On 2017-11-04 18:56, smf wrote: >> ...AFAIK, not even the Amiga is fully correct in this respect... it still >> produces scanlines of integer multiple lengths of the (respective) >> standard color subcarrier cycle. ... > > That is true in non interlace, but interlaced should be standard. > Obviously the c64 couldn't do interlace, while the Amiga can. I don't > think the c64s problems ended at interlace though (especially the early > NTSC 64 clock VIC2). Reading some sources up and thinking it over again, it looks like I was incorrect here... First, no, the problem I was referring to has been a different matter from interlacing (although that one also applies). These color standards usually require an odd multiple relation between half (or, in the case of PAL, the quarter) line rate and color subcarrier frequency. This (amongst other, more serious issues) is usually neglected in early microcomputer video designs. However, speaking of the Amiga, The Amiga Hardware Reference Manual ( http://ada.evergreen.edu/~tc_nik/files/AmigaHardRefManual.pdf ) states, that NTSC mode produces scanlines of 227.5 cycles. Now, NTSC requires the color subcarrier be 455/2 times the line frequency... a.k.a 227.5 color clock cycles per line. This is a perfect match. For PAL, as the manual describes, the line length produced by the video chip is constant 227 cycles, where the base is 4/5 of standard PAL color subcarrier frequency. In PAL, the color subcarrier is required to be 1135/4 times the line frequency + 25Hz. Now, 227*5/4 = 1135. A very small offset aside (--> line rate is divided from the color subcarrier frequency, not the color subcarrier frequency minus 25Hz), this is also a perfect match. Levente Message was sent through the cbm-hackers mailing listReceived on 2017-11-05 11:02:04
Archive generated by hypermail 2.2.0.