I should refactor it into a supershader with functions rather than a bunch of frambuffers.

The translator function in Forum failed on this one.

I'm getting old - I have not a clue what you just said!

😉😊😆

Lol @pickle..now where is that 'blind leading da blind' gif ?

I should re write the scene and its shaders into a more compact form. i can take the code from seperate shaders and combine them into one. That will save using so many framebuffers as i did in the scene we are talking about and may give a speed boost (not much).

@pickle1

I'm getting old - I have not a clue what you just said!

Yes i'm saying the same thing to me, but i will try to Help you :)
For Refactor maybe he means Rescaling,
https://en.wikipedia.org/wiki/Feature_scaling
and here is the Framebuffer
https://learnopengl.com/Advanced-OpenGL/Framebuffers

refactoring is basically re-writing code to make it more efficient/smaller... i think. well... thats what i meant anyway.

The frambuffer link is more of what istripper is doing for us when we add something like this to the scene.scn file

framebuffer
{
id: DoSomethingWeird
size: 3840, 2160
pos: 1920, 1080

sprite
{
source: Girl
shader: TurnInsideOut.fsh
}
}

refactoring is basically re-writing code to make it more efficient/smaller... i think. well... thats what i meant anyway.

You guys carry on enjoying yourselves and don't worry about old fossils like me. 😆

I find your conversations (and @Dorsai6 with his playlist generator) fascinating but the truth is - after 40 years in IT one thing I learned is sometimes it pays not to tamper - so I am happily running a vanilla version of iStripper though I did recently go mad and installed the Belka skin from @lukaszr.

But I'm happy to sit rocking in my chair and let it all go whoosh over my head. 😊

Keep being creative.

Keep working it @Z22.
methink the orig Hue.fsh shader is hauntingly beautiful btw 👍

To be fair most of it is above my head too. I just throw enough s**t at the wall some of it sticks.

@ET, cheers dood. Some of the colour combinations do look somewhat ghostly. Perhaps i should have saved it for halloween... You could do a remix... nudge nudge wink wink.

having to use a default empty directory to call the clipsprite tends to screww
with any textural .scn re-mixing cos it will try to grab those other textures
but worth a try.. tks for whats there maestro !

You can have other directories for textures. i did with hue3

texture
{
id: Feedback
size: 3840, 2160
source: empty/
}


texture
{
id: Tex
size: 3840, 2160
source: Textures/natural-fence.jpg
}

The problem i thought we had in the past isn't actually whats going on with the "undocumented feature" (aka bug)

I had thought the feedbackbuffer was grabbing the entire camera node as its input when it tried and failed to load a texture from the empty directory, but it looks to only be grabbing the last sprite in the camera node, and most importantly that sprite can be invisible (opacity: 0.0) I only figured it out the other day.

You can scale and reposition the girl as much as you like in the camera node and it wont affect the feedback.

change your .scn to say ...

camera {

type: 2D
size: 3840, 2160
pos: 1920, 1080

sprite
{
//source: Invert
}

sprite
{
pos: -900, 0
scale: 0.6, 0.5
source: Refract
}

// Last sprite will be the input to the Feedbackbuffer next frame.

sprite
{
opacity: 0.0 // If only using for feedback you can set to 0.0.
source: Hue
}

}

so close yet so far..ok will try

If you turn invert back on you can see what is being refracted to make the girl, and if you change it to hue then you can see the buffer that is eating it's own tail (last frames output is this frames input).

https://www.istripper.com/forum/thread/29408/35?post=661985

Всё о iStripper / Share your FullScreen - Member Created Scenes here

edit: post deleted by me

this version above doesn't work either on my rig..just noise & vague model as before.
only orig Hue.fsh works. I tried alpha transparency mod of hue to no avail.
It just all went white
if it is somethang peculiar to gpu or memory, someone needs to chime in & say Hue2 works
OK on their end

Which of the hue scenes was it (hue.fsh) from (there may have been changes between them)

only thing i can see atm that isn't commented out or just a referance is the size of the search area

y01<4.01;

in the older version its

y01<3.01;

I am struggling to work in this fashion
.
You have different code values (according to the Beyond Compare app) for the same shader name
Hue.fsh used in Z22_Hue2 and Z22_HueGlass_b
imho modded shaders need to have unique names...or we be barking up the wrong trees, especicially for peeps who use common shader folders over many presets ...ahmen

:(

point taken. Will try to remember to do that in future.

Noticed something weird

if i have

dist == (distance(G, H));

rather than

dist = (distance(G, H));

i get a different result.

Anyony know whats going on with that?

@Z22

The == operator is a comparison for equality that evalutes to 0 for not equal or to 1 for equal. It is not an assignment.

The statement dist == (distance(G, H)); is valid because the shader language's syntax is based on that of C which allows expressions anywhere that you would expect statements. In the example the function distance will be called, its return value compared with dist and the result discarded, in particular dist will not be updated by this statement and if distance has no side effects the whole statement can be optimised away because it has no visible effect.

In any sensible language such statements would be errors, but they are valid in C or GLSL.

The following is a valid, if rather pointless, function

void func(void)
{ 42;
}

It is also easy to make the opposite error and use = where you intend to ==, e.g.

if ( a == 0 ) do_something();

will compare a with 0 and only call do_something if a is equal to zero

but

if ( a = 0 ) do_something();

will set a to the value 0 and will never call do_something() whereas

if ( a = 1 ) do_something();

will set a to 1 and will always call do_something()

in shader it gives a higher result that the propper "=" version.

Thankfully atom ***** if i put "if( a = 0 ) do somthing();" so i catch those mistakes. It was just weird that it didn't catch it in the dist = . Thanks for explaining why.

https://drive.google.com/open?id=1rhUh0T1fqlD0Wa3-tKrkYmlmVM8FwluY

Waterfall
This is a first pass so will change without filename changes for every little difference hence releasing it in this thread.

I'm not happy with the top of the screen so need to find a better way than baseing it off the background texture.
Think you can see where i'm heading though.

https://drive.google.com/open?id=1KvU7y7b9ikOjdRS_-FWKBKci9_USupT2

Just the foam effect.

still a work in progress, i need to reduce the gpu load, think i know how.

https://drive.google.com/open?id=1Lg_zXyznhLYQLbuuZLs0Z25uKg8o-EQK

As above but the "particles" are being refracted.

hmm...
this...

for (float y=(0.0); y<5.01; y+=1.0)
{
for (float x=(0.5*-y); x<0.5*y; x+=0.5)
{
x *=2.0;
vec2 p1 = vTexCoord.xy + Offset * vec2(x, y);

crashes istripper but
this ...

for (float y=(0.0); y<5.01; y+=1.0)
{
for (float x=(0.5*-y); x<0.5*y; x+=0.5)
{

vec2 p1 = vTexCoord.xy + Offset * vec2(x*2.0, y);

doesn't

err dafuk? Something with amd's compiler perhaps?

nm, i have realised why.

I think, in the first case x itself is multiplied by 2.0 in each loop pass, whereas in the second case only x*2.0 is passed to the function vec2.
As a programmer in the last millennium I think...right @z22 ?

Yeh, i think it's because it's the var in the for loop, so the first pass it's fine but when it gets back to do the next for it has a spaz as the value is lower than the starting x.

Have found something else weird though.

for (float y=(0.0); y<25.01; y+=1.0)
{
float xd = (0.25*y);
float xdm = (0.26*y);
for (float x=(-xd); x<xdm; x+=(xd))
{

works

for (float y=(0.0); y<25.01; y+=1.0)
{
float xd = (0.25*y);
float xdm = xd+0.1;
for (float x=(-xd); x<xdm; x+=(xd))
{

crashes.