology I've seen let someone convert their own vocal perfmance into a representative vocal timbre belonging to Paul, George,l John, Ringo which has been programmed as an algorithim for use as a processing filter, but the process could also be carried farther along to the point where a computer, programmed with both the timbre and mannerisms of individual singers, could then SING IN THE VOICE OF THAT PERSON ON NEW MATIERAL. Right now, a human being mimicking, say, Paul's vocal mannerisms and singing style can save the step of having to program those aspects into a computer, but I think an entire voice timbre, singing style, and accent/pronunciation could probably be 'learned' by AI from a catalog of existing recordings and then--for better or for worse--Beatles voices could be coupled to instrumental performances by other people for new songs written in Beatles style whereby the computer reads the lyrics and sings in their own voices and vocal characteristics. You'd in effect be able to create new Beatles-like recordings of original songs written by anyone else. I'm not sure doing this would be legit even if acknowledged for what it is and how it was done, and of course there's the artistic and moral issue of using someone's likeness--here, a vocal likeness rather than a visual likeness of an actor--without their prior permission. But technologically it could probably eventually be done. And it could also be easily enough misused in various ways--e.g., for propaganda or other criminality--which I DON'T advocate.
AI stem seperation is really handy. Currently working on an overdubbed and restructured version of Paul's "4th Of July" demo using said tech. Lots of fun really.
This is perfect so I could learn the drums for each song! Thank you Andrew!
Wow, a great tool! Thanks for posting!
Thanks for watching!
Thanks for this Andrew, I have hundreds of hours of cassette recordings of old band rehearsals, there's some genuinely good takes in there that were never that nice to listen to as the mix was just an ambient mic in the corner of the room, I look forward to splitting and remixing them, some may even be worth releasing.
Have fun trying it out!
I have old tracks of my own music that I've lost the original project files for. Now I can remix them!
Again, a fun and interesting video. I can't see myself using the software, but it looks fun. I can see a person using Ringo's drum tracks for their own songs and so forth. Writing a song and having Paul's voice backing you up, like you said all the stuff you could do. Fun. I don't mind a hawking video for Beatles stuff for sale, find it interesting. Thanks again.
I'd imagine we'll want to make sure to use the most up to date mixes for better separation? I tried loading a Beach Boys tune mixed in mono and the results were less than stellar. Remixed stereo masters will most likely work the best...for now.
Very informative video. Thanks so much, Andrew !
Glad it was helpful, Scott!
Great video, thank you! :-)
I already have some wonderful fan-made AI remixes of the Beatles albums that beat the Giles Martin mixes by some degree. But I look forward to applying this software towards non-Beatles albums I love that have less-than-ideal mixes.
I also wonder how well this software will work for live recordings. Many of us have live recordings in our collections that have great performances, but poor mixes. If this can fix (or significantly improve) those recordings, that would be huge.
You say you you want a Revolution? Demixing / remixing is revolutionary. Based on the number of comments you’ve gotten, everyone wants to try it! Do another video in 6 months to let us know your experiences and tricks.
Will do, Lawrence!
A.I. used here is amazing to hear the refinement of the instrument and vocal tracks of
"Penny Lane" and many more songs that anyone would like to dissect. It's
just a field that one must have a good scheduled time to perform.
This is amazing!! Thank you very much for the info and links!👍🏼😎🤩
You are so welcome!
I like spliter ai myself. I just seperated Elvis Presley's vocal from 1968 Comeback and added to a song recorded for the 1977 CBS Special.
Does LALAL .AI split vocals
Great job on this. Like you, I'm all for pulling tracks apart from what's already done and mixed in the tracks. It's another to create Beatles "filters" to program a Michael Jackson song and make it sound like John Lennon was singing it!
I've been using LALAL for a few months. They named the system MAL as a tribute to Malcolm Frederick Evans
Espero que traduzcan mi comentario, porque hablo mal inglés. Hace poco descubrí este canal, y me encanta. Soy fan de The Beatles también, saludos desde Argentina !
Thanks for telling us about the lalalalai
This sounds like the exact type of video I would make! Another more advanced software you can use to do all of this is RipX. I find it does a far better job than anything online. It also allows you to manually go in and edit stuff wich is very useful for seprating vocal harmonys and such. it also gets each individual part of the drums so you can have the kick, percussion and other parts of the drums mixed differently. Its a lot of fun! Im hopefull that one day we will get access to tech as powerful as MAL so we can truly make our own mixes! Great video as always! Love it!
Glad you enjoyed it!
EVERY part of the drum tracks are percussion. Technically, so are pianos and guitars.
That little snippet of "Penny Lane" without the piano reveals other instruments in the background. It's those extra elements, I feel, that make Beatles songs so interesting to this day.
Agreed. There's so much going on in that song which is buried in the final mix.
Interesting video Andrew. I have seen this, or something very similar featured in different video's for different uses. My concern is the abuse that it could be open to by unscrupulous users. Personally, I am not a great fan of the remixes and love the originals in all their beauty, whether flawed or otherwise, and the snippet of Penny Lane you used left me breathless at the sonic perfection of Paul's voice, in both tone and musicality. Not a pitch correction in sight and what a stunning vocal. But also, where would people stand on a legal level as regards copyright? I may be thinking on too deep a level, because as a piece of fun software, it sounds great for use at home, but you know as well as I do how the world works nowadays and there is sure to be abuse of this somewhere along the line and as for copyright infringement or even theft of works?? Does not even bear thinking about. I know we have been there with sampling, but this is a bit different. Anyway, I will get down off my soapbox now! LOL Great work as always
At last we'll be able to remove Yoko Ono from John's solo output!
Great, another video! In the pub today, so no tea and confectionery bar. In fact, it is a bar of another kind! 🥴
Fantastic.
I wonder if this technology s good enough yet to remix more historical recordings, such as early jazz 78s.
Imagine recordings from a hundred years ago in true stereo without any surface noise or crackle.
There are some examples on YT which, once cleaned up and 'treated', sound truly remarkable.
@adiebarrett
Yes, they sound great.
But has anyone separated the instruments and produced stereo versions yet?
I expect it won't be long until someone does, if they haven't already.
The fact you can do this on a handheld device is mind boggling.
Beatles Fan Mixing UA-cam is my favorite side of UA-cam.
Nice episode. Thanks!
A slightly non-specific question on the Beatles. I get shot down regarding this wherever I breach the topic: number of released studio albums. I claim the number to be 11. Most claim 12 and some even go for 13. I claim MMT and Yellow Sub ain't real studio albums, where the guys went into the studio with the aim to produce a new album with new recordings.
MMT music was released on a double EP, then released as an album with previously released singles. Hardly a studio album. Yellow Sub was a soundtrack where they were obliged to record music to sell the a contractually obligated film. This music was then hashed together with George Martin's, very nice, soundtrack. However, hardly a Studio Album...
Amazing technology! I’m not an
Ai hater at all. I love it and use it everyday for my creations: art, music and writing. It’s a tool!
Question: where did you get your audio tracks from? Seems the companies like 🍎 don't allow recording of their songs...
Do remember it is a breach of copyright on recordings, songs and moral rights.. of artists, record companies and songwriters. Personal use yep.. but you cannot sell the results etc. unless it is pre-63 UK recordings (though songs performed still have their own copyright). I have used LaLa.. to separate my late father's vocals from mono records he made way back, something that was impossible before. Some recordings work better than others, as you can get artefacts, drop outs and phasing, or the instruments don't separate as the stems say,. Modern stereo is certainly more effective for the technique. It is rather amazing to hear a dead loved ones' voice come back to you singing as if they are in the room with you! Also you can as you say remix..but don't tamper with copyright thinking you are exempt due to your own creative use of ready made recordings!
There is an AI of John’s voice on free as a bird, which removes the high hiss of John’s voice and sounds natural…
There's (free) DJ software that does the same thing (though not as well), like Virtual DJ. Supposedly the latest iteration of Audacity does it, too, though I haven't delved into it yet.
Andrew, you are the man! Thank you so much for sharing this information with us.
Glad it was helpful, Jeff!
That AI technology looks fab but I am also happy to listen to the Fabs on mono and a bit of stereo too!