Calculating H samplerate.

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • #45670
    Achelon
    Participant

    Perhaps a stupid question…

    Could someone help me calculate the correct H sample-rate for a device?

    I searched the forum and it seems that the calculation is something like the pixel clock of the source divided by the line rate = the optimal number of samples per line.

    The pixel clock of my source is 6.650428 MHz (PAL). How do plug this into the calculation?

    Regards,
    Achelon

    #45671
    Harrumph
    Participant

    Pixel clock / horizontal freq = sample rate
    Horizontal freq = vertical refresh x scanlines

    You can read both horiz freq, vert freq and line count from OSSC info window.
    Using example of common PAL 288p linecount and 50hz refresh, 6650428/(50×312) = 426.3
    Please replace with exact values for true figure.

    #45681
    Achelon
    Participant

    Thanks for that. I could do with some more technical advice, but I’m not sure you can help to be honest.

    I am using the OSSC in Line 2x (Generic 4:3) mode with a PAL 288p linecount 50hz source and an unusual artefact is visible on the TV (Panasonic TX58DX902B) screen. Essentially there is some kind of vertical banding visible on the screen which I have determined is not caused by brightness variations in the panel itself. I have also ruled out the panel’s FALD feature as a cause.

    This evidently has something to do with the samplerate as the frequency of artefact changes as I alter it. Do you have any advice? See screenshots.

    Example 1
    Example 2

    #45693
    b0nz0beavis
    Participant

    Have you tried adjusting sample rate fraction, sampling phase and changing the aspect ratio of the display? There’s also a ‘panasonic hack’ under compatibility settings, not sure what it does but could be worth a shot.

    #45696
    Harrumph
    Participant

    If it’s the same source as you mention in OP, then in 2x generic set h.samplerate to twice the value you got in original calculation. And you may well need to adjust sample phase.
    Also don’t forget that your TVs scaling may do some funny things, especially with such an unnatural image as a pixel grid. Set your display to unscaled if possible to compare.

    #45706
    Achelon
    Participant

    Panasonic hack = on = no picture on the Panasonic!

    The effect is least pronounced at a H samplerate of 854. Going below 856 requires setting an HDMI VIC of 10! What a puzzler. I would love to test VIC > 31 but this does not seem allowed by OSSC firmware – which is a shame.

    I am coming to the conclusion that while this TV is very good for Film content it is terrible for gaming (not least because the motion blurring is severe). I ran some tests using my Anybeam laser projector and I think that its laser scanning raster system seems to more closer emulate a CRT and this may be the real answer.

Viewing 6 posts - 1 through 6 (of 6 total)
  • You must be logged in to reply to this topic.