The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Tips for Raw Processing with the Fuji X-T1

archiM44

Member
Jono,
I am using the Exif Editor workaround, with X-E2 as the Camera.
Having used Capture One since my first digital camera (Olympus E-1) and subsequently M8 and M9 I find the program very easy to use.
I catalog in LR and take a first look there but do the developing in Capture One or occasionally Irridientand send the result back to LR catalog and sometimes do some work in Phootoshop CC.
With the X-Trans sensor I find Capture One is the best compromise, detail miles better than Lightroom/ACR, slightly less sharpness than Irridient but also less to zero of the artifacts.
The (technical) quality of what I was able to produce in the darkroom on slides and B&W
from 1954 until 2006 using mostly Leica's from M3 to MP and enlarging on Cibachrome for color compared to what the "worst"programs can make of the X-Trans files are reason for me to put things into their proper perspective.
 

douglasf13

New member
Jono,
I am using the Exif Editor workaround, with X-E2 as the Camera.
Having used Capture One since my first digital camera (Olympus E-1) and subsequently M8 and M9 I find the program very easy to use.
I catalog in LR and take a first look there but do the developing in Capture One or occasionally Irridientand send the result back to LR catalog and sometimes do some work in Phootoshop CC.
With the X-Trans sensor I find Capture One is the best compromise, detail miles better than Lightroom/ACR, slightly less sharpness than Irridient but also less to zero of the artifacts.
The (technical) quality of what I was able to produce in the darkroom on slides and B&W
from 1954 until 2006 using mostly Leica's from M3 to MP and enlarging on Cibachrome for color compared to what the "worst"programs can make of the X-Trans files are reason for me to put things into their proper perspective.
If it were an issue with grain, tonality, etc., I'd agree, but my X-Trans camera's files issues and general look, particularly in my usual converter choice of LR or Aperture, are my least favorite looking files that I've had since getting into digital.

I've been shooting the X100 over the RX1 these days, so I'm not clamoring for the best IQ, since just about any digital is good enough these days, but the "weirdness" of certain X-Trans files was enough for me to loose money by going back to the X100 from the X100s. Of course, YMMV.
 

jonoslack

Active member
If it were an issue with grain, tonality, etc., I'd agree, but my X-Trans camera's files issues and general look, particularly in my usual converter choice of LR or Aperture, are my least favorite looking files that I've had since getting into digital.

I've been shooting the X100 over the RX1 these days, so I'm not clamoring for the best IQ, since just about any digital is good enough these days, but the "weirdness" of certain X-Trans files was enough for me to loose money by going back to the X100 from the X100s. Of course, YMMV.
Hi Douglas
Like you I'm not clamouring for the best IQ; I'm trying to decide which to keep out of an A7, an E-M1 and the X-T1, but now I'll be on my toes, looking for weirdness!
 

douglasf13

New member
Hi Douglas
Like you I'm not clamouring for the best IQ; I'm trying to decide which to keep out of an A7, an E-M1 and the X-T1, but now I'll be on my toes, looking for weirdness!
I certainly hope you're not bothered by any IQ weirdness, because it really seems that the XT-1 and lens lineup is about is good as it gets in a mirroless camera system these days.
 

GlenC

Member
Hi Douglas,

I'm wondering if you might have saved any of the files that caused the strange colors that you could share. It would be interesting to try them in the latest version of C1 or others to see if any improvements have been made to the algorithms.


Glen
 

JonMo

New member
I can't comment specifically to the x-t1 files as mine has not arrived yet; but the x-e2 files that I have processed have none of the weird stuff that you guys are finding.

No green sparklies.
No mushy green foliage.

I have only used Capture One.
I have only owned an X cam for 6 months and only the latest x-e2.

Now the only reason I have used C1 over the years is that for .mos files from my leaf back; they were the only Windows based game in town.

C1 7.2 has a decent DAM section with catalogs but only if you are starting from scratch. Bringing in an archived library is absolute crap.

Before the C1 need I had Aperture then Light room, both of which have really nice DAM.

Outside of this; why is it necessary that a person has to use only one Raw Processor?

I am not the youngest person here but I remember my darkroom having a vast array of developers and procedures to get the look I wanted.
The amount of money I spent on various films, chemicals and papers was a bit obsessive.

When I switched to digital I spent just as much on various printers, inks, third party inks, papers and when they became available; different developing programs for images.

I know my bias for money spent is off, I have purchased a single tech cam lens for $6000.00 and though it was a great deal, but why is it so important to only use one system for Raw processing?

A single DAM is important as well as back ups, but all the Raw programs I use all have similar controls, settings, and layouts. The beauty is that they all render different and sometimes; different is good.

And what the heck is "best image quality"? I had always assumed that was a marketing term for "please buy my camera, pretty please".

Well, when my cam FINALLY shows, perhaps I will join in on the weird stuff. :)

My procedure for initial Raw for Landscape X files in Capture One 7.2;
1. In the details menu; Set both noise reduction sliders to 0. (default is 50)
2. color menu; Set the Curve applicable to the subject (I usually like the "film high contrast")
3. then all the typical settings for exposure, high lights etcetera.
4. I like to take the Clarity settings with images of fine detail as "Method" Neutral. Commonly seem to be around the 15 value. Its very easy to make the image look stupid with this slider, but with practice it is exceptional with foliage.
5. Finally back to the Details slider. The X files seem to require very little noise reduction. I will simply nudge them up till acceptable.
6. SHARPENING; I have no set sharpening that I have found that will always get the look I want. But zoomed to 100%, start with low Rad (.2) High Amt ( as much as 800) Then adjust the sliders towards each other till the micro contrast shows up and the halos are controlled.

That's it other than prepping the image for print; most the image is perfected in C1 then off to Photoshop to Rez (if necessary) and sharpen for format ( Web, Print, Proof or off to CMYK)

I know it is not the most scientific list but seems to work for me.
 

douglasf13

New member
Hi Douglas,

I'm wondering if you might have saved any of the files that caused the strange colors that you could share. It would be interesting to try them in the latest version of C1 or others to see if any improvements have been made to the algorithms.


Glen
I've tested C1, LR, Accuraw, Aperture, Iridient and the camera jpegs, and I wasn't really thrilled with any of it (Accuraw seemed best, since the conversion process is adjustable.) C1 doesn't have the artifact issues of Aperture, but there also isn't as much detail. It seems somewhere between LR and Aperture. Besides, like Jono, I didn't really want to change my workflow, and I'm pretty married to LR at this point, anyways.

Outside of the weird IQ things that pop up, even perfect raw conversion with X-trans can never yield more than half of the chroma resolution of Bayer, so, while luminance resolution is high in X-trans (like with test charts,) color just looks oddly smooth to me. I'm just not a fan of the look of the files, and LR is particularly bad to me.

This all used to be mentioned more in the forums, but I think that most of us that were disenchanted with this sensor have already moved on. Since I've recently come back into the Fuji world with the "old" X100, I'm the only one getting sucked into these threads again. LOL

Either way, if you like the look of the X-trans files, the X-system really looks great.
 

kdriceman

New member
We still need Adobe to get raw working for Lightroom and also Capture One and others. If this camera is as popular as we hear, what is taking so long???


You can download the Adobe DNG Converter 8.4 Beta here: Download Adobe Photoshop Camera Raw 8.4 Beta - Adobe Labs

It will convert the XT1 raw files to DNG (in batches) which can be imported and edited in LR. Only slightly inconvenient, but works fine until LR gets it's update out.

I've converted 500 or so XT1 files and imported them into LR.
 

turtle

New member
You've nailed it, Douglas. I too find the Xtrans file 'look' a little artificial under too many circumstances and see no problem with bayer cameras. I think Fuji solved a problem that's not really there and in the process created a bunch of their own. It will be interesting to see what sensor appears in the X-Pro 2....
 

raist3d

Well-known member
Outside of the weird IQ things that pop up, even perfect raw conversion with X-trans can never yield more than half of the chroma resolution of Bayer, so, while luminance resolution is high in X-trans (like with test charts,) color just looks oddly smooth to me. I'm just not a fan of the look of the files, and LR is particularly bad to me.
I would really like to know where is this idea that Xtrans can't yield more than half the chroma resolution of a Bayer sensor. That seems rather odd to me even after taking into account the less red, less blue, more green of Xtrans over a Bayer.

Also one thing where Xtrans really shines over is B&W photos and it does do this better than Bayer (because it has more green). The luminance resolution is not just about "test charts"- this is real world.

Here's one of the issues of AA less Bayer- this issue doesn't show up on the Fuji in this area. The Fuji can still get color moire but it's much more rare.

(Ricoh GR - a phenomenal camera, this is not a completely regular occurrence this bad and a non issue in B&W)

I find the Xtrans has a vibrancy of color that nails a lot of skin, and others really rather well in mixed light I am not seeing as easy to get with other equipment. This could be partially the lens & coatings but could also be the Xtrans and how Fuji massages the data.



One thing Fuji mentions is that they find they can keep color accuracy/tone better due to the way the Xtrans samples the color. I would imagine this is at a slight expense of resolution (chroma) but gradients and general tone may very well look better.

How Xtrans samples color vs Bayer:


My conclusion here is both Bayer and Xtrans have their own artifacts- and I see a lot of Xtrans examination of a shot where some artifacts showed up without having the exact same shot on a Bayer to see what it would do, which I believe it's not fair to the Fuji engineering solution.

Fuji traded some artifacts for others, with the mind of eliminating the AA filter risking less color moire than an AAless Bayer solution. I do think it pays off in them keeping a bit of a "3d bite' a step towards Foveon, while actually making B&W photography better.

I also agree getting a good RAW converter can be a pain, and if your workflow depends on Adobe, then it's totally understandable why not get Xtrans.

But many pros are getting awesome landscapes full rich in color and at this point I don't think the sensor is really overall an issue taking exception with having to use a pure Adobe workflow.









- Ricardo
 

raist3d

Well-known member
One more thing for Douglas- on first pass it seems there isn't much detail with Capture One 7. I discovered proper tweaking with "Structure" slider and the usual iteration of sharpness and luminance noise reduction can get you some very sharp files.

- Ricardo
 

douglasf13

New member
One more thing for Douglas- on first pass it seems there isn't much detail with Capture One 7. I discovered proper tweaking with "Structure" slider and the usual iteration of sharpness and luminance noise reduction can get you some very sharp files.

- Ricardo
These posts by Joakim "theSuede" sum things up nicely. He works in the industry and is probably among the most knowledgeable tech guys on just about any camera forum, so he's a good guy to ask. You could probably send him your questions directly.

X-Pro 1 tested by Pop Photo - FM Forums

My X-E1 vs D7100 High ISO comps - FM Forums

http://www.fredmiranda.com/forum/topic/1223542/1#11653796

While I'm at it, here is a pretty interesting recent thread comparing Bayer to X-trans, and, even with some of the more highly thought of raw converters for X-trans, like Photo Ninja, there is a difference in "look" with X-trans files in some scenes. My X-E1 vs D7100 High ISO comps - FM Forums
 

raist3d

Well-known member
These posts by Joakim "theSuede" sum things up nicely. He works in the industry and is probably among the most knowledgeable tech guys on just about any camera forum, so he's a good guy to ask. You could probably send him your questions directly.

X-Pro 1 tested by Pop Photo - FM Forums

My X-E1 vs D7100 High ISO comps - FM Forums

My X-E1 vs D7100 High ISO comps - FM Forums
I would probably ask about the 50% chroma resolution, but I think that's about tit.

He can be pretty knowledgeable but so can the people at Capture One, and I am not going to take one man's explanation over what I am seeing myself as output from Capture One. Theory is nice, but doesn't mean he's infallible, and a lot of the talk I am seeing here does not reflect the reality of the output I am getting.

Do note the does mention that there are other more computationally heavy options to get more detail as he sees it (which could very well be what Capture one 7 is doing).

I still stand by what I said- I see picking on Xtrans artifacts, which it has, without looking at AA Bayer which has its own set of issues, or AA less bayer which introduces other issues.

Finally keep in mind some of the links you are quoting are pretty old by now.

While I'm at it, here is a pretty interesting recent thread comparing Bayer to X-trans, and, even with some of the more highly thought of raw converters for X-trans, like Photo Ninja, there is a difference in "look" with X-trans files in some scenes. My X-E1 vs D7100 High ISO comps - FM Forums
I am only seeing LR 4.x there. I am vouching for Capture One 7.
If you can point to where they are using Iridient or Capture One 7 in that thread, I would appreciate it.

- Ricardo
 

douglasf13

New member
I would probably ask about the 50% chroma resolution, but I think that's about tit.

He can be pretty knowledgeable but so can the people at Capture One, and I am not going to take one man's explanation over what I am seeing myself as output from Capture One. Theory is nice, but doesn't mean he's infallible, and a lot of the talk I am seeing here does not reflect the reality of the output I am getting.

Do note the does mention that there are other more computationally heavy options to get more detail as he sees it (which could very well be what Capture one 7 is doing).

I still stand by what I said- I see picking on Xtrans artifacts, which it has, without looking at AA Bayer which has its own set of issues, or AA less bayer which introduces other issues.

Finally keep in mind some of the links you are quoting are pretty old by now.



I am only seeing LR 4.x there. I am vouching for Capture One 7.
If you can point to where they are using Iridient or Capture One 7 in that thread, I would appreciate it.

- Ricardo
Sorry, I linked the wrong link for the recent thread comparing Bayer and X-trans: Fujifilm X-Trans Infinity Scene RAW Processing Comparison - FM Forums

FWIW, Joakim's opinion about X-trans hasn't changed since those threads, and his statements about the cfa arrangement won't change, anyways.

I'm happy that you're satisfied with C1's output and X-trans. I certainly wish I was, but, like some others, I've tested it along with many other converters, and I wasn't happy with X-trans, so I went back to the Bayer in my lowly pair of X100 cameras. To be honest, among the converters I've tried with X-trans, I found C1 to be middling.
 

raist3d

Well-known member
Sorry, I linked the wrong link for the recent thread comparing Bayer and X-trans: Fujifilm X-Trans Infinity Scene RAW Processing Comparison - FM Forums

FWIW, Joakim's opinion about X-trans hasn't changed since those threads, and his statements about the cfa arrangement won't change, anyways.
That's fine, if that's his opinion- he's more than entitled to it. My point is that since then progress has been made in raw conversions. Keep in mind again, he does mention that more computer intensive algorithms can do a better job than what he was showing.

I'm happy that you're satisfied with C1's output and X-trans. I certainly wish I was, but, like some others, I've tested it along with many other converters, and I wasn't happy with X-trans, so I went back to the Bayer in my lowly pair of X100 cameras. To be honest, among the converters I've tried with X-trans, I found C1 to be middling.
That's cool. In either case, I have from my view posted evidence that the so called mush in green is a non issue with proper converter use. I can also post a couple of shots with greenery I have that show the point. But I don't mean to convince you of otherwise. I just mention it in the general terms of the discussion because I have seen several issues pointed out that have a solution.

- Ricardo

PS: I looked at the thread you pointed out. If anything, several of those examples the Xtrans look better. In some others they are comparing images solely on Xtrans without a Bayer equivalent when Xtrans is showing quite a bit of detail.

And who knows what Capture One 7 settings they are using, etc.

Anyhow, thanks for the links. Keep in mind I am going by my direct experience too, not some link that some guys over the net are doing.

I leave with this link to a scene with a lot of greenery. I don't see any "mush" and there's quite a stress on DR in this scene too:

(full size, converted with Capture One 7)
http://raist3d.typepad.com/files/fullsizegreen2.jpg

Note: the burned areas in the following shot are recoverable. I left them as is for an overall impression of more contrast.
http://raist3d.typepad.com/files/fullsize_green3.jpg
 

Chuck Jones

Subscriber Member
I don't have a dog in this hunt since my X-Trans sensor Fuji is now an antique X-E1, but here are a couple of thoughts to keep in mind from the research I've done.

First thing, the X-Trans sensor is different from the Bayer based sensors and DOES require a different decode scheme that according to Brian at Iridient is a serious bit of work to program. Not a trivial task. So over a year ago when TheSwede wrote his comments, most of the RAW converters at that time had NOT implemented any change to their decoding algorithm. This resulted in a lot of the problems at that time.

I can't confirm all RAW converters have now been updated to accommodate the new sensor. Some may never, it is after all a lot of work. I would caution that even those that have may still need more work for optimization to get everything "dialed in" with any new camera release such as the X-T1. Those pixel herders are hard at work pushing around the digits, but good code takes time, so we all need to be a bit patient and give them some.

I can state for a fact that anyone using an older RAW converter is NOT going to see the advantages (or disadvantages as some see it) of the X-Trans sensor, even IF the image will open up. Most chose to just implement the same Fuji processing code used for the previous sensor when the X-Pro1 was released. While possibly a mistake in conversion accuracy, at least they had support for the new Fuji cameras. (Their thinking, not mine!)

Capture One 7.1 first implemented the correct algorithm changes. Even with the current 7.2 release I believe Capture One are still calling the X-Trans support "preliminary" not something they consider fully stable.

I would NOT suggest renaming files as some do to fool any RAW converter it into processing your X-T1 files and then comparing them with anything expecting the results to look the same as they will once the technical boys in the various back rooms get their thing done. There is an old saying in the software business that applies to exactly this type of Tom Foolery. Garbage in - garbage out.

Adobe has included preliminary support as well in the current Beta for the X-T1, but again I caution this is a new camera system, and beta software is just one small iteration past Alpha, the first try. Beta is NOT A FULL RELEASE - so things can, and often do, change prior to final release.

As far as Aperture, God and Steve Jobs only know what they are up to, and where they stand on support for the X-Trans sensor decode change. Apple seem to keep to themselves and rarely talk to outsiders like me, so I have no information to share here.

RAW Developer, or Iridient Developer as Brian now calls it, does have full and accurate support for the X-Trans products. Brian is one of the few who actually writes his own decoding algorithms, one of the major differences Iridient Developer has always had. Personally, of the many RAW converters I have evaluated I find Iridient to be my own first quality choice, though I use Lightroom for most quick processing as it is my Digital Asset Management system as well.

Many RAW converters rely on a GPL licensed public domain set of Objects for their actual conversion. This obviously would not work well for X-Trans sensor based cameras, as to my knowledge presently none of them have done the X-Trans changes. These libraries see no difference pre-X-Trans and post, so impossible to get even close to perfect conversions.

Uwe Steinmueller hasn't weighed in on this thread yet, and he has over two hundred and fifty seriously great X-T1 images posted here last I counted:

Fuji X-T1 - Digital Outback Photo

I believe Uwe is still using C1 and his own sharpening/finishing plugins, but as you can clearly see he is not experiencing smearing, color shifts, or anything else in the way of distracting nasties, nor has he commented to me about seeing any problems. He seems to really love the X-T1. I'll ping Uwe and see if he can shed any light here. It may be that a change in sharpening or workflow makes the difference....
 

douglasf13

New member
The comments from Joakim are not outdated, because they are primarily in regards to the inherent trade offs of the cfa scheme, rather than simply due to raw converter issues. He said that, even with perfect raw conversion, X-Trans can't reach more than 50% of Bayer's chroma resolution.

The problem with X-Trans comparisons is that you can't simply show a file that doesn't have an issue and say everything is fine. Some files don't have issues, others do, so it is highly dependent on the scene. Either way, if you can look at the recent thread I linked above and actually prefer the X-Trans output to the Bayer output, than there isn't much more for me to say. Our eyes and preferences are simply different. Nothing wrong with that.
 
Top