[A83] Re: ZPic83 compression
[Prev][Next][Index][Thread]
[A83] Re: ZPic83 compression
>
> > > > Writing both is a good idea to check sizes. but a vertical one
would
> > > > also be much larger..
> > >
> > > I don't see why thay would be..
> >
> > Because the bytes are read left-to-right. If you do a bit-level vertical
> > compression, you'll have to read the 1st bit of the 1st byte, then the
1st
> > bit of the 13th byte, then the 1st bit of the 25th byte, etc.
>
> Ah, a lot more traffic from the lcd driver..
>
>
Noooooo.... the data is not read from the LCD driver at all. It is read
directly from RAM.. the point is that it is much easier and faster to manage
the data the way that the calc was built to (in consecutive blocks of 8
bits). With horizontal compression, you can read each byte the normal way
because all the bytes in the picture are consecutive horizontally. So to
read 8 bits, you just do one ld a,(hl) or something. With vertical
compression, to build a byte of the picture, you must read a single bit
from 8 bytes. That is 8 reads with some masking/rotating to get each bit.
So basically when compressing pictures that are stored in the conventional
method of 12 columns, 64 rows of bytes, vertical compression requires a lot
more overhead.
(If none of that makes sense, It's because I have a killer headache..)
- - Joe
joe@joepnet.com
References: