LMS 240 scanlines starting adress question

[Heaven/TQA]

when I am using a fullscreen (240 scanlines) antic e screen where normal 40 byte mode, where should the screen ram start to avoid the "4k boundary" issue? ($6140, $6010 the standard ones are working in 48 byte mode)

[analmux]

You cannot avoid it without a second LMS and reprogramming the screen driver you're using:

240 scanlines * 40 bytes (per scanline) = 9600 bytes = 2*4096+1408 bytes.

Thus, it uses (parts of) at least 3 blocks of 4kB, so there's at least 2 boundaries of 4kB.

The middle one of the 3 blocks is always used. This can't be divided into an integer number of 40 byte blocks...etc.

[Heaven/TQA]

You cannot avoid it without a second LMS and reprogramming the screen driver you're using:

240 scanlines * 40 bytes (per scanline) = 9600 bytes = 2*4096+1408 bytes.

Thus, it uses (parts of) at least 3 blocks of 4kB, so there's at least 2 boundaries of 4kB.

The middle one of the 3 blocks is always used. This can't be divided into an integer number of 40 byte blocks...etc.

 

 

Well forgotten to say that i am using each scanline an lms command. And it Works with $6020 f.e.

 

[Rybags]

You could use LMS and skip 64 bytes per scanline, but that's a bit wasteful even in Wide (48 byte) mode that's 3840 bytes not being used.

 

Easier way might be to just use table-lookup when addressing the screen RAM, then you can put the skipped screen RAM wherever you want (obviously has to be part of the middle 4K chunk being used).

[MaPa]

Simply you can't have continuous video ram if you have 40 bytes screen width and >8kb video ram.