Oona (windytan)

@windyoona

Programmer interested in signals and sound waves // blog:

Finland
Vrijeme pridruživanja: prosinac 2011.

Tweetovi

Blokirali ste korisnika/cu @windyoona

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @windyoona

  1. Prikvačeni tweet
    10. ožu 2018.

    Thought I saw a sine wave sculpture. Turns out it was just the real part of this complex sinusoid.

    Prikaži ovu nit
    Poništi
  2. 26. sij

    The right border can look messy in some fast YUV modes. Idea: optional zoom-in to hide 1% overscan (only the top-right corner shown here)

    Two versions of the same TV test card, one of them showing a jagged green right border.
    Prikaži ovu nit
    Poništi
  3. 26. sij

    Moderate levels of noise could optionally be removed by adapting the FFT windowing function to SNR (based on F2DC's article in the May/June 2003 issue of QNX: "Some Thoughts on Real-Time SSTV Processing")

    Prikaži ovu nit
    Poništi
  4. 26. sij

    Experimental command-line SSTV decoder used with iTerm2's imgcat

    Screenshot of a command-line program showing textual debug info followed by a picture of a TV test card within the terminal.
    Prikaži ovu nit
    Poništi
  5. 15. sij

    I did try this with a stereo photo pair and color SSTV earlier. There was a perceptible stereo effect in the audio but mostly the colors just all sounded shifted to the right wrt the sync signal. So a slight disappointment. Maybe a simple line drawing would sound more fun.

    Prikaži ovu nit
    Poništi
  6. 15. sij

    The reason I think this would sound interesting is that the depth illusion comes from the horizontal offset of corresponding objects in the two images. In a radiofax transmission, a horizontal offset corresponds to a time delay. We can use inter-aural delays to hear directions.

    Prikaži ovu nit
    Poništi
  7. 15. sij
    Prikaži ovu nit
    Poništi
  8. 15. sij

    Shower thoughts: If you converted a stereoscopic image pair into two radiofax transmissions so that the left image was played to the left ear and vice versa, could the inter-aural time differences tell your brain something meaningful about 3D depth in the image?

    Prikaži ovu nit
    Poništi
  9. 8. sij

    Some waveforms (no spectrograms) recorded in 1993: "A New Type of Furniture ESD and Its Implications"

    Prikaži ovu nit
    Poništi
  10. 8. sij

    Has anyone managed to record the kind of EMI a gas lift chair would emit?

    Tweet je nedostupan.
    Prikaži ovu nit
    Poništi
  11. 30. pro 2019.

    This could be a bug, but I don't normally (npi) work with 3D software so could be expected, too? 1. Scale (s) the default cube to -1.5 so that normals flip 2. Export as PLY 3. Import the PLY 4. scale transform is correctly applied to size but the two objects have opposite normals

    Prikaži ovu nit
    Poništi
  12. 30. pro 2019.

    If you ever encounter a mysterious problem of Blender exporting PLY with inverted normals, check that your object doesn't have a negative scale transform by mistake. (At least it made me write a lot of unit tests!)

    Prikaži ovu nit
    Poništi
  13. 25. pro 2019.

    Another one based on Chu et al. (2002): Detailed Directivity of Sound Fields Around Human Talkers

    Prikaži ovu nit
    Poništi
  14. 25. pro 2019.

    Reading fascinating research papers at 2am again. Here's one about the frequency-dependent directivity of the human voice.

    Prikaži ovu nit
    Poništi
  15. 21. pro 2019.

    Found a more densely sampled database of head-related impulse responses to use in 3D experiments. The 2001 paper has friendly-looking diagrams as well.

    Poništi
  16. 20. pro 2019.

    You can still hear how the HRTF model is sampled only at 15° intervals. Perhaps I could interpolate.

    Prikaži ovu nit
    Poništi
  17. 20. pro 2019.

    Doppler effect emerges when the acoustic path tracing update rate is high enough

    Prikaži ovu nit
    Poništi
  18. 18. pro 2019.

    Piinaako pohjoisen pimeys? Ensi viikosta kirkastuu.

    Poništi
  19. 13. pro 2019.

    Random idea: reconstruct original audio from vinyl scratching

    Poništi
  20. 11. pro 2019.

    A little video of said project (note, surprisingly high dynamic range in the audio)

    Prikaži ovu nit
    Poništi
  21. 9. pro 2019.

    Coding something that uses a FIR filter with a 32-kpoint impulse response that changes on the fly a few times per second. But I want the changes to happen smoothly. So I'll filter the filter with 32,000 smaller filters...

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·