Swimming to the top of today’s hyper-crowded pool of talented Djs can be a menacing task. Everyone’s grandma is now djing with auto beat matching software and music blogs are serving up endless supplies of hot tracks that everyone has within a few weeks. In this increasingly level playing field, one of the best ways to set yourself apart from the crowd is to produce custom edits and remixes of popular tracks. You are probably wondering how that might be possible without acapellas or stems (parts of a song). In this article we are going to look at one way to extract stems from a finished song you already have and use those stems to create a custom remix.
Edits, Remixes, Stems Oh My!
Outside of the production world most people think stems come with flowers attached. A stem though, is a group of similar tracks that have been mixed or “bounced” down to a single stereo track. Drums are a good example. Say you started out with five tracks of drums, one for each drum sound. A drum stem is the composite of all those tracks mixed down to a single track. So if you were to be commissioned by a label to do a remix they would most likely send you a folder with files labeled Drums, Vocals, Synths…ect. Those are the Stems with which you make a “remix”. If you don’t have those files you are technically making an “edit”. In our case we are going to make our own stems, then remix them.
Take a look at the Frequency Chart above. This is your key to tearing apart a song. You will also need some sound editing software with a nice EQ. The EQ is your main weapon in separating specific instruments and sounds from a track. Most of what you will be doing is using the high and low pass shelving filters to cut off the sonic ends of your loops which will leave you with just the sounds in that frequency band. For example if you just wanted the Congas you would cut the lows from 30hz to 300hz and the highs from 20khz to 1khz, roughly. If you are not able to get a sound isolated from the rest of the track, try playing around with effects plugins to mangle the loop into a unique sound that can them be mixed back into the group. The goal is to get as many individual “stems” that can then be re-mixed together to create a new result. If possible, avoid boosting the EQ to minimize phasing issues. I’m using Logic Pro but you can get most of this done in Audacity or Ableton.
The first step is to listen to the track very carefully and note anywhere there is a solo instrument cleanly separated from the group. Common examples might be an intro/outro beat or a characteristic synth in the breakdown. Most popular songs have receptive hooks that weave in and out of the mix- your goal is to isolate each hook so they can be recombined in any order you wish.
Scalpel please
There are an infinite number of ways to dissect a track and each song is going to require specific tools to achieve the results you want. The photo above is from the logic session I used for remixing Miike Snow’s “Animal” without actually having the official stems. I picked this track because there were parts where the vocals would be easy to extract with very little instrumentation under them.
Step 1. Separate
First, I went in and sliced the track up into all parts I thought were interesting and might be useful in the remix. These loops are then placed on different tracks so they can be blended together and re-arranged independently.
Step 2. Isolate
After separating the loops out, I used the frequency chart above to isolate the parts of the loops I needed most. One example, cutting the lows and highs from the vocal with a shelving filter.
Step 3. Duplicate
I wanted to replace the main synth hook with something less adult contemporary future muzak sounding so I replayed the main melody into a loop of midi notes that could be used with any synth and bounced these to their own audio track . By filtering out the original synth and placing your own custom creation on top, its easy to create a much fatter sound.
Step 4. Beef it up.
I now had the original track separated into stems and an original synth sound cleanly seperated but I still wanted an animals gone wild sound, sorta like a riot at the petting zoo. I focused on accomplishing this by building more dance floor worthy drums and layering them on top of the parts.
Step 5. Press it!
Miike Snow “Animal” (Nisus Remix) by Nisus
To give you an inside look into this process, I brushed off my production chops and created a quick remix of Snap’s “Rhythm is a Dancer”. Using the technique’s described above I pulled a few choice loops and vocal phrases from the original, using filters to separate the parts. Then I brought in a newer song, Sander Van Doorn’s “Back by dope Demand”, for extra umph and some much needed build ups. With both songs I was able to create room and engineer fake stems just by using filters creatively.
Download the “Rhythm is a Dancer” remix here:
[audio:http://djtechtools.com/Rythym%20is%20a%20Dancer%20%28Ean%20Golden%27s%20Edit%29%201.mp3]Download the Original Ableton Session + Samples Here
Of course, ears are your most important tool… always be listening! The stereo separation of headphones is handy sometimes, but you’ll need a pair of studio monitors to produce a properly balanced mix.
We suggest the Rokit 6 monitors, available in our webstore!
https://soundcloud.com/user-615149736/snap-rhythmisadancer-m-sound-bootleg
Hello, I show you site http://www.ACAPELLAS.eu – Music Producer Area – this is a great source for download acapellas, remix packs/parts/stems, multitrack, Fruity Loops Projects, presets etc.. Every day new uploads.
[…] Remixing Without Stems – DJ TechTools – DJ TechTools The largest community for DJ and producer techniques, tutorials, and tips. Traktor secrets, controller reviews, a massive MIDI mapping library … […]
[…] Remixing Without Stems – DJ TechTools – DJ TechTools The largest community for DJ and producer techniques, tutorials, and tips. Traktor secrets, controller reviews, a massive MIDI mapping library … […]
I know this is probably a little late in the game, but if you didn’t already know you can get a bunch of stems to remix at indiloop.com
both links are dead, please re-up
i have a track that i want to take a keyboard synth part out of…. it is a disco track with funk guitar playing, some drums, and this out-of-tune sounding keyboard synth thing…. is that even possible to do seeing as the piano synth is like right in the middle of the EQ register?
Good tutorial mate..But 1 question : What’s the frequency range for vocals ???
Would love a copy of the links.
Please fix the links.
Mirror for the dead link here:
http://www.2shared.com/file/ZKmCiN7X/Snap_Remix.html
The link to ableton live project and samples is dead…
Very good tutorial. I’m fairly new to Ableton Live, but thank’s to the included tutorials inside Live, I could understand and quickly apply what is taught in this tutorial. Furthermore, I think this mashup could be prepared to be played live by preparing the clips with some enveloppe modifiers and FX directly in the Session view.
And thank’s for the sources, it’s a good idea to make it available.
Can you please send me the ableton project since the link is down ?
Thanx
This is related. I’m a newish DJ and want to get started creating some of my own Edits. The thing is I’m a little confused as to what exactly an edit is. Does anyone have any good recommendations on how to get started or perhaps DJTT has already posted a tutorial, and I missed it. Advice? Thanks.
This is related. I’m a newish DJ and want to get started creating some of my own Edits. The thing is I’m a little confused as to what exactly an edit is. Does anyone have any good recommendations on how to get started or perhaps DJTT has already posted a tutorial, and I missed it. Advice? Thanks.
Hey Everyone!
@ Lee I used Logic Pro’s audio to score function to get the basic midi notes pulled but it took some work getting what Logic missed.
@ Riotthedj I think M/S extraction is a brilliant way of doing that better than phase cancellation since you don’t need the original plus an instrumental to extract the accapella. I’m gonna give that a try next time around.
How did you extract the midi notes from an audio file?
he didn’t thats impossible
Not exactly, there’s Melodyne (though it’s not perfect).
in Live you can slice audio to a midi track. If you set warp markers on each of the main hits of the loop and then slice it, you can play the hits in different orders.
Thanks for sharing this, and yes, as “Guy” says you could do a video!
I second what Guy said – a step by step video for this in Ableton Live 8 would be amazing!
I would love it if you could do a step by step how-to-video on this. Or on creating a remix with Ableton Live in general.
Thanks for the great information.
Guy
Awesome post! I love the Rhythm Is A Dancer remix.
Excellent post. I wish more people shared stuff like this!
What Are The names of the Programs Used i See no mention of this? BTW Great tutorial nisus!!! PM AT loganntx@yahoo.com Thnaks In Advance
[quote comment=”26213″]http://www.youtube.com/watch?v=2GMhViOJVQA
I’m glad you posted this. Unfortunately I had to learn everything on my own.
Here is a example of what ean golden is talking about above. I did this 5 months ago I think.
I would say this kind of work is part of the future of djing.
In a time where everyone plays the same songs each week , this is essential for standing out.
Old style headliners will be left in the dust by new guys if they don’t learn as well.[/quote]
You made me laugh, thanks from an “old” dj 😉
Phase cancellation!
Are you planning to an addendum to this, explaining how to use phase cancellation to generate stems from audio?
http://www.mediafire.com/?nhicqdaqjvh
http://www.mediafire.com/?dzhwgemj4nm
can’t leave out download links oops! :3
http://www.youtube.com/watch?v=2GMhViOJVQA
I’m glad you posted this. Unfortunately I had to learn everything on my own.
Here is a example of what ean golden is talking about above. I did this 5 months ago I think.
I would say this kind of work is part of the future of djing.
In a time where everyone plays the same songs each week , this is essential for standing out.
Old style headliners will be left in the dust by new guys if they don’t learn as well.
Wow, I have the same FM chart taped next to my mixing console
stems are out there, u just have to look for remix comps, etc,
same for accapellas, however I have a techique with isolating parts by muting the stereo or mono sections
with a mid/side eq or compressor, Its not perfect, but I was able to pull out a good accapella on many nu school hiphop tracks, where the vocal is over hyped in wide stereo, and the low end is mainly an 808 in mono,
Great stuff in the comments. Thank you everyone!!! I’m stoked to be able to contribute to such an awesome community. You guys rock. Also feel free to send me your stuff on soundcloud.com/djnisus I’d love to hear what you did with this article. – Nisus
and final p.s. – nice article Nisus, a big thanks to you too 🙂
oh and Symbiz and Christian – I can’t say thankyou enough for the pointer. I obviously need to spend more time reading and less mixing 🙂
Melodyne’s Direct Note Access technology is available now as as feature of their Editor software. I’ve just spent some time reading up about it and it is the most exciting thing I’ve seen in years.
It’s not the full de-mixing solution – as cloudcodex says correctly it is designed to pull apart isolated instruments but it is revolutionary and there is a world of opportunity to take this technology and properly f*ck about with it. I can imagine some crazy results.
Race you!
Melodyne Direct Note Access only aims to work separating the notes of an already isolated instrument, like a guitar chord OR a piano chord etc… so it will be useful, if it succeeds in this modest task (note it hasn’t been released yet) but don’t get your hopes up for separating all the original tracks from a mixdown!!!
Good article and great information 🙂 Although the techniques probably work better on certain tracks than others, but still good 🙂
Thanks for this. Great information! I assume any of the common DAW suites can accomplish this? (I use Sonar).
@ Anon – our brains locate sounds by measuring the slight difference between the time a sound arrives at one ear and the time it arrives at the other. Below 80 Hz there’s not a lot of that, below 60 there’s pretty much none. The overtones you’ll be able to locate, but not the fundamental.
As far as a filter, I use either the built-in one with Live or the Fabfilter one. Mad love for fabfilter in my laptop. Here’s a link to a free one I found by looking at kvr, not endorsing it, but it’s free, multi-mode and should give you an idea. http://www.grupvega.net/tugrul/
That “Rhythm is a Dancer” remix was awesome! I’ve been working towards separating steams out like this, but I have been less than impressed with Traktor’s EQ & filters. I will have to try some other programs like this.
Thats some good ISH!
Good article – and I’ve messed with this technique before.
Isn’t it just another way to “sample” to a degree??
I know we’re cutting the sounds out, but they are not original (created from scratch by YOU the producer)?
Just curious on any copyright issues doing this..
@ Full Stop – oops, the above question was from me iwaiwa. thanks !
@ Full Stop
Thanks for the info / advice. I agree that 320 MP3’s are sufficient to maintain sound quality (versus the much larger uncompressed WAV files).
Really just wondering about the Low End Bass (say as in DubStep) at around 60/80Hz and lower. By default some MP3 encoders tend to “mono” the bass below a certain frequency. Anyway, I guess I can just manually config the MP3 encoder to keep the stereo image and not do any more compressing at the low end. As long as it sounds fine (not wobbly / or messing with the original low end “phase”).
>> using resonant, synth style filters instead of standard bandpass EQ filters
can you give us an example of such as filter (vst?)
thanks Full Stop !
Word up Ponyboy! Soulwax (Dewaele Bros) are the greatest of all-time, hands down.
Wow nice tutorial – I`d like to see this in a bit more detail,
And maybe put into a nice pdf!
@ iwaiwa – if you start with 320 kbps mp3 then the core sound of your mix/remix can’t get better than that. Internally your recording/remixing app is going to convert the MP3 into a floating point audio file, process it at some ridiculous resolution internally, and then write out a file. 320 kbps MP3 sounds awfully good these days and if you start with that quality of file only the f/x you add to the mix is going to even have a chance of sounding better than 320 kbps MP3. Bottom line, if you go w/ WAV you’re probably just losing hard drive space. I’d be shocked if even under super high quality listening situations anyone could consistently pick one file over the other.
Nice article, btw. One thing I might have brought up is using resonant, synth style filters instead of standard bandpass EQ filters. I’ve found that particularly when trying to extract bass and vocal lines sometimes the resonant filter adds some oomph if you want that.
[quote comment=”25918″]Great post Nisus, this is a real help. I’m only learning but I love the logic pro eq, for me it has to be one of clearest and more easy eq’s for beginner’s to get to grips with.
+++++
I was fortunate to see Soulwax and 2many dj’s just before xmas, I was in great position to study what they were doing behind the decks, which was pretty much nothing more than very basic mixing. virtually all tracks they played had the soulwax remix treatment, every time I thought some live mashup via fx or loops were being used it would catch my attention and I’d have a good look, but there was sweet f a going on. It was great set to listen to but visually it was just watching two guys beatmatch for two hours. I didn’t leave disappointed, but actually educated in how important having songs or remix’s that no one else can get hold of, there set was a dj musical masterpiece to listen to, just not very interesting to watch. A strong reminder that its no always what you do but what you play.[/quote]
IMO it’s ALWAYS about what you play. No one cares how you are doing it, really, they just want to hear great music and have a great time.
Who uses stems?! learn to play the music yourself… you none musically educated DJ’s..
another one afraid of losing his job…what a shame
your a prick, and make sense when you are doing a Negative comment.
Cheers.
if they played the music themselves it wouldnt be a RE-mix…..you RE-tard
Quick question, not to be Buzz Killington, but wouldn’t extracting these tracks in this manner and enabling the filters to isolate the frequencies compress the sound waves to hell and back? Not that there is anyway around it if you do not have the original session tracks. The remixes sound great by the way. I just thought it was a good idea to point out that all of this seems only able to be achieved by proper use of FX after the filtering process. Great post!
This website just keeps getting better. Excellent content.
Fantastic article. 10/10. More like this please.
[quote comment=”25927″][quote comment=”25916″]Related Question on WAV versus MP3:
I would keep it in wav personally. Why loose quality when you can avoid it. Tag your files in the filename itself if you really need more info.[/quote]
Sounds good. Thanks Nico for that suggestion 🙂
guys, you have to check out Melodyne DNA Editor (DNA . Direct Note Acces). With this unique piece of software you can seperate a track into its single instruments again.
http://www.celemony.com/cms/index.php?id=news&L=1
thx again for the great article in here, greetz !
ps. i still wonder why you haven´t reviewed the Vestax – TR 1 yet, i got that wonderful piece of hardware for 3 weeks now, and i´m really curious to buy another one very soon.
greetz
You can only separate the instruments if the note is different for each instrument. If not you get one huge blob for all instruments with the same note.
No. Our Direct Note Access technology isolates polyphonic notes by pitch, not by instruments. If two instruments are playing the same note at the same time, you will see one blob only that includes the sound of both instruments.
taken from Celemony’s website!
It will surely be possible in a middle-term future thank’s to the advances of Artificial Intelligence (I work in that field so I can say I am pretty sure it will happen, just not now).
Nice post
Where is the 3.0 tsi mad for trying it?
yeah and i also need help on isolating the vocals, help me please
[quote comment=”25930″]hi im having trouble understanding Step 3. Duplicate
I replayed the main melody into a loop of midi notes that could be used with any synth and bounced these to their own audio track . By filtering out the original synth and placing your own custom creation on top, its easy to create a much fatter sound.
how is this done in Ableton??[/quote]
check out my ableton session above- in which I did the same thing. There is a midi- track there that was bounced down into 3 versions of the main bass line.
hi im having trouble understanding Step 3. Duplicate
I replayed the main melody into a loop of midi notes that could be used with any synth and bounced these to their own audio track . By filtering out the original synth and placing your own custom creation on top, its easy to create a much fatter sound.
how is this done in Ableton??
Very useful article. I’ve attempted creating edits like this before, but I always thought I was doing something ‘wrong’ or amateurish because I didn’t have the original stems. But now I know there are others creating in the same way.
Thanks Ean.. I love this community.
Hopefully bootleg remixes will increase in quality in the near future.
Although I’m pretty new to ableton, I’ve been trying to use similar techniques with Tpro for a while. Its really useful to see this laid out for me, and I hope to see more great articles out of DJTT in the future. Thanks a lot guys!
[quote comment=”25916″]Related Question on WAV versus MP3:
It’s temping to put the resultant remix into 320 MP3 format (for a variety of reasons), but in doing so the “quality” of the song will now have gone through it’s second compression from the original WAV. (The purchased MP3 being the first conversion). I’m wondering how remix-to-320-MP3’s sound on a large club system versus keeping the remixes in WAV (WAVs not having to go through the second conversion/”loss” that occurs when saving the remix to MP3 format).[/quote]
I would keep it in wav personally. Why loose quality when you can avoid it. Tag your files in the filename itself if you really need more info.
I’ll post one up I did of timbalands apologize when I had the original release. Little house twist.
with that software EVERYTHING will be remixable!!
i guess this will change music production and remixing forever:
http://www.youtube.com/watch?v=jFCjv4_jqAY
Great post Nisus, this is a real help. I’m only learning but I love the logic pro eq, for me it has to be one of clearest and more easy eq’s for beginner’s to get to grips with.
+++++
I was fortunate to see Soulwax and 2many dj’s just before xmas, I was in great position to study what they were doing behind the decks, which was pretty much nothing more than very basic mixing. virtually all tracks they played had the soulwax remix treatment, every time I thought some live mashup via fx or loops were being used it would catch my attention and I’d have a good look, but there was sweet f a going on. It was great set to listen to but visually it was just watching two guys beatmatch for two hours. I didn’t leave disappointed, but actually educated in how important having songs or remix’s that no one else can get hold of, there set was a dj musical masterpiece to listen to, just not very interesting to watch. A strong reminder that its no always what you do but what you play.
Impressive Article. This is really useful information. Gonna try it out myself! Thanks DjTT you Guys really rock 🙂
Related Question on WAV versus MP3:
I download all my orig songs in 320 MP3 format. Then “deconstruct/reconstruct” to make something new. Even though, say with Ableton, the output of the remix is saved in a WAV file, do people later convert these save remixes to 320 MP3 format ? (to make it easier to enter and save ID3 tags, and have the ID3 tags “portable” / embedded in the MP3)
It’s temping to put the resultant remix into 320 MP3 format (for a variety of reasons), but in doing so the “quality” of the song will now have gone through it’s second compression from the original WAV. (The purchased MP3 being the first conversion). I’m wondering how remix-to-320-MP3’s sound on a large club system versus keeping the remixes in WAV (WAVs not having to go through the second conversion/”loss” that occurs when saving the remix to MP3 format).
Thanks, great article..
Is this feasable also with Cubase?
yeah. you guys are on fire. thx.
This is exactly the kind of thing I want to learn more of. What schools would the DJTT team recommend for making this step from plain DJ to to Producer?
Haven’t read all of this yet but the idea is fantastic, exactly the reason I use this site. Keep it up fellas
this technique is how I first got into production. Me and friends would pull out whatever we could and make “remixes” of our favorite tracks. thanks for this. Takes me back…
Lol, this frequency chart is hilarious, especially “Rising Sounds and Elevators”.
Very excellent post, came up just as I was recently reading a few other remixing articles.
A while back I generally had the mentality of “don’t mess with the original”. You know the train of thought where you biasedly treat the original as some untouchable perfection but after a while it does need to be touched up and played around with, it makes it come back to life.
One track that really got me interested in remixes was when I heard Sasha & Digweed’s remix of Ladytron – Destroy Everything You Touch. It was a perfect night, it started up with this rolling bassline and then bam, lyrics picked up and I had the biggest grin on my face.
So yeah, remixing is pretty awesome.
Thanks for this tut, I’ll pass it along to some of my friends as well 🙂
I completely devoted a Soundcloud group to the editting/reming principles described above. Be sure to check it if you’re into disco & nu disco. Lots of edits/remixes to be found there, all 320kb free downloadable 🙂 http://soundcloud.com/groups/deep-disco
For anyone not familar to Soundcloud: It’s thé place to be nowadays for artists as well as dj’s who want to stand out using original tracks & edits not found anywhere else. Tons of downloadable originals and remixes to be found there in literally every genre.
It actually completely surprises me that DJTT has not given this social music medium any serious coverage so far. Soundcloud is quickly becoming the most important place to share and promote music/mixes for digital dj’s and artists alike.
Be sure to check it out if you wanna take a sidestep from your average Juno/Beatport download ranking list.
Forking brilliant!!! Seriously this really help beginner like me, thanks a million!!!
Wow this is pretty awesome… Thanx for this tutorial!
THATS A REALLY NICE POST
ALSO I WANT EVERYONE TO CHECK OUT THIS VIDEO
These articles recently have been phenominal. Thanks so much for this website, I love it more every day.
Good post, but should have been called “remixing 101” because this how most un-offical remixes are done…especially if you’re crossing genres.
It’s very rare to get the hookup on real stems.
Also worth mentioning is that most of these production techniques were pioneered by hiphop producers…using EQ/filtering to isolate specific instruments, adding new original elements, and re-arranging the sections…except we didn’t have a fancy chart or instructions. You had to just figure it out on your own. Try doing this with a tiny LCD screen and 10 seconds of mono sampling.
I don’t miss those days at all.
omfg, thnx!!!!