1
The biggest part of this all I interested is a realisation of short pause before appearance of new character's reply text, when game must not to react to buttons pressing - i. e. I need a functional analogue of RenPy's $ renpy.pause () function. I tried to use the code like the textTimer from my RPG engine, but for some reason it doesn't work and interferes to execute a sprites clearing code in the basic function even.

Maybe, you'll help me to fix a background colour glitches with a character's pictures and stringe pallete bug of unique title screen tile at the same time.

You can see the first results of my work here.
The source is here. Warning: this project uses the DATlib 0.3!!!
avatar
2
Proper color for transparency on sprites and fix is Fuchsia (RGB 255,0,255)
3
That's magenta smile
avatarZeroblog

« Tout homme porte sur l'épaule gauche un singe et, sur l'épaule droite, un perroquet. » — Jean Cocteau
« Moi je cherche plus de logique non plus. C'est surement pour cela que j'apprécie les Ataris, ils sont aussi logiques que moi ! » — GT Turbo
4
In main.c move waitVBlank(); in line 572 up under SCClose(); 554. Also I think that the function Control_Sub(); should not be executed between SCClose and waitVBlank. There are other places in your code this happens.

The color code you need is ff0ff I believe.

Hope this helps. I'll look through the code again later.
avatar
5
Zerosquare (./3) :
That's magenta smile

Nope, Fuchsia embarrassed
6
grin
avatarZeroblog

« Tout homme porte sur l'épaule gauche un singe et, sur l'épaule droite, un perroquet. » — Jean Cocteau
« Moi je cherche plus de logique non plus. C'est surement pour cela que j'apprécie les Ataris, ils sont aussi logiques que moi ! » — GT Turbo
7
hine62 (./4):In main.c move waitVBlank(); in line 572 up under SCClose(); 554. Also I think that the function Control_Sub(); should not be executed between SCClose and waitVBlank. There are other places in your code this happens.

I am sorry, but these changes had no effect. You can test the source version below as a proof.

The color code you need is ff0ff I believe.

I know it, and therefore I tried to place these pictures on the ff00ff coloured background through Photoshop.

tromb Fichier joint : main.c
avatar
8
Moving SCClose(); was only a common practice suggestion for the NeoGeo. Program code should go between waitVBlank() and SCClose(). Like this:
while(1) { waitVBlank(); /*Do stuff*/ SCClose(); }
It can be this way also...
while(1) { SCClose(); waitVBlank(); /*Do Stuff*/ }
I did get your program to run, but for some reason I couldn't compile any changes? Weird.... anyway here is my suggestion.

Put this in the variables section of introduction()
int pauseTimer = 0; //keep track of time int isPaused = 1; //keep track of pause state

Then use a timer to pause the program in an action state. I believe that the neogeo updates 60 times per second. So you could pause for 60 ticks or 1 second. To pause and action just set isPaused in the action before the one you want paused.

if(action==0) { if(isPaused == 1) { pauseTimer++; if(pauseTimer >= 240) //4 seconds { isPaused = 0; //reset pauseTimer = 0; //reset } }else { fixPrint(1,3,0,1,"Мне продолжает сниться этот сон."); fixPrint(1,4,0,1,"Каждую ночь одно и то же."); fixPrint(1,5,0,1,"Ну, сон и сон! По-настоящему странные"); fixPrint(1,6,0,1,"вещи происходят, когда я просыпаюсь."); } }
One last thing. Move the Day1() function above introduction(). The compiler will like this better smile
avatar
9
Thank you very much! Now everything works greatly!
avatar
Glad I could help!
avatar
I'd like you help me with a selection of background files colour reduction way - I decided to process these files with IrfanView (reducting to 256 colours) before giving them to DATimage. But for some reason, althoght DATimage successfully reduces a colours number per tile, it increases a colour number in the whole file.

Here is a sample for comparsion.

tromb Fichier joint : lqTX
tromb Fichier joint : oizS
avatar
Hi VasiiiyFamilya,

EX:

Image is 320x224 (Neo Geo Screen Size) = 71680 pixels
16x16 tile = 256 pixels
Tiles per 320x224 image = 280 tiles

DATimage with this flag -mtc15 will convert an image to the neo geo format (15 colours per tile)

The total number of palettes that can be used on Neo Geo is 256 (0-255) (There is another palette bank I haven't used it yet....probably can only access one bank at a time)
If all 280 tiles have a unique palette you are over the total number of palettes just to load that image.

It's not that there are more colours added its that each tile can have a unique palette.
Even if the colours repeat throughout the image I believe the compiler is looking at each individual tile as its own mini image with its own palette based on the order it sees the colours.
There is probably some cool compression going on under the hood that HPMAN hooked up but that is the breaks as I know them.

My advice is you must compress your image more until you get a reasonable number of palettes at compile time.
Mega Shocked (./12):My advice is you must compress your image more until you get a reasonable number of palettes at compile time.

How should I do it?
avatar
I have pretty lame method but I use PNGquant and run my image through it several times.....there can be alot of trial and error involved depending on how obsessed you get.

https://pngquant.org/

DATimage can also reduce colours "Change -mic256 to -mic128 , mic64 etc....

DATimage pathWhereFilesAre\*.png -mtc15 -mic256
Mega Shocked (./14):
I have pretty lame method but I use PNGquant and run my image through it several times.....there can be alot of trial and error involved depending on how obsessed you get.

https://pngquant.org/

DATimage can also reduce colours "Change -mic256 to -mic128 , mic64 etc....

DATimage pathWhereFilesAre\*.png -mtc15 -mic256

I just tried this way of color reduction for a most of time, and tried to reduce the mtc maximally for economy of palettes allotted for characters portraits. But more often the result of direct reduction through DATimage (-mtcX or -mtcX & -micX) looks much less visually neat than a result of reduction to 8bpp through IrfanView and PNGquant, despite of larger nunber of colours.

Here is a -mtc15 -mic256 version of previous picture.

tromb Fichier joint : E4ZJ
avatar
in my opinion DATIMAGE does a very good job but I think you get the best results when you don't use pre-edited images.
maybe you can upload the original image so that we can do tests together for a better result?
blastar (./16):
in my opinion DATIMAGE does a very good job but I think you get the best results when you don't use pre-edited images.
maybe you can upload the original image so that we can do tests together for a better result?

Here is it:

tromb Fichier joint : Xl7N
avatar