Step 2 - Orientation adjustment

You can put your scanner aside now. The next steps will all be conducted using the bitmap graphics editor.

ZunTzu requires the counter sheets being scanned to be perfectly oriented on the glass pane, because the counters are assumed to be aligned horizontally and vertically. The purpose of this step is to correct the tiny misalignments that are likely to occur.

Load your scanned counter sheets in your bitmap graphics editor, and apply a micro-rotation to adjust the orientation of the counters.

You should do the same with the scanned fragments of large boards. It will make compositing much easier.

Choosing the best downsampling method.
1 - Using the "Bicubic" method (normal size and enlarged 200%).
2 - Using the "Bicubic Sharper" method. Note how the numbers are sharper but how the blue background is less soft.

Step 3 - Downsampling

The purpose of this step is to resize the images by 50%, to obtain images with a 150 dpi resolution, the standard resolution used for ZunTzu graphics.

If your computer has plenty of memory to spare, you can apply this step after the compositing step.

If you're creating a "deluxe" game box (featuring 300 dpi graphics), you can skip this step altogether. Be warned: a deluxe game box file is three times as large as a regular one, and there's a risk some users won't have enough memory to use the game box on their computers.

There are several methods to choose from to resize an image. I usually use the "Bicubic Sharper" method to downsample counter sheets, because the edges get slightly accentuated and so the numbers on the counters are slightly easier to read. For boards on the other hand, I prefer the "Bicubic" method, which produces softer results.

Creating the composite image.
1 - Topmost layer containing an hexagon grid for reference.
2 - A fragment (borders accentuated for this demonstration).
3 - Another fragment, overlapping with the other one.
4 - Overlapping area being erased using the eraser tool.

Step 4 - Compositing

The purpose of this step is to assemble a single composite image from several image fragments.

Estimate the size of the composite image and create an empty image of at least that size. Then copy each fragment into the image, as a separate layer.

Translate the layers until all fragments are at the correct position relative to each others. Normally, the orientation of each fragment has been finely tuned at step 2, so it is just a matter of translating the layers.

Here's a good tip for compositing wargame boards: use an additional layer containing a transparent hexagon grid. The grid will make the positioning of each fragment much easier. You can find tools to produce hexagon grids on the web [Follow link]. Don't forget to hide the grid layer in the final composite.

When each fragment is properly positioned, use the eraser tool to remove overlapping areas.

Finally, save the composite image as a PNG file.

Hand optimization
Hand optimization of the monochromatic areas.
1 - The Magic Wand tool is used to select monochromatic areas in the blurred bottom layer.
2 - The selection is moved to the top layer, then the area selected is erased.
3 - The result image.

Step 5 - Optimization

This step is optional as it is very time consuming. Its purpose is to eliminate excessive grain in large monochromatic areas, to enhance the quality of the image and as a preparation for the compression step.

I optimize my images by hand, as I'm not aware of a filter that could achieve the same results automatically.

Create a new layer containing a duplicate of your image. Blur the bottom layer: the monochromatic areas will get very soft. Using the Magic Wand tool, select all the monochromatic areas. Move the selection to the topmost layer, and erase the area selected so that it becomes transparent. The result image will combine the sharp edges of the original image and the soft backgrounds of the blurred image.

This technique can significantly increase the compression ratio for large monochromatic boards. And the result looks better too.

Step 6 - Cropping

The purpose of this step is to remove the borders of the image, in order to decrease the image size and thus the memory footprint in ZunTzu.

It is a good practice though to keep the copyright notice and the art credits.

Step 7 - Compression

Although ZunTzu can directly use images stored as PNG files, it is a good idea to store the final image as JPEG, because the size of your game box will be decreased dramatically. Note that this compression step has no effect on the memory footprint in ZunTzu: it only affects the number of bytes required to store the game box file on the disk.

In your bitmap graphics editor, you shouldn't hesitate to set the quality setting for JPEG compression to "High" or even "Very High". You'll still have good compression ratios compared to PNG, and you'll keep the JPEG artifacts to a negligible level.

Further reading

Wayne Fulton has a web site giving good scanning tips [Follow link].

Sean T. McHugh's photography site doesn't provide scanning tips as such, but gives generic image processing techniques and tutorials [Follow link].