A89: Re: Re: Re: Grayscale troubles

[Prev][Next][Index][Thread]

A89: Re: Re: Re: Grayscale troubles


First off, Sebastian and James: please do not send HTML-formatted messages
to this list.  Not only is it annoying and highly disruptive to the digest
form of this list, but it is against ticalc.org's list policy.

>   HW2 is something I hadn't considered, what tricks will be necessary to =
> make this routine work on HW2 calculators?  What is it about HW2 =
> calculators that makes grayscale not work?  Is the LCD address port =
> moved?

No, the LCD address port does not EXIST.  That's why it took so long for me
to even get a quick hack working that provided grayscale on HW2 calcs.

It sounds like James has a good idea, but it is simply not feasible on a HW2
calc unless you can get an interrupt synched with the screen refresh (and I
don't even think we know that speed).

> TI didn't release some timings for HW2, if I'm right.

TI didn't release ANY timings for HW2.  Nor did they release any for HW1,
IIRC.  Unless you're refering to the chip speed, which is 10 MHz on HW1 and
12 MHz on HW2.

> That way it will =
> flicker a lot more, because there is no way to completely synchronize it =
> with the LCD.

That is the major issue.  With the LCD address port, you can assure that the
actual transition between two frames will not overlap the screen refresh,
since it is instantaneous.  Since you have no such port on HW2, you would
need to synch the code with the screen refresh.  The only way I can imagine
to do that would be to adjust the speed of int 5, and by then you have a
huge waste of effort, if it can possibly be pulled off at all.

> I don't have an HW2 calculator, but Zeljko Juric says in =
> the documentation of his library that the "flicker bug" has been =
> corrected.  Maybe you could take a look at his grayscale routines.  The =
> file is gray.c, but they are written in ASM.

Whether the flicker has been corrected is a matter of opinion and open to
interpretation.  Having written the original HW2 grayscale code and being a
HW2 owner (since my HW1 was stolen), I can tell you that the amount of
flicker is, in my opinion, certainly NOT acceptable compared to what it
could be.  I've proposed some changes before on this list, although I
believe Zeljko rejected most of them on the basis that his testers said it
was sufficient.

Zeljko, take my word for it - there's a very prominent level of flicker even
when nothing is happening, such as TI-Chess' idle periods.  And it's not
just me; when I got my HW2 and would play TI-Chess in school everyone
commented that the flicker was annoying.  I'll rewrite the code myself, but
I don't think that the current code is satisfactory. . .

    -Scott
Follow-Ups: References: