Fuji x-E1 Color Bleeding Locked

I have had the Fuji X-e1 with 18-55 kit for several weeks now and have noticed an annoying problem that I hope is fixable with a firmware update and that Fuji is working on.

The problem is color bleeding.  I noticed immediately that all of my out of camera jpegs seemed to have a "color cast" that seemed to be greenish hue.  Upon closer inspection I noticed that white areas of an image are saturated with surrounding green or reds.  Red stop sign. The white letters saying "STOP" are full of red bleeding.  In other images where there are green painted objects...the green bleeds into surrounding white.

I am quickly losing my love for this camera since jpeg out of camera was what I was hoping for as a quick lightweight camera to be used for general purposes so I could leave heavy DSLR gear behind for family outings and general shots.

Anyone else experiencing this issue?

--
Regards,
David
www.photographybydmichael.com

No

RAW or JPEG? Cause Ive seen nothing like that from JPEG. Neither from RAW processed in DCraw based converter (which are only ones that actually sorta work).

You should post an example.

Here is 100% crop OOC jpg standard settings in camera jpg



Notice white letters on stop sign have red bleed



--
Regards,
David
www.photographybydmichael.com

I have experienced no such problem.

--
Steve A. Kleinheider

Whoa!  That is bad.  What level of sharpness/noise reduction are you using?

I noticed almost the same thing on my first tests shooting a test chart with an xpro 1 included in a collection of objects and printed matter  seven or eight months ago. Red or white text seemed to be especially bad in terms of "filling in". (raw files). I included a coloured map in that test and it was reproduced so badly that I decided not to use the camera for copywork after all.

None of my other cameras do this.

I have two x-e1 and an x pro 1 and have not experienced those problems in raw or jpeg.

--
EPS

So they say - caused by the X-Trans color filter array.

It hasn't been a problem with any of my photos so far. I guess if you shoot a lot of stop signs, pixel peep, and get hung up the issue, it might annoy you. For the time being, and speaking for myself, I am tremendously happy with the output of the X-E1. Maybe you can get back to enjoying the benefits of this camera as well ...

http://www.dpreview.com/galleries/59028774/photos/2052349/xprodetail_dsf6057

http://www.dpreview.com/galleries/59028774/photos/2052347/x100detail_dsf0114

Example photos here of an x100 sensor and an xtrans sensor taken in the same light of the same  scene. Be sure to click on the "view original" and then click again to be sure you see the 100% view.

The text on the book, the map and the knurled knob on the camera in the bottom left highlight problem areas.

Hi Justin,

thanks for the links. Are both OOC Jpegs?

Thanks in advance.

Vlad

I had to open two pages to actually get what you are talking about plus a third here to respond. Upon a casual look the X-Trans appears to be the better image. It is unfortunate that you did not shoot these closer in timing of light and angle of camera. The sun/shade on the map makes comparing these difficult due to exposure differences. Lettering on the binder of the book are of a different angle, the XP1 is lower and to the left and reflectance makes it also difficult to compare other than again the X-Trans looks better. I don't see the color bleed there.

As far as the camera, it sure would have been nice to have included it in both. The knob, not a very obvious detail. The obvious flaw I can see is the false patterning on the focusing ring which I would loved to have seen compare to the X100.

OK, I went back for one more observation. I can see on the map especially with the "s" in "Peninsula" has been blotched. I don't see the complete letter. An exact comparison would have been much easier for me to view. I do appreciate you posting these in any event for they were helpful. Like I said overall I'd go for X-Trans, but for copy work, you may have a point.

You trivialize an observation the OP brings up and with attitude suggesting he quit getting hung up on minutia and get on with photography, as you see it. If he finds examples in a test that could come up in other scenes that he deems important then maybe what you flippantly point out, "shoot a lot of stop signs" may be shortsighted.

Another poster brought up their concerns for applying this system to copy work. Should that be deemed trivial as well just because you may not include that in your themes of capture?

I'm please that these issues are brought to light. They do make a difference to me and if someone finds including the rendering of a stop sign in there observation that can possibly lead to problems in their subject matter down the road, I'm glad to hear it. I have seen where with some this poses a real problem and with others, none what so ever and they are successful with their results. I don't assume that I know better than them what they should do about it.

Just because these observations don't seem to affect your shooting style doesn't mean they are invalid to others.

Thanks for those examples. They really show how bad the issue can be when using Adobe processing. They also show that the problem is not limited to foliage; it smears many kinds of fine detail like the map here or fabrics in other examples.

both images are from raw captures.

yep, I've seen that behaviour, although rarely that bad. What is your noise reduction set to?

Read the four or so ChromaSoft blog posts about difficulties demosasicing X-Trans raws.  Bleeding is one of the issues that needs to be addressed.

Images aren't in identical light conditions - makes comparison difficult. The 'bleed' is not consistent (in other red/white or light color areas) either. Finally - what level of sharpening? The XPro file is of far higher contrast and - if my aching eyes are any judge - over-sharpened, particularly on edges.

default sharpening  ..... i forget actually. the images are not identical ... different focal lengths and  sensor sizes making matching hard. however the images are a reasonable comparison and i have seen the same results  over and over in many different shots. if you do your own tests on the right subject matter you will see the same thing.

it interests me that the x100 file looks cleaner and sharper despite smaller pixel count. i also have encountered moire with the xtrans on distant brick walls shot slightly off level. visible when processed in silkypix but smeared away in lr4.