I have been working with the Kodak DC-290 in several different modes to get the best quality (highest res - lowest artifcats - truest color) and am having some problems.

I shoot images of finished needle art designs (cross stitch, hardanger, cut threads) and am disappointed with the detail I get. If I shoot closeups (12" from subject at 3X zoom) I get more than acceptable results (I can clearly see the threads and the fabric). However if I shoot the entire subject (typically 36" from subject with no zoom) the image is not sharp.

I have tried the following settings:

1. Uncompressed (TIF) with flash, high res (1792 x 1200 pixels), Auto exposure.

2. Utlra mode (2240 x 1500 pixels interpolated) JPEG, Auto Exposure

When I use any of the JPG modes - they import into Photoshop at 72 dpi and then I convert them to 300dpi. The images gets "lossy" after that.

I have tried taking the images without flash using several "daylight" lamps and reflectors to completely fill the subject with light.

I use a tripod to ensure I do not introduce errors from camera motion.

Any ideas or suggestions are appreciated.

Mark Steele

-- Mark Steele (msteele@edgix.com), June 20, 2000

Answers

Several things come to mind.

1. When resizing uncheck the resampling box on the menu in photoshop? At least I think you uncheck the box to maintain the same file size and just adjust the DPI to 300 vs. the 72 all images are generally assumed to be rendered at. That might help? Photoshop Gurus are you listening?

2. It might simply be that an image of only 1792 x 1200 resolution just isn't sufficient to record all the detail you wish to preserve. If you can see the detail fine in a 3X zoomed closeup captured from 12" away, then you're probably only imaging a very small area of the subject, effectively using a lot more pixels to depict the detail. When you zoom out and move back to 36" away you're using the same number of pixels to capture a much greater amount of detail. What you might have to do is put the camera on a tripod and take a number of overlapping images and stitch them together to create a larger image.

You could estimate how feasible this would be by taking an image of just over a quarter of the object(maybe a bit larger, most programs want 30% or so overlap on images) and seeing if it held enough detail, if so, then 4 images could probably be stitched together to provide the detail you need. Do a search on this forum or a web search and you'll probably find a bit of stitching info and program recommendations. I think there's also an article on stitching software on the Imaging Resource.

Good luck!

-- Gerald M. Payne (gmp@francomm.com), June 20, 2000.