NTSC variable-length scanlines

[TheDarkCoder]

the HRM says that, while in PAL all lines last for 227 "color clocks", in NTSC scanlines alternate between "short" (227 clocks) and "long" (228).
My understanding is that on long line the horizontal posizion read via VHPOSR reaches the value $e2 in short lines and the value $e3 in long. Is it right?

Now I have 2 questions for which I can't find answer in HRM.
1) Which lines are short and which are long?

2) What happens in "programmable scan rate modes"? Are all lines the same lenght as in PAL, or do they alternate as in NTSC, or is it configurable by some bit in BEAMCON0?

[TheDarkCoder]

Originally Posted by TheDarkCoder

the HRM says that, while in PAL all lines last for 227 "color clocks", in NTSC scanlines alternate between "short" (227 clocks) and "long" (228).
My understanding is that on long line the horizontal posizion read via VHPOSR reaches the value $e2 in short lines and the value $e3 in long. Is it right?

Now I have 2 questions for which I can't find answer in HRM.
1) Which lines are short and which are long?

2) What happens in "programmable scan rate modes"? Are all lines the same lenght as in PAL, or do they alternate as in NTSC, or is it configurable by some bit in BEAMCON0?

I think to have found the answer for question 2: it should be configurable.
The HRM does describe the programmable scan rate modes very concisely, in particular BEAMCON0, so one has to interpret a bit. However the description of bit 11 of BEAMCON0 is "Disable long line/short toggle. This is useful for DUAL mode where even multiples are wanted, or in any single display where this toggling is not desired."

I think this what I have asked.

In the AGA guide I also found a partial answer to question 1.
The description of VPOSR given in the aga.guide is:
VPOSR p 004 R A Read vert most sig. bits (and frame flop) VPOSW 02A W A Write most sig. bits (and frame flop) BIT# 15 14 13 12 11 10 09 08 07 06 05 04 03 02 01 00 USE LOF I6 I5 I4 I3 I2 I1 I0 LOL -- -- -- -- -- V10 V9 V8 LOF = Long frame(auto toggle control bit in @{"BPLCON0" link BPLCON0}) I0-I6 Chip identitication: 8361 (Regular) or 8370 (Fat) (Agnus-ntsc) = 10 8367 (Pal) or 8371 (Fat-Pal) (Agnus-pal) = 00 8372 (Fat-hr) (agnushr),thru rev4 = 20 Pal, 30 NTSC 8372 (Fat-hr) (agnushr),rev 5 = 22 Pal, 31 NTSC 8374 (Alice) thru rev 2 = 22 Pal, 32 NTSC 8374 (Alice) rev 3 thru rev 4 = 23 Pal, 33 NTSC LOL = Long line bit. When low, it indicates short raster line. v9,10 -- hires chips only (20,30 identifiers)
So, the LOL bit is what I was looking for. But:
1) from the above description it is not clear which is the LOL bit:
number 7? 8?
2) this bit is not mentioned in the HRM. Is it a new feature of AGA,
or simply an undocumented one? It sounds strange that on OCS and ECS
there was no way to tell if a line is short or long.

[Toni Wilen]

Yes, BEAMCON0 bit 11 is LOLDIS (disable NTSC short/long toggle). VPOS LOL-bit (bit 7) can be used to set/get current line's "length". (Only works in NTSC mode, always reads as zero in PAL mode)

"LOL" is ECS Agnus feature, documented in 3rd edition HRM ECS chapter.

How to confirm/test NTSC line toggle easily:

Create copper list like this:

wait vpos x, hpos close to right border (for example 0xDF)
<lots of dummy moves here, at least 20 or so, skipping horizontal blank>
move to background color (some visible color)
move to background color (clear)
repeat above copper code (vpos+1) to see alternate line toggle

Now change LOLDIS and LOF bits to see how the pattern changes.

[TheDarkCoder]

Originally Posted by Toni Wilen

Yes, BEAMCON0 bit 11 is LOLDIS (disable NTSC short/long toggle). VPOS LOL-bit (bit 7) can be used to set/get current line's "length". (Only works in NTSC mode, always reads as zero in PAL mode)

"LOL" is ECS Agnus feature, documented in 3rd edition HRM ECS chapter.

How to confirm/test NTSC line toggle easily:

Create copper list like this:

wait vpos x, hpos close to right border (for example 0xDF)
<lots of dummy moves here, at least 20 or so, skipping horizontal blank>
move to background color (some visible color)
move to background color (clear)
repeat above copper code (vpos+1) to see alternate line toggle

Now change LOLDIS and LOF bits to see how the pattern changes.

Thanks a lot Toni, I had checked 3rd edition HRM but missed the LOL description. Didn't imagine it was also settable