Olympus C-5050 Review on Dpreview!!!

However, it does appear that lowering
the in-camera sharpening decreases the apparent noise pretty
significantly. A good example of this can be found in the review of
the C-5050 at
http://www.dcresource.com/reviews/olympus/c5050z-review/index.html .
In addition, this example at DCResource illustrates how lowering the in-camera sharpening also reduces the "jaggies".

Just thought I'd state more of the obvious ;)

--
Kevin

*********************************************************
Olympus C-3040Z
Tiffen MegaPlus 2x teleconverter
Promaster 5750DX flash
 
Overall, I think this was a very fair review and the way I read it
the conclusion was extremely positive--I would think the fact that
the resolution matched the vaunted 717s would raise more than a few
eyebrows among those who expected little from the "same old lens"
and a an supposedly overcrowded ccd. Factoring in the price of this
camera only makes it look like a better value.

I was surprised however that Phil didn't really shed light on the
"aggressive default in-camera sharpening vs. noise" issue. He
admits that the default sharpening is too aggressive and states his
preference for a setting of -2. But it appears the noise tests are
done with sharpening at 0. I would have liked to see the "noise"
test done at different levels of sharpening. The point, as someone
mentioned above, is to see how the camera performs at it's optimal
settings. And of course if I am comparing across camera's I want to
see how they compare at their optimal settings--the mere fact that
I can adjust the settings makes the deafult limitations irrelevant.
I realize this just adds to the reviewer's time burden, but it
seems pretty simple to do I'm disappointed this reveiw didn't clear
it up.

As for CA, I guess it's just less of a problem for me in practice.
I have yet to attempt to portray aluminum foil in my shots so maybe
I just have a different preference in subjects from some of you.
But seriously, after taking probably 20,000 pictures with a C3040,
c2100 and now the c5050 (all cameras with perhaps above average
propensity to show the purple fringing whether it's technically
"CA" or "blooming"), I can honestly say that it has yet to ruin a
really high quality picture of mine. It does show up in many
pictures that are over-exposed or try to capture scenes too high in
contrast. But to my eyes, these are pictures that suffer from flaws
in composition and or exposure, anyway, and the purple fringe is
but a symptom of that. Anyway, if you take a great shot and you get
too much for your tastes there are several very easy ways to
correct for it. In fact, I often find that pictures with a lot of
purple fringing tend to look much better in B&W than in color
anyway because of the areas of extreme contrast. Also the amount of
frining in a printed picture is much, much less noticeable than the
100% or 2000% screen view.

One last observation. The acid test of any camera's picture quality
has to be prints rather than 100% screen enlargements. Many of the
imperfections you see at 72 dpi are simply not visible at 300dpi.
It is impossible to show print quailty in an online review, but I'd
at least like to hear the reviewers opinion of the same size prints
of identical scenes from a single printer across camera. My
experience using an Epson 1280 is that the 5050 makes excellent
prints--I've done up to 11x14, so far. They are very sharp; color
is very good and I just haven't seen any p[roblem with noise
(especially since much of the "noise", I believe, is in fact,
sharpening artifacts which always look worse on screen than in the
print because the appearance of oversharpening compensates for the
softness induced by the digital printing process).

Anyone who wants to see what good pictures from the 5050 look like
in actual use rather than contrived test shots is wlecome to check
out mine (alebit in 800x600 size) on pbase:
http://www.pbase.com/jfeinleib/olympus_c5050_samples

In sum, I find it hard to believe that anyone who pays $700-800 for
this camera will be disappointed with the quality of pictures it
can produce to say nothing of how much a pleasure it is to actually
operate.

http://www.pbase.com/jfeinleib
http://www.photosig.com/viewuser.php?id=23754
--
Joel,

I just returned from your pbase gallery and was truly impressed. I'd buy this camera based on your photos regardless of what all the reviewer's have to say.........

'Everything has its beauty but not everyone sees it.'
Confucius
 
I do not have words to say , all that i know is that the most common F stops , starts from the lower until F5 F6 as max , this is at list what the cameras selects most of the times , settings like over F6 ,goes to manual operations , and the shutter speeds over the F6 is not fast , so at the end all this settings effects the usability allot by a negative way ..

Something that i do not aware is if the ISO 200 or 400 helps any for faster shutter speeds !!
CA is the result of the lens, but the complaint that Phil had
really isn't CA, it is blooming.

Blooming occurs when a photosite on a CCD completely saturates in
other words goes to white and the adjacent photosites are still
dark. When a photosite completely saturates the CCD must do
something with the excess charge. So on the CCD there are
"electronic gutters" that carry the charge off the chip. However,
these gutters can only carry so much charge. When they can carry
no more, the charge floods into adjacent photosites and causes
blooming or more commonly known as purple fringing. The chip that
is used on this camera has a very high pixel density. In other
words it's a small chip with a heck of a lot of pixels. And
because of this the "gutters" are very small and not able to deal
with excessive charges.

You are right in that this is a hardware issue and cannot be fixed
by a firmware up date, but the problem actually lies in the CCD and
not the lens.

Now there is one way to prevent the purple fringing and that is to
use a very small aperture like f7 or f8. This slows the rate at
which the photosites recieve the charge and therefore make it
possible for the "gutters" to do their job.
--



Olympus..C21OOUZ since 28/1/02 A-2OO tele, Oly P-4OO ..
HP97OCxi ZI0_USB Sanyo16OO Manfrotto 19ODB+141RC
My mini gallery http://www.sigma-sa.com/kt/pelion/index.htm
' My primary language is Greek '
--



Olympus..C21OOUZ since 28/1/02 A-2OO tele, Oly P-4OO ..
HP97OCxi ZI0_USB Sanyo16OO Manfrotto 19ODB+141RC
My mini gallery http://www.sigma-sa.com/kt/pelion/index.htm
' My primary language is Greek '
 
Does the C5050 have gamma adjustment ?

--

C2100UZ, D600L and that 1.45x teleconverter lens(fits both cameras) Sharp VE-CG40, Kenko 3X, DSP9000, PBase Supporter, Dpreview supporter.
 
I'd think comparing C-5050 shots with sharpening set to 0 with F717
or G3 also with sharpening set to 0 is the only fair comparison
because that's the default setting! Do you think setting one camera
sharpening to -5 and the other to +5 will give you a fair or equal
comparison?
No one would suggest testing one camera at -5 against another at +5.

What people are suggesting is the reviewer should attempt to get the best out of the camera.

If that means for some camera you always shoot with sharpening down a notch or two while for others the default setting is best, so what?

Its well known some cameras default is over aggressive e.g. Fuji 602, so if you can cure the problem by turning it down a notch why would you not do so?

I prefer Phil's review format to that of Imaging resource BUT over there, the high contrast shot is shot at default and if that gives a poor result, is re-shot to get the best result possible by altering the settings.

It's not unfair to show each camera tested in its best light. Then we know what it is capable of. Can the same be said if only defaults are used?

The default settings for the test images in Phils reviews are the only fault IMO.

This is a recurring theme with his reviews. Sometimes he mentions what effect changing the settings has, but mostly only in passing.

Suppose a lower setting got rid of the jaggies? Why not shoot the test target at that setting and recommend users use the lower setting as a default?

It seems people are daft enough to write a camera off on the back of Phil's reviews if they don't get highly recommended when in fact, if the problem is with the images, a bit of tweaking of the settings could change that.

Dave
 
(those were Phil's rules, not mine)

Anyways, instead of engaging in character assassination, why not prove that I'm wrong by posting some examples?

To reiterate:

1. I took the crop shown in the upper right and used Photoshop to set the dpi to 300 with NO RESAMPLING. There was NO altering of the image data, just the dpi setting.

2. Printed the image on my 6 color Canon BJC-8200 (predecessor of the S800).

3. Took the printout and scanned it on my Microtek scanner at 600 dpi. This is the top image.

4. I took the crop shown in the upper right and used Photoshop to set the dpi to 150 with NO RESAMPLING. There was NO altering of the image data, just the dpi setting.

5. Printed the image on my 6 color Canon BJC-8200 (predecessor of the S800).

6. Took the printout and scanned it on my Microtek scanner at 600 dpi. This is the bottom image.

NEVER AT ANY TIME WAS THE CROP OF THE IMAGE (straight from the camera) ALTERED BEFORE PRINTING.

The purpose of scanning at 600 dpi is to show clearly that the jaggies are reproduced as-is from the original image - that the printing process does not obscure defects. Note that the question as to whether or not the jaggies are visible to the viewer of a printout is NOT what I was addressing.

Rhetoric is no substitute for proof.

(final response)
He routinely doctors up images in his comparisons of cameras to
prove his point, rather than provide straight examples.
Downsampling, Upsampling, blowing up, sharpening pretty much as he
decides.

In this example which he has provided, I believe he blew up the
photograph a few hundred percent, then printed it out at 300 DPI,
rather than printing it at 100% actual size.

He doesn't acknowledge (or perhaps he doesn't understand?) the fact
that if you take a 5 MP image, crop a small portion and blow it up
so that the resulting image is at say 50 dpi resolution.. even if
he printed it on a 12,000 dpi printer, the resulting image is STILL
at 50 dpi.

Wonderful work, Einstein.

K.
How do 5050 photos look when printed with respect to the jaggies
mentioned in the review?
What you see on the monitor is what you get on the printout.
Today's high resolution printers resolve enough detail that all
image flaws like CA and jaggies are faithfully reproduced.

Example scan of 5050 printout:

http://www.pbase.com/image/7937152/original
 
The purpose of scanning at 600 dpi is to show clearly that the
jaggies are reproduced as-is from the original image - that the
printing process does not obscure defects. Note that the question
as to whether or not the jaggies are visible to the viewer of a
printout is NOT what I was addressing.
I agree. Printing does not alter the character of an image as some would suggest. Just curious though, can you see the jaggies at 300dpi? 150 dpi?
 
Anyways, instead of engaging in character assassination, why not
prove that I'm wrong by posting some examples?
Speaking of examples, I would appreciate it if you would take down those bastardized images off your webpage - considering those pictures were taken by myself, and used by you without any permission or prior consultation. Thank You.
To reiterate:

1. I took the crop shown in the upper right and used Photoshop to
set the dpi to 300 with NO RESAMPLING. There was NO altering of the
image data, just the dpi setting.
Changing the DPI setting in Image Settings changes the Image data - particularly when you're sending the image to a printer. Try it. Get a 600 DPI image, change it to 5 dpi. Print. Now tell me that the image printed at 600 dpi is equivalent to the image printed at 5 dpi.

Your printer is a CMYK device, and the image displayed on your monitor is RGB. Even when it is changed to a CMYK simulation on the screen, it's still RGB. You have R G and B phosphors on the screen. Your Printer drivers convert that to CMYK or CMYK+2 or whatnot when you're printing.

Voila. Altered image data. Rest is moot.
2. Printed the image on my 6 color Canon BJC-8200 (predecessor of
the S800).

3. Took the printout and scanned it on my Microtek scanner at 600
dpi. This is the top image.

4. I took the crop shown in the upper right and used Photoshop to
set the dpi to 150 with NO RESAMPLING. There was NO altering of the
image data, just the dpi setting.

5. Printed the image on my 6 color Canon BJC-8200 (predecessor of
the S800).

6. Took the printout and scanned it on my Microtek scanner at 600
dpi. This is the bottom image.

NEVER AT ANY TIME WAS THE CROP OF THE IMAGE (straight from the
camera) ALTERED BEFORE PRINTING.

The purpose of scanning at 600 dpi is to show clearly that the
jaggies are reproduced as-is from the original image - that the
printing process does not obscure defects. Note that the question
as to whether or not the jaggies are visible to the viewer of a
printout is NOT what I was addressing.
Note that you changed the DPI with resampling off. Effectively, this scales the image 200% without any resampling - ie just makes the square pixels bigger.

If you truly want to see if, or how badly the 'jaggies' are apparent on a printed image, go ahead and print the image at 600 dpi. Then scan at whatever resolution you want. You may think that you're being kind or extremely fair and methodological, but your samples have been created to emphasize a flaw rather than to investigate if the results produced are acceptable, or even workable.

The beef that people have with you is that you are deliberately trying to make-worse an imperfection in a tool, rather than appreciating the tool for what it -can- do. Do you understand?
Rhetoric is no substitute for proof.

(final response)
He routinely doctors up images in his comparisons of cameras to
prove his point, rather than provide straight examples.
Downsampling, Upsampling, blowing up, sharpening pretty much as he
decides.

In this example which he has provided, I believe he blew up the
photograph a few hundred percent, then printed it out at 300 DPI,
rather than printing it at 100% actual size.

He doesn't acknowledge (or perhaps he doesn't understand?) the fact
that if you take a 5 MP image, crop a small portion and blow it up
so that the resulting image is at say 50 dpi resolution.. even if
he printed it on a 12,000 dpi printer, the resulting image is STILL
at 50 dpi.

Wonderful work, Einstein.

K.
How do 5050 photos look when printed with respect to the jaggies
mentioned in the review?
What you see on the monitor is what you get on the printout.
Today's high resolution printers resolve enough detail that all
image flaws like CA and jaggies are faithfully reproduced.

Example scan of 5050 printout:

http://www.pbase.com/image/7937152/original
 
Changing the DPI setting in Image Settings changes the Image data -
not without resampling it doesn't
particularly when you're sending the image to a printer. Try it.
Get a 600 DPI image, change it to 5 dpi. Print. Now tell me that
the image printed at 600 dpi is equivalent to the image printed at
5 dpi.
it's the same batch of dough, whether you make it into a ball or roll it flat - your 5 dpi image will be a mile wide. It will look different, but the image data has not changed.
Your printer is a CMYK device, and the image displayed on your
monitor is RGB. Even when it is changed to a CMYK simulation on the
screen, it's still RGB. You have R G and B phosphors on the screen.
Your Printer drivers convert that to CMYK or CMYK+2 or whatnot when
you're printing.
I have no idea what effect you think this would have on jagged lines.
Voila. Altered image data. Rest is moot.
not really
2. Printed the image on my 6 color Canon BJC-8200 (predecessor of
the S800).

3. Took the printout and scanned it on my Microtek scanner at 600
dpi. This is the top image.

4. I took the crop shown in the upper right and used Photoshop to
set the dpi to 150 with NO RESAMPLING. There was NO altering of the
image data, just the dpi setting.

5. Printed the image on my 6 color Canon BJC-8200 (predecessor of
the S800).

6. Took the printout and scanned it on my Microtek scanner at 600
dpi. This is the bottom image.

NEVER AT ANY TIME WAS THE CROP OF THE IMAGE (straight from the
camera) ALTERED BEFORE PRINTING.

The purpose of scanning at 600 dpi is to show clearly that the
jaggies are reproduced as-is from the original image - that the
printing process does not obscure defects. Note that the question
as to whether or not the jaggies are visible to the viewer of a
printout is NOT what I was addressing.
Note that you changed the DPI with resampling off. Effectively,
this scales the image 200% without any resampling - ie just makes
the square pixels bigger.
It preserves the image data without any interpolation. He is not testing how good a job photoshop does at smoothing out jagged lines with interpolation. I think the entire point was to show a printout of the jagged lines at a larger scale (a point that he makes obvious by showing the original image crop) so that you can easily see the result.
If you truly want to see if, or how badly the 'jaggies' are
apparent on a printed image, go ahead and print the image at 600
dpi. Then scan at whatever resolution you want. You may think that
This is nonsense. printing at 600 dpi is ridiculous and would yield a printed image so small from that little crop, that you wouldn't be able to make ANYTHING out. Scanning at 600 dpi is perfect for making enlargements and is the exact same resolution I would have chosen.
you're being kind or extremely fair and methodological, but your
samples have been created to emphasize a flaw rather than to
investigate if the results produced are acceptable, or even
workable.
I believe his entire point is to emphasize the flaw. I think his methodology is sound once you realize that he is merely presenting an enlarged view of a printout made at the standard print resolution of 300dpi. I find it difficult to flaw his procedure, but find many holes in the attack on it.
The beef that people have with you is that you are deliberately
trying to make-worse an imperfection in a tool, rather than
appreciating the tool for what it -can- do. Do you understand?
I think this is the root of all this - sour grapes. If anyone doesn't think the jaggies are a problem by viewing 5050 pics on screen, I don't think this will put them over the edge. Noise seems to disappear ALOT when printing - jaggies....well....maybe not as much. Not trying to antagonize or step on toes, I just had to step in as I think Jared's getting a bum rap here.
-Wow

after typing all of this I decided I had better run the same test myself so I don't sound like a complete ass - and my conclusion was that on the print at 150dpi the jaggies were very obvious, but at 300dpi I could barely barely notice it and then only because I knew I was looking for it, but....it was still there, slight though it may be. I believe that is all his point was, and I think he even stated that he was not saying that you would necessarily see them in normal prints with your eyes. Okay, I'll shut up now!! :o -> ;l
 
Higher ISO will give you higher shutterspeeds but you'll get more noise.

Regards, Maxven
CA is the result of the lens, but the complaint that Phil had
really isn't CA, it is blooming.

Blooming occurs when a photosite on a CCD completely saturates in
other words goes to white and the adjacent photosites are still
dark. When a photosite completely saturates the CCD must do
something with the excess charge. So on the CCD there are
"electronic gutters" that carry the charge off the chip. However,
these gutters can only carry so much charge. When they can carry
no more, the charge floods into adjacent photosites and causes
blooming or more commonly known as purple fringing. The chip that
is used on this camera has a very high pixel density. In other
words it's a small chip with a heck of a lot of pixels. And
because of this the "gutters" are very small and not able to deal
with excessive charges.

You are right in that this is a hardware issue and cannot be fixed
by a firmware up date, but the problem actually lies in the CCD and
not the lens.

Now there is one way to prevent the purple fringing and that is to
use a very small aperture like f7 or f8. This slows the rate at
which the photosites recieve the charge and therefore make it
possible for the "gutters" to do their job.
--



Olympus..C21OOUZ since 28/1/02 A-2OO tele, Oly P-4OO ..
HP97OCxi ZI0_USB Sanyo16OO Manfrotto 19ODB+141RC
My mini gallery http://www.sigma-sa.com/kt/pelion/index.htm
' My primary language is Greek '
--



Olympus..C21OOUZ since 28/1/02 A-2OO tele, Oly P-4OO ..
HP97OCxi ZI0_USB Sanyo16OO Manfrotto 19ODB+141RC
My mini gallery http://www.sigma-sa.com/kt/pelion/index.htm
' My primary language is Greek '
 
REality is that phil always seems to side with Nikons, but we also have to recognize that olympus has never been too strong in image quality either...

Every brand has its market and every market has its auidence, olympus has made its way into aspiring cameara that are not quite top nodge, but cameras that do become popular and sells a lot because of their friendlyness and reliablility. Olympus = practicality but not top quality...

Also, just like every music artist has its one big hit album-song, also every brand has its one camera-model that will always be remembered, for Nikon that was the 990, it was top nodge image quality, superior then the 3030z any time any where, only if you could manage to use it and get it to work.

Based on the fame of the 990 I got the 4500, and I have never seen anything that produces so hurrible image quality then this camera. It is auful, I wasted $500.

So life is full of ups and downs, I wish that I would always catch the ups but with the nikon 4500 I got a down...

The 5050 was a good effort in practicality and usuability, it offers versatiltiy that no other camera can match, but it sure does lack in the image department just like we all had spected. It is an Olympus thing.

Some people like me preffered practicality over top image quality.

it is life...
Only 'Recommended'
Phil has never given an Olympus camera a highly recommended review.
He should have tested the C4040Z...that was a terrific camera, and
would smoke many another camera in it's class for image quality.
Actually, it seems the opposite is true...the 4040 was smoked by
the Canon in Phil's tests:

http://www.dpreview.com/reviews/canong2/page18.asp
http://www.dpreview.com/reviews/canong2/page19.asp
 
For your question, the answer should be YES.

In-camera sharpening is a post-exposure processing procedure, so noise is a hardware issue that nothing can elimitate it. The electronics of that camera was physically design, arranged and calibrated in such a way that it automatically adds noise to the image information at the time of exposure, so the noise will always live in the image taken.

The difference is that you do not have control of the in-camera sharpening procedure as you do in photoshop and second, the in-camera sharpening procedure algorythym may be of less quality as the one in photoshop this way harming your image in a worse way.

So, the noise will always be there, is just that the in-camera sharpening may affect the image in such way that it may make the existing noise more visible. Best thing to do is to avoid in-camera sharpening so it will not disturb the image existing noise, and deal with the noise at a later time when a more apropiate tool is availble, like photoshop...

hope this helps...
Wow-

I think you're right that aggressive sharpening appears to simply
"amplify" existing noise, and that the noise would still be there
at lower sharpening levels...However, it does appear that lowering
the in-camera sharpening decreases the apparent noise pretty
significantly. A good example of this can be found in the review of
the C-5050 at
http://www.dcresource.com/reviews/olympus/c5050z-review/index.html .

My question is: If you lower the in-camera sharpening, and use the
unsharp mask in post-processing, will the noise "reappear", and
look like it does with the default level of in-camera sharpening?
[I hope that makes sense :p]

I in no way consider myself to be an expert. I simply state what I
think I've learned through the years--and a lot of what I've
learned has come from you guys & gals :)
--
Kevin

*********************************************************
Olympus C-3040Z
Tiffen MegaPlus 2x teleconverter
Promaster 5750DX flash
 
All I can say is that the noise level I see on all of your images in the your 5050z galleries are a lot more acceptable the what I get from the 4500.

http://cloud.prohosting.com/~poder2k/4500/

http://cloud.prohosting.com/~poder2k/facenoise.htm

I am totally desapointed and sorry that I did not invested my money on an olympus. I currently own the 3030z and I have not found anything else that could beat this camera...
Anyone who wants to see what good pictures from the 5050 look like
in actual use rather than contrived test shots is wlecome to check
out mine (alebit in 800x600 size) on pbase:
http://www.pbase.com/jfeinleib/olympus_c5050_samples

In sum, I find it hard to believe that anyone who pays $700-800 for
this camera will be disappointed with the quality of pictures it
can produce to say nothing of how much a pleasure it is to actually
operate.

http://www.pbase.com/jfeinleib
http://www.photosig.com/viewuser.php?id=23754
 
Joël, the excellent quality of your 5050 pictures made me a believer !

Whatjado wit all die blue in the snow scenes everybody's talking about in the Canon forum :)

Portait of the child - great - soft lighting, subtle and great dynamic range and all the detail one could wish for. Was the original taken at 3 mega ?.

Geo Paris/FR
http://www.pbase.com/geo_paris_fr/roscoff21072002
 

Keyboard shortcuts

Back
Top