STEMS: condemned to failure?

When I first heard about STEMS I was amazed - amazed that no one has thought of that before. Or is it rather that no one wanted to actually take the effort?

I am really surprised about that rather "unsmooth" launch of STEMS. The STEMS forum is close to dead, although labels did offer STEMS a few weeks earlier than the official launch no one really advertised it too much and rather big artists still don't seem to care about it at all. Also, big props to NI here, no one "warned" me about advertising my Creator Tool here, NI (again, that's really cool) didn't remove my forum post and there was little to no interaction with @ NI usernames in the STEMS forums so far.

Now as STEMS officially launched, its range seems lower than I expected, of course the activity in that forum increased a lot but still not as much as e.g. the Maschine board will ever be.

Now to my introduction: I honestly think even NI has kind of given up a bit on this one, It took me a few hours to write a very very basic STEM creator tool which then worked almost bug-free on recent versions OS X. There were 2 things that were criticized: 1) the app didn't launch if an older version of Java was installed. 2) the app didn't work if file paths had spaces in it.
Why do I mention that?

1) https://www.native-instruments.com/forum/threads/missing-msvcp120-dll.260924/
2) https://www.native-instruments.com/forum/threads/error-exporting-stems-file.260931/ (DJ kad post)

These 2 bugs are the exact same ones I accidentally made. Please consider that they presented the creator tool a few WEEKS ago and that NI is a f*cking huge company, even having offices on different continents where a few hundred people work at. How big can the stem team be? 4 coders and 1 musician? How could they not 1) check the framework version so the user would at least obtain a readable error code 2) see this as a potential problem? 2) is something I noticed myself after a few hours of pushing the first version. I. Myself. In less than a day. They got it in their hands for A BIT longer than.

What I am saying is that I think NI only puts the littlest of their workforce into STEMS which MIGHT be an indicator that they already stopped believing in the potential of STEMS. I am not saying I would've done better, but even in a team of a few people, over that timescale, I can't understand how stuff like this can happen.


My 2 cents, I really hope this format will eventually catch up because I really believe in it, but maybe you should wait a few weeks before spending money on STEM gear. I also saw potential in Maschine, something that I still wait for since 2012 but that's another story.

 

I guess you are right on most aspects, it appears as if NI tried to find another unique selling point (usp) for S8 and D2.....
The low interest and feedback in the forum was a suprise too - but not completely unexpected. IMHO there are few issues to be regarded:

1. No interesting tracks available
To have a real impressive start well known/popular stem tracks should have been published so anybody may recognize the difference and options a DJ may have with them. Playing unknown tracks with expected life- /playtime of 4-6 weeks is useless to demonstrate the possibilities as no one recognizes it. As long as no major label shares the system there will be no lasting success....

2. Hardware connection
A lot of Traktor users are disappointed realising that the separate waveforms are only available on S8 or D2. Not an essential feature but also not a good promotion..

3. Missunderstanding
A lot of questions show the expectation of naive(?) users: How can I do my own stem file, do I need separate wave files (only have mp3), how can I separate an existing track and so on. This may give a good impression if contemporary DJing (or missing knowledge) but explains too the absence of interest.

In my conclusion stems may become interesting if all corresponding condtions would have much better third party support - but I doubt it. At the moment I see stems taking the same way as remix decks, only quicker.

An interesting container - with boring content...

 

Yea bro you're right, didn't know the latter tho but like, no one managed to get it into a "major" software.

For DAW's you could've just exported/imported several files which would work pretty much universal for all of them but in the field of DJing, no one's gonna use anything if it's not supported by Traktor, Serato or VirtualDJ due to the fact that it is extremely unhandy to have a second app on the screen in order to be able to experience a STEM-like behavior if not natively supported, which just isn't worth it.

NI can literally turn **** into money but I think even here they will fail for some reason. STEMS will stay in Traktor but NI should've cared much more about who to cooperate with as mainstream music is still far from being released. Their plan to sell exclusives before everyone gets his hands on them sure made them a few thousands, but on long term it would've been better the other way around. So for now, they failed both technically and content wise, hence the very low interest. Let's see if they'll be able to rescue that one.

 

Good discussion here and some very interesting perspectives on STEMS.

NI definitely sold me upfront on the idea, which is why I picked up two F1s and purchased some stems. However, I became deflated when I was unable to integrate stems into my workflow the way I hoped, or in the way that NI advertised them. Stems just flat out don't work properly for me in Traktor 2.9 (windows) right now, and without functional software, stem files are useless and the F1s become just midi devices. Regardless, I don't have qualms about investing a little money in trying to believe in something that is new and exciting. I've always trusted NI, and I love that they are always trying to innovate, but sometimes I think they get a little ahead of themselves in pushing out an idea, hardware product, etc to sell something before they know that all the software that drives that idea (or hardware) works properly. We all know that fixing bugs on the back end of software is costly and detrimental to user appeal, which has been a constant weakness with NI since I've become to know them.
That said, I think there is a lot of work to do on the technical (software) and marketing aspects of the stems format for it to sustain. The potential is definitely there, but potential is nothing without execution, drive and perseverance, and this is a big challenge if you ask me.
I'm hoping a Traktor 3.0 is where we see major improvement, and what brings the stems format into the lime light that NI has advertised to be . I'm also hoping that we see more mainstream and other genres of music adopt the format, because the availability of stems now is limited to a lot of "boring" content if you ask me. So for the time being, my stem files will probably sit in a lonely folder while my F1s collect dust.

 

Question: Which DJ has real need for stems?
When you´re on Deep House, Trance, Techno, etc. there are already some nice tracks available to deal with - but the audience would never appreciate or even notice what you´re doing. No one cares if the actually played track is original, remixed by design or "re-arranged by the DJ himself".
When you´re on mainstream, party, wedding, etc. the crowd loves the tracks they know well - no special remixes (please for we cannot sing along the refrain). Even the best mashups can fail on these occasions.

I love the idea of having my "all time favourite" dance classics as *.stem.mp4 format to do some impressive surprises on an average party but that is just my very personal view. Depending on the situation it may be an interesting option - but not a need.

Most important is the crowd enjoyes my performance and has a lovely evening. If stems help to pay more attention to my work it´s ok. But they won´t change the audiences flavor...

 

I would somewhat agree with you about the deep house type stuff. But when it comes to mainstream stuff, remixes can be great if you have good remixes. Also I have recently started creating my own hip hop stems with the main, instrumental, clean acapella and dirty acapella. So now one file takes the place of 5 files. That alone is worth it. You know all of this has to mature. You know when they first started adding cue points and seemless looping ability to stuff, Alot of the same things were said back then...now most DJs can't live without it.

This is one of this biggest problems. NI needs to address this widening gap they are creating either through marketing or product development that is sending the message that they only care about dance music. This is why Serato is getting ahead of Traktor, they are trying to be more inclusive...

 

I don't know of a single genre that couldn't make use of the stem format, it just hasn't reached them yet. I don't think Native Instruments could have reached all genres at the release of the stem format just generally down to the technology used. The dance music labels & digital stores embraced the new format which has helped it in those areas but I think it just needs someone to lead the way and show the other genres that it is worth doing.

Hip-hop/R&B Djs have always mixed instrumentals and acapellas, they'd definitely benefit from having both those elements within the same track loaded onto one deck - can you imagine the crazy new routines?

It would also be ideal for commercial DJs who need to get in and out of tracks very quickly. They pack so many tracks into a set, playing just the best parts but generally the difficulty with that is coming out of track too early and it clashing with the next. With the stem format that issue would mostly be taken away, dropping out the elements that clashed and providing quicker transitions.

It's very very early days yet, all it needs are for creative producers within the other genres who want to push the live performance of their tracks. If they embrace it and demonstrate it, it wouldn't take too long to catch on.

As for whether the crowd would notice, the whole point is them not necessarily noticing. Everything we've ever done as DJs is to blend or cut records seamlessly together and NI have just given us tools to help towards doing that better. Who knows whether it'll still be around in 5 years but then who knew that of the mp3? It's whether it gets embraced or not, fingers crossed other manufacturers, record labels and genres will pick it up!

 

Don´t forget that stems are nothing new for all labels and producers! As matter of fact it is already the basic music recording routine for decades (https://en.wikipedia.org/wiki/Stem_mixing_and_mastering) f.e. on multitrack tape recording machines even long before any computer program was used.
So the technique obviously is available for more than 50 years - why no remarkable interest or success??

Every single genre could have already participated for long time - but didn´t for decades......

And the reason was never the missing tool!

 

This could have been done earlier, very true. It's always about the implementation though!

It takes someone to come up with the software that can process it in real time, the hardware that can perform with it and NI just happened to have been the ones to do it. You could have EASILY done the same thing with Ableton years ago, it has the power and capabilities but not the simplicity with which NI have implemented it for DJs in this case. You only have to look as far as Deadmau5, he's been using Ableton to perform with for years but his setup is far from simple. I can't imagine a hiphop DJ would go for a setup like that?

But the stem format is something you can see easily fitting into DJs setups. It's simple, makes use of most of the equipment they already have and software they're used to DJing with (assuming they're a Traktor DJ). As with most technology, it's not often about creating something brand new but rather taking something existing and thinking of a different way to use it.

PS> By stem format not reaching them I meant NI STEMS, not DAW audio stems

 

NI is releasing their STEM creator tool SDK kit (if they haven't already) as open source. So expect STEMS implementation in future releases of DAWs and other independent developers.

 

Ok, so let´s play the game:
If I were marketing or sales manager of any major label: How could you convince me to release a (additional) *.stem file to the next upcoming commercial track ?
(And please bare in mind: My task is making maximum profit - not loosing any money. I´m neither artist nor producer.)

Come on: Answers rsvp!

 

If I had that role at a major label, I'd 1) create a cost model for selling STEMs and 2) generate a new revenue stream based on that cost model.

From the business perspective maybe charge 2-5x of the regular price. As a paying customer, I would gladly pay that amount for the freedom to remix at my leisure. I spend way too much time trying to grab bits and pieces from music, especially hard to find acapellas.

 

Acapellas4u is a pretty decent resource. I can't get my head round all this stem malarkey.

 

I think this is why dance music labels have embraced the format, generally down to the lack of revenue from normal digital releases that they're looking into many other ways to add to that revenue. As noted before, stems are there already for every track its just a case of exporting them fit for the STEMS format. This is a 10 minute job for the sound engineer, I did this myself for my latest release and was no more of a chore than preparing remix parts (in fact easier).

The problem arises in the fact that most commercial labels target the listening public, not DJs, so they might not put stock in format squarely aimed at DJs. The hope comes from producer-DJs dictating to labels that they want their material in that format.

 

As sales manager I would tell you that our company does not make great margin out of sales but by playcount and related fees payed via collecting societies all over the world according to TIS/CISAC agreement.
But these agreements do only cover the song (or justified remixes) not single layers/parts of it. We would not earn a single cent if someone just plays the vocals, bassline or drums on radio, club or internet.
So even 5x prices payed by comparably low number of customers would never compensate this loss.

As marketing manager I would tell you: Why should I offer the ingredients of our latest success for undefined and unsafe use?
In addition to that we would loose all our control if somebody integrates these parts in his own track.

 

Richie, your point is valid. Would the cost model support the time involved to make the STEM? Is it worth it? Something the labels will have to work out themselves.

What you said is where I was going with all this. STEMs are definitely a niche product and your average customer will be purchasing the master full track anyway. All great point... yes, yes AND yes!

 

Probably you´re right. But this is not the kind of our mindset......
Maybe we could get people using our raw material to court but it is much cheaper to stop any possible abuse from the start.

Avoiding any conceivable risk is todays managers holy grail!

 

they wont notice, what they will is have heard something they havent heard elsewhere and they will go away having had a good time and isn't that the point

 

I don't think it has anything to do with money like some other people said since it's an open-source format. The problem is that there is no incentive to have stems because you can only play that format on one type of equipment. Since its open-source, it will take time before other companies like pioneer decide to implement it in their gear but it's not impossible. The problem is that NI did not cooperate with other companies first and invest a little bit with them to share the wealth. As a result stems is isolated and playable only on one type of devices. Do I even get a copy of the real song along with a stems purchase? (don't even know). And think about the mess this creates for an artists if you have one database for two softwares (Traktor + Rekordbox) and then if your rekordbox cannot read stems then you're missing on tons of files to manage properly. The truth is that DJs need to be adaptable on any hardware, something NI seems to not be paying attention to with all their proprietary stuff. Otherwise their idea is good.

In addition, I see NI lately releasing tons of products that go against the concept of DJ adaptability. For example the jog wheel are missing on their new controllers! WTF! I real DJ in his career will sometimes need to do back-to-back with someone else. And just recently (1 month ago) they came out with the Beatport Pro software to export to Rekord box and avoid the problem of playing your tagged songs from Traktor only. But I waited 3 years for Beatport Pro software on PC! I mean WTF, this is really not professional thinking! They are just not thinking about adaptability even right now with stems. With rekordbox DJ I now have no choice but to convert. Sorry NI, but you could have prevented this by paying attention to what makes real DJs tick. Not bedroom DJs. If NI is reading this, please put it in the feedback bin. Make your stuff less complicated to handle and easily adaptable.

 

Sorry buddy. You miss the point. They must do whatever it takes to have other companies implement it if it needs to work. So while you're talking to me about who's responsibilities it is. I'm talking about what needs to be done. As of NOW, I can't play stems on any hardware. That's why no body cares. Period. Also, look at how NI didn't make HID available on every CDJ it was compatible with just in order to sell more controllers. Why should others companies be more attentive to NI's technology. The truth is that's it's not a "responsibility" issue. It's a marketing issue. If you want people to bite into your product you have to sell yourself for it.

We can sit here discuss alternate theories all day long on how we perceive things and who's responsibility it is. But if stems cannot be played on other hardware, marketing-wise it won't work. Period. So the only solution is to market it stronger with Pioneer, etc... It's not rocket science, it's still NI's responsibility to make it work if they want it to work.