How Come X-E1 IQ Lags Behind Many Others In Studio Shooting?

This is a serious question.

It is obvious from my posting history that I support Fuji and their innovativeness, as well as the results from the finished, marketed products.

Here are screenshots from the JPEG and RAW comparison sections in the X-E1 review. The results are essentially the same as for the X-Pro 1.

Why are even the X-E1 JPEGs inferior, markedly, to these three other cameras. One is even just a µ4/3 sensor but it is noticeably sharper!

Look at what happens with the green feathers on the Fuji, compared to the others.  First, the JPEG:

The RAW result is no better:

It looks as if Fuji cannot resolve green detail. Is that why Fuji seem to have trouble with foliage?

What is wrong here?

--
Cheers
Trevor G
Silkypix tutorials at: http://photo.computerwyse.com

There is no free lunch I am afraid. Fuji x trans sensor runs into problems with color bleeding in some situations. The upside is increased per pixel sharpness and lower noise levels. It is fujis first generation camera vs years of Bayer filters. I am sure once algorithms improve for third party raw developers and new processors arrives on new cameras this will improve.

Does it look like trouble with real world shots of foliage? look at originals

Good god !  They are appalling.  I don't know where to begin.

--
Without the darkness the light would be nothing !

Do you like this one better? I have sharpened the Fuji in Photoshop: Amount 90, Radius 1 and Threshold 2.

Same again, but this time with a radius of 2:

--
john carson

I really mean that it is much more detailed, primarily, in the green. The perceived "sharpness" is probably a result of the extra detail shown.

How can the OM-5 with a much smaller sensor be so much better?

How can the dpreview test chart show such amazing results in terms of black and white resolution, but miss out in colour, in green in particular, in the studio shots?

What would be beneficial is if someone with an X-E1 would load the studio RAW at 200ISO and do an in-camera RAW to JPEG conversion.  Just copy it into the working folder ifor the camera (on the SD card) and then open it via the Play button.

I would like to see one with default settings, and one or more with increased sharpness via the conversion menu. Surely this will produce better results?

--
Cheers
Trevor G
Silkypix tutorials at: http://photo.computerwyse.com

You probably think Silkypix can produce comparable output to the other 3 cameras?:-D

DP samples are not quite reliable, but it's the real problem for sure. Silkypix helps a bit here, but still bayer sensors are superior, here is the comparison between
NEX-7    Pentax K-5iis
NEX-5n   Fuji X-Pro1

OOC jpegs from fuji are even worse.

1. What RAF conversion was used for the interactive studio shots? The Image Quality Compared (Raw) states the following:

For a (more) level playing field for comparison we also shot our studio scene in Raw mode with each camera and converted it using Adobe Camera Raw.

I downloaded the X Pro1 RAF of the studio shot and viewed with RFC 3.2.9.1* and fiddled with the Tone, Sharpness and Highlight controllers.  I got very noticeable differences in the feather details with very little effort.

2. The Image Quality Compared (JPEG) does not specify the conversion method. Based on my observation of the RAF conversion in RFC, if these images are not from in-camera engine then results depend upon a non-optimised algorithm; on the other hand, if these are OOC JPEGs then the actual camera settings (default) may not be the best for this shot.

*  RFC Ver. 3.2.12.1 is still not available for download.  The link to 3.2.12.1 connects to the Ver. 3.2.9.1



Silkypix (saturation adjusted to match other samples, sharpened with unsharp mask).

Dpreview used a non m43 lens on the OMD E-m5 so il looks better.

--
www.flickr.com/ma_muse

Paraphrasing, "It's just a simple question of software".

Though I'm not a software engineer, I work with a bunch of them.  There are some problems that software can't solve. I'm hopeful that demosaicing X-trans Raw files will get better. The quality of the OOC jpegs helps found this fond hope.

-Framus      😳

Andy, The bus is leaving for school. Get off the internet and pursue your adolescent education or you will be even further encumbered in becoming an adult.

-Framus

--
The sky is full of holes that let the rain get in, the holes are very small - that's why the rain is thin.
Spike Milligan

Did you even look at the pictures at 100% ? Just take the first one and look at the color bleed around the edges of the flower...

The second one is really bad like its completely out of focus or their is some camera shake or something, but I think its just massive color bleeding, the green from the leaves is rubbing off all over the place.

Anyway, the point is I completely agree with the OP, don't get me wrong, I love my Xpro1 and the lenses. The form factor, ergonomics, controls, size, weight, lenses... all of these things are great with the Fuji X system. And of course if you don't pixel peep the pictures are great.

But I don't get why people go on and on about how the X-Trans sensor is revolutionary and everything, IT'S NOT ! It clearly has many problems and I'm starting to doubt that any RAW software will be able to fix these.

Fuji claims they had 4 years of R&D for this sensor and their OOC jpegs still have problems with color bleeding, imagine how long it will take the RAW software developer to reach the same level !

I will continue having fun with my Fuji gear and try not to pixel peep too much

Wrong, Framus ... Andy is already out in the school playground - playing games and annoying the other Fuji kids!

(As usual!)

I've been closely following the X-Pro 1 (and X100 and X-E1) since the X-Pro 1 was released.  I shoot exclusively raw with my current cameras (Canon 1Ds2 and Panasonic GF-1) and process with Adobe Camera Raw.  Two questions: (1) Because Fuji apparently knows how to process the X-Pro 1 and X-E1 raw data into wonderful jpegs, isn't it just a matter of time before Lightroom/ACR or some other raw processing software can, too?  (2) Is it reasonable to shoot raw plus jpeg with the Fujis, process the raw to tiff through Silkypix (or whichever raw converter is doing the best job), then move the tiff into Lightroom/ACR (or whatever you prefer)?

Trevor - pixel peeping comparisons can often be quite unhealthy and rarely gives the viewer meaningful information about the real-world usage of any camera. I totally agree that the image quality of the OMD is sensational ( I own one, and I love it) but it does not necessarily mean that other cameras are "better" than the Fuji. Instead, expect them to have different image characteristics because they have different sensors and different processing engines. This brings me on to RAW - as I'm sure has already been said currently there is no viable mainstream RAW support for the Fuji X Trans sensor, Silkypix does an okay job most of the time, but I am steering well away from XE1 RAW images for now.

I regularly shoot the OMD and the XE1 side-by-side and I would not say that the IQ of one is notably better than the other, it's just a little different, and will suit people with different preferences. The OMD files are a little sharper, but sharpness is not everything and can be added in postproduction. What is far more important (to me at least) is the overall performance of any given camera, and of course the handling and general ergonomics. Far too much emphasis is put on pixel peeping in my opinion! Just buy the camera you like best, the IQ is good on all of them.

I would have thought it was obvious by now that with this colour filter array its inevitable that whilst B&W resolution is really strong you will always get slightly lower colour resolution. This is just basic physics/maths and those waiting for some mythical magical algorithm to solve this issue will be waiting a very long time indeed.

There is an obvious point to all this that no one seems to get. If X-trans is so clever and "better" then why after all these years of Bayer sensor cameras did it take little Fuji to come along and do it? I wouldn't be surprised if the engineers at CaNikon or Sony had thought of this years ago and rejected it.

Don't get me wrong I love my X100 and will likely buy an X pro for myself for Christmas but this X trans is pure marketing hype type technology to create a differentiator. Even the fantastic 35mm lens on the X-pro shooting a resolution test chart can't actually get any more resolution than the Panasonic GX-1 with the PanyLeica 25mm mounted. The numbers are all there over on photozone.de.

What is RFC Software? I have not heard of it before.

I found it - its Fuji's Raw File Converter.

Greg.