PDR-M1

While my CP900 was being repaired, I bought another camera on closeout at Office Depot. This is a Toshiba PDR-M1, which is almost identical to the Fuji MX-500. Here are some observations on the camera. This isn't a review, just some interesting highlights. There are enough sites on the net with sample pictures, so I am just going to show interesting excerpts. Also, I'm almost out of space at my ISP!

Digital zoom, those are worthless, right?

Instead of a real, optical zoom, the camera has a 2X digital zoom. All this does is crop out the center part of the 1280.1024 image in the camera, and save it as a 640x480 file. The question arises, why not just do this yourself later? Comparison of the same scene under controlled conditions shows no resolution advantage to either method. However, it does appear that the camera does make exposure adjustments when using the zoom. Apparently, it only looks at the zoomed area in setting the exposure. Below is an example. (These are resized to 320x240 to save space). On the left is approximately the same area as in the zoom photo on the right, cropped out of a larger 1280x1024 shot. The center part of the larger photo was underexposed because of backlighting through some windows above. This has made the entire cropped area too dark. On the right, you can see that the digital zoom feature did a better job with the exposure. So I would say that there is at least some small advantage to using the digital zoom as opposed to just cropping a larger shot.

Picture
Picture

Cropped from 1280x1024 picture

2X digital zoom

Picture

Soft, Normal, Hard.

The maual makes it sound as if the sharpness settings are just different degress of sharpening. Actually it looks like the difference between the settings is that Soft is the unprocessed image, Normal is the original image with some software sharpening, and Hard is that same degree of sharpening, but with the contrast enhanced. You can really see the contrast boost if you print out a histogram of your shot, and use a tripod to make sure that exactly the same area is included. The shots below are actual size cropped from the original fine mode 1280x1024 images.  You can see the richer color in the hard shot due to the increased contrast.  The hard shot might have more sharpening too, but it is hard to tell. I actually prefer the untouched "soft" images, since I hate artifacts more than I like sharpness. You can start to see artifacts in the Normal shots; look at the black outlines around bright objects. Even at Normal the noise is accentuated also. I prefer to use an adaptive unsharp mask when I want a little more sharpening.

Soft

Normal

Hard

Picture
Picture
Picture
Picture

Infra Red Response.

"Buddy, can you spare some change for some coffee?" Below, on the left, visible light (sunlight). On the right, I used an infra-red filter over the PDR-1. This filter looks black to the eye, but passes IR light. You can see down to the bottom of the cup, and see that I dropped in a penny. The IR response is much more sensitive than the Nikon CP900. That might be bad for outdoor shots. The response looks purple white, and should add some purple white to outdoor shots. I think that this works because the coffee (instant) is really a colloidal suspension of tiny particles. The longer wavelength of the IR light can penetrate further without interacting with them. I had read a report that the macro mode blocked IR, but that wasnt't the case here because I used it. Maybe the macro lens is made of plastic in some cameras or something.

Hmmm, I wonder if this will work with clothes... ha.

Picture
Picture
Picture

Macro Mode and the "Shaking hand" icon.

I've had trouble with shots in macro mode. Sometimes you can get nice and close, sometimes the shot will look good on the LCD screen, but be blurry when you look at it closely. This really had me going for a while. Sometimes the "shaking hand" icon comes on in macro mode, and nothing you do can get rid of it and your shots are blurry.

Here is what is happening:

The camera has two possible f stops that it can use, f3.2 and f8.0. Ony at f8.0 can it get close enough to focus to the 9cm distance advertised. If you are in macro mode, and the camera thinks that it is going to use f3.2, the "shaking hand" comes on. It has nothing to do with image motion as implied in the manual. If you take a closeup when the "shaking hand" is on, it will be blurry.

Prove it yourself! Try this:

Set an item on a table. Get about 6" away. Lower the lights if needed, the shaking hand will come on. Put your camera on a tripod, still the "shaking hand" is on.  Take a picture and look at the f stop with your favorite viewer. f3.2.  Now force the flash on in setup. Take a shot. f8.0, not blurry.

Want more proof? Try this:

 Point your camera at a (lit!) light blub from about 6" away. A desk lamp works well.  No "shaking hand" now, right? Shake the heck out of the lamp and or camera. Still no shaking hand! Take a shot. f8.0, in focus. Now turn off the lamp and use your room lighting. The "shaking hand" will come on (unless your room is very bright). Take your shot, blurry at f3.2.

More insight into this: 03/27/99

Here is an experiment. Get the camera and a flashlight. Turn off some room lights. Put the camera on record. Look at the lens, you can see the shutter in there. The shutter will be closed. Turn on the display. The shutter will openl; the shutter opens when you turn on the LCD. The iris aperture will be "large" (f3.2) due to the low room lights (if not, make it darker). Now hold the flashlight up to the lens (turn it on!). The shutter will stop down. This is the f8 setting. Cool.

OK, now the experiment.  Point the camera at an object 6" away in regular mode. Confirm that the shutter is open and the iris is at large aperture by looking at it. Now press the release half way to try to focus. You can't, you are too close. OK, now switch on macro mode. The focus will get better, but not perfect. Confirm  that the aperture is still large. Setting macro mode did not set f8 automatically. Now shine your light on the test object so the camera stops down. You can check for the small f8 setting visually. The focus will be even better now. This proves that macro mode involves something more than just stopping down.

I think that there IS an extra adjustment for macro mode, but you have to go to f8 to be able to focus closer than 10". From 10 to 20", just switching to macro mode will do it. Closer and you need the f8 setting (can you say pinhole camera?)

What is true is that I can't see any lenses moving in there when you switch macro mode on and off (by looking at the reflections).

I think that what is happening is really stupid or very clever, depending on your point of view. All macro mode does is move the CCD back farther than the normal range. (In this camera, the CCD moves to focus, not the lens) The designers may exclude travel in that range normally because it would increase the focus time when looking for best focus.  (you can hear it doing something, sounds like the regular focus sound when switching to macro). That lets it focus to, say 10". Then to get
closer, you have to be f8, to get the pinhole effect. If it isn't f8, it puts up the shaking hand.

What to do:

This isn't horrible, but it points out that you need a lot of light if you want to use macro mode very close up. Try turning on the flash! The camera seems to do a pretty good job of squelching the flash in macro mode so that you don't overexpose.

Low Light PDR-M1

Well, I don't expect too much from this camera as far as low light goes, because of the 1/4 sec maximum exposure. You can still take a neat low light shot though. I've put up a couple showing the moon and Venus taken March 19, 1999 and March 20, 1999. Resized way down from the original shots. You need a tripod for shots like these! I tried to handhold the one on the left, but the full size shot is blurry. Noise isn't too bad in these shots, but more than the CP900 would be.

Picture
Picture
Picture

Delay between pressing the shutter and taking the shot. ("Lag")

Full press, no flash,  monitor on            ---------------- 1.7 sec

1/2 press prefocus,  no flash, monitor on----------------0.5 sec

1/2 press prefocus, flash, monitor on  -------------------0.5 sec

1/2 press prefocus, no flash, monitor off ----------------0.2 sec

So, if you want the picture to be taken as soon as possible, turn off the LCD monitor, and do the prefocus trick. Having the flash on makes no measurable difference.

(Averages of several trials for each).

Picture

Astro PDR-M1

A number of things make it hard to get good shots through the telescope with the PDR-M1. The main problems are the wide 35mm lens, which assures that you will get some vignetting with even the widest eyepiece, and the maximum 1/4 second shutter speed that limits your astro photos to the brightest objects (like the moon!). The small LCD screen makes it hard to check the focus also.

Below is the first quarter moon, taken March 23, 1999 through a Meade ETX astro telescope. Reduced in size from the original shot. I think that it is possible to do better than this, I just need to give it another shot.

Picture
Picture

Mystery White Spots.

"Have you seen a doctor? No just spots" - Old joke.

 Mike Schiltz brought up a problem that he was having with his PDR-M1; white spots on some flash shots. I had noticed these on some of my own shots, but never given it any real thought. Mike seemed to have a particulary bad case of the "spots". Here is one of his shots below, used with Mike's permission.

Picture

Here is one of my own shots( below left) also showing "spots", not as evident as Mike's example.

(Carl demonstrates anti-gravity technique. Don't you wish that you could do that?)

Picture
Picture

Don't take my word for it! You an even make your own shots showing the white spots. Here is how. Set your camera up on a tripod, at night, and shoot into a darkroom from at least 10 feet away. You want the room to be under-illuminated. Use the flash. You might want to use 640x480 "basic" mode to save space. Take at least 20 shots, just fire them off, but don't move the camera. In some shots, (about 1/3 in my testing), you will find the spots. You can also take the camera outdoors and take pictures of the night sky, but the people next door and really going to think that you are strange. See my example room shot above right.

You will find that the spots appear in random locations, and vary in size.

What causes the spots?

Dust. Just ordinary dust in the air, close to the lens, brilliantly illuminated by the flash. Below is an extreme example of this. I took a duster and ran it along the top of my refrigerator. Plenty of dust there! Next I went outside and took some shots of the night sky while shaking the duster downwind. It was necessary to shake most of the dust out, there was just too much on the inital photos.

Picture

Household dust in the air. Cough Chough

You are more likely to see the dust spots in your flash shots where the background is dark, such as in a large room. Especially if a lot of people are moving around, stirring up dust.


Several things *might* make the PDR-M1 / Fuji MX-500 camera more susceptible to the dust effect than other cameras.
   a) The lens on these cameras has large depth of field, so you might notice the spots more on shots taken with this camera vs. other digitals.
   b) Also, the flash is pretty close to the lens, maybe closer than on your typical point 'n shoot camera. This illuminates the dust closer to the lens. For example, consider two hypothetical cameras.   Imagine a dust particle hanging there at the same distance from both cameras. To make the math easy, say that the dust ends up being twice as close to the flash in the second camera. This means that it will get four times as much light from the flash (inverse square law), and will be 4 times as bright as seen by the lens (same distance in both cases). This makes the dust more visible. To make it worse, the depth of field makes the dust visible more closely, so you get closer particles illuminated.
   c) (related to a), above) Also the maximum opening is only f3.2 (this is what it uses for flash) vs. for example f2.8 on the CP900. This just makes the problem worse by increasing the depth of field so that any light reflected by dust focuses into a smaller point.
   d) Reviewers have called the PDR-M1 flash "under powered", so it produces dark backgrounds in large rooms.
   e) Because of the f3.2 max opening, the camera has to use the flash more than other cameras, meaning that you are more likely to have a chance of seeing spots in your shots.
 

That being said, I do have a "flash dust spots" shot that I took with my CP900 last year, and I have seen other examples (even seen people told to send their cameras in for repair for this), so no camera is immune under the right conditions.

 

Cliff Knight sent this really great back of the envelope type analysis of the size of the spots to me, I've just edited the line breaks to make it fit.


The very short focal length of the PDR-M1 lens (7.7 mm) has a depth of ACCEPTABLE focus of 2.8' to infinity when focused at 8' and assuming that an acceptable circle of confusion (COF) is 0.0084 mm (about 1-1/2 pixels). Our dust is not within the range of acceptable focus, but let's see how big an image it would make if it were...

Using a basic magnification formula, a 0.005" (0.127 mm) diameter dust particle, 50 mm from the lens' optical center (57.7 mm from the focal plane) would produce an image of 3.5 pixels diameter--if the lens were focused at 57.7 mm. But it isn't, so we get a larger out of focus "spot". The exact dimensions of the spot can be calculated; I can't remember the formula right now however I can reverse-engineer the thing, sort of.

If we assume an average spot to be about 1/20th of the 1280 pixel image, then the spot is 64 pixels. Each pixel in the 'M1 is 0.0056 mm, so 64 pixels is 0.3584 mm; let's accept this as our acceptable COF. Running the depth of focus calculations again using this COF we find that when focused at 8' the 'M1's lens has "acceptable" focus from about 1-1/2" to infinity.

There it is--our 0.005" speck of dust, 2-1/4" from the focal plane (1-3/4" or so in front of the camera) and well illuminated by the flash (which probably is too close to the lens) will be imaged as a 0.3584 mm (64 pixel) fairly well-defined light spot, washing out whatever else is in the field of view.

I believe the flash unit placement is the main culprit with regard to the 'M1's being PERHAPS more susceptible to this problem. The difference in depth of focus between an f3.2 lens and an f2.8 lens, using the wildly huge COF from above, is only about 0.28"--this does of course assume that the focal lengths of both lenses are the same. I say "perhaps" more susceptible because I have no experience with other models, we need to get user's to use Bryan's "top 'o the fridge" technique to test their cameras.


 

Picture

32 MB card support.

Although Toshiba says that 32MB smart media cards do not work in the PDR-M1, I recieved a msg from Shawn Barham who reports that they do work. He says "I just bought a 32MB SmartMedia for my PDR-M1.  Everyone
says 16MB is max for the camera, but i tried it anyway.  It works
great! I have heard reports of people getting media format error
messages, and some were able to get only 16MB on the 32MB
card.  But, i get a full 32MB on the card, and have had no
problems.  I am using "Olympus M-32PE 32MB SmartMedia".
My camera is a Toshiba "PDR-M1 Ver1.00"."

Picture

FJCOM software for the PDR-M1 and Fuji Cameras.

I've written some DOS software that lets you automatically unload the PDR-M1 and similar cameras through the serial port. This is better than the usual solution of using the Windows (tm) TWAIN driver to to this because that operation requires that you save each picture individually, and requires that the pictures be resaved with loss of quality. FJCOM just copies each file directly from the camera. It is a DOS program because I wanted it to work on my HP200 palmtop.  (News Flash, it doesn't end up working on the HP! The HP is too slow apparently!)

04/08/99 I've added another small program, FJNAME that will rename your files to a date based name based on the date in the EXIF header. It is included in the FJCOM download. Handy to get a grip on all those PDRMxxx.JPG files. (This program does work on other camera files with standard EXIF headers also I have tested it with CP900 files also).

07/10/1999 Version 0.06 adds relaxed timing for the MX700, and a short date option for those with more than 99 pictures on their camera. I have not fully tested the file naming feature for more than 99 files becaue I don't have a big memory card, so be careful!

There is a small .DOC file in the package that explains it all.

Copyrighted freeware.

Click on the link below to get it.

Download FJCOM version 0.06 (50K zip file)