I'll try to answer all the questions in one response.

There's not much going on in the demo apart from proving that it's possible to swap out the NES palette at a measured point on the screen i.e. during the first black gap under the first row of coloured squares. It takes a lot of manual tweaking of cycle-timed code as you only have time to swap 3 colours during horizontal blanking (the time taken by the PPU between drawing the end of one raster line and the start of the next). So, in between each row of square are 3 raster splits that (off the edge of the screen) change 3 colours each, giving you 9 modifiable colours.

I need to code the remaining splits (one for each row of squares). To do that, I can't manually code each one or I'd go insane. Rather, I need to figure an easily repeatable way to achieve the required accurate timing delays, so that's what I'll do next. I don't know much about cycle-timed code so I need to do a bit of research. The demo so far is a bit of trickery with a lot of trial and error smile

The other side of this is actually coding the colour generation/animation. It's not something I've ever done before.

And, more to the point, where to find the CPU time to do the required data movement/shifting/manipulation - the problem with filling the screen with the timed palette splits is that you're not left with much CPU time to do anything else.

I'll make the source code available once I get the timing loops into a repeatable state and tidy the code up somewhat.

Expanded it a bit;

http://dl.dropbox.com/u/5493868/litewall2.nes

This time only one split (after first row) but I'm swapping 9 colours during the 8 black pixel gap.

Top row of colours cycles slowly one way while second line cycles faster in the opposite direction.

Quick proof that it's doable;

http://dl.dropbox.com/u/5493868/litewall.nes

The first, second and third (onwards) rows are all using the same palette but the colours are modified on the fly. Just need to modify more colours per split and then replicate the timing for the remaining rows.

smile

NO CARRIER wrote:

Sounds interesting. I was inspired by the Blip wall when I made this last year: http://vimeo.com/5250274

That's on the C64, of course. However, I did use MMC3 to do some screen splitting for this: http://vimeo.com/6740666

I'd be up for figuring out timed code to further split each line, maybe. Would be a cool challenge. smile

Don, pop over to nesdev, I'm "talking" about this with blargg on there.

I reckon it's do-able, just tricky and a bit manual list of timing code to do all the splits.

:S

Heh, that was exactly what I was thinking smile

It's an interesting idea and a nice little challenge smile

The core idea is to be able to independently animate the colour of each square which is going to be very difficult on the NES. You'd need to control 90 (in your illustrations) unique colours onscreen simultaneously - the NES can't do anywhere near that.

You have 4 "palettes" of colours, each palette containing 4 colours (in practical terms only 3 because the first colour is the background colour). The screen is divided into "attributes" which are blocks of 2 x 2 characters (8 x 8 pixels). Within these 2 x 2 blocks, all of the characters have to use the same 4-colour palette.

I'd think the only way to do it is to have timed splits in the screen, one for each row of squares and then swap out all 4 palettes at each split. Tricky smile

455

(39 replies, posted in Nintendo Consoles)

Brutal : Paws Of Fury

wink

Someone liked it so much they made me a tribute smile

456

(41 replies, posted in Nintendo Consoles)

egr wrote:
RG wrote:

Also, for what it's worth I drink a LOT.

lol.  That's my stock explanation for almost everything.

It's the only way I get most of this stuff done!

smile

457

(41 replies, posted in Nintendo Consoles)

RG wrote:

So far so good on the DPCM patcher but I'll be working with it more over the next few days. I noticed that the newly created rom still had the original .dmc sounds even though the original .dmc sounds were no longer listed in the patchfile.txt. Should this happen or did I miss something?

Ah, good point. Yes, the old data will still be there. Easy enough to change it so that it clears the data before patching though. Would you prefer that?

458

(13 replies, posted in Nintendo Consoles)

@bucky : hopefully I did the original justice. It's a bit ropey in parts but I think it came out pretty well.

@low-gain : if only I lived near you......i'd move....

459

(13 replies, posted in Nintendo Consoles)

NTRQ apprentices might find this interesting/useful.

http://blog.ntrq.net/?p=312

smile

460

(41 replies, posted in Nintendo Consoles)

So, any feedback from anyone using the NSF exporter and DPCM patcher?

461

(18 replies, posted in Nintendo Handhelds)

Oh, so the sample output could be corrupt/broken....leaky?

I'll record some output and post it up for opinion.

462

(18 replies, posted in Nintendo Handhelds)

I think I'm getting the hang of it now.

And people said NTRQ was complicated! wink

463

(18 replies, posted in Nintendo Handhelds)

Ah. I'll carry on by ignoring them then smile

464

(18 replies, posted in Nintendo Handhelds)

Hmmmm. Do LSDJ samples just sound really poor? They seem to sound OK if I only have the WAV voice playing but at soon as any other channel is used they sound awful.