OG Xbox Optimal Timings (PAL/NTSC)

NewHome Forums OSSC, OSSC Pro and DExx-vd isl OSSC – Discussion and support OG Xbox Optimal Timings (PAL/NTSC)

Tagged: 

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #59372
    Sestain
    Participant

      This was made on 0.89/0.90 firmware (Still haven’t done the hardware modification for 1.0x firmware).
      I use RGB Scart cable and my xbox is 1.4 revision (PAL).

      Video LPF: Auto
      Allow TVP HPLL2x: Off [If this is on it might flip phase by 180 degrees]
      Allow upscample2x: On [If interlaced signal], Off [If progressive signal]
      Analog sync LPF: 2.5Mhz (max)
      TX mode: HDMI (RGB) [Most displays supports this nowadays]

      480p [NTSC-M / NTSC-J]:
      R/Pr offset: 128 / 128
      G/Y offset: 142 / 143
      B/Pb offset: 128 / 128
      R/Pr gain: 26 / 26
      G/Y gain: 20 / 18
      B/Pb gain: 26 / 26
      Pre-ADC Gain: 7 / 7

      Timings [640×480 / Normal]:
      H. samplerate: 762.65 [762.67] (or 762.70, 763 should be good too)
      H. synclen: 62
      H. backporch: 42
      H. active: 640
      V. synclen: 6
      V. backporch: 30
      V. active: 480
      Sampling phase: 180deg [adjust if you see crosstalk, best phase I found]

      Timings [720×480 / Widescreen]:
      H. samplerate: 858
      H. synclen: 62
      H. backporch: 55
      H. active: 720
      V. synclen: 6
      V. backporch: 30
      V. active: 480
      Sampling phase: 123deg [adjust if you see crosstalk, best phase I found]

      480i [NTSC-M / NTSC-J]:
      R/Pr offset: 125 / 143
      G/Y offset: 125 / 143
      B/Pb offset: 125 / 143
      R/Pr gain: 23 / 23
      G/Y gain: 23 / 23
      B/Pb gain: 18 / 18
      Pre-ADC Gain: 8 / 7

      Timings [640×480 / Normal]:
      H. samplerate: 762.65 [762.67] (or 762.70, 763 should be good too, I have it on 762.75)
      H. synclen: 62
      H. backporch: 44
      H. active: 640
      V. synclen: 3
      V. backporch: 16
      V. active: 240
      Sampling phase: 236deg [adjust if you see crosstalk, best phase I found]

      Timings [720×480 / Widescreen]:
      H. samplerate: 858.00
      H. synclen: 62
      H. backporch: 57
      H. active: 720
      V. synclen: 3
      V. backporch: 16
      V. active: 240
      Sampling phase: 157deg [adjust if you see crosstalk, best phase I found]

      576i [PAL]:
      R/Pr offset: 143
      G/Y offset: 143
      B/Pb offset: 143
      R/Pr gain: 28
      G/Y gain: 28
      B/Pb gain: 23
      Pre-ADC Gain: 7

      Timings [640×576 / Normal]:
      H. samplerate: 768.00
      H. synclen: 63
      H. backporch: 55
      H. active: 640
      V. synclen: 3
      V. backporch: 19
      V. active: 288
      Sampling phase: 292deg [adjust if you see crosstalk, best phase I found]

      Timings [720×576 / Widescreen]:
      H. samplerate: 864.00
      H. synclen: 63
      H. backporch: 70
      H. active: 720
      V. synclen: 3
      V. backporch: 19
      V. active: 288
      Sampling phase: 135deg [adjust if you see crosstalk, best phase I found]

      • This topic was modified 1 year, 7 months ago by Sestain.
      • This topic was modified 1 year, 7 months ago by Sestain.
      • This topic was modified 1 year, 7 months ago by Sestain.
      • This topic was modified 1 year, 7 months ago by Sestain.
      • This topic was modified 1 year, 7 months ago by Sestain.
      #59373
      Zacabeb
      Participant

        Good work, but your sample rate seems a little bit too high for 480i. 🙂

        For 480i/p, you can get a perfect match to the square pixel timings for 640×480 games with the following settings:

        https://junkerhq.net/xrgb/index.php?title=Xbox

        This is especially useful if disabling Upsample 2x to avoid moiré. It should however work perfectly with it enabled.

        #59374
        Sestain
        Participant

          Seems like I was doing something wrong the last time I checked my NTSC timings, thanks.

          #59437
          Dekaid
          Participant

            I only really have Outrun 2006 to test, but on my pal 1.4 og xbox (softmodded to run in ntsc) the timings in this thread are actually better, using a samplerate of 780 makes the image too wide and cuts off a bit, 763 seems close to the sharpest I can get too and fills 4:3 properly. Just to add, 720×480 properly fills the screen at a samplerate of 858 too for me.

            On my xbox 360 the samplerate of 780 does work correctly though, so maybe it’s something with the Focus Chip?

            #59439
            Zacabeb
            Participant

              That’s an interesting question. My timings that are at junkerhq are based on the Xcalibur video encoder and apply to the Xbox 360 ANA/HANA as well for 640×480. I also misremembered which revision was first to have the Xcalibur, so I didn’t realize at first we’re talking about the Focus Enhancements video encoder.

              When I went through the datasheets for the Conexant CX25871 and Focus Enhancements FS454 encoders from the earlier revisions of the Xbox, I noticed that their default timings are a bit strange and may not use the same 12.273 MHz square pixel mode. Both have scalers built-in and are designed to take a PC resolution and convert it to 480i/p or 576i/p, with passthrough available for progressive and HDTV modes.

              It’s possible that they neither of them, or at least the Focus Enhancements encoder, output square pixel mode natively when set up by the Xbox, but convert everything to BT.656 timings for 480i and 576i. This is suggested by your timings – if the 640 pixel raster is scaled up the 720 pixel active area by the encoder, then attempting to sample it as square pixel will require (640÷720)×858 samples total, resulting in 762.67 samples per line.

              Interestingly, both the Conexant and Focus Enhancements encoders handle the signal as YCbCr 4:2:2 internally and the Conexant encoder doesn’t output very high quality YPbPr or RGBS at standard definition because the signal passes through the lowpass filters needed for the CVBS signal. The Xcalibur, ANA, and HANA output YPbPr and RGBS with full color bandwidth. At 480p and HDTV resolutions they pretty much route the input as-is to the video DACs.

              Each encoder probably needs its own optimal timings. The optimal ones for the Conexant and Focus Enhancement encoders may be different between interlaced and progressive 640×480 rather than just double the clock as in the Xcalibur.

              • This reply was modified 1 year, 7 months ago by Zacabeb.
              #59465
              Sestain
              Participant

                I created a UnleashX skin that draws vertical lines and used that to get H. samplerate close to correct, later calculated correct, and then used the skin to set the sampling phase closer to optimal.

                #59471
                Zacabeb
                Participant

                  Nice work! I wonder what timings are used by the Focus Enhancements video encoder with 640×480 at 480p.

                  • This reply was modified 1 year, 7 months ago by Zacabeb.
                  #59473
                  Dekaid
                  Participant

                    Sadly I can’t test this as fractions don’t seem to be an option on the dexx_vd with 480p onwards, but thanks for the testing!

                    #59495
                    Sestain
                    Participant

                      480i and 480p uses same H. Samplerate but needs different phase (tested this with force 480p bios and the same scart cable)

                      • This reply was modified 1 year, 7 months ago by Sestain.
                      #59566
                      Sestain
                      Participant

                        It was obvious that PAL and NTSC has different colors but NTSC-M and NTSC-J has difference in colors too. Otherwise the timings are the same on both NTSC modes.

                      Viewing 10 posts - 1 through 10 (of 10 total)
                      • You must be logged in to reply to this topic.