Wide Playfield

[TGB1718]

I'm currently writing a little demo, using modes 9 and 11

Currently my screen is 64 bytes wide with an LMS for every line and I can coarse scroll

about the top half with a joypad so I can see results.

 

However if I set the playfield to WIDE using SDMCTL

 

    lda SDMCTL ; try wide playfiled
    ora #3
    sta SDMCTL

the output is somewhat garbled, I know when you set the playfield to wide you need extra bytes (8 if I remember correctly)

but I thought as I had an LMS on every line it would be ok.

 

I did set the playfield 8 bits wider just to see if it made any difference, but it didn't, am I missing something else I need to

do to use a wide playfield ?

 

thanks in advance

[Wrathchild]

Garbled in what way? Have you zero'd out the full display RAM?

You say 'course scrolling' so presumably you are not setting scroll bits in the DL instructions?

 

Other possibility is that your screen RAM at 192x64 would be 12K, so you cross a 4K boundary twice, e.g. at offsets $1000 and $2000.

Within a line you cannot span that boundary or the data retrieve will loop back to the start of the 4K. 

[TGB1718]

I zero the screen RAM, been playing with the routines and got ride of the garbage, little hiccup on

loading the pixel colour, however still seeing a little problem, the video is NORMAL screen, when I set

to wide screen it looks like the first few box edges are missing from the left of the screen and the left side of the screen is 

at the far right of the screen, all I did was set WIDE, do I have to adjust for the extra 8 bytes of a WIDE playfield. ?

 

 

[Wrathchild]

How do you calculate when the right scrolling stops? Maybe if you take 8 from that then you'll be ok?

[TGB1718]

19 minutes ago, Wrathchild said:

How do you calculate when the right scrolling stops? Maybe if you take 8 from that then you'll be ok?

Just limit the coarse scroll to 24 bytes, normal screen 40 + extra 24 bytes to 64.

Also doesn't explain why the first few vertical plots are missing from the left side of the screen.

 

Think I'll adjust it again to compensate for the extra 8 bytes of a wide playfield as I've sorted the

garbage on the screen, maybe it will fix it

 

So far quite happy with the normal screen, I have a plot routine that will use a 128 pixel wide x 192 height

display in modes 9 and 11 and have a 'draw box' routine also working

 

1 hour ago, Wrathchild said:

Within a line you cannot span that boundary or the data retrieve will loop back to the start of the 4K

I think think I've hit that at about line 64, but above that the display should be ok, actually I've just looked and all

I get is one black line through the colours, the display starts at $4000 so I think I will have to find a nice screen width

that allows the LMS instructions to fall on the right boundaries, thanks for reminding me about that limitation.

[Wrathchild]

1 hour ago, TGB1718 said:

Also doesn't explain why the first few vertical plots are missing from the left side of the screen.

If you think of a gr.0 screen you'd get 4 chars extra per side of the screen, IIRC these don't all fit (and hence show) on a normal TV and help in terms of scrolling things in/off the screen more seamlessly.

1 hour ago, TGB1718 said:

Just limit the coarse scroll to 24 bytes, normal screen 40 + extra 24 bytes to 64.

So if the screen is 48 wide your limit is now 64-48 = 16

 

1 hour ago, TGB1718 said:

the display starts at $4000 so I think I will have to find a nice screen width

that allows the LMS instructions to fall on the right boundaries

You're ok as you were as each line is $40 and multiples will align to a $1000 boundary.

The above course limit will stop a row at, say, $4FC0 from wrapping.

Edited January 14, 2022 by Wrathchild

[Rybags]

Altirra can show stuff that a TV cuts off - the typical maximum is about an extra 2.5 characters each side of the standard 40.

 

In wide mode you have 4 bytes extra on the left that will be partially obscured and 4 on the right.

The bus noise you see at the extreme right will never be visible on most TVs.  Even a modern 16:9 set to 4:3 AR will cut off the display edges.

[MaPa]

On the left side ATARI in wide shows only 2 characters more then in normal (no time for DMA reads).

[TGB1718]

Thanks for the replies, I'm going to complete the full screen scroll in normal first then

have a play with what I need to do to correct the screen in wide mode.

 

[Rybags]

If you have fine scrolling enabled, normal essentially becomes wide anyway, except the borders remain - the thing to note of course is the character offset at HSCROL=0 will hide those first 4 characters.

[pps]

That garbage on the right hand is the 48th of the chars, so in reality the ATARI can show a maximum of 47 chars plus one that is half way garbaged. On good old TV-stes you can ignore that, as they can't display the whole extended overscan mode ALTIRRA opts to you. So widescreen option is more to what they can show. Extended view is possible on modern TV-sets. So with one of it you can see the right border garbage on real ATARI, too.

[TGB1718]

2 hours ago, Rybags said:

If you have fine scrolling enabled

Only doing coarse scroll, I tried enabling fine scrolling on something different, but even without

scrolling anything, just create a display list with hscroll enabled and it seems to mess the display up.

Remove the hscroll bit and everything is fine (except I can't scroll  )

 

Not sure if this is the bug mentioned in the Altirra hardware manual (I'm only using graphics mode 9 and 11).

54 minutes ago, pps said:

That garbage on the right hand is the 48th of the chars,

Not worried about that, I see it a lot of the time, I realise why it's there, thanks

[pps]

I see now, I misunderstood what your problem was