Tag Archives: technology

Digital Music Landscape I : Recommenders

[This is the first part of a three-part post that provides a high level overview of the digital music landscape where Mavrix and MySwar fits in.]

Recommender: specific type of information filtering system technique that attempts to recommend information items (movies, music, books, news, images, web pages, etc.) or social elements (e.g. people, events or groups) that are likely to be of interest to the user. – Wikipedia

Recommendation engines work as blend of many algorithms and approaches, to find similarities between what you find interesting , and what you may potentially find interesting. Often people use a Collaborative filtering model, or ‘wisdom of the crowd’  approach to generate lists of  music, movies, news and other items you wouldn’t have come across in the mess of information around.

Recommendation services have evolved over the decades as I’ve tried to outline below

  •  The idea of collaborative filtering was derived, when developing an automatic filtering system for electronic mail called Tapestry, over at  Xerox Palo Alto Research in 1992. They needed to handle the large amounts of email and messages posted to newsgroups. Users were encouraged to annotate documents , and these annotations could be used for further filtering.
  • Grouplens began as a research group in the University of Minnesota where the students made a system to recommend Usenet News. It collected ratings from Usenet readers and used those ratings to predict how much other readers would like an article before they read it. This recommendation engine was one of the first automated collaborative filtering systems in which algorithms were used to automatically form predictions based on historical patterns of ratings. The research project would eventually spin out the Movielens project in 1997 and be featured in a Malcolm Gladwell column.
  • Engineers from the MIT Media labs created a email-based collaborative music recommendation system called RINGO. The community around this project eventually became known as the Helpful Online Music Recommendation Service (HORM). In 1999, it eventually spun out into a company called Firefly which was acquired by Microsoft where it was killed suddenly.
Today technology has advanced into a stage where recommender systems have become ubiquitous.
  • Amazon is well-known for its item to item recommendation system. All recommendations are based on individual behavior. Whether you like to buy something because it is related to something that you purchased before, or because it is popular with other users, you have a list of social recommendations – what other users bought, or personal recommendations-based on your purchase history.
  • Netflix encourages subscribers to rate the movies they’ve viewed, and their CineMatch program recommends titles similar to those well liked — regardless of a film’s popularity at the box office.
  • Google news serves a personalized news feed by assimilating the user’s genuine news interests as validated by click history and influences of local news trends, together with a collaborative filtering method. The result is that you view articles that align to your interests.

Music discovery is the new keyword on the digital block. To put it simply, an event of listening to a song by accident, having it play in your head, get you to like it and have you realize you want to hear it again is simplified to a website/app doing all that work for you. The music recommendation world today is vastly different from the Ringo email system where you rated some songs on an absolute scale and emailed it to the system, which would reply with songs/albums it thought you would like.

Let’s look at some awesome platforms that are driving this new experience in the second part of this post.

Thank You Steve Jobs

For opening up a whole new world of music to me.

For giving me wonderful experiences through your magical products.

For being a source of inspiration and teaching me important lessons. Through your Stanford commencement speech. Through your Thoughts on Music. Through your many keynotes. Most importantly, through the act of building a great company that built great products.

What I Learnt From Steve Jobs’ “Thoughts On Music”

Steve Jobs wrote a piece called “Thoughts on Music” a few years ago. If you haven’t read it, please do.

It’s amazing how dated this post reads, even though it was written just 4 years ago. Think about it – Just 4 years ago Steve Jobs proposed that DRM be abolished and DRM is already a distant memory (except in countries that are not big Apple markets, like India).

The article may be dated but the lessons I learnt from it are not:

  1. Think big. I have read this article many times and it is only during one of my revisits that I realized that there was a big disconnect between the title and the content. The title of the article should really have been “Thoughts on DRM”. I don’t think the  “Thoughts on Music” title was the result of PR spin. I believe that the title is really a reflection of how Jobs’ thinks. To him, it was not abolishing DRM. To him, it was about making it easier for people to listen to music – the end, as opposed to the means.
  2. To communicate clearly, think clearly. I totally “got” this article the first time I read it, and I thought “This guy can write!”. Over time I realized that the only way to write so clearly about such a complex topic (for most of us) is to think through it and know exactly what we want to say. Steve Jobs is a brilliant communicator because he is a brilliant thinker.
  3. Doing it is more important than talking about it. iTunes was completely DRM free 2 years after he wrote this article. Soon after he wrote this piece, we saw the wheels in motion as one label after the other signed up for DRM-free music on iTunes. The adulation for Steve Jobs is not because of his keynote speeches but because of the accomplishments they represent.

Steve Jobs – through an ecosystem of iPods and iTunes – opened up a whole new world of music to me and millions of other people. He also taught me important lessons on thinking, communicating and doing.

Discovery vs Search

“Discovery vs Search” may be a much-discussed topic in tech circles but it’s still not understood very well by people who use these services. I’ve tried to explain the objective of music discovery in our About page, and I still get questions about it.

So when I saw this question posted on Twitter, I thought I would give a shot at a non-geeky explanation for “Discovery vs Search”. Here it goes:

  • You use Search when you are looking for a specific item. You use Discovery, when you are looking for ideas or recommendations (for books, songs, movies, restaurants, etc.). You may want to Search for a particular song you’ve heard on the radio but you would want to Discover new songs you have never heard of.
  • Search assumes you’re in a hurry – it wants you to find the thing you’re looking for and move on. (How many times have you gone beyond the second page of Google search results?) Discovery assumes you have some leisure time. It encourages you to explore and find more and more things you may like (the omnipresent “Customers Who Bought This Item Also Bought” section on Amazon).
  • Search wants you to be satisfied. Discovery aims at higher emotions – Awe (How did they know I would like this item?), Excitement (This is an amazing find!) and Trust (If they tell me I might like it, I probably will.)
  • For the most part, Search doesn’t care about your preferences. Discovery is driven by you what you like (or don’t). In that sense, Search is transactional and Discovery is relationship-oriented (Thej’s description!).

Search and Discovery also have something in common – both depend on you to be successful. Search depends on you to provide a good description of what you are looking for. Discovery depends on you to provide clues of what you might like.

“Search vs Discovery” is not a debate about what’s better, it’s about understanding how the two are different and how they can complement each other.

Musicians Are A Lot Like Technology Startups

  1. Life experiences (especially early ones) inspire the musician’s music and the entrepreneur’s business. John Lennon used song-writing as an escape from a troubled childhood. Richard Branson started a school newspaper when he got frustrated by rigid school rules and regulations.
  2. Success for both musicians and startups come after a lot of hard work and learning from experience. The Beatles had played together more than 1200 times before they got noticed in 1964. Bill Gates had done 10,000 hours of programming by the age of 13.
  3. Performing cover songs help bands launch careers but they need to deliver originals to sustain their success. Red Hot Chilli Peppers may have gotten attention through their cover of Stevie Wonder’s “Higher Ground” but it is their unique style of funk infused rock that gained them a large following. Similarly, startups can be inspired by existing companies but need to do something very different to deliver more value and be successful. Facebook may have started off as a Friendster clone but scaled better and innovated to become the premier online social network even as Friendster closed shop.
  4. The musician’s first song and the startup’s first product release is always sketchy. You can’t be perfect the first time.
  5. Both music and technology products are a result of teamwork. The band doesn’t take off till the right set of musicians come together. Indian Ocean started off with two people and saw several shake-ups till it got to the line-up that delivered success. Apple‘s early success was a result of teamwork – Steve Jobs’ marketing skills and Steve Wozniak’s engineering prowess.
  6. Most musicians start with free gigs and startups with free products/services.
  7. Musicians and startups make it big by persisting and continuing to do what they love and believe in. Susan Boyle made it big with Britain’s Got Talent, at age 48. Tim Westergren went through years of struggle before Pandora become the much-loved music service that it is today.

iCloud in India

Our website got a bunch of Google hits with various permutations of the “iCloud India” keywords. I guess that means that a lot of us are wondering if iCloud will come to India. Will it? I think yes, and here’s why.

  1. Apple says so! Apple’s US website is very well coordinated with product releases. The announcements at WWDC (iCloud, iOS5, OS X Lion) were up on their US website within half an hour of the keynote’s end. Apple’s India website tends to play catch up and does not always give visitors a clear picture of Apple’s India offerings. For example, Apple India’s iTunes page gives no clue that you can’t really buy iTunes music in India. This is revealed to you via a tiny, indirect disclaimer on the download page. By contrast, their new iCloud page is extremely relevant to India:
    1. They made an effort to use “Coming this autumn” for India (not “Coming this fall”).
    2. There is an explicit disclaimer on the iCloud page that the awkwardly named “iTunes on the Cloud” service starts in the US and that music cannot be purchased in India.

    To me, this explicit, well-coordinated update to Apple India’s iCloud page means that India is definitely getting all iCloud features except some iTunes on the Cloud features.
  1. Part of iCloud is replacement for MobileMe, Apple’s erstwhile paid service for syncing email, contacts and calendar. Apple has been selling MobileMe in India for some time now (though with a weird Singapore Dollar pricing). Apple may delay launching new products in India, but it’s unlikely it will leave existing customers without an alternative.

To summarize – It is very likely that India will get the iCloud extension of all Apple services/features it already has. Since iTunes music purchases are not available in India, it is obvious that the music related features of iTunes on the Cloud (syncing songs and iTunes Match) will also not be available in India.

Why Won’t Apple’s iCloud Stream Music?

Since Apple’s big iCloud announcement a couple of days ago, the internet has been abuzz with people expressing disappointment at the iCloud not going far enough when it came to music (like here, and here). What did Apple miss? Streaming.

Broadly, there seem to be two camps of people when it comes to music listening preference – one that prefers listening to streaming music (via services like Rdio, Spotify, Pandora, Raaga, Saavn) and the other that prefers owned music played locally on their computers, mobile phones, media players or other devices. But even those who prefer owning music (like I do), probably use streaming as an option to discover and sample before they buy.

So, regardless of your listening preferences, you may feel that iCloud missed the bus by not allowing streaming of music via subscription. Amazon Cloud and Google Music Beta also missed that same bus, by the way. They allow streaming but only of the music you already own. In fact, Amazon actually charges you for it. (Duh?!) Google has conveniently not revealed its pricing.

Coming back to iCloud. Streaming of purchased music may not make sense, but why not give customers the option to stream music they don’t own as well as buy songs if they choose to?

My guess is that it could be due to the following reasons:

  1. Profitability – Apple has a predictably profitable model of selling music through iTunes. On the other hand, streaming services have struggled for profitability, whether it’s the ad-supported Pandora, or subscription-based Spotify. And while streaming services are becoming increasingly popular, they still do not represent a sizeable enough market for Apple to be interested (not yet at least).
  2. Risk – iCloud is going to put considerable pressure on Apple’s resources (including their cloud infrastructure) and they know it. Why do you think Steve Jobs showed off the pictures of their huge server farm? Apple wants to come back strongly after MobileMe’s failure (Steve Jobs described it as “Not our finest hour” in the keynote at WWDC), and is not willing to take on the risk associated with the burden of streaming music.
  3. Concerns around bandwidth usage – A few days ago, I wrote a post about why 3G economics don’t work for streaming music in India. According to Paul Lamere, a music+tech guru and a passionate supporter of music subscription services, people in the US can get an unlimited 3G data plan “for the cost of a good meal”. That may be the case today but even in the US, people are becoming more aware of their rapidly increasing mobile bandwidth usage and carriers are shutting down unlimited data plans. Perhaps Apple believes that the consumers’ and the carriers’ increased sensitivity to bandwidth usage may adversely impact the streaming music market.


Apple’s iCloud Is Transformational

What is iCloud?

iCloud is Apple’s service that allows consumers to shift their storage hub from local hard-drives to the internet. iCloud takes care of syncing a variety of content and information (contacts, mails, calendars, music, photos, videos, documents, etc.) across multiple devices. The following apps are available on iCloud:

  1. Contacts, Calendar, Mail are available on iCloud for free. Apple is shutting down MobileMe, the product that used to do the same at $99 per year.
  2. Apps Store
  3. iBooks
  4. Device backup – Backs up important settings and loads on new device
  5. Documents in the cloud – Supports availability of Pages/Numbers/Keynote across devices.
  6. iCloud storage APIs – For developers to build iCloud apps
  7. Photo Stream – Stores the last 1000 photos on the cloud. Allows access across devices. People with more than 1000 photos can move older ones from Photo Stream to their device.
  8. iTunes
    1. Anything bought on iTunes can be re-downloaded on 10 devices
    2. Automated download to all devices. Download starts when you plug in your iPhone for charging.
    3. For non-iTunes music, consumers can
      1. Sync devices
      2. Buy the songs from iTunes, so it’s available on iCloud
      3. Buy the iTunes Match service at about $25/year. The service lets you match your non-iTunes tracks to iTunes’ 18 million song catalog. Matched songs have the same support as iTunes songs – 256K downloads (even if the original track is less than 256K) available on 10 devices. Unmatched songs will be available on the cloud as is. Apple claims that the matching takes minutes (as opposed to the “weeks” it takes to upload music to Google Music Beta or Amazon Cloud).

How much will iCloud cost?

Some details are not known but for the most part, iCloud is free. Storage for purchased music, apps, books and the 1000 Photo Stream photos are free of cost. 5GB of storage is available for free for mails, documents and backup. Apple indicates that 5GB is more than enough and does not even address the possibility of the need for more than 5GB. The only component of iCloud that costs money is iTunes Match at $24.99.

When will iCloud be available?

iTunes on iCloud is already available. The remaining components will be available this fall.

What about streaming?

No streaming. (I will be writing another post with my theories on why Apple did not roll out streaming).

Why is iCloud transformational?

  1. Comprehensive – It is the first and only service that manages such a wide gamut of “stuff” on the cloud.
  2. Invisible – It is so well integrated with the Apple ecosystem, that consumers may not even notice it. Stuff just becomes available across devices.
  3. Big impact – Except for iTunes Match, it is free. At that price point, adoption of iCloud by anyone with an Apple device is a no brainer. With millions of Apple customers using iCloud, the standard for managing multiple devices has changed permanently. For the better.


Mavrix Monthly Update May-2011

  • We finished cataloguing information for 2 decades of Bollywood songs, from current day to 1991. Good news – We were making good progress! Bad news – We have to started listening to the songs from 1980s – probably the worst ever to come out of Bollywood. Please excuse our team members if they’re posting weird videos of like this on FB – I think they find it cathartic. Good songs are few and far-between and hog the office music system, when we do find them.
  • Online stream vs CDs. Since we analyze songs at a fairly detailed level, we find it difficult to do a good job of analyzing songs when the stream quality is not good (“Was that a dholak or a tabla?”) – which is a lot of the time. We therefore try and buy CDs if they’re available. We didn’t have any trouble sourcing CDs for the 2000s (we used Flipkart for the most part) but beyond that, sourcing CDs has been a pain. Recently though we were able to get in touch with some helpful folks at Saregama who put us in touch with a distributor who is now supplying us with a large number of CDs. We are not getting the entire set but it’s a lot more than the small numbers available on Flipkart. The conversation with Saregama was interesting – they made sure we were not pirates (or just plain dumb) before they gave us the distributor’s info.
  • Cover Art. We’d like to be able to display album cover arts on our website but we still haven’t been able to figure out a way to secure rights to do that. If you have any pointers, please send us an email at admin AT mavrix DOT in.
  • Tech work . After a few iterations, we are getting to the UI we want. Finalizing the designs next week. Thej is working fast and furious on finalizing the architecture and technology decisions. (Check out his post on Github vs Bitbucket). We’re shooting to start development work on Jun 20. Our quest to find a smart and dynamic Python/PHP/MySQL developer is still on.

3G Cost In India Will Burst Music Cloud Bubble

Even though Google and Amazon have not yet released their cloud music service in India, the excitement around it is palpable in cities where 3G has been rolled out. One of our team members, Rakshith, can’t wait for Google Music to come to India because he wants access to music anywhere, anytime. To quote him “Have a 3G connection and u have all ur songs in one click. And its FREE if im not wrong.” He was referring to the free 20000 song upload for the Google Music Beta. I asked, “What about 3G costs?”. We had a good debate and that was that.

The matter resurfaced over the weekend though, when a visitor on our blog sent me an email about streaming music. He seemed very keen about it because “anyway i pay for 3G”.

For some reason, people seem to be putting blinders on when it comes to 3G costs. Sorry to burst the bubble, my fellow music buffs but here is something you should know about the prevalent 3G plans:

  • The more you use, the more you pay. There is no unlimited plan. You cannot justify your music streaming with “anyway I am paying” rationale.
  • 3G plans are not cheap.

Based on your listening habits and assuming music streamed at 128kbps, here’s what your 3G cost would be:

Hours of streaming per month510152025
Bandwidth used (MB)281.25562.5843.7511251406.25
Approx annual cost based on current tarrifs (Rs)240054006000810010125

If your argument then is that you’re willing to pay this price for the convenience of having access to your music everywhere, I would say you can pay a lower price for a portable media player like an iPod, which you can truly access everywhere unlike over 3G which is only available in some cities and is unpredictable even in cities it’s rolled out (I’ve been a 3G user for almost a year). You actually save money if you buy a media player:

If you listen to..…you could buy…at Rs…and save over 5 years (approx life of device) Rs
5 hrs of music per monthiPod shuffle32006628
10 hrs of music per monthiPod nano 8GB1070011413
15 hrs of music per monthiPod nano 16GB1270011870
20 hrs of music per monthiPod Classic1520017970
25 hrs of music per monthiPod Classic1520026263

The savings calculation factors in a 10% reduction in 3G tarrif every year over 5 years. Of course, iPod is a premium product and the savings would be much higher if you bought a cheaper media player.

The only compelling argument I can think of for streaming music on the cloud is that your music collection is so big that it does not fit in any portable media player (more than 160GB if you consider iPod).

So, before you jump on to the cloud music bandwagon, you might want to figure out what your 3G cost works out to based on your listening habits and compare it against the cost of a suitable media player.