Reading mode: Light Dark

Comparison of 3 pics from Phibus

Go to last post
J

Joel13861

Well-known member
Messages
127
Reaction score
0
Phibus;

My order from best to worst would be #3, #1, #2. Two is definately out; horrible stairstepping along edges of car. On image #1, the car may be a little sharper than #3, but the trees in the background don't look as rich or real, especially the bare trees. Did you prefocus each camera at the same spot? I noticed that on one of the pics you took the shot a little more to the left of the car, which gave different shadowing. Anyway, I hope the Canon came out on top. E-mail me with the results if you would.
Joel

PS: I couldn't respond to your origanally post. I kept getting an error message.
 
I kept getting an error too, my order would be 1,3,2. 1 and 3 are close but I think 1 has better focus and less chromatic aberation on the trees.
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. On image #1, the car may be a
little sharper than #3, but the trees in the background don't look as
rich or real, especially the bare trees. Did you prefocus each camera at
the same spot? I noticed that on one of the pics you took the shot a
little more to the left of the car, which gave different shadowing.
Anyway, I hope the Canon came out on top. E-mail me with the results if
you would.
Joel

PS: I couldn't respond to your origanally post. I kept getting an error
message.
Click to expand...
 
Let me give it a shot.
  1. 1 is S10: I own a S10 and know that the camera tends to wash out details. Look at the trees in the distance. They look washed out and the image is slightly out of focus. Overexposed and washed out are two main characteristics of S10.
  2. 2 is Mx2900: This is easy to tell since it uses a 3:2 aspect ratio.
  1. 3 is Mx4: This is easy as well.
Mx4's image seems to be the best: well-exposed and good color.
Mx2900 is good as well, sharp and good color.
S10 is the worst! Overexposed, washed out, bluish and out-of-focus.
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. On image #1, the car may be a
little sharper than #3, but the trees in the background don't look as
rich or real, especially the bare trees. Did you prefocus each camera at
the same spot? I noticed that on one of the pics you took the shot a
little more to the left of the car, which gave different shadowing.
Anyway, I hope the Canon came out on top. E-mail me with the results if
you would.
Joel

PS: I couldn't respond to your origanally post. I kept getting an error
message.
Click to expand...
 
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. On image #1, the car may be a
little sharper than #3, but the trees in the background don't look as
rich or real, especially the bare trees.
Click to expand...
I disagree, but that might be because my monitor (Sony) is set up differently from yours. As far as I'm concerned,
1) gives the most accurate color, best image detail (but needs add'l sharpening)

2) gives color much too blue, and image detail inferior to (1); foliage is a blur if you look at details of the fir trees, the branches look as if painted in by an impressionist...

3) the color is way oversaturated, in my view this shot has the weakest image detail of all three; foliage is overly colored blur.

From purely the info above, I'd guess 1) Canon, 2) Fuji and 3) Toshiba
Why? Canon has the best lens detail, but some purple fringing (seen in 1 )

Fuji usually yields cool images, pretty good detail (decent lenses...but not quite in the same league as Canon)

As for the third image, the quality of the lens is so crappy that it HAS to be Toshiba.
BTW I own Nikon digicams, none of the above.
regards
Robert Jeantet
 
Phibus:

Great exercise! I wish there were more of this blind testing done - it might lead to getting rid of the strong biases some have . I think that I should be largely unbiased inasmuch as I do not as of yet own any digicams by any of these makers.
  1. 3 appears to be the best overall and the pavement seems to be the most natural color although I would feel more comfortable making a judgement if the perspective here were more like 1 and 2 (#3 was shot low). There appears to be some 'blue fringing' in the small tree limbs in the upper part of the image.
  1. 1 is not far behind 3 but color rendition is slightly off and I don't like the looks of the 'purplish' sky although it may be the sharpest of the three.
  1. 2 is out - the color rendition seems totally out of line with what I would expect in such a scene (appears to be in the south SF bay area - that looks like it might be a street sign for Zanker road in image 1).
Fred H.
 
Hello all.
Keep them inputs a coming!

Thanks for the responses. Sorry about the xoom site not able to D/L the big images ( 900KB).....give it time & just hit refresh if it stalls. Should I use a smaller size next time ? I will do this test again but will include a different scenery shot & add 2 more cameras ( coolpix 800 & 950 ) to the list.

I will solicit any recommendations on how to carry out the next test to make it more objective/useful test.
I will post the answer on Monday of the 3 mystery cameras.

cheers,
Phi
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. On image #1, the car may be a
little sharper than #3, but the trees in the background don't look as
rich or real, especially the bare trees. Did you prefocus each camera at
the same spot? I noticed that on one of the pics you took the shot a
little more to the left of the car, which gave different shadowing.
Anyway, I hope the Canon came out on top. E-mail me with the results if
you would.
Joel

PS: I couldn't respond to your origanally post. I kept getting an error
message.
Click to expand...
 
I would reccomend putting up a tripod and placing all the cameras on it without moving it because 1 of the shots here was from a pretty different angle. Also it would be nice if you posted the settings from each camera after you reaveled their names. Finally if you could take a 35mm shot too that would be great for comparison.
phibus said:
cheers,
Phi
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. On image #1, the car may be a
little sharper than #3, but the trees in the background don't look as
rich or real, especially the bare trees. Did you prefocus each camera at
the same spot? I noticed that on one of the pics you took the shot a
little more to the left of the car, which gave different shadowing.
Anyway, I hope the Canon came out on top. E-mail me with the results if
you would.
Joel

PS: I couldn't respond to your origanally post. I kept getting an error
message.
Click to expand...
Click to expand...
 
Member said:
I will post the answer on Monday of the 3 mystery cameras.
Click to expand...
After posting my answers (somewhere lower in the thread) I was devoured by curiosity as to the accuracy of my guesses; so I looked into the raw data of the 3 files and was delighted to discover that my guesses were 100% right...! It's interesting that different camera brands should have such typical characteristics that one can readily identify the make (if not the model) by the sharpness, detail color quality and intensity of the image.
cheers
Robert Jeantet
 
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. On image #1, the car may be a
little sharper than #3, but the trees in the background don't look as
rich or real, especially the bare trees. Did you prefocus each camera at
the same spot? I noticed that on one of the pics you took the shot a
little more to the left of the car, which gave different shadowing.
Anyway, I hope the Canon came out on top. E-mail me with the results if
you would.
Joel

PS: I couldn't respond to your origanally post. I kept getting an error
message.
Click to expand...
 
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. ......
Click to expand...
Joel:

Your comment with regard to the stairstepping you see on the edges of the car is not an inherent characteristic of the image but is rather an artifact of the display process. It is very likely that you have your display mode set to 800x600 and this leads to the stairstepping you see in image #2. On a display monitor, one will see stairstepping on any image in which the pixel dimension is not an integral multiple (or submultiple - ie 1:2) of the display mode. Image #1 and #3 are an exact multiple of the SVGA display mode at 2:1 (1600/800) while image #2 is a non-integral 2.25 (1800/800).

Fred H.
 
Member said:
car is not an inherent characteristic of the image but is rather an
artifact uof the display process. It is very likely that you have your
Click to expand...
That too. But also: i. insufficient resolution (camera). ii. jpeg compression (camera).

Too high JPEG compression can lead to aliasing (technical term for jaggies), as can be seen from the camera reviews at imaging-resource.com
 
I never considered that. Thanks for the info.
Joel
Fred H. said:
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. ......
Click to expand...
Joel:
Your comment with regard to the stairstepping you see on the edges of the
car is not an inherent characteristic of the image but is rather an
artifact of the display process. It is very likely that you have your
display mode set to 800x600 and this leads to the stairstepping you see
in image #2. On a display monitor, one will see stairstepping on any
image in which the pixel dimension is not an integral multiple (or
submultiple - ie 1:2) of the display mode. Image #1 and #3 are an exact
multiple of the SVGA display mode at 2:1 (1600/800) while image #2 is a
non-integral 2.25 (1800/800).

Fred H.
Click to expand...
 
Fred H. said:
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. ......
Click to expand...
Joel:
Your comment with regard to the stairstepping you see on the edges of the
car is not an inherent characteristic of the image but is rather an
artifact of the display process. It is very likely that you have your
display mode set to 800x600 and this leads to the stairstepping you see
in image #2. On a display monitor, one will see stairstepping on any
image in which the pixel dimension is not an integral multiple (or
submultiple - ie 1:2) of the display mode. Image #1 and #3 are an exact
multiple of the SVGA display mode at 2:1 (1600/800) while image #2 is a
non-integral 2.25 (1800/800).
Click to expand...
Regarding the "checkerboard" effect along the license plate, for example, this is Fuji's 'trademark', it's not due to display properties. (You can check the reviews of earlier Fuji cameras, like the DX-10, on Imaging-resource.com).

Misha
Member said:
Fred H.
Click to expand...
 
robert jeantet wrote:
I'd guess 1) Canon, 2) Fuji and 3) Toshiba

I fully agree with your guess.

Misha
 
Member said:
I fully agree with your guess.
Click to expand...
Misha,

At this stage of development of the technology, for a good observer it's like shooting fish in a barrel. It's interesting that the different manufacturers have produced brand-characteristic "looks" ; one has to wonder who the techs are in their respective companies who judge the results.

Interesting also to note that the perception of color is slightly different between men and women, between Caucasians and Asians, for example. Although most cameras are produced in Asia, I imagine that the color characteristics of some models are biased for different populations, as determined by the companies' marketing division. Or at least, I would hope that they've figured this out by now...

It would be interesting if someone could run a test to see which color rendition various segments of the population preferred... might there someday be a "women's camera" that captures colors the way women see them or prefer to see them?

I think of this because of the variations in color rendition in the very small (3-image) test from which this thread has grown. Then again, I realize that an awful lot of people really don't care all that much about color accuracy, as long as faces look healthy, the sky, vegetation, street surfaces be damned!
regards
Robert Jeantet
 
Fred H. said:
jkiblen said:
Phibus;
My order from best to worst would be #3, #1, #2. Two is definately out;
horrible stairstepping along edges of car. ......
Click to expand...
Joel:
Your comment with regard to the stairstepping you see on the edges of the
car is not an inherent characteristic of the image but is rather an
artifact of the display process. It is very likely that you have your
display mode set to 800x600 and this leads to the stairstepping you see
in image #2. On a display monitor, one will see stairstepping on any
image in which the pixel dimension is not an integral multiple (or
submultiple - ie 1:2) of the display mode. Image #1 and #3 are an exact
multiple of the SVGA display mode at 2:1 (1600/800) while image #2 is a
non-integral 2.25 (1800/800).

Fred H.
Click to expand...
My first impression is that this was a clumsy resizing of the image in an attempt to equalize the sizes for comparison. But no, it's the camera file, says Phibus.

The stairstepping is quite visible on monitors that are solid state and much higher resolution than the internet file. The stairstepping is quite visibly evident.

It isn't a product of the display process but of the original process that made the image. Scaling, interpolation... that sort of thing. Since the rules of the game were stated as No Post Processing, the only assumption we can make is that this was presented as a 1:1 posting, not an "odd pixel of original" versus "displayed image" issue.

Given this stairsteppy result, one would be hard pressed to consider other factors about this individual camera as having any balancing effect.

-iNova
 
Joel:

The comments I made to you regarding the display mode is not the whole story and depending on the image viewer you use, you may not get the results you would expect. However, you can see the effects of, for example, zoom level on the stairstepping. If you experiment a bit with your image viewer, you will sometimes see the stairstepping and sometimes you won't. Try this:

load image #3 into an image viewer (I use both ACDSee32 and Irfanview) and concentrate on the diagonal parking divider strips. In ACDSee32, what I see (800x600 mode) is that when the image is displayed at 50% level (as denoted at the bottom info bar) the parking strips have a straight edge. However, when I zoom out to 40% (you may have to reload the image - F5 in ACDSee) a bad case of stairstepping is obvious. For image #2, I can get the stairstepping to go away if I go to full screen mode (however, the image doesn't fit completely on the screen and have to pan right and left to see that part of the image).

Fred H.
 
Fred H. said:
Peter iNova said:
It isn't a product of the display process but of the original process
that made the image. Scaling, interpolation... that sort of thing.

-iNova
Click to expand...
Peter:
You are quite wrong on this issue. It IS a product of the display
process. Check this out and prove it to yourself:
http://photo.askey.net/forums/read.asp?forum=1010&message=111306

Fred H.
Click to expand...
I'm sure I don't know what technology in the camera is creating the stairstepping effect but when all three camera images are brought into photoshop at their native, original sizes and pieces of each are dropped onto the same new image file without changing anything about them, the stairstepping persists. Each was a 200 x 300 pixel slice.



This is NOT a display problem.

-iNova

Sorry, file kinda large. I didn't want "compression" to be accused of causing an artifact.
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top