Video-equivalent of "pitch-shifting."

Discussion in 'Amateur Video Production' started by Radium, Aug 22, 2007.

  1. Radium

    Radium Guest

    Hi:

    I started a new thread because the previous one started to go into
    tangent of digital vs. analog but was filled with emotions and
    personal vendettas rather than science and logic. So I changed the
    thread.

    Anyways, Adobe Audition and voice-changers allow the frequencies of an
    audio signal to be shifted w/out low-pass filtering or changing the
    tempo. There are two video-equivalents of this because, while audio
    has only one frequency component [temporal], video has two [temporal
    and spatial].

    The temporal video-equivalent would be changing the rate of back/
    forth, up-down or other repetitive/cyclical movement [such as wing-
    flapping or flickering of lights] of the video signal without high/low-
    pass-filtering, separating any portion of the video signal, or
    changing the speed at which the video-signal -- just as voice-changers
    can lower the frequency of audio without changing the speed of the
    audio. Using a voice-changer to decrease the pitch your voice will not
    cause your speech to slow down.

    The spatial video-equivalent would be changing the "sharpness" of a
    still image without high/low-pass-filtering or changing the size of
    the image.

    Below is an example of low-pass-filtering in the spatial domain:

    Here is an original picture:

    http://www-dse.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/sab/report.normalimage.jpg

    Here is the picture after low-pass filtering:

    http://www-dse.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/sab/report.lopass.jpg

    I obviously do not want this at all. Low-pass filtering involves
    removing high-frequency components while preserving the low-frequency
    components. Once again, this is not what I want. If a device cannot
    handle high-frequencies, then I would like all the frequencies of the
    signal to be down-shifted until the highest frequency is low-enough to
    be acceptable to the device. This down-shifting should be done w/out
    slowing the speed of the signal -- or in the case of spatial
    frequency, w/out increasing the size of the image.


    Thanks for your assistance, cooperation, and understanding,

    Radium
     
    Radium, Aug 22, 2007
    #1
    1. Advertisements

  2. Radium

    BobG Guest

    If you play the tape or record faster, the pitch shifts up. If colors
    are analogous to pitch, speeding up would be a shift to the blue,
    slowing down would be a red shift.
     
    BobG, Aug 22, 2007
    #2
    1. Advertisements

  3. Radium

    Jerry Avins Guest

    A video signal consists of a succession of still images that follow one
    another at fixed intervals. What you call tempo is determined by how
    different each image id from the ones before and after it. That makes
    what you write next wrong.
    Sharpness is altered by applying a filter. There are sharpening and
    softening filters.
    But it together and be more specific.

    Jerry
     
    Jerry Avins, Aug 22, 2007
    #3
  4. Radium

    Radium Guest

    Okay. How would you correct it?
     
    Radium, Aug 22, 2007
    #4
  5. Radium

    AnthonyR. Guest

    So basically you want software that analyzes each frame the way mpeg coverts
    video and instead of eliminating file size by compressing the images, you
    want to have it eliminate most of the static images and thus reduce file
    duration instead. Is this correct?
    So if a scene has movement and talking you want that left alone but if the
    person is just standing still or not much action going on in the scene have
    it eliminated and blended so it shortens the duration but not affecting the
    movement speed thus making the scene shorter in length.
    I don't know of anything that would do this to video but I could see a use
    for it.
    DVD player software can speed up movement by say 10% and allow the sound to
    not be effected as far a pitch goes, this is useful for watching say a 2
    hour and 15 minute movie on a flight that is 2 hours long.
    Is this what you are trying to find?

    AnthonyR.
     
    AnthonyR., Aug 22, 2007
    #5
  6. Radium

    Jerry Avins Guest

    I would leave out the rest of the paragraph. It's based on a false
    assumption.

    Jerry
     
    Jerry Avins, Aug 22, 2007
    #6
  7. Radium

    Stuart Guest

    Sound is physical - Light is electromagnetic radiation

    That would make a great Sci-Fi effect to depict 'beings' in a different
    temporal dimension co-existing with us but what you describe is the effect
    of motion linking the audio analogy to video which is not valid. Sound is an
    air-pressure wave whose speed changes depending on the medium whereas light
    is part of the electromagnetic spectrum whose speed is fixed to the speed of
    light and except for some very high-end academic experiments never changes.
    Never-the-less it's a good special effects used in a modified way in the BBC
    production Ultra Violet.
     
    Stuart, Aug 22, 2007
    #7
  8. Radium wrote:

    (snip)
    It would be reasonably similar to a cyclical intensity of a lamp,
    or something similar. A moving object is different.
    -- glen
     
    glen herrmannsfeldt, Aug 22, 2007
    #8
  9. Radium

    Jerry Avins Guest

    You have very little influence on the speed of sound in the medium in
    which you live.

    The speed of light depends on the medium it travels through and
    sometimes on the frequency. Consider a prism's dispersion.

    Jerry
     
    Jerry Avins, Aug 22, 2007
    #9
  10. Radium

    Radium Guest

    I don't think so. Any thing in the video with a temporal/spatial
    frequency component that is too high for a low-bandwidth device to
    accept, should have all of its frequencies downshifted until the
    highest frequency is low-enough for the low-bandwidth device to accept
    without any aliasing or other artifacts associated with a frequency
    exceeding the limits.
    No. The length of any parts of the movie should not be affected at
    all.
    Not really. This change in video-frequency has nothing to do with
    speeding up a video. The movie should remain exactly the same length.
    Two hours should stay two hours.
     
    Radium, Aug 22, 2007
    #10
  11. Radium

    Stuart Guest


    This is only true in the domestic experience using a gramophone record or
    tape but in the world of TV, Film and Music Industries if a producer asks
    me to speed up some dialog I will do so without affecting the pitch,
    likewise I can take the pitch up without affecting the tempo. The chipmunk
    sound is of course achieved by both increasing tempo and pitch. In the
    analog domain this has been possible and widely used to tighten up
    commercials etc since 1958 via the EMT Pitch and Tempo Regulator ( a German
    invention of 8 playback heads in a rotating drum either in the direction or
    in counter rotation to the linear motion of the tape) and in the modern
    digital domain with a simple plug-in for programs like Adobe Audition or
    Wavelab.
     
    Stuart, Aug 22, 2007
    #11
  12. Radium

    Jerry Avins Guest

    Visible light extends for less than one octave, nominally from 400 to
    700 nanometers. Those are wavelengths more or less centered around
    500,000,000,000,000 Hz. You can't shift the band much and still see it.
    Describe the action of a person taking a one-mile walk and some stations
    along the way.

    Jerry
     
    Jerry Avins, Aug 22, 2007
    #12
  13. Radium

    peter Guest

    Is this about technical discussion of signal processing theory, or is there
    a video effect you are trying to achieve?

    If it is the later, please restate it without using audio as an analogy.
     
    peter, Aug 22, 2007
    #13
  14. Radium

    Radium Guest

    You're talking about color-frequency. Totally irrelevant to my
    discussion of video-frequency. I am talking about temporal and spatial
    frequency, not color-frequency.

    Color-frequencies = frequencies of electromagnetic radiation visible
    to the human eye, which as you pointed out, corresponds to wavelengths
    that are at least 400 nm but no more than 700 nm.

    Once again, by "video frequency", I am referring to the temporal and
    spatial frequencies of the video signal, not the color-frequencies.
    Huh?
     
    Radium, Aug 22, 2007
    #14
  15. Radium

    isw Guest

    I didn't want to post this on the original thread because it would have
    been lost in the noise.

    There is a very fundamental difference between a data stream
    representing audio, and one representing video. Analog or digital; makes
    no difference.

    A stream representing audio is a continuous stream of information, with
    every part of it temporally related to a specific time in the audio. It
    can be slowed down, sped up, or have pieces cut out or added at will.
    These last are the basis for the "speed up/slow down without changing
    the pitch" algorithms.

    There is absolutely no parallel to this for video, as it is handled
    today. Unlike audio, video is a series of still images, equally spaced
    (hopefully) in time -- i.o.w. it is always temporally quantized.

    It is possible to change the spatial resolution (temporal spacing
    between successive images) and the spatial resolution (within each
    individual image) totally independently, and in fact, the two need to
    bear no particular relation each to the other. You can create high
    temporal resolution but low spatial resolution -- or the other way
    around -- easily. You can also increase or reduce the temporal
    resolution by interpolation/decimation while leaving the spatial
    resolution totally unaffected, and you can also do the "reverse".

    The fact that there is essentially no relation between these two
    entities -- i.e. the data stream is comprised of a sequence of
    descriptions of a series of still images -- is the reason why what you
    want to do is almost certainly impossible.

    If you really want to try, the first step will be to devise a method of
    recording video that does not quantize the temporal axis; i.e. not using
    a sequence of still images.

    Good luck, and great fame awaits.

    Isaac
     
    isw, Aug 22, 2007
    #15
  16. Radium

    Stuart Guest

    So much Schatten-lieben
     
    Stuart, Aug 22, 2007
    #16
  17. Radium

    stratus46 Guest

    I can't believe I'm jumping into this muck. Totally useless point in
    that any changes in frequency leaves you with a signal you can't use.
    No recorder can record it and no monitor can display it. Now if
    you're willing to live within the frame/line rate definitions, then
    changing frequencies _within the line frame boundaries_ would be
    similar to a DVE zoom but there is much more capability than that. 25
    years agon the Ampex ADO and Quantel Mirage were literally twisting
    pictures into screws, making them into spheres, rolling them up. Tape
    machines have been slo-mo and speeding up for over 30 years. Stuff is
    way cooler now. Go look it up.

    GG
     
    stratus46, Aug 22, 2007
    #17
  18. Radium

    BOING!! Guest

    [

    Oh dear how busy you are, making up nonsense trolls! If you were serious,
    you would give a couple links to youtube showing examples of spatial &
    temportal "pitch shifting" video equivs
     
    BOING!!, Aug 22, 2007
    #18
  19. Radium

    Chaz Guest

    You lost me?????? http://chaz.2ya.com
     
    Chaz, Aug 22, 2007
    #19
  20. Radium

    Chaz Guest

    OK I saw it lol
     
    Chaz, Aug 22, 2007
    #20
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.