Jump to content
  • Advertisement
Sign in to follow this  
nsn

OpenGL Scale and Bias before texture mapping

This topic is 4113 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hello, I am trying to use OpenGL for visualize 2D 16bit image data. I would like to change brightness (bias) and contrast (scale). In order to achieve this I can upload the texture repeatedly and use GL_RED_SCALE, GL_RED_BIAS, ... But I prefer not to upload the texture each time I want to change brightness. So I am trying to use the texture combiner functions to achieve this. To achieve this, I use RenderScene() { ... glColor3f(scale_val,scale_val,scale_val); GLfloat bias[4] = {bias_val, bias_val, bias_val, 1.0f}; glTexEnvfv(GL_TEXTURE_ENV, GL_TEXTURE_ENV_COLOR, bias); glTexEnvf(GL_TEXTURE_ENV, GL_TEXTURE_ENV_MODE, GL_COMBINE); //For bias glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_SUBTRACT); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB,GL_TEXTURE); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB,GL_SRC_COLOR); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB,GL_CONSTANT); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE2_RGB,GL_SRC_COLOR); //For scale glTexEnvi(GL_TEXTURE_ENV, GL_COMBINE_RGB, GL_MODULATE); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB,GL_PREVIOUS); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE0_RGB,GL_SRC_COLOR); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE1_RGB,GL_PRIMARY_COLOR); glTexEnvi(GL_TEXTURE_ENV, GL_SOURCE2_RGB,GL_SRC_COLOR); //After setting up the texture environment render quad glBegin(GL_QUADS) ... glEnd(); glutSwapBuffers(); } This works as long as my 0.0< scale < 1.0f Any help is much appreciated. -nsn

Share this post


Link to post
Share on other sites
Advertisement
Whoops!!, I forgot to ask the question though

I am looking for way to use value of scale larger than 1.0.

Share this post


Link to post
Share on other sites
I am not sure what you are trying to do, but if you want values outside of 0-1 you will need a floating point buffer setup instead of a fixed point.

Share this post


Link to post
Share on other sites
If I'm understanding right what you want to do, then your input values are clamped to 0.0 to 1.0?
If so you should use Shaders, beceause with Shaders you can completely replace the fixed Function pipeline and do input values larger then 1.0

Share this post


Link to post
Share on other sites
I'm not sure if all OpenGL functions or most have several versions that take different arguments. For example:

glColor3f(scale_val,scale_val,scale_val);

The F at the end of the function name is what type of arguments it takes. In the above example, you pass floats for value: 0.0 - 1.0.

So if you do:

glColor3i( ... )

...the values you pass to glColor3i would be integers. The integers work 0 - 255 instead of 0.0 - 1.0. The way the floating point works is every number between 1.0 and 0.0 is the same as 0 - 255. so 0.5 is the same as 128. 0.25 would be the same as 64. 0.75 is the same as 192. (rounded up estimates) They both are the same, just different ways of representing the value. I personally like to use the 0 - 255 method.

Share this post


Link to post
Share on other sites
Hi MARS_99

Thank you for your response.

What I am trying to do is if I have an image I(x,y) having RGB values

RGB(I(x,y)) = {R,G,B}

After I upload the data using glTexImage2D, I want to change RGB values
according to equation

RGBnew = {R,G,B}* scale + bias;

Note that openGL/GPU automatically clamps the new color values from (0 to 1) prior to display. This is effectively increasing the contrast.


Hi Ingrater,

If I wrote a shader then yes this operation is simple but I am hoping to use the fixed pipeline and hack this one operation.

The reason is that I am really using this to display a 3D volume, using the glBlendFunc in my bigger scheme. What I am doing is using texture env to scale and bias my data before my texture gets applied to the quad. Then draw each slice of 3D data as a quad. Then glBlendFunc is doing the blending to visualize my 3D volume.

Since I donot want to move the data from CPU memory to GPU memory repeatedly, I want to use something in the pipe after loading the data to GPU memory to scale and bias my data.

I read that I could use glCopyTexImage2D can be used, but I donot know how that would work.

Thank you soo much for any suggestions.

-nsn


Share this post


Link to post
Share on other sites
I wanted to add the comment for the equation

RGBnew = {R,G,B}* scale + bias;

Here the problem I having is that scale is being used as a color and it is being clamped to the range (0,1) becaue it is represented as color.

Can I do something clever using alpha value. If so how can I do that.

effectively I am thinking if I can use alpha to scale my colors like

RGBnew = {R,G,B}/(alpha) + bias;

where my alpha = 1/scale. Then I have achieved what I want.

thanks,

nsn

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!