"Get Back" fitted nicely into my current standard arranging style, it's somewhat difficult to play, but worked out well. I'm pleased with the arrangement. If you are new to fingerstyle guitar check out how do I get into fingerstyle? As for arranging songs to fingerstyle, check my previous posts such as Arrangement concepts or Can you arrange a song for me? or How far do you take your arrangements?
I wanted to mix up some creativity with technology, I go through spurts of trying out "the fancy new stuff" and putting a bit extra into my video productions. Generally, fancy new stuff and a bit extra in video means only one thing - a lot of time. Quite a lot of time. Extra time spent recording, organising, producing and waiting for processing. So what's new this time around, why did I spend so much time?
- Got a bit dressed up. As part of Movember, which I have been doing for a long time now, I dyed my hair (and moustache, and eyebrows) black. Well, okay, it's not dressed up, kinda.
- I recorded in the spirit of The Beatles Rooftop Concert - I played guitar standing up (which I never do), amplified, on a balcony.
- I recorded 6 different angles, with the help of my oldest daughter on the camera. Yes, each was a separate play through, I don't have six cameras.
- ...but I do have two cameras, and I used them paired to be able to create a 3D video.
- I spent some time learning yet another video editing package, my usual VirtualDub video editor of choice isn't a multi-scene editor. I produced two videos - a 2D and a 3D.
What about the audio? No, I didn't use any audio from the day. It was all home studio recorded (okay, my kitchen table). Not only was it not live, I recorded the audio a month after I recorded the video! Not only that, the audio was bits and pieces from about 6 different takes! Why oh why JAW would you fool us like this? Look I'm not fooling you. In the video, yep, it's me playing. The audio, also me playing. All real. But both were riddled with imperfections. It is a hard song to play, I've only been playing it for about five or six months. I wanted it to look and sound pretty cool, so I went a lot further in capturing video and audio than I normally would. I didn't aim for absolute perfection (impossible) - you will hear plenty of missed notes, fret buzz, and I can clearly see some places where I played a wrong note in the video but miraculously the audio is correct. I just wanted it to look and sound pretty cool.
3D is an interesting thing. I have two cameras which I welded up a bracket to connect them together - one camera records the left eye, one records the right eye. From previous experience the cameras are too wide to be side by side, that tends to "overdrive" the 3D depth. To reduce the distance between the lenses, I mounted the cameras bottom to bottom. This gave a separation of about 80mm - I was after 65mm, but this was better than the previous 120mm. Not many humans have an eye separation of 120mm...
This introduced a problem - the footage was tall and narrow. I still wanted it to be a normal wide screen video. So, I rotated the videos, and cropped them hard, sadly throwing away most of the picture, to be 4:3. That means it's not full HD video, which I kinda like. Oh well. I then had to further crop the videos to "align" the left eye and right eye. In my experimentation with 3D video I found you should find a point of interest - something in the middle around the centre point of focus - and crop the videos so that point is at the same x,y location in both the videos. This helps make up for the too wide lens spacing and any alignment error in the camera mount. Light physics takes care of the rest - the left and right videos have an outlook that creates the spacial difference that your brain translates to depth. Your brain doesn't seem to be too fussy - the variety of distances and angles it has dealt with (tilt your head - depth perception still works!) I guess means it's prepared to accept - it wants to accept - most stereo inputs and turn them to a depth image.
Putting the images side by side via software (more synchronising - this time synchronising left eye video feed with right eye) will create a "standard distributable" 3D video. Then chop them up and match them to the audio. It seems like putting two HD video streams side by side (3840x1080 pixels) is too much for most systems to bear in 2014, whereas reducing the width by half (1920x1080) seems to be accepted by most 3D playback systems. The 2D video is simply one half of the original video.
I haven't worked out the deal with Youtube and 3D videos, so I uploaded both the 3D and 2D one. I'd hoped that Youtube would automatically play the correct video based on your playback system, but I haven't worked out how, if it even exists.
I've plenty more to say about the creation of the video, if you need more information about some part of the process drop me a comment. Until then, check out the 2D or the 3D video!
Get Back - 3D