Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Alternative Gear & Lenses | Join Upload & Sell

1      
2
       3       4       end
  

Archive 2012 · Fuji XPro1 vs Nex 5n Comparision Images
  
 
Emacs
Offline
• •
Account locked
p.2 #1 · p.2 #1 · Fuji XPro1 vs Nex 5n Comparision Images


sebboh wrote:
third, bayer filter provides higher color resolution all other things being equal, not necessarily higher luminance resolution as i understand.

There's no "luminance" resolution, green is still color, and it's just 60% of perceived light, not 100%, to be real luminance representative (and it's poorly distributed in X-Trans to gain any noticeable profit). And the resolution is degraded to the worst resolved channel anyway during demosaic, because otherwise we'll have tons of color artifacts in contrast places due to unmatched variations between channels.



Dec 18, 2012 at 06:24 AM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.2 #2 · p.2 #2 · Fuji XPro1 vs Nex 5n Comparision Images


flashinm wrote:
I'm not sure I agree with the first part. At any rate, the differences are so minor, they are irrelevant. RPP definitely does pull out more detail, but the color artifacts are all over (the Nex/Lightroom combo interestingly produces some color artifacts as well - look at the fence on the third sample). There is definitely some smoothing in the jpeg file, but these are 100% crops. Not sure how much of a difference this would actually make, and most people so far think it looks the best. Anyway, I just hope that this puts to rest the idea that
...Show more

Thanks for the raws. This doesn't put the low ISO thing to rest, for me. Below are both the 5N and Fuji crops in LR4 and RPP with all sharpening and NR turned off. In the LR version of the Fuji file, the shrubs are a smeary mess, and the bricks look oddly "painterly." There are tons of artifacts in the RPP version of the Fuji file.

Also, keep in mind that not using any sharpening on the 5N file, which has an AA filter, is already an unfair disadvantage, because AA-less cameras create false details through artifacts, which give the impression of more detail.

Off all of these crops, I'd say the 5N processed in RPP looks the best. I don't really find either of the Fuji conversions acceptable, at least in this scene, although it would depend on the print size.

X-E1 file in LR4:






5N file in LR4:






X-E1 file in RPP:






5N file in RPP:






Edited on Dec 18, 2012 at 04:12 PM · View previous versions



Dec 18, 2012 at 07:01 AM
sebboh
Offline
• • • • • •
Upload & Sell: Off
p.2 #3 · p.2 #3 · Fuji XPro1 vs Nex 5n Comparision Images


Emacs wrote:
There's no "luminance" resolution, green is still color, and it's just 60% of perceived light, not 100%, to be real luminance representative (and it's poorly distributed in X-Trans to gain any noticeable profit).


yes there is, it's what you have left when you throw out all the data on what color filter covered each pixel, and yes it's inaccurate in both designs (the reason foveon exists). how well they perform depends on the type of light.



Dec 18, 2012 at 07:28 AM
flashinm
Offline
• •
Upload & Sell: On
p.2 #4 · p.2 #4 · Fuji XPro1 vs Nex 5n Comparision Images


Douglas, there's no question that we are still dealing with less than optimal Raw conversion. Put the artifacts from RPP aside for a moment and just look at the fine details. They are as good or better than the 5n file (despite being at a lower magnification in your samples). If there were problems inherent in the sensor design that prevent the camera from resolving as much detail at low iso, as you've stated multiple times, this wouldn't be possible.

It would be perfectly accurate to say that we have no good way to get the most out of a fuji raw file at the moment. I would agree with this completely and it's something people should consider. But to say that this is because of a limitation inherent in the sensor design is pretty clearly proven false. We are dealing with a software issue.



Dec 18, 2012 at 03:50 PM
FlyPenFly
Offline
• • • • •
Upload & Sell: On
p.2 #5 · p.2 #5 · Fuji XPro1 vs Nex 5n Comparision Images


I don't think you can simply set the 5N Sharpening at zero. Bayer filter sensors are designed to be sharpened in post or in JPEG processing. So the fuji files are benefiting from JPEG sharpening in their internal algorithm while the 5N files are not receiving that benefit.


If or when Fuji gets very good ACR support and Fuji releases really great lenses, I'll take a hard look at that system. After handling Fuji lenses, they just seem ho hum and the test results are nothing worth bragging about.



Dec 18, 2012 at 03:57 PM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.2 #6 · p.2 #6 · Fuji XPro1 vs Nex 5n Comparision Images


flashinm wrote:
Douglas, there's no question that we are still dealing with less than optimal Raw conversion. Put the artifacts from RPP aside for a moment and just look at the fine details. They are as good or better than the 5n file (despite being at a lower magnification in your samples). If there were problems inherent in the sensor design that prevent the camera from resolving as much detail at low iso, as you've stated multiple times, this wouldn't be possible.

It would be perfectly accurate to say that we have no good way to get the most out of a
...Show more

Sorry about the magnification difference between the 5n and X-E1 RPP files. Not sure what happened, but I fixed it in my post above.

The problem with the Fuji RPP files is that you're seeing aliasing/false details, and you can't equate false details to an actual resolution improvement. Those bricks, railings and window frames look like christmas zippers. The shrubs still don't look as good as the 5N, either.

As Flypenfly mentioned, all of these crops are putting the AA-filtered 5N at quite a disadvantage by not using sharpening, and yet it still competes in terms of details.

TheSuede (and a few others) have provided explanations as to why these X-trans problems exist, and there's really no better source, but that's just the science. Whether we notice the issues in actual use is what I've been curious about. So far, whether it be Fuji's jpegs, or any 3rd party converter, the files are still a bit of a mess, comparatively. Maybe that gap will close with a better raw converter in the future, but that's all speculation at this point, because X-trans has been around for nearly a year, and no one has stepped up to the plate, yet. Heck, better raw converters for Bayer may come, too, and widen the gap more. The future is tough to predict, and it'll be interesting to see what Capture One does.

I appreciate you taking the time to provide samples, but they've shown me the same thing as all of the various samples that I've looked at from other sources. In good light, I'd likely choose Bayer, and, in lowlight, I'd likely choose X-trans...at least for now.



Dec 18, 2012 at 04:45 PM
flashinm
Offline
• •
Upload & Sell: On
p.2 #7 · p.2 #7 · Fuji XPro1 vs Nex 5n Comparision Images


The 5n is not at much of a disadvantage at all. It's known for having a very weak AA filter. We are staring at 100% crops trying to find minute differences in detail. Either way, which ever you prefer, there's not enough of a difference to make a choice on which camera you would buy based on that. That's exactly the case when you normally bring up these issues (someone choosing between nex and fuji) and I don't think its a fair assessment.

Raw support is a real issue though, and one that will likely determine the ultimate success of the Fuji system. Remember, bayer processing algorithms have benefited from over a decade to fine tune and perfect them. Processing the Trans x files is not only way more complex, it's also very new by comparison and from a company that has a tiny sliver of the market share.



Dec 18, 2012 at 05:20 PM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.2 #8 · p.2 #8 · Fuji XPro1 vs Nex 5n Comparision Images


flashinm wrote:
The 5n is not at much of a disadvantage at all. It's known for having a very weak AA filter. We are staring at 100% crops trying to find minute differences in detail. Either way, which ever you prefer, there's not enough of a difference to make a choice on which camera you would buy based on that. That's exactly the case when you normally bring up these issues (someone choosing between nex and fuji) and I don't think its a fair assessment.

Raw support is a real issue though, and one that will likely determine the ultimate success of
...Show more

An unsharpened 5N barely competes with the detail of the 12mp, AA-less GXR M module.

I'm going to bring up this X-trans issue whenever I see a thread where someone inquires about the camera, so that they can investigate the issue, because many users do find it to be an issue after buying the camera. Even you told "Lan11" on another thread that, "If you weren't aware of these issues before you purchased your camera, you must have been hiding under a rock," so I'm apparently not bringing these issues up enough.

Either way, while raw conversion can certainly be improved, I'm not sure that it should be assumed that it will happen anytime soon. Foveon has been around for quite a while, and there still isn't good 3rd party raw support. Here's a pretty good insight about 3rd party raw conversion: link



Dec 18, 2012 at 06:20 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.2 #9 · p.2 #9 · Fuji XPro1 vs Nex 5n Comparision Images


It's very hard to compare different types of camera setups this way, since they react to different difficulties in the targets differently.

The Fuji jpg's look very good to me. But since unsharpened output isn't how people use the other files, I guess the most equal setting is to sharpen the compared cameras to equal edge acuity, and THEN look at detail. Edge acuity is not detail, detail is not pure luminosity contrast, fine detail is not just high-contrast acuity.

No need to jump at the OP anyway, I think he explained the exact workflow and what he meant to show very well. That it doesn't show what many would call "equal settings" is something entirely else, and not the OP's problem.

My way to do this would be to sharpen the NEX with an appropriate amount, and then compare detail contrasts at different levels. But that leaves a lot "up to taste", which would nullify the base comparison. Different people would arrive at different settings, and then where is the "equal" part of it?

A camera gives you a file. That file isn't the end result in many cases, at least it leaves room for a lot of adjustments - even you rely on the built in raw-converter in the camera (in-camera-jpg). What the best result of that end-product is is what defines my overall assessment of a camera - unless it's a total mess to get there.



Dec 18, 2012 at 06:44 PM
FlyPenFly
Offline
• • • • •
Upload & Sell: On
p.2 #10 · p.2 #10 · Fuji XPro1 vs Nex 5n Comparision Images


And that's precisely why I can't jump in with Fuji. They are just not compatible with my favorite tools to extract the most out of an image.


Dec 18, 2012 at 06:50 PM
 

Search in Used Dept. 



flashinm
Offline
• •
Upload & Sell: On
p.2 #11 · p.2 #11 · Fuji XPro1 vs Nex 5n Comparision Images


Here's a sharpened copy (more sharpening for Nex) with a little PP to remove the color artifacts. Some zippering artifacts are still there.





?nc=578

I guess the difference between what I'm saying and what you and Lan11 are saying is that I believe this is 100% software related. You keep telling people that even with proper conversion, trans x can never compete with bayer at low ISO. I think this thread has done a good job of proving that false, even without proper raw support. If you want to keep bringing up chroma resolution in every thread that involves Fuji, that's fine but the real issue you should be pointing out is Raw support, imo.

Edited on Dec 18, 2012 at 06:54 PM · View previous versions



Dec 18, 2012 at 06:50 PM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.2 #12 · p.2 #12 · Fuji XPro1 vs Nex 5n Comparision Images


Oh, I agree Joakim. I'm not jumping on the OP. I think we're friendly. This thread was kind of born out of me bringing up the potential issues of this sensor in other threads, and how the raw conversion is going.


Dec 18, 2012 at 06:52 PM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.2 #13 · p.2 #13 · Fuji XPro1 vs Nex 5n Comparision Images


flashinm wrote:
I guess the difference between what I'm saying and what you and Lan11 are saying is that I believe this is 100% software related. You keep telling people that even with proper conversion, trans x can never compete with bayer at low ISO. I think this thread has done a good job of proving that false, even without proper raw support. If you want to keep bringing up chroma resolution in every thread that involves Fuji, that's fine but the real issue you should be pointing out is Raw support, imo.


Joachim was the one who mentioned the chroma resolution thing in another thread that I've been referencing, but, that is theoretical, and I'm not sure how that will pan out in real world shooting, if better conversions do arrive. It may very well be fine.

So, I'll change my tune a bit and say that, for now, LR4 and RPP conversions look pretty bad, and I'll leave it at that. I'm not sure I'd recommend a camera to others right now based on its raw conversion potential in the future.



Dec 18, 2012 at 07:00 PM
flashinm
Offline
• •
Upload & Sell: On
p.2 #14 · p.2 #14 · Fuji XPro1 vs Nex 5n Comparision Images


That's fair.


Dec 18, 2012 at 07:12 PM
aleksanderpolo
Offline
• • •
Upload & Sell: Off
p.2 #15 · p.2 #15 · Fuji XPro1 vs Nex 5n Comparision Images


Is the left one Fuji btw? Now I can see that there are tons of artifact on the window frame on the left panel...

Anyway, thanks for the exercise, I guess in the end the user has to decide for themselves whether they are bothered by these issues or not. I look at 100% crop when I am deciding on sharpening and noise reduction, so I think these artifacts/water color are going to stab me in my eyes.


flashinm wrote:
Here's a sharpened copy (more sharpening for Nex) with a little PP to remove the color artifacts. Some zippering artifacts are still there.

http://img.auctiva.com/imgdata/1/5/2/6/6/4/4/webimg/632919753_o.jpg?nc=578




Dec 18, 2012 at 09:47 PM
michaelwatkins
Offline
• • •
Upload & Sell: Off
p.2 #16 · p.2 #16 · Fuji XPro1 vs Nex 5n Comparision Images


douglasf13 wrote:
An unsharpened 5N barely competes with the detail of the 12mp, AA-less GXR M module.


Having owned both the 5N and GXR/M and shot with the same lenses on both, I have to agree with Douglas and by extension FlyPenFly as well.

There are multiple good to great raw processing options for Bayer cameras but the same can't be said for X-Trans equipped cameras.

The bottom line for me is that while X-trans raw file processing remains in this state I find it impossible to get excited about their cameras. That's a shame because I'm rather looking forward to a new and improved X100 but fear that if Fujifilm is working on an update for the X100, it'll incorporate a X-Trans sensor.



Dec 18, 2012 at 10:02 PM
uscmatt99
Offline
• •
Upload & Sell: On
p.2 #17 · p.2 #17 · Fuji XPro1 vs Nex 5n Comparision Images


Even the GXR files are quite variable regarding the degree of color artifact depending on what is used for RAW conversion. When I started using the GXR, Aperture 3 was giving me fits to the point I was stopping down to f/11-16 to get diffraction softening, then resharpening. RPP put this to rest for the GXR, but still appears to have big issues with the X-trans sensor. I've settled into using LR4 for the GXR files, but it's not as good as RPP for conversion and I'd probably use RPP for anything I'd want to print large.


Dec 18, 2012 at 10:33 PM
Emacs
Offline
• •
Account locked
p.2 #18 · p.2 #18 · Fuji XPro1 vs Nex 5n Comparision Images


flashinm wrote:
Here's a sharpened copy (more sharpening for Nex) with a little PP to remove the color artifacts. Some zippering artifacts are still there.

I guess the difference between what I'm saying and what you and Lan11 are saying is that I believe this is 100% software related. You keep telling people that even with proper conversion, trans x can never compete with bayer at low ISO. I think this thread has done a good job of proving that false, even without proper raw support. If you want to keep bringing up chroma resolution in every thread that involves Fuji, that's
...Show more
You don't get the idea. Both sensors has the same resolution in your terms. But in practice we need relatively artifact-less picture with as much details as possible. And X-Trans is inferior in this aspect, because it's resolution is limited in general by lower resolved blue and red channels.

Just my two cents: same level of details, but much less artifacts with NEX.


Of course, this picture doesn't prove anything, unlike this one:








Dec 19, 2012 at 11:15 AM
Lan11
Offline
• •
Upload & Sell: Off
p.2 #19 · p.2 #19 · Fuji XPro1 vs Nex 5n Comparision Images


doug13,
Unfortunately, TALKING about the Fuji problem is NOT the answer. You have to show how it looks. Most buyers, even experienced like myself, will not see it even when they heard about it. I didnít expect Fuji to release another prototype camera and wasted time testing it. I offered to post the picts. on the other thread, but there were no takers. It turns out that another person has examples and wrote about it earlier:
http://www.diglloyd.com/index.html

The links to the pictures show this ugly mosaic which exists in EVERY file including JPEGs and RAWs. It looks like a paint peeling off the wall or a leperís skin. Once you saw it youíll recognize it, albeit in some files it is more visible than in the others. The Fuji purposefully smears details to mask the problem. Otherwise these cameras would be unsalable.

I recently spent few weeks testing cameras to select a small system. The Sony NEX-7 was eliminated due to lack of hi quality lenses, though on the 16MP sensor the lenses they currently offer may be acceptable to some people.
The X-E1 at first glance looked fine (at least it has dioptric adjustment), until I discovered the hidden, ugly truth. I only managed to test the zoom for optical centering and resolution distributions across the frame at various zoom/f settings and gave up. I didnít want to waste more time. You canít evaluatew the lens as long as this problem persists. The properly working converter, when/if it appears, will be another $300 extra and would mess up my workflow and for me it is a show stopper.

Iím very satisfied with Oly E-M5 results and 12,45 and PL25 lenses, although I had to go thru 11 in total to find what I liked. The QC is poor in the Oly China and the P-L Japan lens factories.

On a final note to prospective buyers. If the ergonomics are important to you, and it should be in case for such small cameras, donít waste time ďtryingĒ cameras at the store. Rent what you need to assess, preferably at the same time, and give it a good spin.
Good luck.



Dec 27, 2012 at 10:42 PM
Mescalamba
Online
• • • •
Upload & Sell: Off
p.2 #20 · p.2 #20 · Fuji XPro1 vs Nex 5n Comparision Images


Can you share those RAWs please? (from original comparsion..)


Dec 27, 2012 at 11:02 PM
1      
2
       3       4       end




FM Forums | Alternative Gear & Lenses | Join Upload & Sell

1      
2
       3       4       end
    
 

You are not logged in. Login or Register

Username   Password    Reset password