SamuKata
vihart
vihart

patreon


Feb. update and Ambisonic test

Happy February! Thank you all for sticking around while I work on this project. It's wonderful to be able to pay my bills whether I want to work on something easy or something a bit more outside my usual.

The video is looking pretty good at this point, after finally breaking down and getting a Premiere CC subscription. It's a bit different from CS6 but I'm quickly getting up to speed.

Which means I'm starting to focus on the audio! As I mentioned before, I have my heart set on ambisonic audio for this video, and have been struggling through many Reaper tutorials to try and figure out how to produce it, without much success there as of yet (lots of old plugins are deprecated now).

Luckily Premiere CC has some very basic ambisonic capabilities built in, as one of you kindly pointed out :D

Unfortunately, they're really rough and I could not find any tutorials or examples of how to actually use it for assembling mono sources. The only hint that its even possible seems to be an adobe person saying they tested it and that yes mono sources can be panned, so I'm working in the space between "does it work" and "how does it work". It's not enough to pan a sound randomly, I need to pan it to the right place!

There's no in-software ambisonic audio source monitoring that can sync up with the video, and none of my usual spherical video players seem to support it, so basically I have to export a video and upload it to YouTube every time I want to test something, which takes several hours, and then another mystery amount of time before YouTube finishes doing whatever it is YouTube does to make the sound start working (which is what I'm waiting on right now and why I'm taking some time to write this post).

As you might be able to hear in the above test video if you use headphones and look around on a phone, or by clicking around in a browser, the sound position does indeed change with your head movements! I will totally be able to put sounds in places! 

You might also notice the sound panning settings don't seem to correspond to the video panning settings.

So to try and get it right without too many days lost guessing, exporting, uploading, and testing, today I went back into research mode and made the above video to help me pinpoint just what Premiere is doing with my mono sources when I apply an ambisonic panner. 

I wrote my interpretation in the video description but if you want to independently try to figure out the function between the audio and video panning numbers we can see if we get the same conclusion!

Ah new technology.

I think after this video I'm going to have to have fun doing a bunch more spherical video experiments just to not waste all this time spent learning how to do this all. And because sound art is fun :D

And that's a taste of the kind of detour this video has been taking me on! 

I'm looking forward to sharing some more of those interesting detours in the making-of video, which I will be working on next week while the final versions of this video get finished up and go through their many exporting/uploading phases to test.

The background image/video is from a previous collaboration with Henry Segerman, which you can find here along with some info on Möbius transformations: http://elevr.com/spherical-video-editing-effects-with-mobius-transformations/ 

Vi

Feb. update and Ambisonic test

Comments

Step 1: Organizing a huge batch of metronomes Step 2: Making that video Step 3: Selling off those metronomes Step 4: World domination <3

Arnim Sommer

That would be very cool, and more than that, I really want a pile of physical metronomes. I always just barely manage to have the will power not to buy a fancy metronome but if it were for a video... ;]

Vi Hart

its difficult to hear the verticals without headphones/headset that allow a proper tilt!

Vi Hart

Nice that you are taking Premiere CC for a spin. I tried really hard to map what Premiere was doing with the sound in the video and how it mapped to the perceived visual directions. Some of it was straightforward -- like 0 was actually left -- and you noted your observations as well in the video description. However, a lot of it was very hard to discern, especially because of changing volumes of the ringing sound. It became especially more difficult in the top-bottom axis. Also, one part of the video mentioned 180, followed by -180, and they were supposed to originate from the same location, but they did not at all seem to do so.. Sorry not very helpful -- this is a super late comment too though. I hope you were able to figure out the ambisonic sound mapping! It sucks that you had to upload to Youtube each iteration to see if things worked. That sounds very cumbersome.

Rajat Tibrewal

How about position some metronomes (different sizes and tones) in the room and then let the viewer look around as he likes.


More Creators