Digital Cameras - Toshiba PDR-M5 Test Images
(Original test posting: 11/5/99)
We've begun including links in our reviews to a Thumber-generated
index page for our test shots. The Thumber data includes a host of
information on the images, including shutter speed, ISO setting, compression
setting, etc. Rather than clutter the page below with *all* that detail,
we're posting the Thumber index so only those interested in the information
need wade through it! ;) |
|
Movies: We put this at the top of the
page, because it's both a neat and unusual capability of the PDR-M5: It
can capture AVI-formatted movies for playback on the camera, or on your
computer. We've seen "mini-movie" options in digicams before,
but not thus far anything like the PDR-M5, which records up to 120 seconds
of (silent) video on the standard memory card, proportionately more on larger
cards. We suspect this would be a very popular feature in any household
with kids, and can also imagine it being useful for things like studying
golf or tennis swings, etc. You can record in two different image sizes
(160x120 or 320x240), at any of three different compression settings. The
two samples below were "filmed" at the small/high quality and
large/medium quality settings. (Depending on how your browser is set up,
and whether or not you have QuickTime(tm) installed, these links may or
may not work for you. If the browser doesn't know what to do with it, try
downloading to disk and then double-clicking on the file, in case your computer
has QuickTime, but it just isn't connected to your browser properly.).
|
|
||||||||||||||||
|
|
Outdoor portrait:
(838k) Good color, very good resolution and detail.
Tonal range is good, but the darkest shadows have somewhat higher noise
levels than we've seen in competing units. Some cameras have pixelation
problems between the strong red and blue elements in the flowers, but the
PDR-M5 shows no sign of these. In common with many digicams, it has some
problems with the blue hues in the flowers and the model's pants, leaving
them slightly purplish. Color saturation overall is just a touch low, and
the image has a slightly bluish cast. We took our main
shot (838k) using the "Daylight" white
balance setting, and with +0.6EV of exposure compensation. We also tried
the "auto" white balance setting, which produced a slightly bluer
cast overall. (Sorry, didn't save the shots for reproduction here. As with
most cameras, the exposure meter responded to the very strong highlights
in the model's shirt, dropping the default exposure
(816k) somewhat. We did experience some odd behavior
with the M5's EV adjustment though, in that it wasn't always consistent
from shot to shot: Two shots with the same setting could easily come out
with 0.3EV difference in the actual exposure. The table below shows the
results of a range of exposure compensation settings, from 0EV to +1.5EV,
all shot with the Daylight white balance setting.
Exposure compensation series:
|
|
|||||||||||||||
|
Closer portrait:
(813k) This shot generally requires less exposure
compensation than the one above, since the model's face fills more of the
frame. With the PDR-M5, it looked like it actually wanted about 0.6EV of
compensation again, but there was very little difference between the +0.3
and +0.6 EV settings, and then a large jump between +0.6 and +0.9 EV. We
chose the +0.9 EV shot (813k)
shot because it had better fleshtones than the +0.6EV
one (794k), but would have been happier with just
a bit less exposure. We shot this test using the "auto" white
balance setting, so you could see the slightly more bluish cast it produced
as compared to the "daylight" setting used above. Good color and
excellent detail, good tonality as well. (No odd tonal breaks in the creases
of the model's face.) The table below shows a range of exposure compensation
settings from 0 to +1.5EV, all shot with the auto white balance setting.
Exposure compensation series:
|
|||||||||||||||||
|
Indoor portrait,
flash: (641k) Only one shot here, as
there weren't really any variations to try. The basic flash exposure worked
very well though, with a nice tonal balance, and good color. The color is
very slightly "warm", probably a result of the strong incandescent
lighting in this scene, but the overall result
(641k) is one of the more color-accurate shots we've
seen of this test using on-camera flash. Overall, a very good performance. |
|||||||||||||||||
|
|
Indoor portrait,
no flash: (619k) This shot is a very
tough test of a camera's white balance capability, given the strong yellowish
color cast of the household incandescent bulbs used for the lighting. With
the white balance set to "incandescent," the PDR-M5 did a credible
job, as seen in our main shot (619k),
although it needed the exposure compensation set up by 1.2 EV units. By
contrast, the auto white balance setting produced a very yellowish image,
as seen here (619k). The
table below shows a range of exposure compensations, from 0 to +1.5 EV,
all shot with the incandescent white balance setting
Exposure compensation series:
Contrast in the images was somewhat low, although a full tonal and color range was present. For the fun of it, we tried adjusting the images, both in Photoshop(tm) and our favorite easy-to-use, ultra-cheap image adjuster PhotoGenetics(tm). The results were very interesting, as shown in the table below:
Image-adjustment examples:
We went through quite a few "generations", but the results were worth it, and never required more than deciding whether we liked image "A" or "B" better. Significantly better color than our own laborious tweaking in Photoshop (IOHO), and the best part is the "genotype" is ready to apply to other images from the same camera! Pretty cool! |
|||||||||||||||||
|
House shot:
(949k) Always a tough test of camera resolution, the
PDR-M5 turned in results that were about average of the 2 megapixel field
that we've tested so far (October, 1999), although to its credit, the sharpness
doesn't fall off in the corners as much as we've seen on some cameras. Color
is good, with a slight yellowish tinge, contrast is slightly low. Our main
shot (949k) was taken with the white balance set
to auto. Here are a pair of low-resolution shots, taken with auto
white balance (233k) and daylight
white balance (232k). (The daylight version resulted
in a more pronounced color cast.) Overall a good, but not exceptional performance.
We also observed that the low-resolution (800x600) images of the PDR-M5
aren't as sharp as you might expect: For smaller images, you'd probably
be best off capturing the image at the larger file size with maximum compression,
and then reducing the size in an image-editing program after the fact. The
table below contains samples of all the resolution/quality modes: Resolution/Quality Series:
We also shot versions of this image in the largest/highest-quality mode, testing the three "sharpness" settings. The results are in the table below: Sharpness Series:
|
|||||||||||||||||
|
Far-Field shot:
(933k) This image 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 window detail, 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. Another tough resolution test, in this one, the PDR-M5 captures all the detail, but the in-camera sharpening doesn't do as good a job as it could making it evident in the final image. The upside is that the resulting image takes sharpening unusually well in Photoshop. Color is quite good, in our main shot (933k), taken with the auto white balance setting. (For comparison, here are two low-resolution images, taken with automatic white balance (219k) and daylight white balance (215k) ) The strongest highlights in the house are blown out, but that's not entirely the camera's fault, because: 1) The house was repainted just a month or so prior to this shot, with brighter white paint than previously, and also without the mildew tinge present in many earlier shots. 2) Since the camera's exposure system responded to the extremely bright white, darkening the rest of the image, we applied exposure compensation of +0.3EV to brighten the image in our main exposure series. This exacerbated the loss of highlight detail. The table below contains samples of all the resolution/quality modes, all shot using the auto white balance setting: Resolution/Quality Series:
We also shot versions of this image in the largest/highest-quality mode, testing the three "sharpness" settings. The results are in the table below: Sharpness Series:
The PDR-M5 has a variable-ISO option, which we also exercised on this shot. As usual, boosting the ISO value increases noise, which is fairly easy to see in the relatively flat-tinted gray of the roof on the house. The results are in the table below: ISO Series:
|
|||||||||||||||||
|
|
Lens Range (new): We've received
a number of requests from readers to take shots showing the lens focal length
range of those cameras with zoom lenses. Thus, we're happy to present you
here with the following pair of shots, showing the field of view with respectively,
the normal shooting mode, and with the "digital telephoto" option
engaged. Note that both images here are shot in low-resolution mode, to
ease download times. In normal shooting mode, the larger image size is also
an option, but in digital telephoto mode, only the smaller file size is
possible.
|
|||||||||||||||||
|
"Musicians"
poster: (898k) The detail in this shot
is quite good, and the color is accurate, but rather muted by lower contrast
overall. As we noted earlier, the detail is there, but the image is a bit
soft overall. Sharpening in Photoshop brings out much more detail than is
evident in the raw camera image. Again, we chose the automatic white balance
option over the daylight setting as being the more accurate. The table below
shows the results obtained with all resolution/image quality setting combinations: Resolution/Quality Series:
We also shot versions of this image in the largest/highest-quality mode, testing the three "sharpness" settings. The results are in the table below: Sharpness Series:
|
|||||||||||||||||
|
Macro shot:
(898k) PDR-M5 macro performance is good, if not microscopic:
As seen in this sample (898k), the minimum capture
area is 3.0 x 3.9 inches (75 x 100 mm). The digital
telephoto (194k) can be used in macro mode, adding
a 2x magnification, but only by halving the image size. The flash is only
rated to a minimum distance of 90cm, but we found was very capable of throttling
down for macro work, as seen here (894k). |
|||||||||||||||||
|
"Davebox"
test target: (784k) In this test, the automatic
white balance option again won out, producing the exceptional color seen
in our main shot (784k).
Here's a low-res example of the daylight white balance
option (150k), which you can see produced rather
yellowish results. This was really an excellent performance, with very accurate
colors, proper saturation, excellent separation of the tough red/magenta
colors, and good tonal range, with good detail in both highlights and shadows.
Overall, very impressive! The table below shows the results obtained with
all resolution/image quality setting combinations: Resolution/Quality Series:
We also shot versions of this image in the largest/highest-quality mode, testing the three "sharpness" settings. The results are in the table below: Sharpness Series:
|
|||||||||||||||||
|
|
Low-Light Tests Along with its little brother the PDR-M4, one of the PDR-M5's real strong points is low-light shooting: Toshiba has taken advantage of the very powerful processor they've built into the camera to do some very sophisticated processing to "subtract" the noise and work around "stuck" pixels in the sensor when shooting at very low light levels. (They do this by taking two shots for every picture when the camera is working in "bulb" mode. The first shot takes the picture for you, the second is taken with the shutter closed, and provides a noise calibration. By essentially subtracting the second from the first, they cancel-out a lot of the noise that would otherwise be produced.) In our testing, we felt we obtained very good results down to a light level of only 0.5 foot-candles (5.5 lux, the level we've been referring to as 6 EV in previous reviews.) Even at a light level of 0.25 foot-candles (2.7 lux), we obtained this shot (948k). - It has a very strong magenta cast, but a little levels-correction in Photoshop produced this result (948k). That's pretty amazing! 0.25 foot-candles is so dark we have a hard time finding our way around the studio. For comparison, a typical night scene lit with streetlights is about two full EV units (f-stops) brighter, at 1 foot-candle! The table below has links to images shot at light levels ranging from 8 foot-candles (88 lux) down to 0.25 foot-candles (2.7 lux). The thumbnail index page contains links to other low-light shots, taken with different exposure settings, for those who might be interested.)
|
|||||||||||||||||
|
|
Flash Range Test (New) (This test was added in August 1999, so cameras tested before that time won't have comparison pictures available. As we go forward though, all the new models will have similar tests available.) Toshiba specifies a maximum range for the on-board flash of 2.5 meters, or about 8 feet. We found that the flash was indeed brightest at that distance, but felt it could be considered usable to at least 10 feet. We were surprised though, that it doesn't appear possible to use the flash in conjunction with a boosted ISO value. - This would have roughly doubled the maximum usable distance. The table below shows the results we obtained with the PDR-M5's onboard flash, at distances ranging from 8 to 13 feet.
|
|||||||||||||||||
|
ISO 12233 ("WG-18")
resolution target: (787k) The PDR-M5's results
on the resolution target generally supported it's behavior on the other
tests we performed. With a horizontal visual resolution of 700 lines per
picture height, and a vertical visual resolution of 600-625, it's about
midrange in the current crop of 2 megapixel digicams. As the test pattern
frequency increases beyond its resolution limit, it shows moderate color
aliasing. As noted earlier, its half-resolution mode doesn't provide the
best results: Sharpness is way off, and it's very prone to color aliasing.
- If you need to conserve memory space, we'd recommend shooting the higher
resolution, only in the lowest quality mode, and shrink the images down
in the computer after the fact: The results will be far superior. The tables
below show the usual array of images shot in various combinations of size,
quality, and sharpness setting, at both wide angle and telephoto settings: Resolution Series, Wide Angle:
Resolution Series, Telephoto:
We also shot versions of this image in the largest/highest-quality mode, testing the three "sharpness" settings. The results are in the table below: Sharpness Series:
|
|||||||||||||||||
|
Viewfinder accuracy/flash uniformity target:
The PDR-M5's optical viewfinder proved to be rather "loose" in
our parlance: That is, considerably more of the subject ended up in the
final image than was seen through the optical. The LCD finder was about
average. At the wide-angle end of the lens' range, the optical
viewfinder (225k) shows about 75% of the final
view, while the LCD viewfinder (215k)
shows about 90%. At the telephoto end of the lens' range, the optical
viewfinder (214k) shows 72.5% of the final view,
while the LCD viewfinder (202k)
again shows 90%. (Surprising as it may seem, most LCD viewfinders show less
than the full image area, so the 90% result for the D-M5's LCD is about
typical.) In digital tele mode (56k),
the LCD finder showed 88% of the final view. Flash uniformity is better than most cameras, with only very slight light falloff in the corners at the wide-angle end of the zoom range. We've recently begun testing cameras for optical distortions, such as barrel/pincushion distortion and chromatic aberration. The PDR-M5 showed moderate barrel distortion at the wide angle end, measuring 0.9% deviation across the width of the frame, and about 0.65% barrel distortion in telephoto mode. Chromatic aberration was almost non-existent at both telephoto and wide angle lens settings. (At the wide angle end, there's slight flare in the blue channel that could be mistaken for chromatic aberration, but even this is very minimal.) |
Back to Toshiba PDR-M5 Review
Jump to Comparometer(tm) to compare with other cameras
Or, up to Imaging Resource Cameras Page.
Follow Imaging Resource: