Difference between revisions of "Commute"

From CCRMA Wiki
Jump to: navigation, search
 
Line 36: Line 36:
 
binaural rendering available at:
 
binaural rendering available at:
  
https://ccrma.stanford.edu/~walker/222/binaural.html
+
https://ccrma.stanford.edu/~walker/222/surroundcommute.html
  
 
This project has been incredibly fun so far.  I got to ask my friends to either record for me, or send existing things they have been working on.  Mixing 30 seconds or less of a particular genre is also fun because it rarely gets stale before you are finished.  I learned a few things, or perhaps formed some opinions, while doing this project, they are:
 
This project has been incredibly fun so far.  I got to ask my friends to either record for me, or send existing things they have been working on.  Mixing 30 seconds or less of a particular genre is also fun because it rarely gets stale before you are finished.  I learned a few things, or perhaps formed some opinions, while doing this project, they are:

Latest revision as of 13:46, 15 June 2017

for my final project, i am going to make a 5 minute piece for listening that is an impressionist take of my morning commute. audio will be spatialized for playback in following formats: 8-channels, binaural, and 7.1 surround. audio included in this piece will consist of sound effect samples, field recordings, studio recordings, vst’s, and sample manipulations. i may borrow various facets of compositions(and/or audio snippets) from sources that i find particularly inspiring. sections of this immersive experience will employ different spatial/compositional techniques I have ponderer, theorized, and did my best to realize.


5/24/17

project milestone 

Initially, I planned to do a piece in one style of music that lasted roughly 4 minutes. Now, I am going to do a virtual commute that contains music from multiple styles, all of which are mixed for multiple surround formats, which should last roughly 5 minutes.

To ‘move’ objects in horizontal space, I have developed a matrix/routing scheme using seven instances of Ableton Live’s Autopan, in a root or tree like fashion. This system is rather complicated to describe in prose, but I am happy to demonstrate or draw it out for anyone interested in using it. It will be part of presentation on Thursday 5/25/17.

As for the content of my multiple musical styles, I am in the process of creating these scenes in Ableton, recording friends and other musicians, gathering stems of existing tracks from friends/previous musical collaborators, and capturing some field recordings to simulate driving in a car. There will be roughly 10 different musical genres presented to the listener over the course of the 5 minute piece; half of these segments are complete, half are just concepts, only 1 or 2 are spatialized so far.

This piece will be completed by Monday of Week 10 (for 8 channels). All audio should be recorded by mid-next week. Spatializing/mixing these projects will be exciting and presumably not terribly time-consuming. Later in Week 10, or perhaps during finals week, I will have separate mixes for binaural and 5.1 surround.


List of people involved(so far):

Chris Chafe

Jesse Green

Jay Kadis

Justive Njoku

Dylan Nyguyen

Davide Tiso

Nakul Tiruviluamala


6/10/17 binaural rendering available at:

https://ccrma.stanford.edu/~walker/222/surroundcommute.html

This project has been incredibly fun so far. I got to ask my friends to either record for me, or send existing things they have been working on. Mixing 30 seconds or less of a particular genre is also fun because it rarely gets stale before you are finished. I learned a few things, or perhaps formed some opinions, while doing this project, they are:


To spatialize in binaural, objects must move angularly. If objects are panned at all different angles around the listener's head on a horizontal plane, yet they remain static, the effect is not too different than stereo. Movement does not need to be drastic, an oscillation of 30 deg back and forth does the trick.

Moving objects radially inward is challenging mostly because it involves a steep increase in gain. If you have found a gain stage that is nice for a sound object at the perimeter of the space, moving it inward quickly undoes all of that work.

Without using any ambisonic format, going from 8-channels to binaural is easy, but the opposite is not. In Ableton Live, the user can specify the number of mono outputs and pan between two of them. In Logic Pro X, outputs are always locked to stereo pairs, but their binaural and surround panner calculate everything for you. The last radio station before the listener exists the car was the only piece spatialized in Ableton Live. I recorded mono audio from the 8 outputs in Ableton, opened these tracks in Logic Pro, and place them in the respective angular positions using binaural panner. Track 1 was between 315 and 360/0; Track 2 was between 0 and 45; Track 3- 270 and 315; Track 4- 45 and 90, etc. As far as I could tell, this conversion worked well as no phasing was introduced, and spatialization was very similar. I am curious to see how it will sound in 5.1. I cannot image how to go from a binaural/5.1 format back to an 8-channel at this point....

Short sounds do not move well quickly. The illusion that the percussion sounds are swimming around you becomes too unrealistic and does not sound good.

Compression after a binaural pan can minimize the spatial effect. Presumably Logic's calculated ILD(intramural level difference) is compromised the gain reduction is introduced.

Unnatural sounding things can be interesting. In the metal radio station, the drums physically bounce around the listener. Although it does not feel correct, it still interesting in my opinion.

Mixing in Logic is perhaps better than Ableton. I had not used Logic Pro X very much until this project, so this was a great way to dive into it. Logic Pro X shows the effects channel strips for every track where as Ableton only shows you the effects on the selected track. Logic Pro contains the binaural and 5.1 panners. Logic Pro has several stock compressors and genre/instrument specific presets; the Ableton compressor is nice too, and perhaps better at side-chaining, but Logic's quickly feel warmer after you get comfortable with them. Automation is a little different, but perhaps easier to do across many tracks in Logic Pro. (Almost every stem in eery track I did for this project has multiple automation curves changing spatial placement, volume, etc.)

That being said, Ableton is better at quick editing, sample manipulation, arrangement, and side-chaining. I developed a work flow that basically looked like this this: Record in Ableton, edit, arrange, bounce individual stems to Logic Pro X, spatialize, bounce to 30 second clip of genre, insert in Surround Commute narrative. I am happy to show anyone my project files for the individual tracks(by genre) or the final radio narrative.


Again, this project was incredibly fun. I am excited to mix it in 5.1 in the next few days. Thank you to everyone who helped!

The people involved are as follows:

Chris Chafe on cello - Bach Chorale

Davide Tiso on guitar - Metal

Marvin Paez on bass - Latin

Jay Kadis on guitar and bass - Rock

Jesse Green on many things here - Electronic

Justice Njoku on vocals - RnB (‘Location’ cover by Khalid)