Date:  03-Nov-95 14:14 CET
From:  Christian Bauer > INTERNET:bauec002@goofy.zdv.Uni-Mainz.de
Reply to: Re: $3fff and sprite priorities

Sender: c64-hackers-owner@nic.funet.fi
Received: from nic.funet.fi by dub-img-2.compuserve.com (8.6.10/5.950515)
        id IAA00301; Fri, 3 Nov 1995 08:11:12 -0500
Received: from funet.fi ([130.230.1.1]) by nic.funet.fi with ESMTP id <90044-4>;
Fri, 3 Nov 1995 15:08:04 +0200
Received: from trudi.zdv.Uni-Mainz.DE (actually user
root@trudi.zdv.Uni-Mainz.DE) 
          by funet.fi with SMTP (PP); Fri, 3 Nov 1995 14:58:19 +0200
Received: from [134.93.174.20] (dialin20.zdv.Uni-Mainz.DE [134.93.174.20]) 
          by trudi.zdv.Uni-Mainz.DE (8.6.12/8.6.12) with SMTP id NAA01350 
          for <c64-hackers@nic.funet.fi>; Fri, 3 Nov 1995 13:58:04 +0100
Date:         Fri, 3 Nov 1995 13:58:04 +0100
X-Sender: bauec002@pop.uni-mainz.de
Message-Id: <v01510102acbfcfa22bdc@[134.93.174.20]>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: 64 hackers <c64-hackers@nic.funet.fi>
From: bauec002@goofy.zdv.Uni-Mainz.de (Christian Bauer)
Subject: Re: $3fff and sprite priorities

Hi!

>-Pasi
>(Oh, and I am one of the first people on the list to actually
> send a message.. <g> Nothing to ask or share ?)

Thanks, Pasi! I almost forgot that I have subscribed to this list... :-)
Now, here's something I want to share:

We know that the VIC has three graphics modes that only result in a black
screen: ECM+MCM, ECM+BMM and ECM+BMM+MCM. But in all these modes the display
data sequencer generates valid output that sprites can collide with! I have
included a small "scanner" program below that uses a single-point sprite
to scan an invalid mode and records the results into a bitmap at $e000 that
is shown afterwards. With this, you can get an impression of what the invalid
modes would look like if the VIC designers hadn't decided to disable the
video output. Of course, with sprite collisions you can only tell apart
foreground and background pixels, all color information is lost.

The ECM+MCM, ECM+BMM and ECM+BMM+MCM modes look very much like multicolor text,
standard bitmap and multicolor bitmap modes, respectively, but the ECM bit
causes
bits 9 and 10 of the video address bus to be low during chargen/bitmap accesses.
As a result, the character set in ECM+MCM is limited to 64 chars (as in normal
ECM mode) and the two invalid bitmap modes show some repetition in the graphics
data (hard to explain by words, have a look at it yourself).

Enough talk, here is the program (start address $c000). It is assembled for
scanning the ECM+BMM+MCM mode. If you want to analyze other modes, you have to
change the program bytes at $c03d and $c042:

begin 644 scanner
M`,"I?XT-W*GIC10#J<"-%0.I`*J=0`/HX#_0^*F`C4`#J0V-^`>I`8T5T*D`D
MC1?0C1O0C1S0C1W0J0"-$M"I>XT1T*D8C1;0J0&-&M"I`(VSP8VTP8VUP2"7$
MP:T?T*T1T!#[K1'0,/NM$M#)(-#Y(/7`KK3!T`SNL\'0(^B.M,%,G<#NL\&M\
ML\')0-`2J0"-L\&-M,'NM<&MM<')R/`0()?!K1'0$/NM$=`P^TQFP*D_C0+=S
MK0#=*?R-`-VI2(T8T*D[C1'0J0B-%M!XJ32%`:D0H@"=`-"=`-&=`-*=`-/H-
MT/&I-X4!6$SFP,X9T*T?T(VRP4PQZJVUP4I*2@JHN5W!C;?!N5[!C;C!K;7![
M*0<8;;?!C;?!K;/!*?B-ML$8J0!MM\&%_JG@;;C!A?\8I?YMML&%_J7_;;3!\
MA?^ML\$I!ZJI-(4!O8_!H`"NLL'0!4G_,?XL$?Z1_JDWA0%@``!``8`"P`,`2
M!4`&@`?`"``*0`N`#,`-``]`$(`1P!(`%$`5@!;`%P`90!J`&\`<`!Z`0"`08
F"`0"`:VUP1AI,HT!T*VSP1AI&(T`T*VTP6D`C1#08``````````<.
``
end
size 443



Bye,
Christian

--
  / Christian Bauer, Langenaustr. 65, 56070 Koblenz, Germany
\/ WWW homepage: http://www.uni-mainz.de/~bauec002/




Distribution:

To: 64 hackers > INTERNET:C64-HACKERS@NIC.FUNET.FI

