Bonum Certa Men Certa

Help Fight Patent Bullying From Shazam -- Spread This Code!

Gorilla



Summary: This post looks at patent bullying against Free software and it calls for the spreading of source code which Shazam unlawfully tries to remove from the Internet

EARLIER TODAY we wrote about NetApp's threats against ZFS distributors. As one blogger put it:

Enterprise Strategy Group senior analyst Terri McClure wonders why NetApp didn't hit Nexenta with the same letter since Nexenta supplies its ZFS software to multiple storage vendors.

"If NetApp did it would make sense – stop a number of vendors instead of just one. It certainly makes you wonder why they would single out Coraid, people could read into this that NetApp sees Coraid as a threat. Coraid's NAS product is pretty new but the underlying platform has been on the market a while and is solid, at a really aggressive price point," said McClure.

"[NetApp] just spent a couple of hundred dollars in lawyer's fees and took a competitor out of the market. Quick and easy, but a little disappointing, too. At the end of the day, ZFS is open source, and while there is no way to predict how the settlement talks between Oracle and NetApp will turn out, you can't really un-open source ZFS," she said.

There's still no word from NetApp on the matter.


The "patent troll, NTP, is back, buoyed dosh from RIM," says Glyn Moody, who found this new article.

NTP, a patent-holding company best known for prying a settlement of more than $600 million from the maker of the BlackBerry, is now suing the other big names in the smartphone industry: Apple, Google, Microsoft, HTC, LG and Motorola, writes The New York Times’s Steve Lohr.

The suits, filed late Thursday afternoon in federal district court in Richmond, Va., charge that the cellphone e-mail systems of those companies are illegally using NTP’s patented technology.


We mentioned NTP before and so did Patent Troll Tracker. Speaking of trolls, earlier today we wrote about Shazam's patent bullying. That previous post gave just the gist of it and the discussion at Slashdot ought to say more. From the summary:

"The code wasn't even released, and yet Roy van Rijn, a Music & Free Software enthusiast received a C&D from Landmark Digital Services, owners of Shazam, a music service that allows you to find a song, by listening to a part of it. And if that wasn't enough, they want him to take down his blog post (Google Cache) explaining how he did it because it 'may be viewed internationally. As a result, [it] may contribute to someone infringing our patents in any part of the world.'"


Jan Wildeboer calls it "Patent Infringement Madness" and another post Wildeboer says "is (a) a blog entry or (b) patent infringement? I say (a) Shazam says (b)"

Two readers urged us to make a mirror just in case (other people ought to mirror this too, in order to ensure that Shazam will lose hope of successfully censoring perfectly legal Dutch code).

Patents are supposed to encourage publication of ideas, not to suppress them. The following code is not in any way infringing Shazam copyrights.






Creating Shazam in Java





A couple of days ago I encountered this article: How Shazam Works



This got me interested in how a program like Shazam works… And more importantly, how hard is it to program something similar in Java?

About Shazam

Shazam is an application which you can use to analyse/match music. When you install it on your phone, and hold the microphone to some music for about 20 to 30 seconds, it will tell you which song it is.

When I first used it it gave me a magical feeling. “How did it do that!?”. And even today, after using it a lot, it still has a bit of magical feel to it.
Wouldn’t it be great if we can program something of our own that gives that same feeling? That was my goal for the past weekend.

Listen up..!

First things first, get the music sample to analyse we first need to listen to the microphone in our Java application…! This is something I hadn’t done yet in Java, so I had no idea how hard this was going to be.

But it turned out it was very easy:



1 final AudioFormat format = getFormat(); //Fill AudioFormat with the wanted settings


2 DataLine.Info info = new DataLine.Info(TargetDataLine.class, format);
3 final TargetDataLine line = (TargetDataLine) AudioSystem.getLine(info);


4 line.open(format);
5 line.start();

Now we can read the data from the TargetDataLine just like a normal InputStream:



01 // In another thread I start:


02   


03 OutputStream out = new ByteArrayOutputStream();




04 running = true;


05   




06 try {


07     while (running) {




08         int count = line.read(buffer, 0, buffer.length);


09         if (count > 0) {


10             out.write(buffer, 0, count);


11         }


12     }


13     out.close();


14 } catch (IOException e) {


15     System.err.println("I/O problems: " + e);


16     System.exit(-1);


17 }


Using this method it is easy to open the microphone and record all the sounds! The AudioFormat I’m currently using is:



1 private AudioFormat getFormat() {




2     float sampleRate = 44100;


3     int sampleSizeInBits = 8;


4     int channels = 1; //mono




5     boolean signed = true;


6     boolean bigEndian = true;


7     return new AudioFormat(sampleRate, sampleSizeInBits, channels, signed, bigEndian);




8 }

So, now we have the recorded data in a ByteArrayOutputStream, great! Step 1 complete.

Microphone data

The next challenge is analyzing the data, when I outputted the data I received in my byte array I got a long list of numbers, like this:



01 0


02 0


03 1


04 2


05 4




06 7


07 6




08 3


09 -1




10 -2


11 -4




12 -2


13 -5


14 -7


15 -8


16 (etc)

Erhm… yes? This is sound?



To see if the data could be visualized I took the output and placed it in Open Office to generate a line graph:

Ah yes! This kind of looks like ’sound’. It looks like what you see when using for example Windows Sound Recorder.

This data is actually known as time domain. But these numbers are currently basically useless to us… if you read the above article on how Shazam works you’ll read that they use a spectrum analysis instead of direct time domain data.
So the next big question is: How do we transform the current data into a spectrum analysis?

Discrete Fourier transform

To turn our data into usable data we need to apply the so called Discrete Fourier Transformation. This turns the data from time domain into frequency domain.
There is just one problem, if you transform the data into the frequency domain you loose every bit of information regarding time. So you’ll know what the magnitude of all the frequencies are, but you have no idea when they appear.



To solve this we need a sliding window. We take chunks of data (in my case 4096 bytes of data) and transform just this bit of information. Then we know the magnitude of all frequencies that occur during just these 4096 bytes.

Implementing this

Instead of worrying about the Fourier Transformation I googled a bit and found code for the so called FFT (Fast Fourier Transformation). I’m calling this code with the chunks:



01 byte audio[] = out.toByteArray();


02   


03 final int totalSize = audio.length;


04   


05 int amountPossible = totalSize/Harvester.CHUNK_SIZE;


06   


07 //When turning into frequency domain we'll need complex numbers:




08 Complex[][] results = new Complex[amountPossible][];


09   


10 //For all the chunks:


11 for(int times = 0;times < amountPossible; times++) {


12     Complex[] complex = new Complex[Harvester.CHUNK_SIZE];




13     for(int i = 0;i < Harvester.CHUNK_SIZE;i++) {


14         //Put the time domain data into a complex number with imaginary part as 0:


15         complex[i] = new Complex(audio[(times*Harvester.CHUNK_SIZE)+i], 0);




16     }


17     //Perform FFT analysis on the chunk:




18     results[times] = FFT.fft(complex);


19 }


20   


21 //Done!

Now we have a double array containing all chunks as Complex[]. This array contains data about all frequencies. To visualize this data I decided to implement a full spectrum analyzer (just to make sure I got the math right).
To show the data I hacked this together:





01 for(int i = 0; i < results.length; i++) {




02     int freq = 1;


03     for(int line = 1; line < size; line++) {


04         // To get the magnitude of the sound at a given frequency slice


05         // get the abs() from the complex number.


06         // In this case I use Math.log to get a more managable number (used for color)


07         double magnitude = Math.log(results[i][freq].abs()+1);


08   


09         // The more blue in the color the more intensity for a given frequency point:




10         g2d.setColor(new Color(0,(int)magnitude*10,(int)magnitude*20));


11         // Fill:


12         g2d.fillRect(i*blockSizeX, (size-line)*blockSizeY,blockSizeX,blockSizeY);


13   


14         // I used a improviced logarithmic scale and normal scale:




15         if (logModeEnabled && (Math.log10(line) * Math.log10(line)) > 1) {




16             freq += (int) (Math.log10(line) * Math.log10(line));


17         } else {




18             freq++;


19         }




20     }


21 }

Introducing, Aphex Twin

This seems a bit of OT (off-topic), but I’d like to tell you about a electronic musician called Aphex Twin (Richard David James). He makes crazy electronic music… but some songs have an interesting feature. His biggest hit for example, Windowlicker has a spectrogram image in it.
If you look at the song as spectral image it shows a nice spiral. Another song, called ‘Mathematical Equation’ shows the face of Twin! More information can be found here: Bastwood – Aphex Twin’s face.



When running this song against my spectral analyzer I get the following result:

Not perfect, but it seems to be Twin’s face!

Determining the key music points

The next step in Shazam’s algorithm is to determine some key points in the song, save those points as a hash and then try to match on them against their database of over 8 million songs. This is done for speed, the lookup of a hash is O(1) speed. That explains a lot of the awesome performance of Shazam!

Because I wanted to have everything working in one weekend (this is my maximum attention span sadly enough, then I need a new project to work on) I kept my algorithm as simple as possible. And to my surprise it worked.

For each line the in spectrum analysis I take the points with the highest magnitude from certain ranges. In my case: 40-80, 80-120, 120-180, 180-300.





01 //For every line of data:


02   




03 for (int freq = LOWER_LIMIT; freq < UPPER_LIMIT-1; freq++) {




04     //Get the magnitude:


05     double mag = Math.log(results[freq].abs() + 1);


06   


07     //Find out which range we are in:




08     int index = getIndex(freq);


09   




10     //Save the highest magnitude and corresponding frequency:


11     if (mag > highscores[index]) {


12         highscores[index] = mag;


13         recordPoints[index] = freq;


14     }


15 }




16   


17 //Write the points to a file:




18 for (int i = 0; i < AMOUNT_OF_POINTS; i++) {




19     fw.append(recordPoints[i] + "\t");


20 }




21 fw.append("\n");


22   




23 // ... snip ...


24   




25 public static final int[] RANGE = new int[] {40,80,120,180, UPPER_LIMIT+1};


26   


27 //Find out in which range




28 public static int getIndex(int freq) {




29     int i = 0;


30     while(RANGE[i] < freq) i++;


31         return i;


32     }


33 }

When we record a song now, we get a list of numbers such as:





01 33  56  99  121 195


02 30  41  84  146 199


03 33  51  99  133 183




04 33  47  94  137 193




05 32  41  106 161 191


06 33  76  95  123 185


07 40  68  110 134 232




08 30  62  88  125 194


09 34  57  83  121 182


10 34  42  89  123 182




11 33  56  99  121 195




12 30  41  84  146 199


13 33  51  99  133 183


14 33  47  94  137 193




15 32  41  106 161 191


16 33  76  95  123 185


If I record a song and look at it visually it looks like this:


(all the red dots are ‘important points’)

Indexing my own music

With this algorithm in place I decided to index all my 3000 songs. Instead of using the microphone you can just open mp3 files, convert them to the correct format, and read them the same way we did with the microphone, using an AudioInputStream. Converting stereo music into mono-channel audio was a bit trickier then I hoped. Examples can be found online (requires a bit too much code to paste here) have to change the sampling a bit.

Matching!

The most important part of the program is the matching process. Reading Shazams paper they use hashing to get matches and the decide which song was the best match.

Instead of using difficult point-groupings in time I decided to use a line of our data (for example “33, 47, 94, 137″) as one hash: 1370944733
(in my tests using 3 or 4 points works best, but tweaking is difficult, I need to re-index my mp3 every time!)

Example hash-code using 4 points per line:



01 //Using a little bit of error-correction, damping


02 private static final int FUZ_FACTOR = 2;


03   


04 private long hash(String line) {


05     String[] p = line.split("\t");


06     long p1 = Long.parseLong(p[0]);


07     long p2 = Long.parseLong(p[1]);




08     long p3 = Long.parseLong(p[2]);


09     long p4 = Long.parseLong(p[3]);


10     return  (p4-(p4%FUZ_FACTOR)) * 100000000 + (p3-(p3%FUZ_FACTOR)) * 100000 + (p2-(p2%FUZ_FACTOR)) * 100 + (p1-(p1%FUZ_FACTOR));


11 }

Now I create two data sets:

- A list of songs, List<String> (List index is Song-ID, String is songname)
- Database of hashes: Map<Long, List<DataPoint>>



The long in the database of hashes represents the hash itself, and it has a bucket of DataPoints.

A DataPoint looks like:



01 private class DataPoint {


02   


03     private int time;


04     private int songId;




05   


06     public DataPoint(int songId, int time) {


07         this.songId = songId;


08         this.time = time;


09     }


10   




11     public int getTime() {


12         return time;


13     }


14     public int getSongId() {


15         return songId;


16     }


17 }

Now we already have everything in place to do a lookup. First I read all the songs and generate hashes for each point of data. This is put into the hash-database.
The second step is reading the data of the song we need to match. These hashes are retrieved and we look at the matching datapoints.

There is just one problem, for each hash there are some hits, but how do we determine which song is the correct song..? Looking at the amount of matches? No, this doesn’t work…
The most important thing is timing. We must overlap the timing…! But how can we do this if we don’t know where we are in the song? After all, we could just as easily have recorded the final chords of the song.

By looking at the data I discovered something interesting, because we have the following data:

- A hash of the recording
- A matching hash of the possible match
- A song ID of the possible match
- The current time in our own recording
- The time of the hash in the possible match



Now we can substract the current time in our recording (for example, line 34) with the time of the hash-match (for example, line 1352). This difference is stored together with the song ID. Because this offset, this difference, tells us where we possibly could be in the song.
When we have gone through all the hashes from our recording we are left with a lot of song id’s and offsets. The cool thing is, if you have a lot of hashes with matching offsets, you’ve found your song.

The results

For example, when listening to The Kooks – Match Box for just 20 seconds, this is the output of my program:



01 Done loading: 2921 songs


02   


03 Start matching song...




04   


05 Top 20 matches:




06   


07 01: 08_the_kooks_-_match_box.mp3 with 16 matches.




08 02: 04 Racoon - Smoothly.mp3 with 8 matches.


09 03: 05 Röyksopp - Poor Leno.mp3 with 7 matches.


10 04: 07_athlete_-_yesterday_threw_everyting_a_me.mp3 with 7 matches.


11 05: Flogging Molly - WMH - Dont Let Me Dia Still Wonderin.mp3 with 7 matches.


12 06: coldplay - 04 - sparks.mp3 with 7 matches.


13 07: Coldplay - Help Is Round The Corner (yellow b-side).mp3 with 7 matches.


14 08: the arcade fire - 09 - rebellion (lies).mp3 with 7 matches.




15 09: 01-coldplay-_clocks.mp3 with 6 matches.


16 10: 02 Scared Tonight.mp3 with 6 matches.




17 11: 02-radiohead-pyramid_song-ksi.mp3 with 6 matches.


18 12: 03 Shadows Fall.mp3 with 6 matches.




19 13: 04 Röyksopp - In Space.mp3 with 6 matches.


20 14: 04 Track04.mp3 with 6 matches.




21 15: 05 - Dress Up In You.mp3 with 6 matches.


22 16: 05 Supergrass - Can't Get Up.mp3 with 6 matches.


23 17: 05 Track05.mp3 with 6 matches.


24 18: 05The Fox In The Snow.mp3 with 6 matches.


25 19: 05_athlete_-_wires.mp3 with 6 matches.


26 20: 06 Racoon - Feel Like Flying.mp3 with 6 matches.


27   




28 Matching took: 259 ms


29   




30 Final prediction: 08_the_kooks_-_match_box.mp3.song with 16 matches.


It works!!

Listening for 20 seconds it can match almost all the songs I have. And even this live recording of the Editors could be matched to the correct song after listening 40 seconds!

Again it feels like magic! :-)

Currently, the code isn’t in a releasable state and it doesn’t work perfectly. It has been a pure weekend-hack, more like a proof-of-concept / algorithm exploration.

Maybe, if enough people ask about it, I’ll clean it up and release it somewhere. Or turn it into a huge online empire like Shazam… who knows!





Comments

Recent Techrights' Posts

The Sheer Absurdity of the EPO's Career System Explained by EPO Staff
"Staff representation has previously pointed this out to management, and the career system has been the reason for several industrial actions and litigation cases initiated by SUEPO."
Forget About India's and Pakistan's Nuclear Weapons and Armament Race, They Need to Abscond Windows and Microsoft (Security Swiss Cheese)
Both countries would be wise to remove Windows as soon as possible, irrespective of the local party politics
statCounter: GNU/Linux Rose From 0.2% to Over 3% in Pakistan
GNU/Linux "proper" (i.e. not ChromeOS) has the lion's share
The "D" in Debian Stands for Dictatorship That Extends to Censorship at DNS Level
Of course the registrar, which charged for domains until 2025, just went along with it
 
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Saturday, May 25, 2024
IRC logs for Saturday, May 25, 2024
GNU/Linux in Greenland
The sharp increases for GNU/Linux started last summer
[Meme] Productivity Champ Nellie Simon: It Takes Me 3+ Weeks to Write 6 Paragraphs
Congrats to Nellie Simon!
It Took EPO Management 3+ Weeks to Respond to a Letter About an Urgent Problem (Defunding of EPO Staff)
The funny thing about it is that Nellie Simon expects examiners to work day and night (which is illegal) while she herself takes 3+ weeks to write a 1-page letter
Staff Union of the EPO (SUEPO) in The Hague Taking Action to Rectify Cuts to Families of Workers
they "are active in challenging this measure via the legal system"
Links 25/05/2024: Microsoft Adds More DRM (Screenshot Blocking), Another Microsoft Outage Takes Down Everything
Links for the day
Gemini Links 25/05/2024: "Bill Smugs" and OpenBSD Mirror Over Tor / I2P
Links for the day
Microsoft #1 in Gaming Layoffs, Laid Off Workers Receive Another Insult From Microsoft
Many of them never chose to work for Microsoft
In New Caledonia Windows is Now Below 30% (It Used to be Over 90%)
Microsoft's Windows absolutely collapsing and the measures are relatively stable
Red tape: farmer concerns eerily similar to Debian suicide cluster deaths
Reprinted with permission from Daniel Pocock
Galway street artists support social media concerns
Reprinted with permission from Daniel Pocock
Links 25/05/2024: Section 230 and Right of Publicity Violations by Microsoft (Which Attacks Performance Artists)
Links for the day
[Meme] No Microsoft
For fun!
Microsoft Windows Falls to New Lows in Poland
It may mean people delete Windows from relatively new PC
A 3-Year Campaign to Coerce/Intimidate Us Into Censorship: An Introduction
The campaign of coercion (or worse) started in 2021
Cybersecurity and Infrastructure Security Agency (CISA) Getting Stacked by Microsoft
it lets Microsoft write policies
The Parasitic Nature of Microsoft Contracts
Stop feeding the beast
Gemini Links 25/05/2024: Emacs Windows 2000 Screenshots and Little Languages
Links for the day
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Friday, May 24, 2024
IRC logs for Friday, May 24, 2024
Ireland Exits Microsoft's Vista 11
Microsoft can't be doing too well in Ireland because Microsoft had tons of layoffs in that country last year
A Recognition for Hard Work
Running this site is a lot of work
The Web We Lost...
Vintage War Censorship Poster...
Daniel Pocock (IND) in European Election Debate
In this segment he speaks of the effects of social control media and phones on children
[Meme] Next Target: Sub Domains
Deb.Ian.Community
In Republic of (South) Korea, as of This Month, Android Climbs to Record High of 48%
Judging by statCounter anyway
"Linux" is Second-Class Citizen at IBM
sends the wrong message to Red Hat staff and Red Hat clients
Links 24/05/2024: More Software Patents Invalidated (US), New Fights to Protect Free Speech
Links for the day
"You Touched the Wrong Lady"
What Rianne wrote more than 8 months ago
Links 24/05/2024: Layoffs at LinkedIn and Election Interference Via Social Control Media
Links for the day
Getting a 'Thank You' From Software Freedom Conservancy (SFC) Will Cost You $5,000 to $30,000 (Same as Last Year)
Right now one of their associates (SFC) tries to spend money to censor us
KDE Neon Weirdness
Reprinted with permission from Ryan Farmer
Congratulations to Sirius Open Source, Still Claiming to Employ People Who Left Half a Decade Ago (or More!)
What signal does that send to con men?
[Meme] Bluewashing
Cent OS? No more.
IRC Proceedings: Thursday, May 23, 2024
IRC logs for Thursday, May 23, 2024
Over at Tux Machines...
GNU/Linux news for the past day
Tenfold Increase for ChromeOS+GNU/Linux in Brunei
Brunei Darussalam is a country most people don't know about and never even heard about
Coming Soon: Another Round of 'Cancel Stallman' Chorus
The series required a great deal of patience