#12 slight timing problem with overscan

closed-fixed
None
5
2008-06-01
2008-04-15
Georges Kesseler
No

Hi

this particular demo (low rez ST) has a flickering issue because of timing being slightly off.

http://www.cpu.lu/ulm/OSZI.PRG

source code is available if needed.
Georges

Discussion

  • Logged In: YES
    user_id=207643
    Originator: YES

    might be the same issue on the 3rd screen in this demo.
    http://www.cpu.lu/ulm/demoarchive/ulm_ny.msa
    One instruction has a too long cycle count.

     
  • Logged In: YES
    user_id=1995833
    Originator: NO

    This was fixed in current CVS.
    The confusion was from the following code, which does consecutive switch to 60 Hz before going back to 50 Hz :

    00014c74: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c7a: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c80: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c86: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c8c: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c92: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c98: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c9e: 11fc 0000 820a 11fc 0002 MOVE.B #$00,$ffff820a
    00014ca4: 11fc 0002 820a 201f 4238 MOVE.B #$02,$ffff820a

    In that case, only the last switch to 60 Hz was used to compute border timings, which was not correct.

    Nicolas

     
  • Logged In: YES
    user_id=1995833
    Originator: NO

    This was fixed in current CVS.
    The confusion was from the following code, which does consecutive switch to 60 Hz before going back to 50 Hz :

    00014c74: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c7a: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c80: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c86: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c8c: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c92: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c98: 11fc 0000 820a 11fc 0000 MOVE.B #$00,$ffff820a
    00014c9e: 11fc 0000 820a 11fc 0002 MOVE.B #$00,$ffff820a
    00014ca4: 11fc 0002 820a 201f 4238 MOVE.B #$02,$ffff820a

    In that case, only the last switch to 60 Hz was used to compute border timings, which was not correct.

    Nicolas

     
  • Thomas Huth
    Thomas Huth
    2008-06-01

    • assigned_to: nobody --> npomarede
    • status: open --> closed-fixed
     
  • Thomas Huth
    Thomas Huth
    2008-06-01

    Logged In: YES
    user_id=130746
    Originator: NO

    OSZI.PRG is working stable now.