Adobe, MAKE SOME NOISE

Handling sound in Flash is restricted to setting the volume and panning. We want to create sound and music applications using the Flash platform but as for now the platform restricts our possibilities in many ways.

For all the other Adobe nerds: What do you need to make audio applications that rock? Contact us! info@make-some-noise.info

For Adobe, here is the list of things we wish for our work to be outstanding:

  • Writing sound bytes (amplitudes) continuously to an existing sound object.
    With the current sound capabilities its impossible to broadcast generated or modified audio to a given output. We need to use hacks or other environments (like java) to achive that. This doesn’t only mean we have to be aware of a ton of wicked things to reach the target, but more important is the fact that we loose lot of performance and/or rely to another environment to be necessary for the work.
  • High resolution timer
    For audio processing its important to have a timer that is as exact as possible and capable to process code within very short time distances. With the current situation we face the problem that the timing isn’t really exact enough for decent applications.
  • Access to the microphones amplitudes
    Right now the microphone input is life-encoded with the Nellymoser Codec. This means its impossible for us to record any input in a reasonable quality for audio processing using the flash API.
  • Multi-Threading
    Audio processing requires a lot of performance itself. But if we continue to create applications with a rich user interface we will end with only few processor cycles spared. Any heavy operation in the ui will stop the audio stream from working which is a crucial problem if we ever want to have serious output. The optimum would be to have it at the language level, so we could just require threads using regular code. But we would be satisfied if each .swf would get a seperate thread to work at, within browsers.
  • Lossy audio codec -Vorbis Codex
    In some situations we need to send or recieve audio files. The MP3 codec is outdated and of questionable quality. We would need the implementation of a proper audio codec like the Vorbis codec which is open-source (no licence costs) and produces better results.
  • Lossless audio codec - Flac support
    Recorded music from any device should be raw (as mentioned earlier). To send it to the server its important to have a codec that allows to send raw audio data in a compressed way to reduce the upload time and create faster expiriences.
  • Midi controls IO
    The input device is extremly important for creating music. Midi IO (which has been standardized) would allow to integrate the music devices into our applications.
  • Native FFT (Fast Fourier Transform) in both directions
    FFT would provide us a shortcut to the current frequency domain of sound. We could analyse and modify the spectrum of a discreet time segment. We already implemented it in AS3, but we think a native approach would be much faster. FFT is also very important for image processing.
  • Native random seed
    The current Math.random solution in Flash is not predictable. Well, you may think that random should act like this, but in many cases you like to get the same sequence of pseudo random numbers everytime you start the flash application. It is also great for game-development (predictable, but random-like behavior) and image-processing as well. A seed parameter would alter the sequence.


신고
Posted by Ssirius

카테고리

전체 (92)
Programming (13)
Digital Nomad (2)
Projects (7)
Sound (14)
Travel (45)
Think (9)

달력

«   2017/12   »
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31            

티스토리 툴바