#miniaudicle

[gile]

A great talk by Ge Wang around computer music.

[annmariegcreates]

<Maths and Art (175006): Assignment 2: Constructed Sound - Removieing>

Reflective Statement: Komorebi

I knew from the beginning of this project, I wanted to create an alternative narrative based around the awkward love story of the lead characters in the movie Freaks. I was fascinated with how the scenes juxtaposed so well with the emotionally artistic scenes of Melancholia. I found creating the video clip first helped in having a direction when creating the sound. I knew what sounds I needed to create and the length in which they needed to be. 

I overlaid clips I had recorded from miniAudicle so that there was a somewhat smooth transition for emotion and tone creation. The clips I used were a variation of sample shreds within miniAudicle, of which I overlaid and altered their frequency, pitch and tone where I needed for fitting with the story I wanted to portray. Even small changes in tone and speed, or even volume of the shred, created a significant difference in how it sounded. The more emotional scenes, such as the beginning and end of the clip, present a much slower tone and wavelength, in comparison to the conflict in the middle where there is a higher frequency and at one point an unintelligible sound representing the conflict in the relationship. I ensured there were obvious transitions at each turn in the story, so that there was a flow to portray emotion and romance.

The piece is named Komorebi as the sound reminded me of the Japanese word that describes light that breaks through leaves. As with the composition, there is a level of stop start to the light. The main criticism I find with my soundtrack is that it is too robotic for it to represent a love story. The beginning and end track needed to be the same to reiterate that the lead characters end up back together. After watching the clip many times with the created track, I find myself viewing it as a rather depressing story and perhaps the composition emphasises this. I had originally intended the sound to have a sense of overcoming a struggle, however with such a short clip, it just was not possible. An unintended sound to video correlation was in the cartoon section where the sound almost represents speech. I set out to remove all dialogue as I wanted the music to represent the emotion of each scene, however as this part of the video is significantly different in comparison to the rest I decided to keep this.

I found it challenging to relate to the sound and how it would fit into the story I had created. Using a coding software was a very different process to cutting and chopping clips as done in the studio project. I have had to remind myself multiple times that I am to create sound, rather than music. I found it much easier to record clips and cut them into how I wanted them, however the process seemed quite the opposite in this case. Due to the complexity of coding the music, I had less control and in my limited knowledge of the software, used and modified the samples provided by the program. As an extra exercise to test how the story is able to be developed through sound, I had also created a soundtrack using the music from the videos used. I found this was able to aid me in constructing the final track through miniAudicle as I had an audio guide to work from. The final story, I do believe is portrayed through the composition accompanied by the video. The title cards may or may not be necessary, but for the purpose of continuity of the black and white scenes, I thought it added to the end result.

References:

Example code that was altered/modified was downloaded from: http://audicle.cs.princeton.edu/mini/

MiniAudicle example clips used:

blit.ck

rlp.ck

wind2.ck

chubograph.ck

stereonoise.ck

hp.ck

oft_02.ck

oft_03.ck

oft_05.ck

larry2.ck

unclap.ck

Clips used in creating the video:

Zeduka. (2012, 7 January). Melancholia (2011) - Opening. Retrieved 28 August, 2013, from http://pdl.vimeocdn.com/46765/309/159703635.mp4?aktimeoffset=0&aksessionid=07502ed02f845248f64cc2edd8744d58&token=1377775212_23958a0d16255bd3cf00d943293757e6

Tanya Mills. (2013, April). Freaks (1932). Retrieved 27 August, 2013, from http://pdl.vimeocdn.com/46765/309/159703635.mp4?aktimeoffset=0&aksessionid=07502ed02f845248f64cc2edd8744d58&token=1377775212_23958a0d16255bd3cf00d943293757e6

andythebeagle. (2013, 26 June). Aesop’s Fables Wood Choppers (1929) VanBeuren Studios cartoons. Retrieved 29 August, 2013, fromhttp://www.youtube.com/watch?v=duoQZN5ZQZ0

No copyright infringement intended in the creation of this clip. For educational purposes only. 

[felixb13sonic]

AS 90620: Term Two, Week Ten

The entirety of our code is due in tonight. I am far from worried though, as I have completed all the difficult work and all I need to do is write one last shred (The final five note flourish) and I have done all the coding. I will finish that shred tonight.

I really hope I get to recording this week; a few people have already begun, and it seems like an interesting process. It did remind me of a rather important issue though, that I have neglected to mention throughout these blog posts. Due to the convenience of sporking, the majority of the class are recording their code onto a single audio track in Logic (or Reaper, depending on their OS); the sporks automatically enable the dormant functions, so all one has to do is press record to start and stop. This is slightly more complicated for me: the piece of code I have been using as the basis of my piano parts already contain a function, so to spork it you would need to find some way of enabling a 'function-within-a-function'. Neither my teacher, the almighty Internet nor I have any idea how to do this. We are not sure if it is even possible.

So, I have made an agreement with Terry to hand in all the shreds of code separately and likewise record them as separate tracks into Reaper. I may even compile a directory of the code so that it can all be found in one convenient file. This track-by-track recording method will certainly take time, so I will need to start recording this week.

Listening:

Leon Russell - Tight Rope

Justin Timberlake (Feat. Timbaland) - Cry Me a River

The Sensational Alex Harvey Band - Long Hair Music

Bert Jansch - Poison

Jesse Winchester - The Brand New Tennessee Waltz

[felixb13sonic]

AS 90620: Term Two, Week Nine

I have written almost every shred I need! 

This section of the process is going faster than I had expected, especially considering the mid-process change of concept that could have seriously hindered my progress. I attribute the swiftness to my methodology of re-using the same piece of code and altering it slightly for each new shred; this means I do not have to keep writing in complex new functions. Some would argue that it shows a lack of skill or diversity, but I believe that a consistent sound is more important than the wide-range of code used, given the client-oriented nature of this process.

This week will be tough and primarily maths related. When I drafted the piece in Reason, I automated a number of tempo changes that really built a sense of anxiety and dread. I lack the technical skill to create such automation with ChucK, so I had another idea: I would have two shreds with a time difference of 20% (one shred is standard speed, the other is 20% faster); then I would create an intermediary shred that was 10% faster, so midway between both shreds. This gives the illusion of the gradual increasing incrementation of tempo, as best as I can do with my current ChucK knowledge.

This requires some basic percentage calculations: in Reason, I must figure out the start tempo and finish tempo of the automated sections. I then must find out the percentage increase between the original and end tempi. I must then decrease the values of the note rate/duration by this percentage to give me what should be the correct number. There are a number of possible complications: I set the duration of the bass notes by ear, so there may be synchronicity problems when I change the rate/duration; it will be particularly hard for the bass section where the duration of each note is defined by a float value in the array. I hope that with the power of accurate calculations, none of these possibilities will bother me in the slightest.

Accordingly, my goal for the week is to accurately calculate and execute the myriad speed changes required for this piece.

Listening:

Fleetwood Mac - Rhiannon

Slave - Watching You

G-n Marko Ivanov - Ne Plachi Maiko Ne Zhali

Bruce Springsteen - Born to Run

Justin Timberlake (Feat. T.I.) - My Love

[felixb13sonic]

AS 90620: Term Two, Week Eight

Changes abound!

I have decided to change direction with my piece and follow a music-based route instead of an ambient one. After composing the four ambient elements (the wind, the bassy WOMP effect, the 'Polar Sines' and the attenuating deep bass siren) I was at a complete loss as to where to go next. It's not that I wasn't pleased with these elements - I've showered them with enough praise in previous posts - but that I lacked the inspiration to continue with them. As mentioned in the Key Considerations paragraph 'Concepts', the freefrom nature of an ambient piece meant that I had no real sense of direction. When combined with a lack of fresh ideas, it lead my piece to a temporary dead end.

This was a mere minor setback, though, as I had ideas in reserve. In a frenzy of inspiration, I composed a Halloween-esque piece for piano that I was extremely pleased with. This would be my new piece. After perfecting it physically, I created a draft of the song in Reason 4, where I was able to automate tempo changes that exceeded my own technical proficiency. Again, I was very pleased with how this sounded.

To make the crossover from Reason to MiniAudicle I adapted a piece of example code from the ChucK website, which would form the bones of almost every imaginable shred. The 'Rhodey' class I chose was one of two piano-based classes provided by ChucK, and I much preferred it to the 'Wurley' class.

Separate posts detailing each part of the process will follow. An MP3 of the Reason file will also follow. My goal this week is to write code for every piano part as heard in the MP3, and perhaps start on writing code for the accompanying bass line.

Although I did achieve the goals I had set last week, that is ultimately irrelevant as I am not going to be using them in the final piece.

Listening:

Johnny Thunders - Daddy Rolling Stone

Bruce Springsteen - Ramrod

New Order - Let's Go

Hank Thompson - Cocaine Blues

Yo La Tengo - Autumn Sweater

[felixb13sonic]

AS 90620: Term Two, Week Seven

Another brief post, but one filled with action. I have a clear plan of attack for this week. The structure of my piece will go as follows: Wind fades in, remains for around twenty-five seconds, fades out - whilst wind is fading out, the deep bass siren will fade in and play for half a minute - after this time, the Polar Sines will fade in over the top - gradually, the bass siren fades out and we are left only with the strange Polar Sines. This will probably be around half of the required time, which is for all intents and purposes, a very fair amount.

The problem is the absence of the attenuating bass siren. I have devoted this week to refining that effect; so far, I have identified several possible solutions to the problem. The first I will attempt, on this coming Wednesday, is the BandedWG class, which will use waveguide techniques to model the sound of the SinOsc I plan to link it with. I will update this blog on Wednesday as to whether this was a successful experiment or not.

Listening:

NEU! - After Eight

The Human League - Love Action (I Believe in Love)

Accept - Bound to Fail

Einstürzende Neubauten - MoDiMiDoFrSaSo

Necrophagist - Fermented Offal Discharge

[felixb13sonic]

AS 90620: Term Two, Week Six

Very short post his week. Due to this being Shakespeare Society performance week, I am refraining from doing schoolwork for the entire week. Once this has finished, I will launch back into it, but for now it demands most of my time and attention. This was not unseen, as I had planned for it from this project's outset.

The successes I achieved last week in the creation of the two new functions were joyous indeed; I am not altogether sure where to go from here though. I have enough score for around thirty seconds of the sequence; there may be some more useful examples on the ChucK site. When week seven comes around, I will have a much better idea of where I want the piece to go. I also must not forget about my attempts to create the Morricone-style bass siren.

Listening:

Yello - Vicious Games 

Traffic - Berkshire Poppies

Leon Russell - Pisces Apple Lady 

King Crimson - Pictures of a City (Including 42nd at Treadmill)

Jeff Beck - Psycho Sam

[felixb13sonic]

AS 90620: Term Two, Week Five

I have begun proper work on my audio accompaniment to The Shining, although I am not off to a stellar start by any stretch of the imagination.

My first attempt at creating a Morricone-esque deep bass siren met only with failure. I have not given up though, as there are a few things I may use to salvage the effect: from what I have read on the ChucK website and heard from examples, the BandedWG class may be able to achieve the attenuation I am looking for; the swell function may also help. Additionally, I have heard that my classmate Tom has created some kind of fade function, which is another possibility.

But let us not focus on what I did not get done and focus on what I did get done. Today I created an entire function that I will most likely use in the final piece, the creation of which is dealt with in a separate post. I am very happy with it, and pleased that I am showing some signs of progress.

For the coming week (the last before the Shakespeare Society performs) I want to create at least one other function to use in my piece, as well as refine the failed deep siren.

Listening: 

The Move - Cherry Blossom Clinic Revisited 

Yello - Stalakdrama

Sir Douglas Quintet - I Don’t Want to Go Home

1917 German Naval Band Recording - Unter der Admiralsflagge, Marsch, Op. 32

Skyzoo - Maintain