|
Post by mirkosoft on Sept 25, 2015 9:36:22 GMT
Only to results: Don't look at image quality - it was captured by VGA instead HDMI, there is problem only one: colors. I will contact you Robert (Hydrophilic) for explaining me more 'cause you created intensity integration into converter, what maked from 64 interlaced colors real 136 colors. So, nice solutons, I tried so many converters that I have currently little video studio and can capture from SDTV up to HDTV and scale from/to 2K and everything capture with so many sources that I don't want more converters, only any additional solution.
Thank you all.
Miro
|
|
|
Post by mirkosoft on Sept 16, 2015 17:00:12 GMT
I'm sorry, I forgot to leave it on my domain... they're back!
Miro
|
|
|
Post by mirkosoft on Sept 15, 2015 19:40:06 GMT
Hi! I'm using few converters for VDC display, also C1084 monitor - but only for lightpen/gun input. I tried to adjust VDC colors as good as possible and result is not good extremely in case of dark yellow. Converter cable created Robert Willie (aka Hydrophilic) - so, you Robert maybe can help... Here are images: original result: result of VDC palette: VDC color test original (VICE) and result: and last is 256 color test: Can anybody explain me what to adjust? Thank you for each idea. Miro
|
|
|
Post by mirkosoft on Jun 20, 2015 9:02:10 GMT
Hi!
I'm still amazed how can CP/M monitor access to disk at screen...
I have dream to create something similar:
Something like: drive number: track, sector and by red write, by green read...
writting/reading to drive 8 on track 1 sector 2: <red/green> 08 01 02
Have anybody idea how to do it by fastest way?
Thank you for everything!
Miro
|
|
|
Post by mirkosoft on Jun 4, 2015 18:53:16 GMT
OK, only last post 'cause again hack like VIC2E Color RAM nibble width, summary:
1 MHz of 8502 handles operating & VIC2E screen 2 MHz of 8502 handles operating, VIC2E screen unhandled 4 MHz 2 cycles handles operating, 2 dead cycles can leave VIC2E handle screen 'cause 2 cycles @ 4 MHz = 1 cycle @ 2 MHz without dead cycle(s) Result: C128 can run @ 4 MHz on both screens - VIC2E & VDC but MIPS equals to 2 MHz, but when is VIC2E turned off and 2 dead cycles could be alive MIPS equals 4 MHz = true 4 MHz clock
Only mysteroius reason to use VIC chip in C128... in case marketting only one reason: sell as many as possible
Miro
|
|
|
Post by mirkosoft on Jun 3, 2015 15:11:46 GMT
OK, now other Q:
When Z80 runs @ 4 MHz and has 2 dead cycles, is it possible to make them alive?
Really C128 @ 4 MHz (I know that Z80) can be really fantastic and no matter if VIC2E can handle it, my favorite videochip is VDC - its variability is even more than EGA!
Miro
|
|
|
Post by mirkosoft on May 29, 2015 11:48:02 GMT
Hi!
Really wrong decision of Commodore 128 designers and really only for compatibility with C64 is VIC2E chip as one of two display chips. Meaning that VIC2E can work only at 1MHz, or by IRQ can be system speed up to 1,3MHz approximatelly... When we look at TED or VDC they both are better chips and are not limited by 1MHz. TED - 121 Colors and flash attribute with resolution up to 320×496 at ~1,76MHz and VDC with separate VideoRAM, resolution up to known working 800×600 and color cells variability incl. 4 attributes flash - reverse - underline and alternate...
But can anybody eplain me this behavior of VIC2E when is Z80 active:
In case of active MOS 8502 @ 2MHz VIC cannot handle screen display.
In case of active Zilog Z80 @ 2MHz VIC can handle screen display.
What it means?
Z80 is faster - works @ 2MHz.
Does it its downclocking from 4MHz to 2MHz?
Or handles VIC screen with downclocking even to 1MHz?
Or VIC can work @ 2MHz an maybe even @ true 4MHz Z80 clock?
Can anybody explain this?
Thank you for suggestions.
Miro
|
|
|
Post by mirkosoft on May 23, 2015 18:51:37 GMT
Hi!
I was investigating Robert Willie's (aka Hydrophilic) explaining twice availability of VIC2E Color RAM...
Really it needs to write that in Basic is it impossible but few POKE commands I tried.
VIC and CPU can access areas separatelly
$01 / bit 0 = 0 (CPU access Color RAM 0 / mc-bitmap)
$01 / bit 0 = 1 (CPU access Color RAM 1 / text)
$01 / bit 1 = 0 (VIC access Color RAM 0 / mc-bitmap)
$01 / bit 1 = 1 (VIC access Color RAM 1 / text)
Here are POKE commands to use all choices:
00 = poke1,peek(1)and252
01 = poke1,(peek(1)and252)or1
10 = poke1,(peek(1)and252)or2
11 = poke1,(peek(1)and252)or3
It's all about VIC Color RAM which is not available in banks 0 and 1. But really not true VIC area, really CPU RAM is VIC DATA RAM - better to say default 1024 - 2023 is possible use in both banks by simple switch:
pokedec("d506"),peek(dec("d506"))and191:rem vic ram in bank 0
pokedec("d506"),peek(dec("d506"))or64:rem vic ram in bank 1
Really in Basic is it not possible to use 'cause Kernal controls all, even also VIC registers have their shadows... In assembler is it perfect like buffer, for example: before any change in text editing you save data to bank 1, type anything, or scrolling will erase lines above, you can press keystroke, e.g. CTRL+Z and it move buffer 1024-2023 from bank 1 to bank 0 back and it looks really like UNDO...
I don't know if it is explained understandeable, if not please correct me Robert...
Miro
|
|
|
Post by mirkosoft on May 23, 2015 14:21:14 GMT
OK, Robert.
Here's main part of Booter last revision code. It returns always if is file found or not, bell tone like to ML monitor jump and if is any program executed it always ends in ML monitor with bell tone.... (at mark // >>>> HERE<<<< in code)
.pc = $1800 "BOOTER CODE" // version 1.1.0 revision 1 Build.X2015.0412-83@1857
lda #00
sta $ff00
lda #$73
sta $01
lda #$2f
sta $00
jsr $ff8a // Restore vectors
jsr $ff84 // Init I/O Devices, Ports & Timers
jsr $ff81 // Init Editor & Video Chips
jsr default
start: jsr primm
.text "mPRESS SPACE FOR OPTIONSmESC TO CANCEL SCAN"
.byte 0
lda #$00 // turn off messages
sta $ff00
jsr setmsg
ldy #00
ldx #00
get: lda 212
cmp #88
bne got
inx
bne get
iny
cpy #50
bne get
loop: lda 212
got: cmp #88
beq nokey
cmp #72
beq cancel
cmp #60
bne nokey
jmp menu
nokey: lda #08 // length of name
ldx #<filename
ldy #>filename
jsr setnam
lda #00 // memory bank
ldx #00 // filename bank
jsr setbnk
lda #00 // logical filenumber
ldx $fb // device
ldy #00 // secondary address ($00 uses load address)
jsr setlfs
lda #00 // 0 - load / 1 - verify
ldx #$00 // LO-byte load address
ldy #$0e // HI-byte load addres
jsr load // returns A = error code
bcc file_loaded
not_found:
jsr clall
jsr reset_drive
lda $fc
beq norm
dec $fb
lda $fb
cmp #07
bne loop
jmp err
norm: inc $fb
lda $fb
cmp #31
bne loop
cancel:
err: lda #$c0 // turn on messages
jsr setmsg
lda #00
sta keybuf
ldx $fd
jsr default
lda #$30
jsr pom
rts
// >>>> HERE <<<<
file_loaded:
lda $0e00 // version of BOOT.CFG
cmp #$10 // compare to BOOTER version
bmi err
lda $0e04 // type
bne next
jsr addr_corr
next:
lda $0e05 // length of name
ldx #$06
ldy #$0e
jsr setnam
lda #00 // memory bank
ldx #00 // filename bank
jsr setbnk
lda #00 // logical filenumber
ldx $0e03 // device
ldy #00 // secondary address ($00 uses load address)
jsr setlfs
lda #00 // 0 - load / 1 - verify
ldx $0e01 // LO-byte load address
ldy $0e02 // HI-byte load addres
jsr load // returns A = error code
bcc ok
jmp err
ok: lda #$c0 // turn on messages
jsr setmsg
lda $0e04
cmp #02 // Data type
bne skip
jmp err // no action jumps to ERR and causes the same skip: cmp #01 // ML type
bne skip2
jsr default
jmp ($0e01)
// also >>>> HERE <<<<
skip2: jsr default
jmp run // Basic type
// also >>>> HERE <<<<
reset_drive:
lda #2
ldx #<reset
ldy #>reset
jsr setnam
lda #00
jsr setbnk
lda #01
ldx $fb
ldy #15
jsr setlfs
jsr open
lda #01
jmp close
addr_corr:
lda $76
bne alloc
lda #$01
ldx #$1c
store: sta $0e01
stx $0e02
rts
alloc: lda #$01
ldx #$40
jmp storeinsert code here Miro
|
|
|
Post by mirkosoft on May 18, 2015 17:01:35 GMT
Hi! I was discovering lightpen use in The Ace and first steps were in simple on-screen keyboard create to type text. I finished little fun demo, where on your VIC screen you can move cursor, type and delete notes. It has no save option, only X button as exit... In other thread I was asking about lightpen emulation in VICE, no answer, so when I finished "LightPad 40" I tried to use it in VICE and success... I own Turbo Computer lightpen and Defender 64 lightgun... So, I enabled lightpen emulation and selected Datel pen. My demo has also Button check sub-program... and here is Q to all who can/own lightpen: I you can please download Lightpad 40 And test your lightpen with Button check and leave reply of received data. Emulation is simple: Datel Pen - move - press/hold mouse left button - push button - press mouse right button Inkwell Pen - move - move mouse over the screen - push button - press mouse left button - button check of Inkwell Pen is same as my Turbo Computer pen: keypress (212) = 90 [this value has no one key of keyboard] modifier (211) = 4 [Control ] Control port 1 = 251 [direction left] Control port 2 = 127 [no move 'cause lightpen is in port 1] JOY(1) Basic command = 7 [direction left] Commonly are values these (existing exceptions I want to know) PEEK(212) = 90 - keypress (this value has no one key of keyboard) PEEK(56321) = 251 - joyport 1 left JOY(1) = 7 - Basic command joyport left Lightpad 40 is simple Basic 7 program... Next task = LigtPad 80 :) Miro lightpad40.prg (1.99 KB)
|
|