Reading mode: Light Dark

Direct Test Comparison - S2 vs. SLR/n

Go to last post
Sean's story that color moire can be removed quickly and easily with no loss of resolution, that is.

Sean, how about some sample images where you have large-scale moire problems and also important color detail, so that we can see how much color detail has to be given up to remove the moire?

Re the S2 versus Pro/n comparison, that's very interesting. The Pro/n is definitely way sharper than the S2 (as is expected with twice as many samples and no blur filter). But the amazing thing is how it picks up the color details where the S2 thought it was just flat monochrome (just kidding!).

j
jean bernier said:
Jim Herndon said:
Thanks. I'll be studying up on moire removal, and I'm going to be
ordering CS soon (still using 7.0).

Jim Herndon
Sean Keegan said:
http://forums.dpreview.com/forums/read.asp?forum=1026&message=7663398

Regards,
Sean
Jim Herndon said:
Hi,

I've posted a fairly comprehensive test comparison of the Fuji S2
to the Kodak SLR/n. It can be found at:

http://www.onlocationportraiture.com/comp0.htm

I hope this helps answer questions you might have about the two
cameras.

Jim Herndon
Click to expand...
Click to expand...
Click to expand...
--
JohnnyBGood
Hi Jim thanks for sharing this. I was not expecting this
comparison so quick as the SLR/n is not available everywhere yet.
Very well done and some real-world shooting situations.
The thing is I've looked at samples here and there plus your latest
input and I've seen so many issues with these files that I don't
think I'll pursue and even try the camera . I believe in hands-on
experience, but in this case, I feel I'd just lose my time. There
seems to be c-aliasing and moire just about on every shot and I do
not want to have to deal with this. Having to scrutinize every
shot and fix these issues: no thanks !
Add the rumored green color shifts with WA and this looks to me
like: too much trouble.
Would'nt the designer of the striped shirt who commisioned the
photo of the man in the studio scream ? Your wedding publicity is
funny-coloured too. I seldom have any problem with S2 files.
Moire is a rare occurence and fine lines patterns or letters rarely
need fix. Sooo why did I consider the SLR/n? Cause resolution is
there...and it's affordable to more people than these $Digiback$
I think if you did put an AA filter on Kodak's sensor, it would
filter out most of these problem and, yes, it would probably lose
most of it's resolution advantage over the S2. In the S2, small
detail is slightly blurred through the AA filter, and "re-created"
with sharpening. In the Kodak, small details are focused on the
sensor where they are allowed to interfere with the "grid", but
resolution is as sharp as the lens permits. Sharper images are
easier to sell than clean lower-resolution ones right?
BTW the S2 do look focused but blurred by vibration....funny.

So I sincerily hope you won't regret buying this model, it' still a
hefty sum. Thanks for posting this but your contribution put the
final nail in the Kodak's coffin in my view.
Regards JB
Click to expand...
 
Hi Jim,

I find myself asking why you have thrown yourself into this fire? I mean this in a nice way and not a suspicious way. One can't help but find themselves in a defensive posture when making these comparisons and then having to defend their results.

I am writing this because I think we should lay it to rest. I think that if your happy with your purchase that this should be all that matters. I know as an S2 user I am very happy with my camera.

Technically speaking the your new camera should be better than my camera on short shots. It has more pixels and unless Kodak has found a way to mess it up, it should deliver superior resolution. Whether it will be significant enough for many to toss their S2 in favor of the Kodak will probably be a personal decision for each and every person.

I know you made your best effort at doing a fair test and I respect that. I think for people using a camera in a Studio environment its a was a tremendous effort. For people like myself, I'd have to see how it performed outdoors with long lenses as well and know a whole lot more about it before I would consider it. Given the S2's 1.5 X crop I doubt that the difference would any greater than what you tried to illustrate in your tests.

To one persons point, and for all those people who do post process their images, a best sample verses best sample would have been more telling in my opinion. Again, for the number of pixels though and the cost, I would expect the Kodak to show a significant difference.

Again, thanks for the effort! I hope you enjoy the camera and that the S2 folks who are happy with their cameras will continue to be like I am.

I am simply amazed at how my investment continues to hold up a year and a half after I purchased my camera. I feel very lucky I chose the S2.

For those worried about the S3, it is going to have it problems like any other DSLR out there. I think that to another posters point, its going to take a lot more pixels to see a huge difference. I think Fuji is on the right evolution path right now and that is filling the remaining gap between Digital and Film for image quality, Dynamic Range, etc. Of course the proof will be in the final product. I think the S4 will bring more Pixels and a better body if the improvements in the S3 pan out. Some companies chose pixels, Fuji chose DR, try not to be too disappointed.

Jim, in closing, I would not subject yourself to the punishment on this, my advice is to go out and enjoy that wonderful camera. Phil will give us a review I am sure. If I had that bad boy, I would be clicking away!

Take care everyone!

--
Thanks!
Mark

Fuji S2 Pro
Tamron's 19-35, 90 Macro
Nikon's 24-120VR & 70-200VR, 80-400VR
Nikon's 20mm F2.8, 50mm 1.8
http://www.radphotos.net
 
Hi,

Thanks for these test, and your correct your opening yourself up to all sort of flack! May if your time allows, you cold try some of the suggestions out!

The comparision look good to me & there is all ways better or at lest another way of doing some thing.

In some way your finding a what I would have expected & therefore no surprise.

I too photograph wedding and the issue around burn out going towards mag. affect me too. I have found that with bright sun light the S2 metering is way off! How does the SLR/n cope in these conditions
Alex
Jim Herndon said:
As I wrote on the 'Intro' page, I'm not going to go into why I
didn't do this or this. I chose standard (and largely equivalent)
settings on both cameras to try and give a baseline to judge
quality issues.

In general, the Kodak defenders are saying that I should have run
RAW conversion in Adobe CS and the Fuji defenders are saying I
should have applied more careful sharpening techniques
post-process. Both techniques WILL draw out higher quality results,
but 1) I don't have time to optimize EVERY single image I shoot, so
quality directly from the camera is important to me and 2) if there
are better ways to optimize the file than what you get from EX or
PhotoDesk, I'd strongly suggest to Kodak and Fuji that they should
improve their software.

A member of the Kodak forum has offered FTP space to upload the
Kodak RAW files. He wants to optimize them and see what results he
is getting. If a member of this forum wants to offer the same for
the Fuji RAW files, I will be happy to upload them to you.

Also, I'm afraid I don't have time to answer every posted question
in the forums. I'll try to answer all emails, but it will take me a
few days. Thanks for your patience.

Jim Herndon
Jim Herndon said:
Hi,

I've posted a fairly comprehensive test comparison of the Fuji S2
to the Kodak SLR/n. It can be found at:

http://www.onlocationportraiture.com/comp0.htm

I hope this helps answer questions you might have about the two
cameras.

Jim Herndon
Click to expand...
Click to expand...
 
Didn't you look at my link, Just looking? Pretty typical of you. It is pretty obvious that it works. Don't you think? I do not see any loss of color detail since I did not have to blur the color channels by much at all to get rid of the color moire. This kind of moire is easy to fix without hurting color detail since it is so small and sparse. Since I know you are lazy, Just looking, here is the link, again, where I removed the moire with no color losses... with the S2 it is even easier to remove because there is less to begin with:

http://forums.dpreview.com/forums/read.asp?forum=1026&message=7663398

The color moire of the SLR/N is minor. It might be all over the image, true, but at least it is not large swaths of color moire like the example below from the Canon 1D.

Large swaths of color moire is harder to remove without hurting color detail, but it is also really rare on the Kodak or S2. But here is an example, keep in mind I have never seen color moire this bad EVER with my S2, this is from a Canon 1D:

http://forums.dpreview.com/forums/read.asp?forum=1021&message=6929495

Completely removed with no loss of resolution or addition of noise. At the expense of a little color, but hardly noticeable unless you compare back and forth directly. But when the image is this bad, a little color loss is worth it.

Just remember, the file from the Kodak is so large that a blur of 2 pixel radius to get rid of the color moire is so slight that there will be no discernible loss of color at all even in a very large print. Even the color moire might not be visible without removal in a 13x19" print. I have not tried it though.

I remember getting into an argument over color moire with you before, Just looking... you said it was one reason bayer really sucks compared to the Foveon method. Well maybe, but not for most cameras, the Kodak and Canons are exceptions. And what little color moire there might be is easily removed. And the Foveon suffers from regular moire as bad as any Bayer sensor. In fact, if you took a photo of that same jacket using the SD9 you would get just as bad moire, just not colored.... but I must admit, at least your colors would still be good as they had to sacrifice the color to remove the color moire. But the aliasing and noise problems of the SD9/10 are much worse than a little color moire to me. You can remove color moire where a viewer would never know it ever existed... even on that jacket. But aliasing is much harder to remove, and removing noise always affects the details. So I much prefer getting rid of the color moire when it shows up, which thankfully is very rare with the S2. As for resolution, I hope Phil compares the S2 to the new Canon 1DMKII using the EX converter! Then maybe people will finally be convinced the S2 has the resolution of 8-9mp of other Bayer cameras.

Regards,
Sean
Just Looking said:
Sean's story that color moire can be removed quickly and easily
with no loss of resolution, that is.

Sean, how about some sample images where you have large-scale moire
problems and also important color detail, so that we can see how
much color detail has to be given up to remove the moire?

Re the S2 versus Pro/n comparison, that's very interesting. The
Pro/n is definitely way sharper than the S2 (as is expected with
twice as many samples and no blur filter). But the amazing thing
is how it picks up the color details where the S2 thought it was
just flat monochrome (just kidding!).

j
Click to expand...
 
Hi Sean

I have a sample print from Phil's test of the 14n (the one of the white lighthouse in bright sunlight).

Even in a 10 x 8 print, there are immediately visible pink, purple, green thread-like colour artifacts all over the white textured surface of the lighthouse building but also all over the grey stone wall that surrounds the building.

Quite disturbing.

Is this the kind of problem your method would solve?

Regards

Dave
Sean Keegan said:
http://forums.dpreview.com/forums/read.asp?forum=1026&message=7663398

The color moire of the SLR/N is minor. It might be all over the
image, true, but at least it is not large swaths of color moire
like the example below from the Canon 1D.

Large swaths of color moire is harder to remove without hurting
color detail, but it is also really rare on the Kodak or S2. But
here is an example, keep in mind I have never seen color moire this
bad EVER with my S2, this is from a Canon 1D:

http://forums.dpreview.com/forums/read.asp?forum=1021&message=6929495

Completely removed with no loss of resolution or addition of noise.
At the expense of a little color, but hardly noticeable unless you
compare back and forth directly. But when the image is this bad, a
little color loss is worth it.

Just remember, the file from the Kodak is so large that a blur of 2
pixel radius to get rid of the color moire is so slight that there
will be no discernible loss of color at all even in a very large
print. Even the color moire might not be visible without removal in
a 13x19" print. I have not tried it though.

I remember getting into an argument over color moire with you
before, Just looking... you said it was one reason bayer really
sucks compared to the Foveon method. Well maybe, but not for most
cameras, the Kodak and Canons are exceptions. And what little color
moire there might be is easily removed. And the Foveon suffers from
regular moire as bad as any Bayer sensor. In fact, if you took a
photo of that same jacket using the SD9 you would get just as bad
moire, just not colored.... but I must admit, at least your colors
would still be good as they had to sacrifice the color to remove
the color moire. But the aliasing and noise problems of the SD9/10
are much worse than a little color moire to me. You can remove
color moire where a viewer would never know it ever existed... even
on that jacket. But aliasing is much harder to remove, and removing
noise always affects the details. So I much prefer getting rid of
the color moire when it shows up, which thankfully is very rare
with the S2. As for resolution, I hope Phil compares the S2 to the
new Canon 1DMKII using the EX converter! Then maybe people will
finally be convinced the S2 has the resolution of 8-9mp of other
Bayer cameras.

Regards,
Sean
Just Looking said:
Sean's story that color moire can be removed quickly and easily
with no loss of resolution, that is.

Sean, how about some sample images where you have large-scale moire
problems and also important color detail, so that we can see how
much color detail has to be given up to remove the moire?

Re the S2 versus Pro/n comparison, that's very interesting. The
Pro/n is definitely way sharper than the S2 (as is expected with
twice as many samples and no blur filter). But the amazing thing
is how it picks up the color details where the S2 thought it was
just flat monochrome (just kidding!).

j
Click to expand...
Click to expand...
 
Hi Alex,

Still too early to tell. Feel free to email me in a month or so, I should have an answer by then. Thanks.

Jim Herndon
uk102 said:
I too photograph wedding and the issue around burn out going
towards mag. affect me too. I have found that with bright sun light
the S2 metering is way off! How does the SLR/n cope in these
conditions
Click to expand...
 
;)

The first question is: is the SLR/n a better camera than the S2? Maybe. In certain applications yes, in certain applications no. Is the SLR/n worth the extra $$ than the S2? Probably not; unless you use it well (and often) for its strengths.

I hope to do this, otherwise my buy is a bust.

Still testing, still evaluating. Whatever I decide, my S2 will still have a prominent place in my work.

Thanks.

Jim Herndon
Mark Abraham said:
Hi Jim,

I find myself asking why you have thrown yourself into this fire?
I mean this in a nice way and not a suspicious way. One can't help
but find themselves in a defensive posture when making these
comparisons and then having to defend their results.

I am writing this because I think we should lay it to rest. I
think that if your happy with your purchase that this should be all
that matters. I know as an S2 user I am very happy with my camera.

Technically speaking the your new camera should be better than my
camera on short shots. It has more pixels and unless Kodak has
found a way to mess it up, it should deliver superior resolution.
Whether it will be significant enough for many to toss their S2 in
favor of the Kodak will probably be a personal decision for each
and every person.

I know you made your best effort at doing a fair test and I respect
that. I think for people using a camera in a Studio environment
its a was a tremendous effort. For people like myself, I'd have to
see how it performed outdoors with long lenses as well and know a
whole lot more about it before I would consider it. Given the S2's
1.5 X crop I doubt that the difference would any greater than what
you tried to illustrate in your tests.

To one persons point, and for all those people who do post process
their images, a best sample verses best sample would have been more
telling in my opinion. Again, for the number of pixels though and
the cost, I would expect the Kodak to show a significant difference.

Again, thanks for the effort! I hope you enjoy the camera and that
the S2 folks who are happy with their cameras will continue to be
like I am.

I am simply amazed at how my investment continues to hold up a year
and a half after I purchased my camera. I feel very lucky I chose
the S2.

For those worried about the S3, it is going to have it problems
like any other DSLR out there. I think that to another posters
point, its going to take a lot more pixels to see a huge
difference. I think Fuji is on the right evolution path right now
and that is filling the remaining gap between Digital and Film for
image quality, Dynamic Range, etc. Of course the proof will be in
the final product. I think the S4 will bring more Pixels and a
better body if the improvements in the S3 pan out. Some companies
chose pixels, Fuji chose DR, try not to be too disappointed.

Jim, in closing, I would not subject yourself to the punishment on
this, my advice is to go out and enjoy that wonderful camera. Phil
will give us a review I am sure. If I had that bad boy, I would be
clicking away!

Take care everyone!

--
Thanks!
Mark

Fuji S2 Pro
Tamron's 19-35, 90 Macro
Nikon's 24-120VR & 70-200VR, 80-400VR
Nikon's 20mm F2.8, 50mm 1.8
http://www.radphotos.net
Click to expand...
 
DMillier said:
Hi Sean

I have a sample print from Phil's test of the 14n (the one of the
white lighthouse in bright sunlight).

Even in a 10 x 8 print, there are immediately visible pink, purple,
green thread-like colour artifacts all over the white textured
surface of the lighthouse building but also all over the grey stone
wall that surrounds the building.

Quite disturbing.

Is this the kind of problem your method would solve?
Click to expand...
If it is color moire, you got it Dave. :) Try it. It does wonders. I cannot open that image for some reason... looks like Phil's server is down so I cannot see what you mean.

Regards,
Sean
 
Jim Herndon said:
;)

The first question is: is the SLR/n a better camera than the S2?
Maybe. In certain applications yes, in certain applications no. Is
the SLR/n worth the extra $$ than the S2? Probably not; unless you
use it well (and often) for its strengths.

I hope to do this, otherwise my buy is a bust.

Still testing, still evaluating. Whatever I decide, my S2 will
still have a prominent place in my work.

Thanks.

Jim Herndon
Click to expand...
I have a friend who's also a shooting pro and we joke about who's going to be the guinea pig when new gear comes out. We usually trade places with each new item. You have saved 'us" approximately $10,000 that we don't need to experiment with, and for that I (and I'm sure many silent others) are very thankful for your efforts. You have presented a compelling case supported by others that there are no valid reasons to make the changes from our beloved S2's. Again, thank you for your effort, Jim, because I truly understand that time is money. We appreciate your sacrifice.

While we eagerly anticipate the next big leap forward in the area of DSLR's the following thoughts come to mind:

The big loser in all this is Kodak. They really couldn't afford to mess up again. And it would seem that they have. Just the traffic and topics on this site will prove that in time. We were all waiting for the next great machine, and this isn't it.

The big winner in all of this is DPReview. What a valuable resource for all photographers. This community and the ability to share information definitely helps all of us to become the best informed consumers and eventually this will work to nudge the manufacturers that serve us to do a better job.

Another qualified winner is Fuji. They hopefully are being guided by their commitment to not introducing their next pro camera until "it is absolutely ready to go". I'm counting on that and I'm betting that over time they will implement a plan to provide us with the camera that gets the best overall photographic quality with the prerequisite increase in resolution. The pixel race must be in the context of a balanced improvement in all aspects of image quality. I'm trusting Fuji to lead in this regard. And I for one look forward to the day that we are able to shoot a lot more and spend a lot less time in front of a computer (postproduction is not photography).

The qualifier is that Fuji needs to learn to communicate better with their audience. They can trust us: we're adults. Information about anticipated pricing and a target introduction date would just build our loyalty and demonstrate their respect for their professional users. It's not hard. And it's predicated on my belief that just like they did in film over a twenty year period Fuji will establish their clear dominance in the digital area (sorry Kodak).

I got carried away, but this is my life and being in the early adopter group on a hot developing technology, while exciting in many ways, is really not that much fun. Just necessary.

Regards, Grog
A great camera doesn't guarantee great photos,
but better glass always makes for better photos.
 
I analyzed your full resolution Jpeg shots, all (or most) S2 Pro shots are out of focus (back focus about 4-5cm, that is a lot for 105mm shots and resolution/sharpness comparison. I made some crop pairs for ISO 160, 400 and 800 were S2 focused from your Jpegs:
http://digicam.narod.ru/kodak14/comparison.htm
Difference is more less than you found...

I think, you focused one camera once, then made shots with different ISO values without refocusing, then change camera to another one with the same sequence. In this case I understand the same focusing error for all S2 shots. Else (if you focusing for each shot anew), your camera needed for AF to be adjusted in the repair center.

--
TeddyBear, Russia
http://digicam.narod.ru editor
 
Hi,

I'll look at your samples tomorrow at work. I'm at home right now with a dinky 28k connection - too slow.

ALL shots for both the S2 and SLR/n were focused individually (and continuously) with my preference of focusing on the eyes with center focusing, re-compose holding shutter halfway down, then shooting. I was also careful to focus on the same point for the studio shots.

It would be coincidence to the point of absurdity that ALL SLR/n images are in focus and ALL S2 images are out of focus.

I'll grant that 'dr_f.jpg' may have suffered from camera shake or mirror vibration.

Please look at the S2 full-res images in Photoshop at 25% and let me know if they look out of focus. Then notice my settings. If I was off by 4-5 cm on the ISO images they would be very, very noticeably out of focus - as I was shooting at open apertures.

I don't know why it's so difficult to think that the SLR/n image might appear sharper because the camera has a higher resolution. I give the S2 high marks because it came so surprisingly (to me) close to the SLR/n.

If this is just impossible to accept, please get your hands on the SLR/n and run some tests. Please let me know if I'm just completely wrong.

Finally, as Fuji people are absolutely correct that the S2 images would be sharper by leaving sharpening off in-camera and carefully sharpening post-process, the Kodak people are absolutely right about converting images in Adobe RAW instead of PhotoDesk.

Adobe's results are 100% better; virtually free of moire (even better than the S2 in many images) and instead of the 'painter' look they have a film-like look. 160 ISO SLR/n images now look for all the world like perfectly scanned 160 ISO medium format film to me.

Thank you for everyone's comments. I'm happy that some might have found the comparison useful. I'm happy that proud S2 owners remain proud; it's still a great camera and the best value on the market. Thanks again.

Jim Herndon
Teddy said:
I analyzed your full resolution Jpeg shots, all (or most) S2 Pro
shots are out of focus (back focus about 4-5cm, that is a lot for
105mm shots and resolution/sharpness comparison. I made some crop
pairs for ISO 160, 400 and 800 were S2 focused from your Jpegs:
http://digicam.narod.ru/kodak14/comparison.htm
Difference is more less than you found...
I think, you focused one camera once, then made shots with
different ISO values without refocusing, then change camera to
another one with the same sequence. In this case I understand the
same focusing error for all S2 shots. Else (if you focusing for
each shot anew), your camera needed for AF to be adjusted in the
repair center.

--
TeddyBear, Russia
http://digicam.narod.ru editor
Click to expand...
 
Sean, did you overlook what I asked for:

"sample images where you have large-scale moire problems and also important color detail"?

I understand that your method of blurring away color works when the problems are small-scale relative to the color details that you care about. It's called a lowpass filter. I'm not denying that, just pointing out that the images you prove it with are worthless for evaluating how well it might work more generally.

Your "safe and effective" claims can fail on either count, depending on the image. That's why AA filters were invented for Bayer sensors.

j
Sean Keegan said:
http://forums.dpreview.com/forums/read.asp?forum=1026&message=7663398

The color moire of the SLR/N is minor. It might be all over the
image, true, but at least it is not large swaths of color moire
like the example below from the Canon 1D.

Large swaths of color moire is harder to remove without hurting
color detail, but it is also really rare on the Kodak or S2. But
here is an example, keep in mind I have never seen color moire this
bad EVER with my S2, this is from a Canon 1D:

http://forums.dpreview.com/forums/read.asp?forum=1021&message=6929495

Completely removed with no loss of resolution or addition of noise.
At the expense of a little color, but hardly noticeable unless you
compare back and forth directly. But when the image is this bad, a
little color loss is worth it.

Just remember, the file from the Kodak is so large that a blur of 2
pixel radius to get rid of the color moire is so slight that there
will be no discernible loss of color at all even in a very large
print. Even the color moire might not be visible without removal in
a 13x19" print. I have not tried it though.

I remember getting into an argument over color moire with you
before, Just looking... you said it was one reason bayer really
sucks compared to the Foveon method. Well maybe, but not for most
cameras, the Kodak and Canons are exceptions. And what little color
moire there might be is easily removed. And the Foveon suffers from
regular moire as bad as any Bayer sensor. In fact, if you took a
photo of that same jacket using the SD9 you would get just as bad
moire, just not colored.... but I must admit, at least your colors
would still be good as they had to sacrifice the color to remove
the color moire. But the aliasing and noise problems of the SD9/10
are much worse than a little color moire to me. You can remove
color moire where a viewer would never know it ever existed... even
on that jacket. But aliasing is much harder to remove, and removing
noise always affects the details. So I much prefer getting rid of
the color moire when it shows up, which thankfully is very rare
with the S2. As for resolution, I hope Phil compares the S2 to the
new Canon 1DMKII using the EX converter! Then maybe people will
finally be convinced the S2 has the resolution of 8-9mp of other
Bayer cameras.

Regards,
Sean
Just Looking said:
Sean's story that color moire can be removed quickly and easily
with no loss of resolution, that is.

Sean, how about some sample images where you have large-scale moire
problems and also important color detail, so that we can see how
much color detail has to be given up to remove the moire?

Re the S2 versus Pro/n comparison, that's very interesting. The
Pro/n is definitely way sharper than the S2 (as is expected with
twice as many samples and no blur filter). But the amazing thing
is how it picks up the color details where the S2 thought it was
just flat monochrome (just kidding!).

j
Click to expand...
Click to expand...
 
Jim Herndon said:
Finally, as Fuji people are absolutely correct that the S2 images
would be sharper by leaving sharpening off in-camera and carefully
sharpening post-process, the Kodak people are absolutely right
about converting images in Adobe RAW instead of PhotoDesk.

Adobe's results are 100% better; virtually free of moire (even
better than the S2 in many images) and instead of the 'painter'
look they have a film-like look. 160 ISO SLR/n images now look for
all the world like perfectly scanned 160 ISO medium format film to
me.
Click to expand...
I would like to see a comparison based on the facts above. The best you can get from the S2 (and don't over-expose all of them next time) and the best from the slr/n using ACR. This would give better results to evaluate. I'd do it myself if I had a slr/n and ACR at my disposal. I don't.

Robert
 
Jim Herndon said:
Hi,
I've posted a fairly comprehensive test comparison of the Fuji S2
to the Kodak SLR/n. It can be found at:
http://www.onlocationportraiture.com/comp0.htm
I hope this helps answer questions you might have about the two
cameras.
Click to expand...
Hi Jim

I read your comparison from the post in the Kodak forum and was interested in the results etc.

I have to say that I was dissapointed in the sharpness you achieved in every shot large from the S2, I am not just saying this because I have one but I shoot at 6mp jpeg fine and think my good pics are usually sharper than that on face eye and eyelash detail even in poor ambient light and if I light people with a bounced flash or something they can be well near perfect sharpness wise out of the camera when viewed at 100% ... I find the 12mp output does need more processing to look the same when viewed at 100% and thats just jpeg fine .. not explored raw at all yet.

There is no doubt that in your comparison the Kodak seems sharper.

Interesting the colour issues with the kodak on the shirt, it would concern me that but if you are happy. In your samples it certainly seems the sharper of the two which surprised me I have to say.

Anyhow a lot of work you did and thanks for sharing it because it does make interesting reading.

--
Mark
Samples & Galleries is just about photographs
Comment plenty please - forums dont thrive on silence :-)
http://forums.dpreview.com/forums/forum.asp?forum=1005
 
You're right.

I own both the S2 and Kodak 14n. Your tests confirm what I already know from shooting thousands of images with both cameras. The 14n (and SLR/n) comfortably out-resolve the S2. The S2 is better than any other 6mp camera I have seen, but it cannot compete for pure resolution with the Kodak. It has other strengths of course, such as high ISO capability.

RIL
Jim Herndon said:
I'll look at your samples tomorrow at work. I'm at home right now
with a dinky 28k connection - too slow.

ALL shots for both the S2 and SLR/n were focused individually (and
continuously) with my preference of focusing on the eyes with
center focusing, re-compose holding shutter halfway down, then
shooting. I was also careful to focus on the same point for the
studio shots.

It would be coincidence to the point of absurdity that ALL SLR/n
images are in focus and ALL S2 images are out of focus.

I'll grant that 'dr_f.jpg' may have suffered from camera shake or
mirror vibration.

Please look at the S2 full-res images in Photoshop at 25% and let
me know if they look out of focus. Then notice my settings. If I
was off by 4-5 cm on the ISO images they would be very, very
noticeably out of focus - as I was shooting at open apertures.

I don't know why it's so difficult to think that the SLR/n image
might appear sharper because the camera has a higher resolution. I
give the S2 high marks because it came so surprisingly (to me)
close to the SLR/n.

If this is just impossible to accept, please get your hands on the
SLR/n and run some tests. Please let me know if I'm just completely
wrong.

Finally, as Fuji people are absolutely correct that the S2 images
would be sharper by leaving sharpening off in-camera and carefully
sharpening post-process, the Kodak people are absolutely right
about converting images in Adobe RAW instead of PhotoDesk.

Adobe's results are 100% better; virtually free of moire (even
better than the S2 in many images) and instead of the 'painter'
look they have a film-like look. 160 ISO SLR/n images now look for
all the world like perfectly scanned 160 ISO medium format film to
me.

Thank you for everyone's comments. I'm happy that some might have
found the comparison useful. I'm happy that proud S2 owners remain
proud; it's still a great camera and the best value on the market.
Thanks again.

Jim Herndon
Teddy said:
I analyzed your full resolution Jpeg shots, all (or most) S2 Pro
shots are out of focus (back focus about 4-5cm, that is a lot for
105mm shots and resolution/sharpness comparison. I made some crop
pairs for ISO 160, 400 and 800 were S2 focused from your Jpegs:
http://digicam.narod.ru/kodak14/comparison.htm
Difference is more less than you found...
I think, you focused one camera once, then made shots with
different ISO values without refocusing, then change camera to
another one with the same sequence. In this case I understand the
same focusing error for all S2 shots. Else (if you focusing for
each shot anew), your camera needed for AF to be adjusted in the
repair center.

--
TeddyBear, Russia
http://digicam.narod.ru editor
Click to expand...
Click to expand...
 
Just Looking said:
Sean, did you overlook what I asked for:
"sample images where you have large-scale moire problems and also
important color detail"?

I understand that your method of blurring away color works when the
problems are small-scale relative to the color details that you
care about. It's called a lowpass filter. I'm not denying that,
just pointing out that the images you prove it with are worthless
for evaluating how well it might work more generally.
Click to expand...
Find me an image of what you mean.... I am having trouble understanding how the images I provided do not qualify in this area. Send me the image and I will "Fix" it and post it so people can judge.
Just Looking said:
Your "safe and effective" claims can fail on either count,
depending on the image. That's why AA filters were invented for
Bayer sensors.
Click to expand...
I agree with you there... but the AA filter is just as needed on the Sigmas...it is not bayer specific. Both the Kodak and Sigmas would be better cameras with the proper AA filter. But at least color moire is easy to fix... aliasing is not. The Kodak gets away with no AA filter because of sheer number of pixels. The Sigmas with only 3mp output really could use a AA filter in my opinion. The 6mp X3 chip if it ever comes out might have enough pixels to get away without having an AA filter... we shall see, maybe. But I would still prefer it had one.

Regards,
Sean
Just Looking said:
j
Sean Keegan said:
http://forums.dpreview.com/forums/read.asp?forum=1026&message=7663398

The color moire of the SLR/N is minor. It might be all over the
image, true, but at least it is not large swaths of color moire
like the example below from the Canon 1D.

Large swaths of color moire is harder to remove without hurting
color detail, but it is also really rare on the Kodak or S2. But
here is an example, keep in mind I have never seen color moire this
bad EVER with my S2, this is from a Canon 1D:

http://forums.dpreview.com/forums/read.asp?forum=1021&message=6929495

Completely removed with no loss of resolution or addition of noise.
At the expense of a little color, but hardly noticeable unless you
compare back and forth directly. But when the image is this bad, a
little color loss is worth it.

Just remember, the file from the Kodak is so large that a blur of 2
pixel radius to get rid of the color moire is so slight that there
will be no discernible loss of color at all even in a very large
print. Even the color moire might not be visible without removal in
a 13x19" print. I have not tried it though.

I remember getting into an argument over color moire with you
before, Just looking... you said it was one reason bayer really
sucks compared to the Foveon method. Well maybe, but not for most
cameras, the Kodak and Canons are exceptions. And what little color
moire there might be is easily removed. And the Foveon suffers from
regular moire as bad as any Bayer sensor. In fact, if you took a
photo of that same jacket using the SD9 you would get just as bad
moire, just not colored.... but I must admit, at least your colors
would still be good as they had to sacrifice the color to remove
the color moire. But the aliasing and noise problems of the SD9/10
are much worse than a little color moire to me. You can remove
color moire where a viewer would never know it ever existed... even
on that jacket. But aliasing is much harder to remove, and removing
noise always affects the details. So I much prefer getting rid of
the color moire when it shows up, which thankfully is very rare
with the S2. As for resolution, I hope Phil compares the S2 to the
new Canon 1DMKII using the EX converter! Then maybe people will
finally be convinced the S2 has the resolution of 8-9mp of other
Bayer cameras.

Regards,
Sean
Just Looking said:
Sean's story that color moire can be removed quickly and easily
with no loss of resolution, that is.

Sean, how about some sample images where you have large-scale moire
problems and also important color detail, so that we can see how
much color detail has to be given up to remove the moire?

Re the S2 versus Pro/n comparison, that's very interesting. The
Pro/n is definitely way sharper than the S2 (as is expected with
twice as many samples and no blur filter). But the amazing thing
is how it picks up the color details where the S2 thought it was
just flat monochrome (just kidding!).

j
Click to expand...
Click to expand...
Click to expand...
 
I agree that the samples Jim posted from his S2 are definitely not as sharp as I am used to getting from my S2... but the SLR/n shots definitely have more resolution even when comparing with my S2... of course I cannot compare directly.... but I am used to the sharpness and detail I get from my S2, and while better than what Jim got... it is still not as good as the Kodak. The thing that impresses me is how well the S2 holds up. Same when comparing with the 1DS. I do not see a whole lot more resoltuion with the 1DS either. Just confirms even more to me that the S2 is more like a 8-9mp camera. It will be interesting to see how it stacks up against the Canon 8mp DSLR.

Regards,
Sean
Teddy said:
I analyzed your full resolution Jpeg shots, all (or most) S2 Pro
shots are out of focus (back focus about 4-5cm, that is a lot for
105mm shots and resolution/sharpness comparison. I made some crop
pairs for ISO 160, 400 and 800 were S2 focused from your Jpegs:
http://digicam.narod.ru/kodak14/comparison.htm
Difference is more less than you found...
I think, you focused one camera once, then made shots with
different ISO values without refocusing, then change camera to
another one with the same sequence. In this case I understand the
same focusing error for all S2 shots. Else (if you focusing for
each shot anew), your camera needed for AF to be adjusted in the
repair center.

--
TeddyBear, Russia
http://digicam.narod.ru editor
Click to expand...
 
Of course, SLR/n shots must be sharper and must have higher detalisation (without any tests). I agree that S2 (that have hexahonal structure CCD) resolution is adequate to resolution of about 8Mpixel DSLR with square Bayer structure. Moreover, I measured that number of small details that S2 can captured is about 4.5M (for example, Canon D60 captured about 3.45M details, 1Ds - about 6M details).

I know that S2 cannot capture more details than 1Ds or 14n... but this test have back-focusing error for all S2 Pro shots, I sent crops were S2 focused, in these crops 14n out of focus, then test results are inaccurate, and difference between cameras detalisation will be less than in this test.

Second methodology error - meauring Dynamical range in the lights only. Sensor overload capacity is not a value of DR. Sensor saturation limit determine sensor own ISO speed only and necessity for signal amplification when camera used another ISO index.

For me, this test does not answer to the questions:
  • How much is resolution difference between cameras
  • Is DR of 14/n is better than DR of S2 Pro or not, because without dark samples analysis we cannot determine camera DR.
(DR = Saturation Signal / Dark noise signal),

For me, this test show:
  • AF accuracy with AF-S 70-200/2.8 VR lens for two camera examples
  • Sensor overload capacity for overexposure for both cameras*
    • in other words, from test results I can made conclusion that 14/n sensor have less own ISO speed than selected on camera ISO index (exposure index evaluted for 18% gray, or -2.5Ev from saturation level: if sensor have overload capacity 3.5 Ev from 18% gray point, it's own ISO speed is half from declared value).
In practice this is good news when you need extra overload capability, but it is bad news because signal must be amplificated twice together with noise, then, noise will be more than you want and DR will be less than you want. If camera have one extra Ev for overload, is reasonably always shot with twice exposition (+1 Ev expocorrection) for minimize noise levels. But, this is adequate to half ISO speed using.

--
TeddyBear, Russia
http://digicam.narod.ru editor
Sean Keegan said:
Regards,
Sean
Teddy said:
I analyzed your full resolution Jpeg shots, all (or most) S2 Pro
shots are out of focus (back focus about 4-5cm, that is a lot for
105mm shots and resolution/sharpness comparison. I made some crop
pairs for ISO 160, 400 and 800 were S2 focused from your Jpegs:
http://digicam.narod.ru/kodak14/comparison.htm
Difference is more less than you found...
I think, you focused one camera once, then made shots with
different ISO values without refocusing, then change camera to
another one with the same sequence. In this case I understand the
same focusing error for all S2 shots. Else (if you focusing for
each shot anew), your camera needed for AF to be adjusted in the
repair center.

--
TeddyBear, Russia
http://digicam.narod.ru editor
Click to expand...
Click to expand...
 
to put out such a comprehensive test for your fellow forum members. I've never considered the 14n and I don't think it will change with or without test reviews etc. It's just a camera I am not interested in but I was very interested in your test procedures and the results just out of curiousity.

I'm glad you are happy and satisfied with your purchase. The only 'visual' thing that struck me was that the skin tone from the Kodak appeared to be more natural, neutral. Somehow these days the red in the S2 producing warmer skin tone is appearing to 'get' to me. I don't like it as much as I thought I did.

As you have said yourself, although there are pluses in the Kodak, they aren't significant and I agree at least with what little info/research I have in hand to date. I think it would be hard to justify (for me) the price hike for the rate of returns. But all this is of course subject to also getting to know your camera, optimising on the strengths and the learning curve of the software etc etc. Results may well be worth the extra $ indeed.

I had hoped that this year would be truly exciting for the next generation but I perceive it at this point in time to be so lukewarm.

Good luck and looking forward to your update when you've grown into the 14n :-)

babe

--
LIFE ON THE FLY.

 
Great post... you make some really valid points....
Teddy said:
Of course, SLR/n shots must be sharper and must have higher
detalisation (without any tests). I agree that S2 (that have
hexahonal structure CCD) resolution is adequate to resolution of
about 8Mpixel DSLR with square Bayer structure. Moreover, I
measured that number of small details that S2 can captured is about
4.5M (for example, Canon D60 captured about 3.45M details, 1Ds -
about 6M details).

I know that S2 cannot capture more details than 1Ds or 14n... but
this test have back-focusing error for all S2 Pro shots, I sent
crops were S2 focused, in these crops 14n out of focus, then test
results are inaccurate, and difference between cameras detalisation
will be less than in this test.

Second methodology error - meauring Dynamical range in the lights
only. Sensor overload capacity is not a value of DR. Sensor
saturation limit determine sensor own ISO speed only and necessity
for signal amplification when camera used another ISO index.
Click to expand...
I definitely agree with this one... people complain about the DR of digital, but only consider what happens during overexposure.... If you expose for the highlights digital does REALLY well because of the low noise in the shadows. You proved this nicely on your Film VS S2 comparison, Teddy. For those who missed it they are about about halfway down the page here:

http://www.fcenter.ru/articles.shtml?digitalphoto/5562
Teddy14735 said:
For me, this test does not answer to the questions:
  • How much is resolution difference between cameras
  • Is DR of 14/n is better than DR of S2 Pro or not, because without
dark samples analysis we cannot determine camera DR.
(DR = Saturation Signal / Dark noise signal),

For me, this test show:
  • AF accuracy with AF-S 70-200/2.8 VR lens for two camera examples
  • Sensor overload capacity for overexposure for both cameras*
    • in other words, from test results I can made conclusion that
14/n sensor have less own ISO speed than selected on camera ISO
index (exposure index evaluted for 18% gray, or -2.5Ev from
saturation level: if sensor have overload capacity 3.5 Ev from 18%
gray point, it's own ISO speed is half from declared value).
In practice this is good news when you need extra overload
capability, but it is bad news because signal must be amplificated
twice together with noise, then, noise will be more than you want
and DR will be less than you want. If camera have one extra Ev for
overload, is reasonably always shot with twice exposition (+1 Ev
expocorrection) for minimize noise levels. But, this is adequate to
half ISO speed using.
Click to expand...
Yep... and to get around this, the consensus is that 14N owners routinely overexpose to get the noise down... not something I consider good. I would rather the ISO is accurate and expose for the highlights and use curves to bring up the shadows... then you get optimum noise performance.
Teddy14735 said:
--
Click to expand...
By the way Teddy, I think your test comparisons are among the best I have ever seen from anywhere in the world... Thanks.

Regards,
Sean
 
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