Digital Needle - A Virtual Gramophone


Remember those flat round things you may have found lying around the house? Those that never really worked well as flying saucers. Well, the other day I happenned to have a good look at one through a magnifying glass. I was able to discern something waveform'esqe in the shape of the groove. I then recalled an old legend I once heard concerning these objects. This legend, being of technological nature, entaled a diamond needle that would vibrate when placed atop the revolving flat things. These vibrations, when amplified, would produce music. I thus decided to try and extract something sensible off of this object.


When scanned, some records seemed much denser than others. Dense tracks were much harder to follow.


Vivaldi's "The Four Seasons" was fine




When illuminated in parallel with the groove's direction some sort of depth-modulation became visible.



This made me guess that one of the stereo channels might be encoded in the depth/angle plane while the other was encoded in the radius/angle plane, the later being the one I initially noticed.




The nature of my scanner and of the records reflectiveness caused only the lower half of the record's radius/angle modulation to be exposed on each scan.



Thus, to supply the virtual gramophone at least one complete revolution of the record, four regions had to be separately scanned. These were later stitched together to form an image of several complete inner tracks of the record.




Once the image was ready, writing the decoder was very simple. All it did was rotate a "needle" around a given center at some predefined angular velocity, attempting to keep track of the groove the needle was initially positioned on. The offsets (dr) between this track and the basic radial were bunched into a sequence of samples. these were later converted into wav files.





Here are the few nearly-intelligible decodings that I managed to recover off of the inner tracks of Vivaldi's "The Four Seasons". Listening to the following reference recording might allow you to recognize these as the music-within-noise they actually are!

Reference recording -

gramophone3.mp3

Samples decoded from the scan -

a1.mp3
a2.mp3
a3.mp3
a4.mp3

b1.mp3
b2.mp3
b3.mp3
b4.mp3
b5.mp3

c1.mp3
c2.mp3

Comments appreciated at zbbrox@hotmail.com
Ofer Springer 5/9/2002.



7/9/2002 - Woke up, fell out of bed... rampaged by a Slashdot horde. Some clarifications are in order:

The whole thing was done in a couple of late nights so I didn't really have much time to gather all the technical details concerning phonograph modulations. Moreover, the reverse-engineering aspect was part of the fun. I now know (thanks to some great replies) that the horizontal modulation (the only one I did decode) is a combination of the left/right audio channels (which are each encoded on the sides of a v-shaped groove).

Some repliers seemed to be a bit confused as to what recordings were the actual decodings. I'd like to stress that gramophone3.mp3 is a recording while the rest were decoded from the image.