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!

Kudo's to Phase One on the X-Pro2 work--raw conversion excellence

Paul2660

Well-known member
I posted this on Lula wanted to also post on getdpi

After reviewing C1 9.1.1 on over 200 different raw files from my X-Pro2, I have to say, great job. Phase One, has done a great job on pulling the finer details out without the creating the known issues that Adobe just can't seem to figure out, watercolor or just strange issues with the finer details.

I actually am finding what I consider to be better overall conversions with the X-Pro2 over my X-T1. But C1 does a nice job on those also.

One issue that I was doing on all my Fuji conversions is to slide the details slider to the left (as it's under noise reduction), don't do this move it to the far right as this really pulls up the finer details. The colors are excellent files look very good. Add to this the power of the local adjustments in C1, you have an excellent work space.

LR still leads with the ability to combine and make HDR or Pano's while still raw (dng) for but now the extra details I am seeing in C1, over rule the LR conversions. Just adds an extra step having to import back to LR for the pano, but it works.

Again, Kudo's to David and team, the results were worth the wait.

Paul Caldwell
 

sc_john

Active member
Paul,

Thanks for the tip on the detail slider; I had not understood that. It really does pull out fine detail in the X-Pro 2 file... even in a "torture test" like black feathers. I agree on C1 v9.1.1 vs LR, overall. However, until C1 activates lens profile info for X-Pro2, I do not think it is quite ready for "prime-time". This is especially noticeable in distortion for some lenses (XF 18/2 , XF 50/2) and with fringing. Never experienced these issues with C1 with X-T1 files, so I am confident C1 will resolve.

Thanks again for tip on Detail slider in NR.

John
 

Paul2660

Well-known member
Yes, the details slider is a huge huge difference maker. In older versions of C1, 7.x and older it worked just the opposite. I missed this totally, and with the Fuji files the difference is impressive.

It makes a difference on Phase files also. I just got into the LR process, and moved over C1 and was using the slider the same way. IMO in C1 it's a bit counter intuitive.

Paul C
 

bensonga

Well-known member
Thanks Paul. If it is not too much to ask, I would love to see a comparison on images you have processed with Lightroom and C1 which illustrates the improvements in detail you've mentioned here.

Gary
 

JaapD

Member
Paul,

You did not mention anything about the in-active lens corrections. It looks like a C1 bug to me. What is your opinion on this?
 

Paul2660

Well-known member
I believe it's a work in progress and Phase will add these later. I am assuming the added resolution made it necessary to redo all the corrections. The data is carried over from the camera automatically and it's a matter reading the corrections and applying them. LR does not allow you to turn this off, in the past C1 has given the ability to turn off the corrections, which at times I will select.

Paul C
 

scott kirkpatrick

Well-known member
In email exchanges with Phase on various forums, they explain that they treat each camera model individually. They certainly have to treat the new series of 24 MPx sensor-based Fuji's as a new problem, since the color specifications of the Bayer filter matrix have changed with the new chip, not only the size of the array. It's possible that the specific way in which the distortion information is conveyed in the RAF file has changed and this is taking additional time to unravel. I would expect to see distortion correction in a 9.1.n release one of these days.

By the way, I checked with my resolution test case files, and I find that in C1, setting the detail slider to 100 gives a little better resolution in the center of the image than setting detail to 0. But the effect was slight. At the edges, the slight smearing was due to lens/chip interactions at the edges. The detail slider did not seem to have any effect. There may be even better results at some intermediate value of the detail setting, as that is the default setting.

scott
 

Paul2660

Well-known member
Hi Scott,

For me the difference between details set to the middle and moved more to the right, to me is there, albeit, minor. However the difference from sliding the details all the way to the left (as I was doing) and either the middle or around 85 or so is most noticeable.

Paul C
 

scott kirkpatrick

Well-known member
Hi Scott,

For me the difference between details set to the middle and moved more to the right, to me is there, albeit, minor. However the difference from sliding the details all the way to the left (as I was doing) and either the middle or around 85 or so is most noticeable.

Paul C
Yes, for any file, setting the detail slider to zero smooths things visibly, while 100 crisps them up. I checked again with the usual wall of books, and the effect was bigger than with an outdoor scene. Book spines have a texture which is good for this purpose. This is presumably being done during the de-Bayering stage, before correcting tones and local modifications.

scott
 

scott kirkpatrick

Well-known member
Phase One on the X-Pro2 raw conversion vs AccuRaw

I suspect Sandy McGuffog, developer of AccuRaw (Mac) and PhotoRaw (for the iPad) has gotten as deeply into the problem of demosaicing XTrans files as anyone outside of Fuji. I recently shot a worst case example in which the lens -- my 23/1.4 -- outresolves the 24 MPx sensor in my X Pro2. In this shot taken at f/5.6, distance about 2.5 m, the text on the spines of the books across the room have dimensions in the 1 to 3 pixels range:

DSCF0542 3 by scott kirkpatrick, on Flickr

So I played with both AccuRaw and COne to find the best settings to extract these details. For COne, I could not remove the color artifacts that occur when the text is fine white letters surrounded by an area of solid color, but here's one of the best results. The three sliders for noise management were set at 80, 60, and 40:

DSCF0542 3 copy by scott kirkpatrick, on Flickr

In AccuRaw, with all three sliders set at 100 ( a little less also works), the chroma artifacts are removed, but there are tiny gaps in some of the characters:

DSCF0542_3x100 copy by scott kirkpatrick, on Flickr

I'm not sure this is much of a problem unless you are pixel-peeping, viewing the files at actual size at a screen resolution of about 100 dots per inch.

scott
 
Top