Toshiba PDR-M1 Test Images
|
In our normal fashion, we're posting the test images for the PDR-M1 as soon as they are available. We'll post the full review as soon as possible, although travel plans mean that probably won't be until about July 9. (We'll have these images patched into the Comparometer(tm) by Thursday, 6/18/98, so you can readily compare them with those from other cameras.) Several readers have requested that we include more examples of cameras operating in their lower-resolution modes, and we have responded in this set of test images by exercising more of the permutations and combinations of image size and compression level in key test scenes. We have also added two new images, again in response to reader requests. First, we've added a close-up to the outdoor portrait shot, responding to a request for something that would better show portrait performance. (We don't have the resources or time to set up a true studio portrait system, but this should at least give some idea of attainable detail.) Second, we've added a "live" version of the house shot, to test far-field lens performance. This goes against our philosophy a bit, in that the picture is obviously going to change drastically over time, with the weather, seasons, and growth of the landscaping. Some elements will nonetheless stay the same, and we believe the shot will give at least a reasonable indication of optical performance at infinity. Overview As has been the case with several cameras we've tested recently, our sample of the PDR-M1 was an early pre-production sample. (At the time we shot these tests, our unit was one of only *three* in the country!) We found no limitations in the image quality or performance though, the only obvious lack being the software which hadn't caught up yet. Actually, this proved a boon, as it gave us our first chance to work with the "FlashPath" floppy-disk adapter for the SSFDC SmartMedia(tm) memory cards the camera uses. What a great concept - an adapter that lets you plug SSFDC cards into a standard floppy drive! (With appropriate driver software.) With the FlashPath in-house, we also noticed an immediate up-tick in the use of our personal SSFDC-based camera. (To remain nameless here, this is Toshiba's moment to shine...) As you'll see in the images below, the PDR-M1 takes beautiful pictures. Almost equally important, the menus and options are logically laid out, and very easy to navigate. The camera felt great in the hand, and we really liked little details like the rotating lens cover built into its front. Some scheduled travel will delay presentation of the full review for this camera until ~7/9, but the pictures below will speak for themselves. Other good news: Conversations with the Toshiba folks suggest they've lined up unusually broad distribution for this camera (unusual for a new digital camera, anyway), so you shouldn't have any problems buying one, once the first shipments arrive. Oh, yes - the price: Toshiba is projecting $699 as a "street"
price. - For a 1.5 megapixel, 1280x1024 digital camera! A great deal,
and one likely to cause some "adjustment" of competitors' pricing! |
 |
|
Outdoor portrait: (741k)
Truly excellent color! Sharpness and detail is very, very good, with
no trace of compression artifacts or pixellation to be seen. This image
(as were essentially all we captured) was shot with the PDR-M1's "sharpness"
setting in "normal" mode. It appears the camera in this mode applies
a slight sharpening function, as evidenced by subtle but detectable "halos"
where the dark leaves are seen against the white blouse. We suspect the
odd contrast breaks in the shadows around the eyes are a result of this
as well. Shooting with the sharpness set to "soft" would likely
eliminate these, but we didn't have a chance to go back and experiment,
as we had the camera for such a short time. Overall image quality is superb
on this shot though! |
|
||||||
|
|
Closer portrait: (694k)
Did somebody say "detail"?! There's plenty of it here to
go around! Skin tone and tonal range are excellent, but you'd probably want
to use a camera with a longer focal-length lens for this sort of close-cropped
portrait shot: The wide-angle of the PDR-M1 distorts the model's face when
this close. The in-camera sharpening is playing games in the creases of
the model's face again here, and causing the hair to appear a bit thicker
than it might otherwise. Again, we wish we'd played more with the "soft"
setting of the Sharpness control, as we believe it would have eliminated
this minor flaw. As an interesting aside though, the in-camera sharpening
proves valuable when you print the resulting image out to hardcopy: On our
inkjet printer, the PDR-M1 shots looked sharper than others that didn't
sharpen in-camera. Of course, you can always sharpen in Photoshop or other
image-editing program, but this could make it easier to go directly from
camera to print. (A useful comparison is with the Nikon CoolPix 900, which
has similar resolution, but seems to capture finer detail, due to the lack
of in-camera sharpening. Comparing ~7x9 inkjet prints from both cameras
though, the PDR-M1 looks a little sharper.) |
||||||||
|
|
Indoor portrait, flash:
(311k) The PDR-M1 did a good job of balancing between
its flash and the ambient lighting. The result (saved at "Normal"
resolution here) shows good color, with a slight cast still showing-through
from the incandescent lighting. The M1 has the unusual feature of allowing
the photographer to make manual adjustments to flash exposure, in addition
to the ambient-light exposure compensation common to many other cameras.
This shot shows the ambient exposure adjustment
turned all the way down, and the flash turned all the way up. The result
is a slightly underexposed photo, with less of the room light appearing
in the overall color balance. (While we didn't save any images showing it,
the ability to turn down the flash is particularly useful when taking macro
shots.) |
||||||||
|
|
Indoor portrait,
no flash: (305k) The PDR-M1 also did very well
in low-light situations. In this shot, under fairly bright incandescent
lighting (taken with exposure compensation set up two notches, or 0.6 f-stops),
the detail and tonal range are both excellent as is the basic color. We
liked the ability of the PDR-M1 to adjust the exposure compensation in 1/3
f-stop increments. The white balance never completely removed the yellow
cast of the incandescent though. (The image here was shot with white balance
set to "tungsten," which seemed to remove a little more of the
color cast than did the auto setting.) |
||||||||
|
|
House shot: (691k) Very, very good detail, very good color. A slight yellowish cast is evident in this image, which was shot under our precisely daylight-balanced studio lighting, using the "daylight" white-balance setting on the camera. The "auto" setting tended to produce rather over-warm (magenta-cast) images in the studio, although this wasn't evident in our outdoor shooting under "real" sun. The in-camera sharpening of the "normal" setting for image sharpness is evident here again, in the form of slight halos along the roof edges, and a minor coarsening of the finest detail in the tree branches against the sky. We saw it only here and in the resolution target, but there's an odd pixel artifact visible between the vanes of the louver in the front gable of the house. Overall, truly excellent image quality from a camera this affordable! We shot this image at both sizes the camera supports (1280x960, 640x480), in all compression modes, so you can see how the different compression settings affect final image quality. The entries in the matrix below refer to image size/compression level.
|
||||||||
|
|
Far-Field shot: (695k) This is the second of the new shots we mentioned above. It is shot at infinity to test far-field lens performance. NOTE that this image cannot be directly compared to the other "house" shot, which is a poster, shot in the studio. The rendering of detail in the poster will be very different than in this shot, and color values (and even the presence or absence of leaves on the trees!) will vary in this subject as the seasons progress. In general though, you can evaluate detail in the bricks, shingles, and windows, and in the tree branches against the sky. Compression artifacts are most likely to show in the trim along the edge of the roof, in the bricks, or in the relatively "flat" areas in the windows. Here again, we wish we'd known to try a shot with the "sharpness"
adjustment set to "soft" vs the "normal" setting we
actually. On the other hand, we don't want to harp on this: The detail
shown here is very close to the best we've seen from any camera, even
models costing hundreds of dollars more. At first glance, compression
artifacts are completely invisible, but a closer look at 200% (you'll
have to download the images & view them in an image editor to do this)
shows a few subtle ones here and there. (Look at the inside edges of the
windows at high magnification, also in the bricks above the first floor,
right-hand windows.) Overall, great color, great detail, no falloff of
lens performance at infinity. |
||||||||
|
|
"Musicians" poster: (656k) In this shot, we once again did the full matrix of image size/compression ratio (6 combinations) so you can evaluate how the camera performs across the full range of final file sizes. As with our other studio shots, this one was taken with the camera set to "daylight" white balance. The result is a slightly warm tone overall, but with excellent color balance and saturation, and very natural skin tones.
This was one shot we did take with all three "sharpness"
settings. The main version shown above was shot with this control set
to "normal." Here is the full series, with sharpness set to
soft, normal,
and hard. Notice how much more fine
(but subtle) detail is available in the "soft" version. If we
were going to use the images exclusively for inkjet prints, we'd prefer
the "normal" or even "hard" settings. On the other
hand, for archival purposes (assuming subsequent sharpening in the image
editor), we'd choose "soft." |
||||||||
 |
Macro shot: (651k) For a camera with a wide-angle lens, the PDR-M1 did surprisingly well in macro shooting, although the very close 3.5 inch minimum working distance made lighting a little tricky, to avoid catching a shadow of the camera or photographer's hand in the image. The area captured at closest approach is 3.0 x 3.8 inches (7.7 x 9.6 cm), a very respectable performance. Here, we show you three versions, all at high resolution, varying only in compression setting.
We were surprised at how well the PDR-M1's flash worked in extreme close-up
situations. This image (normal
res, 327k) was shot with the flash at its default setting, at a
working distance of 3.5 inches. The reflection off the silver dollar confused
the exposure system, but we could have compensated by turning up the flash
exposure a step or two in manual mode. |
||||||||
 |
"Davebox" test target: (624k) Excellent color, very good tonal range. The previously-noted
slight yellow cast from the combination of the camera's "daylight"
setting and our studio "daylight" lighting is once again evident,
but overall color is very, very good. Color saturation is just right in
the strong primaries of the MacBeth chart, and the delicate pastels of
the Q60 target (even down to a glimmer of color in row "A")
are preserved well. Tonal discrimination is maintained down to step 18
of the large gray scale, albeit with a fair bit of noise at that level.
Highlights detail in the gauze is preserved well also. (By the way, we
found that an "auto levels" operation in Photoshop(tm) did an
excellent job of removing the slight yellow cast.)
|
||||||||
 |
"WG-18" resolution target: (630k) Resolution is very good, although the behavior is somewhat different vertically and horizontally. Horizontal resolution is a solid 700 line pairs/picture height, while an odd pixel artifact appears in the vertical direction, with closely-spaced horizontal lines. (This "zipper" artifact was also evident in the House shot, in the horizontal vanes of the louvered vent in the central gable.) Vertical resolution appears to extend to 700 lp/ph as well, but the checkerboard pattern of the artifact makes it difficult to discern exactly where the lines of the resolution target finally give out. This artifact only appears in horizontal lines: The camera actually shows resolution approaching 800 lp/ph at an angle of 45 degrees, albeit with slight color aliasing. Overall resolution is close to the best we've tested to date (June, 1998), but we're a little hard-pressed to assign a number in the vertical direction.Virtually NO color artifacts are visible, and geometric distortion is practically non-existent. Herewith the size/compression matrices:
We also shot the resolution target with the "Digital
Zoom" enabled. |
||||||||
 |
Viewfinder accuracy/flash uniformity target: Flash
uniformity on the PDR-M1 is very good, but the viewfinder is a little "loose,"
both optically and on the LCD screen. The LCD
was the most accurate, showing about 88% of the area captured by the sensor,
horizontally and vertically. The optical finder
only showed 80%. The LCD orientation produced final images offset slightly
to the top left of what was centered in the display, while the Optical finder
was nearly centered vertically, but biased slightly to the right. We consider
the overall accuracy of the optical finder to be a trifle sub-par (maybe
4 out of 10, with 5 being average), while the LCD is about typical for digital
point & shoots. |
Follow Imaging Resource: