Native Instruments CTO and President Mate Galic introduces Stems at Miami's Winter Music Conference. Image courtesy Motormouth Media.

Native Instruments CTO and President Mate Galic introduces Stems at Miami’s Winter Music Conference. Image courtesy Motormouth Media.

The path forward is clear: there’s no reason in this age of digital producing and DJing that music needs to be stereo.

The need is there, but so far, not the solution. A file format announced in a press briefing at Miami’s Winter Music Conference and made public today wants to succeed where others failed.

It’s called Stems, and there are a few details that make it different.

It’s simple. “Stems” – the format – include four tracks. So that could be bass, drums, melody, vocal, for instance. (Or bagpipe, castrati chorus, tambourines, and banjo. But the point is, dividing things into four makes a lot of sense.) You can also choose the order, color, and add names to individual tracks.

It’s compatible and built on existing standards. A new file format? Good luck. (See xkcd #927.) But Stems uses an MP4 container format (that’s MP4 only, not MP3). Load your Stems file onto any software or hardware that supports MP4, and you’ll get stereo playback of the mix – including on the Pioneer CDJ. ID3 tags for the track work, too (for the overall mix). Load it into software that supports Stems – which we’re assuming will most likely be some sort of DJ software – and you can play back the individual parts. (And mix them, remix them, add effects, whatever.) It’s really just four MP4s.

It’s free for anyone to use. An official website coming in June will detail how to make the files. There will also be a free Stem Creator for anyone wanting to create their own files. And the file format will be not only detailed on the Stems website in full, but there are no licensing fees for creation, distribution, or use. (No, you don’t have to pay to get the specs, either – I’m looking at you, MIDI.) No word yet on how the Stems branding will work.

It’s backed by some key players. Native Instruments announced that Traktor Pro 2.7.4 or later already have Stems support. (See the public beta.) Beatport, Juno, and Traxsource all promise to sell Stems format starting in June. In Miami, DJ/producer Luciano and KCRW’s Music Director Jason Bentley joined a panel to introduce the idea. 16 labels have chimed in with support, too, including Mobilee, Monkeytown, 50Weapons, Get Physical, and InFiné. I suspect it’s really the labels and stores, combined with Traktor, that could kick-start this thing.

It’ll be easy to DJ with. Any group of four controller faders/encoders can be mapped to the different parts – the structure of NI’s own F1 and new D2 all map logically, and so will a lot of other things.

Now there’s a reason to use it – money. The Stems backers are pretty direct about their appeal: release Stems so you can charge more for your music. And while the pitch is for a “premium” price, the timing is also essential. With Beatport launching its own free streaming service, with listeners more likely to stream, and with DJ apps like djay even adding Spotify support, producers and labels need a format that they can still sell. Vinyl alone probably isn’t enough to keep them afloat.

So…

Who’s it for? The main audience is clearly DJs: the idea is to convince producers to share stems in a standard format that makes it easier for DJs to think about playing with individual tracks and not only the mixdown. The Stems FAQ suggests even producers might want to use Stems to move their own music between their production software and a live set. That’s less of an issue if you’re DJing with Ableton Live, but certainly could be a boon if you use Traktor, Serato, and so on. I imagine this could also target consumer listeners – remix your artist’s favorite tracks. See NinjaTune’s NINJA JAMM for one take on this idea.

Can it be a standard? Now, it’s really technically a de facto standard, and a license free standard, not a standard or open standard in the technical sense – there’s no real governance that I can see on the “standard” side, and it’s “open” only in that it’s published and free to implement. But that could be enough, with the success of MIDI a prime example.

What’s the competition? Interestingly enough, even Wikipedia has an article on stem releases. The reason? There really aren’t that many such releases to begin with. Trent Reznor and Nine Inch Nails deserve some real credit, with remix.nin.com, Year Zero (which also included GarageBand and Ableton Live formats), and Ghosts I–IV. Open licensing means you can remix the last of these and redistribute what you make; the stems in this case will mostly be for DJs to use when playing, but would have to be licensed to sample. And looking through that list, you see a confirmation of what you might suspect – most release stems in simple stereo formats, and just give you more files. It’s even possible to do this with iTunes, easier still with Bandcamp. But this isn’t a terribly common practice, and releases for specific DAWs have tended to be sporadic and typically focus on remix artists, not DJs.

And… is this a good idea? Actually, I was pondering this after writing this, especially after I saw Engadget wrote this:

“If you head to a dance music festival this summer and notice that one or two sets are particularly creative, you’ll know why.”

Well – wait a minute. What makes a creative set is not necessarily doing mash-ups of stems from different songs or cutting out the vocal. So let’s ponder this: it really will be up to both the artist releasing tracks, and the DJ using stems, to determine whether using stems is a good idea. I suppose the hope is that this gives creative DJs additional choices – even if (as you’ll see raised in comments) it might also make bad DJs worse. Then again, lots of things can make bad DJs worse – turning down their master volume may be the way to make them better. The key to making use of Stems may be making use of them selectively, as with any other special technique. DJing has done just fine with mixing stereo tracks since the beginning, often in two-channel configurations, even if product manufacturers increasingly push live remixing as a way of convincing you to buy new hardware and software. So, making use of Stems should be a means to an end. And tasteful DJs may find a way of using this that still shows some deference to the intent of the original track.

What next? It’s not hard to see why Native Instruments would support this, since the ability to mix beyond simple 2-channel stereo is part of Traktor’s market differentiation. At the same time, it’s encouraging to see NI back a reasonably open standard, and not just double down on their own proprietary Remix Sets and the like. (It’s also nice that this is simple – making Stems would be a lot more straightforward than authoring Remix Sets.)

But that also means we’ll need to see support beyond just Traktor for the format to really take off. We’ll see if other app developers get onboard, or decide this is just an “NI thing.” And hardware support would surely lag software. At least Traktor solves some of the chicken-and-egg problems of a format of this kind: it has a built-in audience of producers and DJs.

I’m not even sure what ideal DAW support should look like, since DAWs generally have an arbitrary number of tracks. But that means the Stem Creator has to work pretty well in the interim.

But I have to say, while I met the headline for the media announcement with a heavy dose of skepticism, the more I learn about Stems, the more optimistic I am. There’s clearly desire now for DJs to set themselves apart, particularly in the age of streaming and instant access, so the demand on that side is there. And there’s desire on the part of artists and labels, not least as they increasingly sell music to specialists and DJs. What might make Stems work even given past failures to solve this problem is that it’s actually easy as well as desirable to implement. And that may lead to a “built it and they will come effect” – because it’s both easy to do and easy to justify.

And it’s certainly a lot better than some new proprietary format — or the status quo of random a cappellas here and there or removing vocals with (gah) EQ.

We’ll be watching for this in June. Let us know if you have any questions. (I’ll meanwhile investigate more details of how implementation works on the creation and Traktor DJ side.)

The announcement event. Okay... so, it's hard to make a file format visual. But the future is so bright, Jason Bentley from KCRW is wearing shades.

The announcement event. Okay… so, it’s hard to make a file format visual. But the future is so bright, Jason Bentley from KCRW is wearing shades.

  • Daniel Courville

    “MP3 won’t work because it isn’t multichannel”. See this http://en.wikipedia.org/wiki/MP3_Surround

    • Yeah, I was writing that and feeling like, this is wrong…

      I think the issue is that metadata in MP3 isn’t possible on individual tracks using that format…

      Also, I am almost certain licensing issues with Fraunhofer were involved. Note that the link from MP3 surround takes you directly to licensing. 🙂

      See MPEG Surround here:
      http://www.iis.fraunhofer.de/en/ff/amm/prod/audiocodec/audiocodecs/mpegsurr.html

      But I’m just speculating; I can look into this.

      • newmodernscience

        If you think about it though, a multichannel surround mix isn’t 5.1 channels of stereo, it’s 5.1 channels of mono, so that standard isn’t going to do anyone any favors. Positional audio is a very different beast than multi-TRACK, which is what I believe we’re really talking about here. I’m sure they’ve thought this through, but if MP4 is only carrying 4 mono channel tracks, that’s not really adding much to the standard single stereo or dual mono mp3 standard. My guess is that we’re talking about 4 stereo tracks, or really 8 channels of audio.

        • Actually, I believe 5.1 is just one example of how MPEG Surround works – it’s still a container for multiple tracks, and my recollection of the spec is that it can include an arbitrary number of channels (or at least different surround formats — 8.1 for instance would give the right number). It may not allow you to call those channels “drums” or something, though.

          The other question is whether there was at some point a generic multichannel MP3 spec.

          But the MPEG4 spec here makes more sense, probably – and may be easier to license. I can double-check that, though.

          • newmodernscience

            The open format is obviously a HUGE benefit here. I remember way back in the day having to pay $100 for a Fraunhofer licensed plugin for whatever DAW I was using at the time. I think now even you have to pay $20 for the mp3 plugin to export out of Pro Tools.

          • Daniel Courville

            “I think now even you have to pay $20 for the mp3 plugin to export out of Pro Tools”. Not anymore. Since version 9 if memory serves me right.

          • May be different licensing for the surround format, though.

      • Daniel Courville

        Up to a few days ago, I had a Fraunhofer MP3 surround encoder and a separate Fraunhofer MP3 surround player on my computer and I got them for free from the Fraunhofer website. I deleted both because I wasn’t using them…

        • But that’s from Fraunhofer. The licensing is to do with patents. So they can release something free of charge, of course.

          I don’t know, though. What I’m saying is, any number of reasons they might have chosen MP4 over MP3.

  • Daniel Courville

    “MP3 won’t work because it isn’t multichannel”. See this http://en.wikipedia.org/wiki/MP3_Surround

    • Yeah, I was writing that and feeling like, this is wrong…

      I think the issue is that metadata in MP3 isn’t possible on individual tracks using that format…

      Also, I am almost certain licensing issues with Fraunhofer were involved. Note that the link from MP3 surround takes you directly to licensing. 🙂

      See MPEG Surround here:
      http://www.iis.fraunhofer.de/en/ff/amm/prod/audiocodec/audiocodecs/mpegsurr.html

      But I’m just speculating; I can look into this.

      • newmodernscience

        If you think about it though, a multichannel surround mix isn’t 5.1 channels of stereo, it’s 5.1 channels of mono, so that standard isn’t going to do anyone any favors. Positional audio is a very different beast than multi-TRACK, which is what I believe we’re really talking about here. I’m sure they’ve thought this through, but if MP4 is only carrying 4 mono channel tracks, that’s not really adding much to the standard single stereo or dual mono mp3 standard. My guess is that we’re talking about 4 stereo tracks, or really 8 channels of audio.

        • Actually, I believe 5.1 is just one example of how MPEG Surround works – it’s still a container for multiple tracks, and my recollection of the spec is that it can include an arbitrary number of channels (or at least different surround formats — 8.1 for instance would give the right number). It may not allow you to call those channels “drums” or something, though.

          The other question is whether there was at some point a generic multichannel MP3 spec.

          But the MPEG4 spec here makes more sense, probably – and may be easier to license. I can double-check that, though.

          • newmodernscience

            The open format is obviously a HUGE benefit here. I remember way back in the day having to pay $100 for a Fraunhofer licensed plugin for whatever DAW I was using at the time. I think now even you have to pay $20 for the mp3 plugin to export out of Pro Tools.

          • Daniel Courville

            “I think now even you have to pay $20 for the mp3 plugin to export out of Pro Tools”. Not anymore. Since version 9 if memory serves me right.

          • May be different licensing for the surround format, though.

      • Daniel Courville

        Up to a few days ago, I had a Fraunhofer MP3 surround encoder and a separate Fraunhofer MP3 surround player on my computer and I got them for free from the Fraunhofer website. I deleted both because I wasn’t using them…

        • But that’s from Fraunhofer. The licensing is to do with patents. So they can release something free of charge, of course.

          I don’t know, though. What I’m saying is, any number of reasons they might have chosen MP4 over MP3.

  • Daniel Courville

    “MP3 won’t work because it isn’t multichannel”. See this http://en.wikipedia.org/wiki/MP3_Surround

    • Yeah, I was writing that and feeling like, this is wrong…

      I think the issue is that metadata in MP3 isn’t possible on individual tracks using that format…

      Also, I am almost certain licensing issues with Fraunhofer were involved. Note that the link from MP3 surround takes you directly to licensing. 🙂

      See MPEG Surround here:
      http://www.iis.fraunhofer.de/en/ff/amm/prod/audiocodec/audiocodecs/mpegsurr.html

      But I’m just speculating; I can look into this.

      • newmodernscience

        If you think about it though, a multichannel surround mix isn’t 5.1 channels of stereo, it’s 5.1 channels of mono, so that standard isn’t going to do anyone any favors. Positional audio is a very different beast than multi-TRACK, which is what I believe we’re really talking about here. I’m sure they’ve thought this through, but if MP4 is only carrying 4 mono channel tracks, that’s not really adding much to the standard single stereo or dual mono mp3 standard. My guess is that we’re talking about 4 stereo tracks, or really 8 channels of audio.

        • Actually, I believe 5.1 is just one example of how MPEG Surround works – it’s still a container for multiple tracks, and my recollection of the spec is that it can include an arbitrary number of channels (or at least different surround formats — 8.1 for instance would give the right number). It may not allow you to call those channels “drums” or something, though.

          The other question is whether there was at some point a generic multichannel MP3 spec.

          But the MPEG4 spec here makes more sense, probably – and may be easier to license. I can double-check that, though.

          • newmodernscience

            The open format is obviously a HUGE benefit here. I remember way back in the day having to pay $100 for a Fraunhofer licensed plugin for whatever DAW I was using at the time. I think now even you have to pay $20 for the mp3 plugin to export out of Pro Tools.

          • Daniel Courville

            “I think now even you have to pay $20 for the mp3 plugin to export out of Pro Tools”. Not anymore. Since version 9 if memory serves me right.

          • May be different licensing for the surround format, though.

      • Daniel Courville

        Up to a few days ago, I had a Fraunhofer MP3 surround encoder and a separate Fraunhofer MP3 surround player on my computer and I got them for free from the Fraunhofer website. I deleted both because I wasn’t using them…

        • But that’s from Fraunhofer. The licensing is to do with patents. So they can release something free of charge, of course.

          I don’t know, though. What I’m saying is, any number of reasons they might have chosen MP4 over MP3.

  • chaircrusher

    Always a good idea to give terrible DJs a new way to be terrible.

    Which is flip and sarcastic, but honestly, if you’re playing quality tracks, then you appreciate the unique work that went into them, and you certainly don’t think you can improve them by trying to remix them on the fly in a noisy DJ booth.

    There’s also the issue that for any genre on Beatport the top 100 tracks have interchangeable drum tracks.

    I think the way labels or sites like Beatport and Junodownload will make money out of this is selling to wannabes. This is the Guitar-Center-ification of DJing — not that it hasn’t already been going on for 20 years.

    • That’s very possible.

      Still, there’s a lot of desire from both DJs and producers here. That is, there may be a possible to use this for good and not only evil. 😉

      I added some more text to address this… I mean, terrible DJs can be resourceful in their awfulness, so really the question is, is there something here for good DJs?

      But I will say, the need for this should not necessarily be accepted as self-evident.

  • chaircrusher

    Always a good idea to give terrible DJs a new way to be terrible.

    Which is flip and sarcastic, but honestly, if you’re playing quality tracks, then you appreciate the unique work that went into them, and you certainly don’t think you can improve them by trying to remix them on the fly in a noisy DJ booth.

    There’s also the issue that for any genre on Beatport the top 100 tracks have interchangeable drum tracks.

    I think the way labels or sites like Beatport and Junodownload will make money out of this is selling to wannabes. This is the Guitar-Center-ification of DJing — not that it hasn’t already been going on for 20 years.

    • That’s very possible.

      Still, there’s a lot of desire from both DJs and producers here. That is, there may be a possible to use this for good and not only evil. 😉

      I added some more text to address this… I mean, terrible DJs can be resourceful in their awfulness, so really the question is, is there something here for good DJs?

      But I will say, the need for this should not necessarily be accepted as self-evident.

  • chaircrusher

    Always a good idea to give terrible DJs a new way to be terrible.

    Which is flip and sarcastic, but honestly, if you’re playing quality tracks, then you appreciate the unique work that went into them, and you certainly don’t think you can improve them by trying to remix them on the fly in a noisy DJ booth.

    There’s also the issue that for any genre on Beatport the top 100 tracks have interchangeable drum tracks.

    I think the way labels or sites like Beatport and Junodownload will make money out of this is selling to wannabes. This is the Guitar-Center-ification of DJing — not that it hasn’t already been going on for 20 years.

    • That’s very possible.

      Still, there’s a lot of desire from both DJs and producers here. That is, there may be a possible to use this for good and not only evil. 😉

      I added some more text to address this… I mean, terrible DJs can be resourceful in their awfulness, so really the question is, is there something here for good DJs?

      But I will say, the need for this should not necessarily be accepted as self-evident.

  • Here I thought everybody used Phase inversion in Audacity to get those Brandi Acappellas 😉 (or actually paid for them)

  • Here I thought everybody used Phase inversion in Audacity to get those Brandi Acappellas 😉 (or actually paid for them)

  • Here I thought everybody used Phase inversion in Audacity to get those Brandi Acappellas 😉 (or actually paid for them)

  • Ted

    Wait haven’t we had multichannel wavs for a while now? How is this different? https://msdn.microsoft.com/en-us/library/windows/desktop/ee418679%28v=vs.85%29.aspx

    • I think there’s standard metadata encoded in the MP4 – that’s what I’m trying to find out in greater detail.

      • Ted

        I don’t know – if we’re all going to “ooh-ahh” over Funktion 1 (which is a deserved reaction!) then I would want to see NI responding by creating a high-resolution Stems…I’m trying to find some info on MP4 bitrates, and it looks like there’s a 160kb/s cap – that’s pretty abysmal. On the other hand, I’ve got a suspicion that they’re simply utilizing MP4 as a container, and that maybe there’s some cool hi-res work going on inside the container?

        • I’ll get back to you on this information.

          Basically, “MP4” says nothing about encoding. I have confirmed at least that it’s AAC, but not a) if there’s a specified bitrate in Stems and b) what the default output of their creation software is.

          • Ted

            I definitely see a need for uncompressed formats, especially given that track counts are arbitrary in the digital world. Uncompressed formats – as I understand it – are easier to “render” in realtime, assuming your hard disk read-speed can keep up your CPU will be fine.

        • Oh, and, given there are very efficient lossless encoding methods, storing things in WAV makes effectively no sense whatsoever. In fact, it’s rather a shame that DAWs insist on keeping all their audio in such a space-consuming format with no option to use lossless compression.

          • B.C. Thunderthud

            Considering that storage is so cheap any additional processing overhead or latency whatsoever seems like a bad choice. Lossless makes sense in a few areas–this streaming business, where people actually spend a lot of money for data–but I don’t think it has any place in studios. Maybe the cost is small but the benefit is essentially nil.

  • Ted

    Wait haven’t we had multichannel wavs for a while now? How is this different? https://msdn.microsoft.com/en-us/library/windows/desktop/ee418679%28v=vs.85%29.aspx

    • I think there’s standard metadata encoded in the MP4 – that’s what I’m trying to find out in greater detail.

      Also, multichannel WAVs – good for producers wanting to remix with stems. Bad for DJs (they take up too much space). Sure, we now have people arguing DJs should start using higher-quality files, but… the reality is still making a compromise to save file space.

      • Ted

        I don’t know – if we’re all going to “ooh-ahh” over Funktion 1 (which is a deserved reaction!) then I would want to see NI responding by creating a high-resolution Stems…I’m trying to find some info on MP4 bitrates, and it looks like there’s a 160kb/s cap – that’s pretty abysmal. On the other hand, I’ve got a suspicion that they’re simply utilizing MP4 as a container, and that maybe there’s some cool hi-res work going on inside the container?

        • I’ll get back to you on this information.

          Basically, “MP4” says nothing about encoding. I have confirmed at least that it’s AAC, but not a) if there’s a specified bitrate in Stems and b) what the default output of their creation software is.

          • Ted

            I definitely see a need for uncompressed formats, especially given that track counts are arbitrary in the digital world. Uncompressed formats – as I understand it – are easier to “render” in realtime, assuming your hard disk read-speed can keep up your CPU will be fine.

        • Oh, and, given there are very efficient lossless encoding methods, storing things in WAV makes effectively no sense whatsoever. In fact, it’s rather a shame that DAWs insist on keeping all their audio in such a space-consuming format with no option to use lossless compression.

          • B.C. Thunderthud

            Considering that storage is so cheap any additional processing overhead or latency whatsoever seems like a bad choice. Lossless makes sense in a few areas–this streaming business, where people actually spend a lot of money for data–but I don’t think it has any place in studios. Maybe the cost is small but the benefit is essentially nil.

  • Ted

    Wait haven’t we had multichannel wavs for a while now? How is this different? https://msdn.microsoft.com/en-us/library/windows/desktop/ee418679%28v=vs.85%29.aspx

    • I think there’s standard metadata encoded in the MP4 – that’s what I’m trying to find out in greater detail.

      Also, multichannel WAVs – good for producers wanting to remix with stems. Bad for DJs (they take up too much space). Sure, we now have people arguing DJs should start using higher-quality files, but… the reality is still making a compromise to save file space.

      • Ted

        I don’t know – if we’re all going to “ooh-ahh” over Funktion 1 (which is a deserved reaction!) then I would want to see NI responding by creating a high-resolution Stems…I’m trying to find some info on MP4 bitrates, and it looks like there’s a 160kb/s cap – that’s pretty abysmal. On the other hand, I’ve got a suspicion that they’re simply utilizing MP4 as a container, and that maybe there’s some cool hi-res work going on inside the container?

        • I’ll get back to you on this information.

          Basically, “MP4” says nothing about encoding. I have confirmed at least that it’s AAC, but not a) if there’s a specified bitrate in Stems and b) what the default output of their creation software is.

          • Ted

            I definitely see a need for uncompressed formats, especially given that track counts are arbitrary in the digital world. Uncompressed formats – as I understand it – are easier to “render” in realtime, assuming your hard disk read-speed can keep up your CPU will be fine.

        • Oh, and, given there are very efficient lossless encoding methods, storing things in WAV makes effectively no sense whatsoever. In fact, it’s rather a shame that DAWs insist on keeping all their audio in such a space-consuming format with no option to use lossless compression.

          • B.C. Thunderthud

            Considering that storage is so cheap any additional processing overhead or latency whatsoever seems like a bad choice. Lossless makes sense in a few areas–this streaming business, where people actually spend a lot of money for data–but I don’t think it has any place in studios. Maybe the cost is small but the benefit is essentially nil.

  • Xebulon

    How would this affect the mastering process? Aren’t most mastering plugins applied on the stereo bus? So I’d imagine it would affect the final sound to have the 4 stems mastered separately rather than mastered together.

    • Eric

      I’m no mastering expert by any stretch, but couldn’t you (or whatever software is mastering) simply apply whatever should be applied to the stereo bus to each individual track?

      • Eric

        Scratch that. Just re-read and that’s essentially what you’re asking lol I guess in this case you’d really need to take that in to consideration and get plugins specifically for each stem. Would definitely add to the mastering process :S

        • Xebulon

          Yeah, effects like limiters and compressors, when triggered by audio that exceeds their threshold, will affect the volume of all tracks when placed on a stereo bus (creating a sometimes-desirable pumping effect). But if there are 4 separate tracks with 4 limiters then they would affect the volumes of each track independently so the end result would certainly sound different.

      • Well, yes and no. Some of those stems releases above really represented the original project stems.

        We may be going about this all wrong. I think the fundamental question is, what is a producer’s concept when releasing music this way – how do they want it to be heard? And do they want to release stems at all? Because while interesting, these aren’t essential to DJs. And if you do want fans to cough up more money for stems, you’d better have some idea in mind of why.

        • Jos Smolders

          Although everything depends on your definition of mastering and its function in my book it’s essentially useless to master stems separately. I see the stems mostly used as an exchange format between artists who want to work online. I suggest to go into mastering after all the creative exchange business has finished. Or not master at all if you want to keep a stem file ‘alive’ forever.

          By the way: there already have been multichannel wavs for years. It ticks all of the abovementioned boxes. So it’s not as innovative as we think.

          • Yes, but the fact that people are DJing with multichannel WAVs suggest there might be an opportunity to create something tailored to the use case.

            Mastering: mastering I think ought to be defined as preparing music for a distribution format. There was a time when we were mastering to mono and not stereo. In broadcast and film, you have more complex scenarios, too – like surround. (Worst of all, in many surround applications you have to contend with automated mixdowns that sum – happily that doesn’t happen here.)

            Look, this is a format for people who want to use this format. If they do, they can find a way to make it work. If they don’t – if no one does – it’s okay to walk away and let this die.

          • Jos Smolders

            Hmm, I don’t think DAW makers will agree. Yet another format… while there already is one that works well. But hasn’t been properly promoted.

          • DPrty

            ambisonic WAVE_FORMAT_EXTENSIBLE files
            http://people.bath.ac.uk/masrwd/ncdream/researchdev/wave-ex/wave_ex.html

            Reaper Change log
            v5.0pre12 – February 9 2015+ WAV reading: support ambisonic WAVE_FORMAT_EXTENSIBLE files

            http://www.ambisonic.net/mulchaud.html

    • I was thinking the same.
      Streaming stems for mixing could be fun, I guess even some end consumers might be happy to pay for that.

    • Right, I think the way to conceptualize this would be that you’re separately mastering for four tracks. So it’s another reason to bus.

      There would be different approaches:
      1. Do a separate stereo master on the master bus, then a separate “Stems” master on four separate buses (listening to the mixdown to hear the stereo result). You might also compare these, depending on how important it was to you for these to match.

      2. Skip the stereo master entirely and master from the stems. True, as others say here, those buses’ dynamics then become independent, but it’s still possible to combine compression on those buses in such a way that you get the resulting dynamic range you want.

      Option #2 isn’t quite as weird as it might sound, in that some mix/mastering engineers will work in certain cases from stems.

      Film production is already dealing with issues like this because of the necessity to distribute in surround, etc. – there are more dimensions to mastering.

      I’m not a mastering engineer, though, so I may poll some folks to see what they think. NI wins twice on this, though, as then they’re also possibly the ones selling you your mastering tools. 😉

      • chaircrusher

        Undoubtably the stems will be individually mastered, For this whole thing to work, they’d not be as heavily effected as a proper stereo mastering job. Otherwise everything would turn to a soupy mess.

        It raises an issue about the wisdom of handling tracks this way. A proper mixdown requires a producer to manage the interaction between all the parts. I often use the trick abused by every EDM producer on earth — I’ll have the drums and percussion on one bus, and all the other instruments on another, and sidechain compress the instruments with the drum bus’s bass. You can really mash the instrument bus and sound like every terrible EDM hit with that off-beat pumping thing.

        But you can also be subtle with it, and it keeps the kick sound slightly forward in the mix without having to turn it up.

        With a stem file, you can’t do that, so the separation and interaction of different parts of the track are lost. That might not matter to most people, but it would drive me crazy.

        • Why can’t you do that with stems?
          You might have to sidechain stuff separately by groups or do it on the tracks. Am I missing something?

          • Matt Mclean

            You could do that with stems, and in fact thats the way I do it with live backings for my electronic band. Disable the output of say kick drum, but the chain will still send to synth bus when exporting, retaining the side chains and grooves.. Yes the master will react differently due to the way a limiter works, so the mixes will not sound exactly the same, but you can get it fairly close.

          • chaircrusher

            We’re talking about something that gets loaded into a DJ app. Of course if you mix down to stems and load them into a DAW you can use sidechain compression. But I doubt that’s going to be a feature of Traktor, or something a DJ in front of a heaving crowd can manage.

          • What I mean is do it before it gets exported, sure when you’ll take off the drum track you still have the pumping effect on the bass or the other stems. And when the tracks are not at 100% level it might sound like it’s not all that glued together. But it surely works.
            I often export groups of tracks from ableton to mix in mixbus or studio one, I do all my sidechaining in ableton because of it’s nice routing capabilities.
            You could even decide to put your kick and everything is not high mids and highs in the same track to smash it right and maybe have just some filtering reacting to the kick on other stems for cleaning further. And you can still export as stereo track for a particular track if your after that.
            Personally I think workarounds are the salt of music production.

        • Tom

          It would matter to me too, but fear not, you can still do it.
          I nearly always use a muted channel to provide the source for a sidechain anyway.
          All you need to do using your preferred technique, pull down the fader on the bus containing the output of all your drums, and use a (pre fade) send to a 2nd bus for your compressors to SC from. 🙂

      • It’s really no different than when a client asks me to master stems for them to use in their live shows I believe. Typically in those cases the stems are treated individually so they sound as good as possible when summed. Then a separate master is done for the stereo file that will also get released separately. There’s still no word from NI if the container format will have a separate stereo file as well as the 4 stems, but either way it shouldn’t be a big deal from a mastering engineers perspective.

      • this is were it gets cool. if you turn the mastering process around we will get some new interesting things happening. in a way making a 2-track export in a DAW and master than feels silly to me. any mastering engineer would fix the mix if he/she was able to do that. so why not do everything in the DAW? I do this for years now. works fine.

    • chaircrusher

      Interesting question. One person to ask would be Stewart Walker, since he masters the NI Remix sets. @peterkirn:disqus?

  • Xebulon

    How would this affect the mastering process? Aren’t most mastering plugins applied on the stereo bus? So I’d imagine it would affect the final sound to have the 4 stems mastered separately rather than mastered together.

    • Eric

      I’m no mastering expert by any stretch, but couldn’t you (or whatever software is mastering) simply apply whatever should be applied to the stereo bus to each individual track?

      • Eric

        Scratch that. Just re-read and that’s essentially what you’re asking lol I guess in this case you’d really need to take that in to consideration and get plugins specifically for each stem. Would definitely add to the mastering process :S

        • Xebulon

          Yeah, effects like limiters and compressors, when triggered by audio that exceeds their threshold, will affect the volume of all tracks when placed on a stereo bus (creating a sometimes-desirable pumping effect). But if there are 4 separate tracks with 4 limiters then they would affect the volumes of each track independently so the end result would certainly sound different.

      • Well, yes and no. Some of those stems releases above really represented the original project stems.

        We may be going about this all wrong. I think the fundamental question is, what is a producer’s concept when releasing music this way – how do they want it to be heard? And do they want to release stems at all? Because while interesting, these aren’t essential to DJs. And if you do want fans to cough up more money for stems, you’d better have some idea in mind of why.

        • Jos Smolders

          Although everything depends on your definition of mastering and its function in my book it’s essentially useless to master stems separately. I see the stems mostly used as an exchange format between artists who want to work online. I suggest to go into mastering after all the creative exchange business has finished. Or not master at all if you want to keep a stem file ‘alive’ forever.

          By the way: there already have been multichannel wavs for years. It ticks all of the abovementioned boxes. So it’s not as innovative as we think.

          • Yes, but the fact that people are DJing with multichannel WAVs suggest there might be an opportunity to create something tailored to the use case.

            Mastering: mastering I think ought to be defined as preparing music for a distribution format. There was a time when we were mastering to mono and not stereo. In broadcast and film, you have more complex scenarios, too – like surround. (Worst of all, in many surround applications you have to contend with automated mixdowns that sum – happily that doesn’t happen here.)

            Look, this is a format for people who want to use this format. If they do, they can find a way to make it work. If they don’t – if no one does – it’s okay to walk away and let this die.

          • Jos Smolders

            Hmm, I don’t think DAW makers will agree. Yet another format… while there already is one that works well. But hasn’t been properly promoted.

          • DPrty

            ambisonic WAVE_FORMAT_EXTENSIBLE files
            http://people.bath.ac.uk/masrwd/ncdream/researchdev/wave-ex/wave_ex.html

            Reaper Change log
            v5.0pre12 – February 9 2015+ WAV reading: support ambisonic WAVE_FORMAT_EXTENSIBLE files

            http://www.ambisonic.net/mulchaud.html

    • I was thinking the same.
      Streaming stems for mixing could be fun, I guess even some end consumers might be happy to pay for that.

    • Right, I think the way to conceptualize this would be that you’re separately mastering for four tracks. So it’s another reason to bus.

      There would be different approaches:
      1. Do a separate stereo master on the master bus, then a separate “Stems” master on four separate buses (listening to the mixdown to hear the stereo result). You might also compare these, depending on how important it was to you for these to match.

      2. Skip the stereo master entirely and master from the stems. True, as others say here, those buses’ dynamics then become independent, but it’s still possible to combine compression on those buses in such a way that you get the resulting dynamic range you want.

      Option #2 isn’t quite as weird as it might sound, in that some mix/mastering engineers will work in certain cases from stems.

      Film production is already dealing with issues like this because of the necessity to distribute in surround, etc. – there are more dimensions to mastering.

      I’m not a mastering engineer, though, so I may poll some folks to see what they think. NI wins twice on this, though, as then they’re also possibly the ones selling you your mastering tools. 😉

      • chaircrusher

        Undoubtably the stems will be individually mastered, For this whole thing to work, they’d not be as heavily effected as a proper stereo mastering job. Otherwise everything would turn to a soupy mess.

        It raises an issue about the wisdom of handling tracks this way. A proper mixdown requires a producer to manage the interaction between all the parts. I often use the trick abused by every EDM producer on earth: sidechain compression. I’ll have the drums and percussion on one bus, and all the other instruments on another, and sidechain compress the instruments with the drum bus’s bass. You can really mash the instrument bus and sound like every terrible EDM hit with that off-beat pumping thing.

        But you can also be subtle with it, and it keeps the kick sound slightly forward in the mix without having to turn it up.

        With a stem file, you can’t do that, so the separation and interaction of different parts of the track are lost. That might not matter to most people, but it would drive me crazy.

        • Why can’t you do that with stems?
          You might have to sidechain stuff separately by groups or do it on the tracks. Am I missing something?

          • Matt Mclean

            You could do that with stems, and in fact thats the way I do it with live backings for my electronic band. Disable the output of say kick drum, but the chain will still send to synth bus when exporting, retaining the side chains and grooves.. Yes the master will react differently due to the way a limiter works, so the mixes will not sound exactly the same, but you can get it fairly close.

          • chaircrusher

            We’re talking about something that gets loaded into a DJ app. Of course if you mix down to stems and load them into a DAW you can use sidechain compression. But I doubt that’s going to be a feature of Traktor, or something a DJ in front of a heaving crowd can manage.

          • What I mean is do it before it gets exported, sure when you’ll take off the drum track you still have the pumping effect on the bass or the other stems. And when the tracks are not at 100% level it might sound like it’s not all that glued together. But it surely works.
            I often export groups of tracks from ableton to mix in mixbus or studio one, I do all my sidechaining in ableton because of it’s nice routing capabilities.
            You could even decide to put your kick and everything is not high mids and highs in the same track to smash it right and maybe have just some filtering reacting to the kick on other stems for cleaning further. And you can still export as stereo track for a particular track if your after that.
            Personally I think workarounds are the salt of music production.

        • Tom

          It would matter to me too, but fear not, you can still do it.
          I nearly always use a muted channel to provide the source for a sidechain anyway.
          All you need to do using your preferred technique, pull down the fader on the bus containing the output of all your drums, and use a (pre fade) send to a 2nd bus for your compressors to SC from. 🙂

      • It’s really no different than when a client asks me to master stems for them to use in their live shows I believe. Typically in those cases the stems are treated individually so they sound as good as possible when summed. Then a separate master is done for the stereo file that will also get released separately.

        There’s still no word from NI if the container format will have a separate stereo file as well as the 4 stems, but either way it shouldn’t be a big deal from a mastering engineers perspective. I plan on fully supporting this with my mastering business once NI makes the tool available, it’s a great opportunity.

      • this is were it gets cool. if you turn the mastering process around we will get some new interesting things happening. in a way making a 2-track export in a DAW and master than feels silly to me. any mastering engineer would fix the mix if he/she was able to do that. so why not do everything in the DAW? I do this for years now. works fine.

    • chaircrusher

      Interesting question. One person to ask would be Stewart Walker, since he masters the NI Remix sets. @peterkirn:disqus?

  • Xebulon

    How would this affect the mastering process? Aren’t most mastering plugins applied on the stereo bus? So I’d imagine it would affect the final sound to have the 4 stems mastered separately rather than mastered together.

    • Eric

      I’m no mastering expert by any stretch, but couldn’t you (or whatever software is mastering) simply apply whatever should be applied to the stereo bus to each individual track?

      • Eric

        Scratch that. Just re-read and that’s essentially what you’re asking lol I guess in this case you’d really need to take that in to consideration and get plugins specifically for each stem. Would definitely add to the mastering process :S

        • Xebulon

          Yeah, effects like limiters and compressors, when triggered by audio that exceeds their threshold, will affect the volume of all tracks when placed on a stereo bus (creating a sometimes-desirable pumping effect). But if there are 4 separate tracks with 4 limiters then they would affect the volumes of each track independently so the end result would certainly sound different.

      • Well, yes and no. Some of those stems releases above really represented the original project stems.

        We may be going about this all wrong. I think the fundamental question is, what is a producer’s concept when releasing music this way – how do they want it to be heard? And do they want to release stems at all? Because while interesting, these aren’t essential to DJs. And if you do want fans to cough up more money for stems, you’d better have some idea in mind of why.

        • Jos Smolders

          Although everything depends on your definition of mastering and its function in my book it’s essentially useless to master stems separately. I see the stems mostly used as an exchange format between artists who want to work online. I suggest to go into mastering after all the creative exchange business has finished. Or not master at all if you want to keep a stem file ‘alive’ forever.

          By the way: there already have been multichannel wavs for years. It ticks all of the abovementioned boxes. So it’s not as innovative as we think.

          • Yes, but the fact that people are DJing with multichannel WAVs suggest there might be an opportunity to create something tailored to the use case.

            Mastering: mastering I think ought to be defined as preparing music for a distribution format. There was a time when we were mastering to mono and not stereo. In broadcast and film, you have more complex scenarios, too – like surround. (Worst of all, in many surround applications you have to contend with automated mixdowns that sum – happily that doesn’t happen here.)

            Look, this is a format for people who want to use this format. If they do, they can find a way to make it work. If they don’t – if no one does – it’s okay to walk away and let this die.

          • Jos Smolders

            Hmm, I don’t think DAW makers will agree. Yet another format… while there already is one that works well. But hasn’t been properly promoted.

          • DPrty

            ambisonic WAVE_FORMAT_EXTENSIBLE files
            http://people.bath.ac.uk/masrwd/ncdream/researchdev/wave-ex/wave_ex.html

            Reaper Change log
            v5.0pre12 – February 9 2015+ WAV reading: support ambisonic WAVE_FORMAT_EXTENSIBLE files

            http://www.ambisonic.net/mulchaud.html

    • Yanakyl

      I was thinking the same.
      Streaming stems for mixing could be fun, I guess even some end consumers might be happy to pay for that.

    • Right, I think the way to conceptualize this would be that you’re separately mastering for four tracks. So it’s another reason to bus.

      There would be different approaches:
      1. Do a separate stereo master on the master bus, then a separate “Stems” master on four separate buses (listening to the mixdown to hear the stereo result). You might also compare these, depending on how important it was to you for these to match.

      2. Skip the stereo master entirely and master from the stems. True, as others say here, those buses’ dynamics then become independent, but it’s still possible to combine compression on those buses in such a way that you get the resulting dynamic range you want.

      Option #2 isn’t quite as weird as it might sound, in that some mix/mastering engineers will work in certain cases from stems.

      Film production is already dealing with issues like this because of the necessity to distribute in surround, etc. – there are more dimensions to mastering.

      I’m not a mastering engineer, though, so I may poll some folks to see what they think. NI wins twice on this, though, as then they’re also possibly the ones selling you your mastering tools. 😉

      • chaircrusher

        Undoubtably the stems will be individually mastered, For this whole thing to work, they’d not be as heavily effected as a proper stereo mastering job. Otherwise everything would turn to a soupy mess.

        It raises an issue about the wisdom of handling tracks this way. A proper mixdown requires a producer to manage the interaction between all the parts. I often use the trick abused by every EDM producer on earth: sidechain compression. I’ll have the drums and percussion on one bus, and all the other instruments on another, and sidechain compress the instruments with the drum bus’s bass. You can really mash the instrument bus and sound like every terrible EDM hit with that off-beat pumping thing.

        But you can also be subtle with it, and it keeps the kick sound slightly forward in the mix without having to turn it up.

        With a stem file, you can’t do that, so the separation and interaction of different parts of the track are lost. That might not matter to most people, but it would drive me crazy.

        • Yanakyl

          Why can’t you do that with stems?
          You might have to sidechain stuff separately by groups or do it on the tracks. Am I missing something?

          • Matt Mclean

            You could do that with stems, and in fact thats the way I do it with live backings for my electronic band. Disable the output of say kick drum, but the chain will still send to synth bus when exporting, retaining the side chains and grooves.. Yes the master will react differently due to the way a limiter works, so the mixes will not sound exactly the same, but you can get it fairly close.

          • chaircrusher

            We’re talking about something that gets loaded into a DJ app. Of course if you mix down to stems and load them into a DAW you can use sidechain compression. But I doubt that’s going to be a feature of Traktor, or something a DJ in front of a heaving crowd can manage.

          • Yanakyl

            What I mean is do it before it gets exported, sure when you’ll take off the drum track you still have the pumping effect on the bass or the other stems. And when the tracks are not at 100% level it might sound like it’s not all that glued together. But it surely works.
            I often export groups of tracks from ableton to mix in mixbus or studio one, I do all my sidechaining in ableton because of it’s nice routing capabilities.
            You could even decide to put your kick and everything is not high mids and highs in the same track to smash it right and maybe have just some filtering reacting to the kick on other stems for cleaning further. And you can still export as stereo track for a particular track if your after that.
            Personally I think workarounds are the salt of music production.

        • Tom

          It would matter to me too, but fear not, you can still do it.
          I nearly always use a muted channel to provide the source for a sidechain anyway.
          All you need to do using your preferred technique, pull down the fader on the bus containing the output of all your drums, and use a (pre fade) send to a 2nd bus for your compressors to SC from. 🙂

      • It’s really no different than when a client asks me to master stems for them to use in their live shows I believe. Typically in those cases the stems are treated individually so they sound as good as possible when summed. Then a separate master is done for the stereo file that will also get released separately.

        There’s still no word from NI if the container format will have a separate stereo file as well as the 4 stems, but either way it shouldn’t be a big deal from a mastering engineers perspective. I plan on fully supporting this with my mastering business once NI makes the tool available, it’s a great opportunity.

      • this is were it gets cool. if you turn the mastering process around we will get some new interesting things happening. in a way making a 2-track export in a DAW and master than feels silly to me. any mastering engineer would fix the mix if he/she was able to do that. so why not do everything in the DAW? I do this for years now. works fine.

    • chaircrusher

      Interesting question. One person to ask would be Stewart Walker, since he masters the NI Remix sets. @peterkirn:disqus?

  • Jos Smolders

    I sympathize. Considering the 90-9-1 rule I don’t think it’s gonna fly high. But I sympathize.

  • Jos Smolders

    I sympathize. Considering the 90-9-1 rule I don’t think it’s gonna fly high. But I sympathize.

  • Jos Smolders

    I sympathize. Considering the 90-9-1 rule I don’t think it’s gonna fly high. But I sympathize.

  • Scott

    Why would they limit it to only 4 tracks? Seems very limiting.

    • foljs

      Compared to 2?

  • Scott

    Why would they limit it to only 4 tracks? Seems very limiting.

    • foljs

      Compared to 2?

  • Scott

    Why would they limit it to only 4 tracks? Seems very limiting.

    • foljs

      Compared to 2?

  • King Britt

    Exactly Why I have been using Ableton to play for past 4 yrs (getting shit from djs (I dont care)…. now they get it lol

  • King Britt

    Exactly Why I have been using Ableton to play for past 4 yrs (getting shit from djs (I dont care)…. now they get it lol

  • King Britt

    Exactly Why I have been using Ableton to play for past 4 yrs (getting shit from djs (I dont care)…. now they get it lol

  • Video?

    Wonder if this might include video files eventually?

  • Video?

    Wonder if this might include video files eventually?

  • Video?

    Wonder if this might include video files eventually?

  • an3

    ..did anyway mention the possibility to make stems for your own tracks and songs ? and available software for this /

  • an3

    ..did anyway mention the possibility to make stems for your own tracks and songs ? and available software for this /

  • an3

    ..did anyway mention the possibility to make stems for your own tracks and songs ? and available software for this /

  • Ries

    I have been playing like this with Ableton and an APC40 for years. Part of the fun is researching samples and reconstructing tracks to play https://www.youtube.com/watch?v=mXZHXWNOrdw&spfreload=10 https://soundcloud.com/misterries/live-at-booty-night-2009

  • Ries

    I have been playing like this with Ableton and an APC40 for years. Part of the fun is researching samples and reconstructing tracks to play https://www.youtube.com/watch?v=mXZHXWNOrdw&spfreload=10 https://soundcloud.com/misterries/live-at-booty-night-2009

  • Ries

    I have been playing like this with Ableton and an APC40 for years. Part of the fun is researching samples and reconstructing tracks to play https://www.youtube.com/watch?v=mXZHXWNOrdw&spfreload=10 https://soundcloud.com/misterries/live-at-booty-night-2009

  • max

    what a nonsense, who wants his/her fine work of art ripped apart?
    these also is no $ to be made here by selling to djs only.

    fail

    • drno

      I think its sad. The original artist is already basically doing everything for free, now it will be almost impossible to get proper credit cuz the ego driven DJ is going to be the focus and co-producer of the music. It’s a rock concert now.

  • max

    what a nonsense, who wants his/her fine work of art ripped apart?
    these also is no $ to be made here by selling to djs only.

    fail

    • drno

      I think its sad. The original artist is already basically doing everything for free, now it will be almost impossible to get proper credit cuz the ego driven DJ is going to be the focus and co-producer of the music. It’s a rock concert now.

  • max

    what a nonsense, who wants his/her fine work of art ripped apart?
    these also is no $ to be made here by selling to djs only.

    fail

    • drno

      I think its sad. The original artist is already basically doing everything for free, now it will be almost impossible to get proper credit cuz the ego driven DJ is going to be the focus and co-producer of the music. It’s a rock concert now.

  • Dan Reifsnyder

    The second this is actually available, some hammerhead is going to say “Hey, why not put up 4 speakers in the corners of the room and send a separate stem to each one of them? We can make it sound like you’re sitting in the middle of the band!”

    Having lived (and recorded) through the alien abortion that was Quadraphonic Sound (http://en.wikipedia.org/wiki/Quadraphonic_sound) back in the 70’s and 80’s, I shudder to see it rising from it’s smoking grave wearing a button that says “Hi, my name is STEMS!”.

    I’m all for technology that actually helps me create music, but I’m skeptical that this does.

    Remember, “Those not old enough to remember the past are going to make the same #)$(*&$ mistakes using shiny new technology”.

    And yes, my beard has been grey for a long time.

    • Actually, I think I probably said that already. 🙂

      Come on, actual quadraphonic spatialization can sound really good – not like bad quad mixing. But no, using Stems for that doesn’t make any sense – it’s not the tool for the job.

      Anyway, it doesn’t appear people are going to try to distort Stems for this application, so rest easy.

      • Dan Reifsnyder

        I think you did say that (well-written article, btw)… I was just feeling particularly deja-vu’d this morning as I read it. Composing for games and video I often work with multiple stems with different orchestrations that get crossfaded during playback… which seems to be the raison d’etre for pushing this new technology. Well, that and selling the STEMS for more money 🙂

      • I would LOVE to distort STEMS for this purpose. How much longer do have to hear stereo tracks blasted in our face every time we go to a club to hear a DJ ? I have been trying to find a viable to way to DJ in surround for years. This seems like one of the most promising developments in sound spatialization for DJs that I’ve seen in recent memory, or ever. Is this really going to be a multi-channel format sent out of 4 outputs or just 4 channels to remix inside TRAKTOR (or DJ program of choice) and then send out a stereo mix like every other gd DJ set ever in the history of DJing ?

    • On a similar note, wasn’t this the idea behind General Midi? I feel like we’ve been here before, and more than once. This smells like a flop to me…. Particularly because a lot of producers go out of their way to ensure their bassline or vocal cut isn’t isolated and reused…

    • Since you have direct, hands-on experience with recording Quadrophonic Sound, you should probably remember that “There was no uniformly compatible system for making Quad LPs. There were three incompatible systems SQ (developed by CBS Records), CD-4 (developed by RCA, no relation to Compact Discs, which wouldn’t be invented for another 10 years) and QS (developed by Sansui).”

      (http://historysdumpster.blogspot.com/2013/07/the-history-of-stereo-and-quadraphonic.html)

      As clearly stated in the article, STEMS has been designed form the outset to be an OPEN STANDARD, or did you gloss over that fact while crafting your grumpy-old-school-audio-guy rebuttal ? Furthermore, no one was really DJing with QS since there were three competing systems each releasing music on different labels. It simply was not feasible. But times have changed – for the better. So has recording, DJ, and club PA technology. Read up on it before you beat up on it.

      • Ahh my Dad’s old Lafayette SQ Surround. Pink Floyd’s Dark Side of the Moon in Quad – Probably the Best Sounding Rock Record EVER, (yeah Stereo was nice but they REALLY knew (or figured out) how to work four channels. (yeah we have 5.1 Surround blah blah but who makes records and records/plays back?)

      • Tom Vear

        100% This.

  • Dan Reifsnyder

    The second this is actually available, some hammerhead is going to say “Hey, why not put up 4 speakers in the corners of the room and send a separate stem to each one of them? We can make it sound like you’re sitting in the middle of the band!”

    Having lived (and recorded) through the alien abortion that was Quadraphonic Sound (http://en.wikipedia.org/wiki/Quadraphonic_sound) back in the 70’s and 80’s, I shudder to see it rising from it’s smoking grave wearing a button that says “Hi, my name is STEMS!”.

    I’m all for technology that actually helps me create music, but I’m skeptical that this does.

    Remember, “Those not old enough to remember the past are going to make the same #)$(*&$ mistakes using shiny new technology”.

    And yes, my beard has been grey for a long time.

    • Actually, I think I probably said that already. 🙂

      Come on, actual quadraphonic spatialization can sound really good – not like bad quad mixing. But no, using Stems for that doesn’t make any sense – it’s not the tool for the job.

      Anyway, it doesn’t appear people are going to try to distort Stems for this application, so rest easy.

      • Dan Reifsnyder

        I think you did say that (well-written article, btw)… I was just feeling particularly deja-vu’d this morning as I read it. Composing for games and video I often work with multiple stems with different orchestrations that get crossfaded during playback… which seems to be the raison d’etre for pushing this new technology. Well, that and selling the STEMS for more money 🙂

      • I would LOVE to distort STEMS for this purpose. How much longer do we have to hear stereo tracks blasted in our face every time we go to a club to hear a DJ ? I have been trying to find a viable to way to DJ in surround for years. This seems like one of the most promising developments in sound spatialization for DJs that I’ve seen in recent memory, or ever. Is this really going to be a multi-channel format sent out of 4 outputs or just 4 channels to remix inside TRAKTOR (or DJ program of choice) and then sent out in stereo like every other gd DJ set ever in the history of DJing ?

    • On a similar note, wasn’t this the idea behind General Midi? I feel like we’ve been here before, and more than once. This smells like a flop to me…. Particularly because a lot of producers go out of their way to ensure their bassline or vocal cut isn’t isolated and reused…

    • Since you have direct, hands-on experience with recording Quadrophonic Sound, you should probably remember that “There was no uniformly compatible system for making Quad LPs. There were three incompatible systems SQ (developed by CBS Records), CD-4 (developed by RCA, no relation to Compact Discs, which wouldn’t be invented for another 10 years) and QS (developed by Sansui).”

      (http://historysdumpster.blogspot.com/2013/07/the-history-of-stereo-and-quadraphonic.html)

      As clearly stated in the article, STEMS has been designed from the outset to be an OPEN STANDARD, or did you gloss over that fact while crafting your grumpy-old-school-audio-guy rebuttal ? Furthermore, no one was really DJing with QS since there were three competing systems each releasing music on different labels. It simply was not feasible. But times have changed – for the better. So has recording, DJ, and club PA technology. Read up on it before you beat up on it.

      • Ahh my Dad’s old Lafayette SQ Surround. Pink Floyd’s Dark Side of the Moon in Quad – Probably the Best Sounding Rock Record EVER, (yeah Stereo was nice but they REALLY knew (or figured out) how to work four channels. (yeah we have 5.1 Surround blah blah but who makes records and records/plays back?)

      • Tom Vear

        100% This.

  • Dan Reifsnyder

    The second this is actually available, some hammerhead is going to say “Hey, why not put up 4 speakers in the corners of the room and send a separate stem to each one of them? We can make it sound like you’re sitting in the middle of the band!”

    Having lived (and recorded) through the alien abortion that was Quadraphonic Sound (http://en.wikipedia.org/wiki/Quadraphonic_sound) back in the 70’s and 80’s, I shudder to see it rising from it’s smoking grave wearing a button that says “Hi, my name is STEMS!”.

    I’m all for technology that actually helps me create music, but I’m skeptical that this does.

    Remember, “Those not old enough to remember the past are going to make the same #)$(*&$ mistakes using shiny new technology”.

    And yes, my beard has been grey for a long time.

    • Actually, I think I probably said that already. 🙂

      Come on, actual quadraphonic spatialization can sound really good – not like bad quad mixing. But no, using Stems for that doesn’t make any sense – it’s not the tool for the job.

      Anyway, it doesn’t appear people are going to try to distort Stems for this application, so rest easy.

      • Dan Reifsnyder

        I think you did say that (well-written article, btw)… I was just feeling particularly deja-vu’d this morning as I read it. Composing for games and video I often work with multiple stems with different orchestrations that get crossfaded during playback… which seems to be the raison d’etre for pushing this new technology. Well, that and selling the STEMS for more money 🙂

      • I would LOVE to distort STEMS for this purpose. How much longer do we have to hear stereo tracks blasted in our face every time we go to a club to hear a DJ ? I have been trying to find a viable to way to DJ in surround for years. This seems like one of the most promising developments in sound spatialization for DJs that I’ve seen in recent memory, or ever. Is this really going to be a multi-channel format sent out of 4 outputs or just 4 channels to remix inside TRAKTOR (or DJ program of choice) and then sent out in stereo like every other gd DJ set ever in the history of DJing ?

    • On a similar note, wasn’t this the idea behind General Midi? I feel like we’ve been here before, and more than once. This smells like a flop to me…. Particularly because a lot of producers go out of their way to ensure their bassline or vocal cut isn’t isolated and reused…

    • Since you have direct, hands-on experience with recording Quadrophonic Sound, you should probably remember that “There was no uniformly compatible system for making Quad LPs. There were three incompatible systems SQ (developed by CBS Records), CD-4 (developed by RCA, no relation to Compact Discs, which wouldn’t be invented for another 10 years) and QS (developed by Sansui).”

      (http://historysdumpster.blogspot.com/2013/07/the-history-of-stereo-and-quadraphonic.html)

      As clearly stated in the article, STEMS has been designed from the outset to be an OPEN STANDARD, or did you gloss over that fact while crafting your grumpy-old-school-audio-guy rebuttal ? Furthermore, no one was really DJing with QS since there were three competing systems each releasing music on different labels. It simply was not feasible. But times have changed – for the better. So has recording, DJ, and club PA technology. Read up on it before you beat up on it.

      • Ahh my Dad’s old Lafayette SQ Surround. Pink Floyd’s Dark Side of the Moon in Quad – Probably the Best Sounding Rock Record EVER, (yeah Stereo was nice but they REALLY knew (or figured out) how to work four channels. (yeah we have 5.1 Surround blah blah but who makes records and records/plays back?)

      • Tom Vear

        100% This.

  • I found out a thing which seems not to be mentioned in your article: it adds the 4 ‘stem’ tracks ALONG with the full mix. So it’s in fact a 6 track protocol, or maybe even 10 tracks if all tracks can be stereo (haven’t found any info about that if it’s true). But even stereo + 4 additional mono tracks is a cool thing.

  • Guest

    I found out a thing which seems not to be mentioned in your article: it adds the 4 ‘stem’ tracks ALONG with the full mix. So it’s in fact a 6 track protocol, or maybe even 10 tracks if all tracks can be stereo (haven’t found any info about that if it’s true). But even stereo + 4 additional mono tracks is a cool thing.

  • Guest

    I found out a thing which seems not to be mentioned in your article: it adds the 4 ‘stem’ tracks ALONG with the full mix. So it’s in fact a 6 track protocol, or maybe even 10 tracks if all tracks can be stereo (haven’t found any info about that if it’s true). But even stereo + 4 additional mono tracks is a cool thing.

  • missed one huge thing: Stems will offer 4 stereo tracks. This is huge! Which is kinda cool for stuff I am doing. For example when I want to send my documentary soundmix to a video-editor and making channels for:1. full music mix
    2. sound effects
    3. set noises
    4 spoken words / dialog

    If Apple starts adding support for this in Final Cut X and Adobe in Premiere this will be killer for many people in my opinion.

    • chif

      why wouldn’t you just do 4 offline exports? do you really think that hollywood is going to use mp4s for a sound mix ?

      • to make sure sync and levels are ok. in old version of Final Cut MONO tracks were lower in volume than STEREO channels. and sync errors are still often the case. adding one synced audio-container could be a solution. although only 4 tracks is a limiting factor.

  • missed one huge thing: Stems will offer 4 stereo tracks. This is huge! Which is kinda cool for stuff I am doing. For example when I want to send my documentary soundmix to a video-editor and making channels for:1. full music mix
    2. sound effects
    3. set noises
    4 spoken words / dialog

    If Apple starts adding support for this in Final Cut X and Adobe in Premiere this will be killer for many people in my opinion.

    • chif

      why wouldn’t you just do 4 offline exports? do you really think that hollywood is going to use mp4s for a sound mix ?

      • to make sure sync and levels are ok. in old version of Final Cut MONO tracks were lower in volume than STEREO channels. and sync errors are still often the case. adding one synced audio-container could be a solution. although only 4 tracks is a limiting factor.

  • missed one huge thing: Stems will offer 4 stereo tracks. This is huge! Which is kinda cool for stuff I am doing. For example when I want to send my documentary soundmix to a video-editor and making channels for:1. full music mix
    2. sound effects
    3. set noises
    4 spoken words / dialog

    If Apple starts adding support for this in Final Cut X and Adobe in Premiere this will be killer for many people in my opinion.

    • chif

      why wouldn’t you just do 4 offline exports? do you really think that hollywood is going to use mp4s for a sound mix ?

      • to make sure sync and levels are ok. in old version of Final Cut MONO tracks were lower in volume than STEREO channels. and sync errors are still often the case. adding one synced audio-container could be a solution. although only 4 tracks is a limiting factor.

  • Mr Zippy

    ‘simple’ to me would be just put 4 stems in a .zip file and make the app load .zip and any stems found inside.

  • Mr Zippy

    ‘simple’ to me would be just put 4 stems in a .zip file and make the app load .zip and any stems found inside.

  • Mr Zippy

    ‘simple’ to me would be just put 4 stems in a .zip file and make the app load .zip and any stems found inside.

  • Rob

    I just came across a new stem distribution site called Dubseed:
    http://www.dubseed.com

  • Rob

    I just came across a new stem distribution site called Dubseed:
    http://www.dubseed.com

  • Rob

    I just came across a new stem distribution site called Dubseed:
    http://www.dubseed.com

  • Tom

    I think this is an awesome idea, i’m imagining the potential of just an instrumental and an acapella, that alone would get me excited about this. The only immediate drawback to it catching on is Pioneer, it would take them 5 years to implement this in Rekordbox. I just downloaded a demo of Traktor to see how well it integrates with the Nexus gear in HID mode, it can’t be any worse than Serato DJ in HID can it!?

  • Tom

    I think this is an awesome idea, i’m imagining the potential of just an instrumental and an acapella, that alone would get me excited about this. The only immediate drawback to it catching on is Pioneer, it would take them 5 years to implement this in Rekordbox. I just downloaded a demo of Traktor to see how well it integrates with the Nexus gear in HID mode, it can’t be any worse than Serato DJ in HID can it!?

  • Tom

    I think this is an awesome idea, i’m imagining the potential of just an instrumental and an acapella, that alone would get me excited about this. The only immediate drawback to it catching on is Pioneer, it would take them 5 years to implement this in Rekordbox. I just downloaded a demo of Traktor to see how well it integrates with the Nexus gear in HID mode, it can’t be any worse than Serato DJ in HID can it!?

  • drno

    lol. the slow death of the original artist and record label. It just got boring having a great DJ play great music released by great labels it seems. Now the DJ is the co-producer who will finished those tracks on stage. wow. I can only imagine what track lists will look like.

  • drno

    lol. the slow death of the original artist and record label. It just got boring having a great DJ play great music released by great labels it seems. Now the DJ is the co-producer who will finished those tracks on stage. wow. I can only imagine what track lists will look like.

  • drno

    lol. the slow death of the original artist and record label. It just got boring having a great DJ play great music released by great labels it seems. Now the DJ is the co-producer who will finished those tracks on stage. wow. I can only imagine what track lists will look like.