Music, The City & You is mostly about new ways of discovering music through your location or via elements of the city in which you live. It’s an ever expanding presentation that also has a blog where I’ll be keeping good examples of the stuff that I find.

Hopefully I’ll be giving the talk a few times this year but Future Music Camp was a wonderful start. Thanks so much to Ryan, Daniel and Eva for being so accommodating and bringing me over here to try this out.

Music, The City & You @ Future Music Camp, Mannheim, Germany, May 2011

May 9th – SF Music Tech Sumit, San Francisco (lanyrd.com)
May 13th – The Great Escape, Brighton, UK (lanyrd.com)
May 19th – Liverpool Sound City, Liverpool, UK (lanyrd.com)
May 27th – Future Music Camp, Mannheim, Germany (lanyrd.com)

I’ll also be attending the SF Music Hackday on May 7th & 8th and dropping in to the demos for Music Hackday Barcelona in June.

For coverage of any of the conferences that I can pull together, check out my profile over at lanyrd.com

OutNow, the hack that I created at the MIDEM Music Hackday in January will be taking on a new life very soon as Releasify.me – let’s face it, I wasn’t ever going to be able to find a variation on OutNow to use in a domain.

The tidied up, non-laptop version that people saw at MIDEM will soon be live in an alpha state, doing what it was intended to do, which is show off the week’s new releases, allow people to check out clips, get links for purchase/streams and compare which of the week’s new releases is actually the hottest.

It’ll still be using all the good data from around the web that it was built with initially, and I plan to iterate new features into it about once a month, or more often if I get on a roll.

If you want to follow the progress, it’s all being documented over at blog.releasify.me.

The format was almost the same as the Music Hackday’s that have gone before, except that there were no presentations of tech and services from companies involved, we just walked in, plugged in, and got right to it.

I’m not sure any of us knew what to expect from the audience who were milling around 3 floors below our windowless hack-den, but the response to the “Meet The Hackers’ panel I was on with Dave (Soundcloud), Matt (Formerly Last.fm), Paul (Echonest), Ben (Goldsmiths Uni), Dan (Extension FM) and Hannah (Last.fm), was extremely positive.

The hack presentations went down even better, mainly due to the hugely creative and amusing stance taken by some of the teams. Some absolute genius came out of the room, you can check out the full list over at cannes.musichackday.org

My hack was based around showcasing how much usable data is available from the companies who took part and I built a site that informs users of new releases each Monday, along with all the data they could wish for about the artist and the track itself. It even includes buy links!

The write up for my hack is on the Music Hackday Wiki, but I’ll write more about that and each of the hacks and some other thoughts about why this take on Music Hackday was so positive when I’m not sitting in a hotel lobby waiting for a taxi.

Working at UMG means that I’m free to choose to work with whatever technology I need to use to get the job done. There’s a lot I could talk about but I thought I’d concentrate on a few that I’m excited about and make life so much easier when you’re working for a huge company on projects that have lots of stakeholders.

Mixpanel (www.mixpanel.com)
When you work somewhere that likes their numbers right now Mixpanel is a dream come true. They provide real time analytics for web apps, android and iOS applications with extremely simple implementation. I haven’t even begun to dig around in the A/B testing uses yet but I’m already a huge fan. They also have full API access, which means that my dreams of dynamically reordering content based on popularity are now so much closer.

Heroku (www.heroku.com)
Again, when time is of the essence and you need a simple/complex web app live by the end of the day, if you can wrap it in Rails then Heroku is going to ensure you’re in the pub at 6pm, not tinkering around in a shell updating your gems on a server somewhere.

Providing simple, quick launch Ruby on Rails hosting is the core of Heroku’s business and for UMG it means that we can be pushing promotions using pre packaged software like Soundcloud’s social download and remix competitions out to the fans in a matter of hours, not days.

Wrapping up a site using a gem CMS like Radiant means that we could be launching sites in a day as well.

Hockey (http://github.com/TheRealKerni/Hockey)
Distributing mobile application betas (ad hocs) to stakeholders at a company like UMG is a nightmare, and often means you’re going around helping people install them more than working on them.
Hockey’s over the air distribution software allows you to distribute self-updating versions of applications to people in the first instance, then, when there’s a new version uploaded to the server, it will update all the handsets it exists on.

Sure, it takes a little setup and you need to provision a server to run it from but the half a day spent setting it up and gathering the UDIDs of all those that require the app is worth the effort once you start distributing the latest versions without needing your iPhone cable.

MobileRoadie API (www.mobileroadie.com/home/api)
MR is one of the most exciting technology companies in the mobile applications field and their platform is invaluable to any industry that requires a quick start to application development that will drive engagement at a time when everyone simply must have an iPhone app.

However, the most exciting part of their platform for me is their API. It offers the opportunity to put an artist centric, simple CMS at the heart of the digital offering, not just the mobile app but their website too. One update takes care of everything and every inbound communication from users is distributed back out to both platforms. Using this as the central CMS also gives you the opportunity to use it to build mobile applications on platforms that MR doesn’t currently publish to.

Photo by deanj on Flickr, used under Creative Commons Attribution 2.0

Universal Music Group’s first release using this data is Timewheel, an iOS application aimed at discovering and purchasing music from years gone by. It’s pretty addictive. It’ll certainly help settle the arguments about who was number one on a particular date. Maybe you’re thinking about putting together a playlist for someone featuring tracks from the day/year they were born, this app will help you do that.

I must have repressed the fact that ‘Seven Tears’ by Goombay Dance Band was number one on the day I was born, but it’s true. If I want the whole chart from that day, I can buy it.

This app is a great example of how presenting a database to users in a way that allows them to explore the data within can be enough to build an app out of.

The reaction in the first few weeks, despite the promotion not having really started yet, has proved that this type of application was already long overdue in this form.

Digging around in music can also be pretty frustrating if you don’t know what you’re looking for so there are additional editorialised elements such as recommendations from Taio Cruz (left). This helps to give the app some Universal Music Group flavour and also allows us to promote the app to fans of his, extending the reach a little bit further.

Biggups to Austin Lee (UMG Central Digital Team) for taking his wonderfully simple idea and turning into a great looking, sticky app, and thanks to Apps on Fire for the development.

Timewheel is available for iPhone right now, so grab it, let us know what you think and what we might add to it for future versions.

As of June 21st I’ve been working for Universal Music Group UK as their Product Innovation Manager. Essentially, I’m there to maximise what Universal Music do with their use of technology both internally and externally. Initially my focus is going to be on mobile applications, something I’m excited about getting properly stuck into.

Coupled with this I’m looking to make Universal Music more a ‘part of the web’ by exposing as much useful metadata around releases and any other connected assets as I can.

I’ll be writing more over the coming weeks about some of the projects I’m working on, but for now, if you’re doing something cool with technology and music, give me a shout (@martynrdavies), maybe there’s something we can do together.

Minus boss, lover of haircuts shaped with a protractor and clicky beats exponent, Richie Hawtin, has developed a plugin for Native Instruments popular Traktor software that posts each new track play to @rhawtin on Twitter, in real time.

So the rumours about DJs playing mixes and checking Twitter in the booth are most likely true! I’m kidding, I know it’s more complicated than that… Firefox makes Ableton crash for a start! True story.

Jokes aside, this is an excellent development and on thinking about it, it really could help artists out, as there are a number of uses for this kind of information that might not be so apparent on first look.

Here’s four that came to mind, and one that’s only there so I could say ‘data set’ and ‘hadoop cluster’:

1. “Hey, Performing Rights Society! Check this list!”

Getting an accurate representation of which records are actually played in clubs for the distribution of royalty payments by the PRS is really hard, but software like this could make things so much easier. I’d love to see a way to register a ‘performance twitter’ account on the PRS that could be scraped for accurate information about which tracks were played, at what venue, on what date and try to match them with the correct artists then allocate payments.

Artists that aren’t matched up properly could be presented in a simple UI that the DJ could check afterwards and clear up discrepancies.

2. “All live, all the time”

What a great way of powering a site this could be! Powering ‘live’ sites with track information allows for some epic extension and excellent levels of music discovery.

A DJ could tweak sections of their official site to react to the tracks they’re playing out in a club. The site pulls in links, pictures, video, mixes, related tracks, literally any data out there that you could match up with an artist or trackname.

There’s a great example of this in the Simon Cross constructed ‘Now Playing vs The Web’ prototype. His version is powered by the tracks currently being played by BBC Radio 1, as opposed to tracks published to Twitter but adapting it wouldn’t be hard.

3. “I so filled up the BBC with my mix the other night”

BBC Radio 1′s Essential Mix is one of the most popular DJ lead radio programmes in the world. You can find MP3 rips of the mixes online less than 2 minutes after the show goes off air and the web is filled with wonderful tracklist and mix archives dedicated to the show.

The tracklists for the mixes are like actual gold dust. No joke. If the tracks aren’t there immediately all hell breaks loose on the comments section.

The tracklistings are currently published in advance and go live right at the end of the show. This has been the norm for as long as I can remember and it definately was when it was my job to do that publishing. However, we always came across the problem of how to turn around tracklistings for Essential Mixes coming live from clubs or festivals. The demand for the list remains the same but turning them round is harder… So let’s publish them live via Twitter instead!*

*I realise that email, CMS for producers and typing up a list when you get back to the hotel are also valid options but humour me a little…

4. “It’s like Dopplr, but for music!”

If we lived in some sort of ideal world where every working DJ had the ability to live publish their sets as they happened then a ‘musical coincidences’ mashup would be lots of fun to work on. Especially for artists that release a new cut of a track on Friday afternoon to a large mailing list of DJs ready for the weekend. You could track the musical journey, play patterns, similarities, location of plays against each other and come up with an uber reaction sheet over just 1 weekend.

One line of information would be enough for me: “Your track ‘Sonic Biscuit Resolution’ was reported as being played 25 times between 6pm Friday 3rd July and 6am Monday 6th July.”

Click that and be taken through a Google Analytics style interface that shows me who played it, where and if anything interesting happened with it – for instance, the track may have appeared 3 times in a set by Annie Mac at Wax On in Newcastle. An error perhaps? It may very well mean the track got three rewinds and is in fact the standout killer track from the weekend.

Hell, if that wasn’t enough, take those plays and their location and put it on a map. If in doubt, always put it on a map!

5. “Is that a massive data set in your pocket or are you just carrying a Blackberry and an iPhone?”

Data sets are the new bevelled button. Having all this data out there would make someone’s Hadoop cluster very happy.

Everyone play nicely and tidy up after yourselves…

I realised whilst writing this that in order for something like this to work you’d need to get all standardy on people and have quite a lot of things in place in order for it to work once at scale, so if everyone could chip in on finishing this to-do list I’d appreciate it:

1. Take Richie’s software and make it work with Serato, Ableton and Pacemaker.
2. Decide once and for all whether it’s Artist Name – Track Title – Label
3. Install wireless broadband in every venue that has ever featured a set of 1210s
4. Do development over at the PRS so we can do that music reporting thing in number 1.
5. Set up separate Twitter accounts for every working DJ just for tracks (@martynplayed)
6. Develop some sort of open source Music Message Queue API type thing so we’re not bothering Twitter with this.
7. This could totally be gamed. Someone think of a way around that.
8. Go to Music Hackday because that’s where dreams like these may come true.
9. Give me the design skill to make pretty mockups of this stuff to illustrate my wordiest ever post.
10. Buy milk

Comments, builds and bitchslaps welcome on @martynrdavies or here.

This post was written to the beat of ‘Rave Side of the Moon’ by AGT Rave Cru, the Cursor Miner remix of ‘We Are Electric’ by Fischerspooner, ‘Bomb Scare’ by 2 Bad Mice and the Sub Focus Essential Mix from BBC Radio 1.

Organised by Rewired State’s James Darling and the man we call Mr. Soundcloud, Dave Haynes, this is the first time a hackday has focussed specifically on building stuff around music.

Tracks for the weekend include traditional hacking, on APIs from Last.fm, Gigulate, Songkick and more, alongside the data sets provided by the BBC’s /music and /programmes sites (or are they services now?). Alongside this, there’s also some hardware hacking happened courtesy of Tinker.it and the magical RjDJ team.

I’m really pleased that Patrick, Nick and Yves are going to be down there with me (I’ll be the confused looking one). They’re the engineering brains behind much of the BBC’s linked data strategy, so if you’re there and you want to know just how much you can flex our data, give them a shout.

There will also be other great BBC tech folks around from the BBC Future Media & Technology teams for Audio & Music and Vision as well as the great BBC Backstage community. I’ll try to make sure we’re all as visible as possible.

The full list of participating companies with data and other business to hack with is as follows:
Last.fm, Gigulate, BBC Music, Songkick, People’s Music Store, Echonest, 7Digital and Soundcloud

I’m expecting a really great weekend that will result in some amazing hacks and hopefully some ideas we can bring back and turn into real elements of the /music and /programmes services.

There’s still time to register, so go do the form over here: http://musichackday.org