Apologies for spending over a week without posting anything . This week was extremely eventful , starting off with my parents and brothers jetting off to London for medical reasons , leaving me on my own in Malta.
From the academic aspect , it was extremely interesting featuring also an article written about the Brain Music System published on the Times of Malta ( article available from : http://www.timesofmalta.com/technology/view/20100415/news/playing-music-by-a-simple-thought )
Research has also done well this week , with Niki Young ( a good friend of mine , and also one of the tested people which took part in the exerimantal stages of testing the system ) .He managed to master the EEG waves and uploading the results on his you tube page ( Below kindly see his EEG to music representation to the Brain Music system ) All this shows that the main aims of the system are being reached , and I am looking forward to develop this further
Of course I need to thank everyone that help in the development of my research further , and I look forward for more people to give the well appreciated help .
Hi Adrian,
What an interesting blog you have here! Really
Regarding your project, I’ve seen a video on YouTube where you play music by brainwaves – it looks ubelievable that you’re doing that just with your EEG alone! I can definitely say conventional brain-computer interface methods would not do that… How long does it take for full novice to train to play music like this using your device?
Comment by ringo-ring — April 22, 2010 @ 1:44 pm
Thanks for your comment ( you got a great blog yourself) . so let me start by stressing out that to collect data and represent them in such a method ( as I clearly explained in my paper on the subjecT) is doable , especially since we modified the LORETA algorithm for 3 channel correlation and using DSP techniques ( we where lucky to find some extremely interesting litterature on the methodology that we could use) . The challenge itself is within ( and maybe its why I am cautious about it) the fact that electrodes give you a limited time of accurate representation of the EEG wave ( so after some time the recording becomes irrelevant) … also we do our best to filter artefacts , by using filters to lower down the known frequencies , however cardiac and muscle artifacts are usually not tackled enough by the algorithm itself . The project still got over a year to be in a “product stage” and the demo you saw is just a trial …Even if the system is giving good musical representations ( as I will have the opportunity to show further at AES convention this May)…it does still need a lot of work on …. and maybe people with blogs such as yours can act as a feedback source to blogs like mine ( which are only set up to gather people and ask for opinions)
I think I weirdly answered the question about how much a newbie would take to produce such music . The algorithm takes only a few times of usage to run effectively , but the artifact glitches ( especially the ones I mentioned) need to be looked at in more details ( the only artifact that are easily erased are the blink ones)
I hope to hear from you
…and from what I read , looks like Ill be following your blog with great interest
Comment by adytrevisan — April 22, 2010 @ 2:25 pm