#spokenweb

[arielkroon]

Podcast – Academics on Air

This spring, I was part of producing a podcast about a UAlberta campus radio show from 40 years ago. Check it out at

Last year I had the privilege of deep-diving into campus radio with some colleagues wearing my academic research assistant hat. What we found was PRETTY NEAT.

[arielkroon]

This is ShortCuts Live! Katherine McLeod sits down with past SpokenWeb Podcast producers Ariel Kroon, Nick Beauchesne, and Chelsea Miya.

"This month, ShortCuts presents another ShortCuts Live! It is a conversation with Ariel Kroon, Nick Beauchesne, and Chelsea Miya about their collaboration in producing “Academics on Air” (May 2022) for The SpokenWeb Podcast. That episode became a paper that Ariel, Nick, and Chelsea co-presented at the 2022 SpokenWeb Symposium and Institute. After that presentation, ShortCuts producer Katherine McLeod sat down with Ariel, Nick, and Chelsea around a microphone in the SpokenWeb Amp Lab at Concordia University. They talked about processes of collaboration and archival listening that shaped their work. Starting with one audio clip as the short ‘cut’ that caught their attention in the archives, they talk about about context of that clip in the Voiceprint archives, the potential for podcasting to be a radical act of unarchiving, and what makes recordings of a radio show a unique task for cataloguers working with literary sounds recordings, and much more."

[kodamacartonniere]

Kodama Cartonera is proud to announce its first collaboration with SpokenWeb. On December 4 and 13, our editor Aurelio Meza will participate in the Ghost Reading series at Concordia University, celebrating the anniversary of readings by poets Ted Berrigan and James Wright in Montreal. Come and join us. Bring your own beanbag chair!

Kodama Cartonera est fière d'annoncer sa première collaboration avec SpokenWeb. Le 4 et 13 décembre, notre éditeur, Aurelio Meza, participera à la Ghost Reading Series à l'Université Concordia, célébrant l'anniversaire des lectures des poètes Ted Berrigan et James Wright à Montréal. Venez nous rejoindre. Apportez votre propre pouf !

[yumsoonhan]

FMOD and Reel-to-Reel

I spent the day today trying to figure out how we could emulate a reel machine on iOS. It's been a while since I've looked for audio libraries on iOS, and it seems the field has changed. FMOD now has a pretty established API -- unfortunately it's in Obj-C++, but I have written wrappers around that.

A reel machine is unique in that it must have both stitching arbitrary sections of audio files and reversal of the entire tape. Thus there is both a 'subsound' playhead position and a 'total' position and velocity. In the future we will need to dynamically splice out user-designated sections of tape while the reel is loaded.

Toward this goal, I successfully tweaked a channel mixing example to reverse a sound and change frequency values in real-time.

The "realtimestitching" example provided by FMOD showed how to perform stitching in real-time with streams. But I've figured out that streams cannot be played in reverse with FMOD, nor can they be seeked to a subframe, so streams are out of the bag (this is because the fancy features are only provided in the FMOD_SOFTWARE mode, which requires sounds to be preloaded). Most likely I will have to change the code to load whole sounds on-the-fly -- whether or not this will work, we shall see. If not the last resort is loading all of the sounds into a single sentence at once. 

The good news is that once this is done, the audio part of the SpokenWeb app will be finished, and we can begin experiments with all  the functionality in earnest.

Visuals will require a 3D tape ribbon simulation, but having already transferred verlet rope code from JavaScript into Obj-C for Jarbles, I might be able to use the (2D) rope code (or generalize to 3d, if it's not too difficult) on top of a sequence of 3D rectangles chained together. The nice part is that the tape audio can remain separate from these visuals. But all-in-all unless I want to code a 2D rigid body physics simulator from scratch again, I need to look into visual libraries with little overhead. In the past, Cocos-2D was overkill and lagged the app when it shouldn't have. But when dealing with 3-d 2-d hybrids there's not a lot of choice to go for.

Anyway -- I'll retool the audio code tomorrow and cross my fingers.