Forums

  • avatar
    68 sounds
    862 posts


    Not contributing to the discussion, but just wanna point out that I am a bit inspired now.
    I will be running solid color files (800x600) through audacity and generate outputs: color red sound, color white sound, color yellow sound etc, and put them all in a pack here on freesound.

    Thanks for the dare!

  • avatar
    103 sounds
    1291 posts


    jeffercake wrote:
    hey, got some good results importing photoshop files into audacity, have a listen:
    http://www.freesound.org/people/jeffercake/sounds/222568/

    Been working on creating spectrograms, editing them as images in photoshop, and the re-synthesising them. Will post up the results shortly.

    Can you tell me the exact format used? PSD (Photoshop native)or PS (PostScript)?

  • avatar
    2067 sounds
    2243 posts


    I am sure some of you are following this thread. Here is a little Dare 26 spin-off:
    http://www.freesound.org/forum/dare-the-community/34991/
    Would appreciate your participation in this experiment! - Thanks

    I want to believe.
  • avatar
    68 sounds
    862 posts


    I was trying to take a file and import it as raw data in audacity. Thing is though, when I import it, there are several options (float, unfloat, 32 bit PCM, un-somethin 32 bit), stereo, endian something. With the different settings, the same file gets rendered completely differently. Why does this happen? (I mean is it just a difference in algorithm, and if that's the case, what's the most common setting used by y'all?)

  • avatar
    103 sounds
    1291 posts


    afleetingspeck wrote:
    I was trying to take a file and import it as raw data in audacity. Thing is though, when I import it, there are several options (float, unfloat, 32 bit PCM, un-somethin 32 bit), stereo, endian something. With the different settings, the same file gets rendered completely differently. Why does this happen? (I mean is it just a difference in algorithm, and if that's the case, what's the most common setting used by y'all?)

    My sounds for dare-26 were made using interpretation of raw input data as 44kHz/16bit/Stereo.

    In Audacity try : Signed 16bit PCM/Little endian/44100 Hz sampling frequency.

    But you can experiment, too smile

  • avatar
    246 sounds
    566 posts


    afleetingspeck wrote:
    I was trying to take a file and import it as raw data in audacity. Thing is though, when I import it, there are several options (float, unfloat, 32 bit PCM, un-somethin 32 bit), stereo, endian something. With the different settings, the same file gets rendered completely differently. Why does this happen? (I mean is it just a difference in algorithm, and if that's the case, what's the most common setting used by y'all?)
    Here's a rough attempt at scratching the surface and giving you an idea:
    Given a file that contains these bytes, whose value can range between 0 and 255 (they are so few that you get the typical cheap toy/8bit console sound),
    1 4 8 4 5 9 4 0 3 4 2 5 7 1 6 2
    if you treat it as 8 bit mono it gets played as is.

    In stereo the bytes will probably get shuffled in 2 tracks:
    1 8 5 4 3 2 7 6
    4 4 9 0 4 5 1 2

    Let's go back to mono for simplicity, but move to 16 bits, the bytes get grouped so that it takes 2 of them to represent a sample, and with 16384 different values for each sample, the sound gets definitely better:
    1&4 8&4 5&9 4&0 3&4 2&5 7&1 6&2

    Let's look at the first couple, 1 and 4; to see those two bytes as one bigger number, you either multiply the 1 by 256 and add 4, or multiply the 4 by 256 and add 1; these are the 2 endian modes, and you can imagine them yielding radically different values.
    Then you can scale it up to 32 bits... and then you have all the possible combos... and then you'll curse me for confusing you even more smile smile

  • avatar
    2067 sounds
    2243 posts


    afleetingspeck wrote:
    I was trying to take a file and import it as raw data in audacity. Thing is though, when I import it, there are several options (float, unfloat, 32 bit PCM, un-somethin 32 bit), stereo, endian something. With the different settings, the same file gets rendered completely differently. Why does this happen? (I mean is it just a difference in algorithm, and if that's the case, what's the most common setting used by y'all?)

    When a program loads a wav file there is a header on the file that specifies how the program should interpret the file.
    Wav is a standard format, so different programs will load and play a wave file in the same way.
    When you import a file as raw data, there is no header. So Audacity asks you to provide the info that would normally be on the header.
    The most important parameters are:
    1) number of chanels (usually mono or stereo, but could be more channels too)
    2) bit depth
    3) sample rate

    Bit depth gives the range of values each point (ie each individual sample can have).
    In a picture, it determines the number of different colours that the image format can represent (or how many shades of grey, in monochrome).
    In a sound it determines how many different levels exist between silence and max volume.
    4 bit depth is like cheap toys or very old computer games (ZX Spectrum and so on), 8 bit would be Commodore Amiga quality, old samplers and synths, 16 bit is CD quality.
    Note that sample rate also influences quality (when recording sounds)

    Sample rate is important when recording - determines the higher frequencies that can be adequately captured in the recording.
    In this case, we are converting data that already exists (so it has already been 'recorded') into sound. Sample rate merely controls the playback speed.
    High sample rate will play the sound faster and at higher pitch, lower sample rate will play the same sound slower and lower pitch. Just like playing the same record at different speeds.

    Of the the above, bith depth has the greater impact on the resulting sound. For example, importing a wav file as raw data at the wrong bit depth will detroy the sound completely.

    The endian parameter sometimes has an effect, sometimes doen not seem to have any effect. I do not know what it relates to.

    I want to believe.
  • avatar
    2067 sounds
    2243 posts


    copyc4t wrote:
    Let's look at the first couple, 1 and 4; to see those two bytes as one bigger number, you either multiply the 1 by 256 and add 4, or multiply the 4 by 256 and add 1; these are the 2 endian modes, and you can imagine them yielding radically different values.
    Then you can scale it up to 32 bits... and then you have all the possible combos... and then you'll curse me for confusing you even more smile smile

    copyc4t, as you may realize I was writting my post at the same time as you.
    Thanks for the explanation above.
    I guess the important aspect is "what is the byte size to begin with?" I am assuming it probably is 16-bit, so 32-bit would combine two 16-bit numbers in the way you describe. I have imported most of my sounds as mono, 16-bit, 44kHz - so maybe the endian selection only makes a difference for bit depth > 16? That would explain why I could not see a difference in most of the files I imported...

    On the other hand, you can also select "no endian"... How would that work?

    I want to believe.
  • avatar
    0 sounds
    1 post


    I'm hoping that someone can produce country gospel music using a washboard words are optional. If you can add a saw fiddle sound anda jug drum sound and a few other hillbilly musical sounds that would be the bomb.

  • avatar
    48 sounds
    10 posts


    AMBIENT v4.1 - updated!

    Descriptor: Unique standalone audio processing software capable of producing radically transformed audio, sound sculpting and design. Paint with sound ...

    http://www.audiobulb.com/Ambient.htm

    Platform: Mac Osx & PC Windows

    AMBIENT – v4 is a unique ambient soundscape generator capable of producing a vast array of ambient textures, from the bizarre to the beautiful. The module has already featured on numerous professional recordings demonstrating its versatility in bring inspiration to the creative process. AMBIENT processes any WAV or AIFF sound you care to load into it. The possibilities are endless.

    Once the artist has achieved the sculpted sound they desire they simply press record to output the new sound WAV – with the option of live recording any tweaks of changes they choose to make. Often the output WAV will be then imported into a DAW of the artists choice.

    AMBIENT 4.1 - next version enhancements

    Thank you for your support of the new software update. The response has been incredible! We've been listening to your feedback and have a version 4.1 for you.

    http://www.audiobulb.com/Ambient.htm

    Enhancements in 4.1 include improved GUI:

    ▪ Hover over dials causes dial name to appear and also value to show
    ▪ File name shows in bottom left when file is selected
    ▪ Dial colour changes to black and white when mapped to a midi controller, goes back to normal when mapping is removed.

    We hope you enjoy!

    www.audiobulb.com - exploratory electronic music