Introduction

This page contains random notes about my experiences with my Kodak DC290 Zoom digital camera. I'm putting this here not just to organize my thoughts, but so that others may benifit. Some will be specific to this camera, or similar models, but others will be general.

Back to Digital Photography and Scanning page.

Sharpness

My pictures are fuzzy!

The first thing I noticed upon trying out the camera in earnest is that the photos were fuzzy, compared to scanned 4x6 prints from 35mm taken with an SLR.

I was able to improve it by using an "UnSharp Mask" filter in PhotoShop. Since then I learned that this is common practice, and that the camera itself may automatically do some sharpening.

The Sharpness Setting

The DC290 has an adjustable sharpness setting. It's not on any menu, but a script (which comes with the camera) can inspect and change the value. I figured I just needed to turn it up from the default level, to get results I liked better. So I tried taking test pictures on different levels, and found out something interesting.

With sharpening turned off, the files are smaller!

The difference between "none" and "sharp" was 150K. The difference between "standard" and "sharp" was only 40K. The more sharpening that was applied in the camera, the larger the resulting JPEG file. This is logical once it is pointed out: the more diffuse images have less high-frequency data, so compress better.

I can apparently get the same results by doing a simple "Sharpen" filter in PhotoShop. So, I can fit 20–25% more on a flash card, and do the sharpening after I download.

I wonder though, if the results are indeed the same. Does sharpening in the camera cause more information to be preserved, as opposed to performing lossy compression on data with a much larger low-frequency component? That is analogous to subtracting 1 from 1.00023 before rounding off to 3 dignificant digits. I don't think so, but if someone knows more, please tell me.

My experiment was a photo of a sharp magazine page. I'll do more tests with uncompressed TIFF, too, to see if this causes JPEG to be more lossy.

The other reason I'm attracted to turning sharpening off inside the camera is because I can better control it myself. If I'm applying some sharpening filters to the image later, applying that on top of the sharpening it already did might actually be more difficult than starting with the original data. I can batch process all the images (and save the results under different names so I don't lose the originals) if I want to simply view them all, and be no worse off than I was if the camera was doing the built-in sharpening.

Flash

External Flash Is Flakey

Sometimes it fired, sometimes it didn't. I think, after exchanging the cord once and much experimentation, that the PC connector on the end of the Kalt "Hot-Shoe to PC Connector with PC Cord" (Part number NP11131) has a poor contact against the jack on the camera body. With the new one, I can get it to work every time if I push on it with my thumb.

The "15 ft. PC-PC Female to Male Extension" (Kalt NT10002) seems to not have the same problem, and by using that I avoid the problem.

External Flash Exposure

When I set the camera to external flash mode and choose the F-stop, the "X-sync" shutter speed appears to be 1/60 second.

This is not in the instructions anywhere, and since it uses a leaf shutter, the "X sync" speed could be much faster. However, the benifit of exposing for 1/60 second anyway is compatibility. Meters will consider ambiant light during the exposure time as well as the flash burst. My flash meter has a switch for 1/60 or 1/250 speed. Flash units themselves sometimes can sense the light and cut off automatically based on an exposure, and that's the same principle. Any such device that meters light will be built with traditional exposures in mind. All cameras with focal-plane shutters have been "60" for a long time. In more recient times, good SLR's do 125 or even 250. So 60 is certainly useful for every meter out there, but it's too bad I can't set it to 125 or 250 as well. Or set it arbitrarily if I don't need a meter. After all, I can see the picture instantly, and can adjust it via direct visual feedback.

The speed was discovered by looking at the EXIF record stored with the picture. The listing of attributes produced by Thumbs Plus somewhat matches the description at TsuruZoh Tachibanaya's site, which seems to be the only online documentation on the subject.

The relevant values I found are:

It's interesting that the values are supplied in two different ways. The aperture value is expressed as an integer number of half-stops, which is probably the "real" value on a physical camera. The Shutter Speed is expressed as an integer number of stops, which is also probably the way real cameras worked, in order to maintain equivilant EVs between different combinations of aperature and shutter. But the Exposure time value, which is a rational number that can be expressed very precicely, is a rather odd value. There is no reason why 1/64 or even 1/60 cannot be represented accuratly in this field. So is that odd value indeed the real exposure time?

Filll Flash Exposure

Notes on fill flash will go here...

Close-Up

This section contains notes on using Close-Up lenses. This flower was shot using a +4 diopter attachment, from a few inches away.

Focusing with Close-Up Lenses

On my Canon EOS 35mm cameras, the auto-focus feature works optically through the lens, so it focuses correctly no matter what kind of lens or lens attachments are in use. I don't know exactly how the Kodak DC-290 operates, but it appears to always use red beams of light. I suppose it's optical though the lens, but only looks for the beam, rather than using whatever is in the scene.

When I tried it, I found that the first picture I took with a +4 diopter attachment auto-focused just fine, but others I took did not work at all.

Playing around with it, I noticed something: the read beams would miss the subject! Unless you are taking a close-up picture of a large flat surface like a wall, this will generally be the case. The paralex error is extreme when you are only a few inches away, and it's best not to consider that auto-focus would work.

So do it by the numbers. Set the focal length manually, and find out what this really means with the "glasses" in front of the camera. Then know how far to hold the camera from the subject. Remember, the large depth of field works for your advantage here, and this is considerably more tolerant than a 35mm. Also, bracket your shots, taking one a little closer and one a little farther. Again, go with the digicam's strenghts, in this case the ability to take lots of pictures without increasing costs.

The best way to find out the focal zone of your camera wearing its “glasses” is to shoot a ruler or piece of graph paper.

Here, the camera is set to its closest focus (1.5 ft or 50 cm), but the +4 diopter attachment is added. I can see in the full-resolution image that 4 inches is still out of focus, 5 is pretty close, 6 is perfectly in focus, and 7 is just starting to get out of focus, but too far is better than too close.

Conclusion: with the +4 diopter, the subject should be between 5 and a quarter to 6 and a half inches away. A flat subject should be at 6, a figure with depth should be within these limits, with the far side fading more naturally than the close side.

Here is a later test, using a test pattern I’ve dubbed the “telephone poles”. The part of the pattern that's clearly in focus is shown in the detail, at it's original resolution (though messed up a bit with aggressive lossy compression). In the full frame, you can correlate this to distance. The 5 and 6 spot dice faces are at 5 and 6 inches from the lens, to serve as a check. This shows a clear focus from a little less than 5 inches to a little more than 6.

Repeating the test (not shown) with different zoom settings indicates that at telephoto settings, the depth of field is significantly shallower, and a little farther away. Specifically, at maximum (optical) zoom, 5 is distinctly out of focus, and 7 is in focus.

The telephone pole pattern, or any scene for that matter where yoursquo;re looking at things to the edge of the frame, could be a little off. If the distance is actually an arc drawn centered on the focal plane, the sides are actually a little farther away than the center. The best accurate tests, then, would be done with a single object that is photographed in different positions along the center line, rather than filling the frame with objects at various distances. However, a real frame-filling object will need to be in focus at the edges, too, so you could argue that the edge distance is also meaningful for real 3-D subjects.

It’s difficult to make a chart stating the focal depth for each available diopter, when it changes with the zoom factor too. That is, if an image is too big, you can”t back up because you'll be off your focal spot, and if you zoom out you change the correct focal spot, too! So, there will be not substitute for trying it out, upon setting up the scene. The numbers below tell you how to initially position everything, but then check your image on the computer screen (not the camera’s LCD) and adjust things. Given that, I don’t need exact numbers. So I’ll measure the center zoom setting (what it is when the camera is turned on).


Page content copyright 2000 by John M. Dlugosz. Home:
http://www.dlugosz.com, email:mailto:john@dlugosz.com