pleblem of using 3d texture

[hwj_just]

hello i define a 3d texture from a stack of dicom images,when using one CT image to difine a 3d texture i can get correct result.but when use two CT images to define a 3d texture i get a wrong result, and above three the result only a white square.anyone who can give me some advice of the problem?thank you very much!

[dorbie]

Above 2 you may be running out of space but remember all axes must be 2^n in size.

The errors you think you see may be the result of interpolation. Try a nearest filter and/or clamping and remember that texture wrap when a coordinate = 0 will give you a 50%/50% weighting between the two samples on any axis that has only two samples stored. Clamping in conjunction with understanding your sampling in 3D will probably resolve your issues.

[hwj_just]

now i know that because of all axes must be 2^n in size ,i only can use 2^n CT images to form 3d texture.and that's right now.but i found when using 2 images, the result seems to be enlarge and the image be break up to four cornor.and image quality become bad .when using 4 images it distort much severity,when using 8 or 16 image to form a 3d texture it seems only render a little area of initial image.what is the matter?thank you !

[ZbuffeR]

screenshots, please

[hwj_just]

how to put the image here?i try but i failed