CRTC³ a new CPC+ cartridge demo

[roudoudou]

Press c i t r o n during the boot.
To see the crocodile after the amiga part press CTRL F9
This wont work with winape of course

[gerald]

Press c i t r o n during the boot.
To see the crocodile after the amiga part press CTRL F9
This wont work with winape of course

Nor on a 128k GX4000 

[roudoudou]

Nor on a 128k GX4000 

You will need a true CPC+ memory card (except RAM7 memory expansion, i don't know another one...)
EDIT: Oups, i miss the keyboard problem 

[Maniac]

Press c i t r o n during the boot.
To see the crocodile after the amiga part press CTRL F9
This wont work with winape of course

Nice bonus content! Thanks for the tip! [emoji6]

[DanyPPC]

Thanks

[roudoudou]

Hey
Eventually the final version is out!
Several bugfixes in order to keep a proper 50Hz at transitions, fix some visual glitches
A new screen conversion for the Serval, more colors, better palette choice, better dithering
And the final Wasp is now truly interactive

EDIT: false alarm, there is still bugs on a real machine

[roudoudou]

one week later, bugs are fixed, pfiouuuuuuuuu

[dthrone]

Out of interest, what sort of bugs were you encountering with a real machine?

[roudoudou]

Out of interest, what sort of bugs were you encountering with a real machine?

music corruption, wrong keyboard inputs, frozen interrupts, ...

[dthrone]

I'm having issues at the moment with DMA music being corrupted by particular CRTC settings on real machine.  No problems on Winape, Arnold etc. emulators.  Any info you've garnered appreciated

[roudoudou]

I'm having issues at the moment with DMA music being corrupted by particular CRTC settings on real machine.  No problems on Winape, Arnold etc. emulators.  Any info you've garnered appreciated

you have to be more specific about those CRTC settings

you may encounter 2 interrupts per line when the screen is "wide" (see arnold specs)

also if your programm run from a cartridge, you cannot avoid PPI initialisation

[dthrone]

I'm having issues at the moment with DMA music being corrupted by particular CRTC settings on real machine.  No problems on Winape, Arnold etc. emulators.  Any info you've garnered appreciated

you have to be more specific about those CRTC settings

you may encounter 2 interrupts per line when the screen is "wide" (see arnold specs)

also if your programm run from a cartridge, you cannot avoid PPI initialisation

Thanks for the reply.  The DMA messes up switching to a 'wide' screen, CRTC R1 =  &30, R2 = &32, so I'm guessing the issue is with R2.

[roudoudou]

DMA occurs during HBL because they read memory when CRTC does not, so R1 is too big

[dthrone]

DMA occurs during HBL because they read memory when CRTC does not, so R1 is too big

Ah, ok, so what is a realistic upper limit for R1?

[roudoudou]

Offset told me a long time ago the limit is flexible becaus ASIC updates many things when electron beam reaches R1
If only one DMA is used, HBL may be shorter
But there is probably also some needs for split register update, interrupt register update, ...

[dthrone]

Offset told me a long time ago the limit is flexible becaus ASIC updates many things when electron beam reaches R1
If only one DMA is used, HBL may be shorter
But there is probably also some needs for split register update, interrupt register update, ...

Thanks very much for this, very helpful, it sounds like I need to do some experimentation!

[dthrone]

Update:  It seems to work ok now on the real hardware with R2 set to &31 (or less) so I suppose the HSYNC rising edge must have been too late for the DMA to function correctly??  What surprises me is it even works with all 3 DMA channels active and LOAD instructions on at least two channels on the same scanline.

ps sorry for the slight thread hijacking

[Zik]

one week later, bugs are fixed, pfiouuuuuuuuu

Great! In addition to the fixes, it is good to have wasp control, finally. So now, the music has an appropriate length!

[FruityFrank]

one part inspired by Nooon Stars?

[roudoudou]

one part inspired by Nooon Stars?

Not at all! Stars displayed true 3D instead of this =>

The technics used in Flirt/Igloo wasp was WAY faster than Nooon bee