Digital Light & Color

Again, you need to switch to Gaussian.

What are you using for the alignment method?

Hard to tell from just the end result - can you post the original image and give me an idea of what you effect you are trying to achieve again?

Unfortunately, the overlay image and its alignment points are not saved.

I can't do much about the positioning of the alignment points, but I forgot to mention that you should change the alignment method as mentioned in the white paper to Gaussian Method (via the Opt menu). This makes the effects of each pair of alignment points more localized.

Here is what I would do in this case. Since you don't really want to warp the overlay image but simply rotate it, you should be able to use 2-point alignment. You might want to start by darkening the reflected log image a little, or this may not be necessary. 1) It sounds like you were able to compo...

The warp tool just does perspective warping. What you need to use is the Composite transformation with multi-point alignment -- there is an example of using this to warp images in the white paper: Registering and Distorting Images.

For details, this technique is covered in the Stack Images white paper.

Feathering is purely geometrical so the falloff rate is constant. If you use mask curves, the falloff rate depends on the image data - a sharp edge will fall off faster than a subtle gradient.

The only two reasons I know of are: 1) Many scanners do not have a 64-bit Twain driver 2) There is no 64-bit support for PhotoCD Most scanners come with stand-alone software that simply encapsulates their Twain driver and lets you save the scanned image as a TIFF or JPEG file. This is what I normall...