Jared Willson
Senior Member
- Messages
- 1,714
- Solutions
- 7
- Reaction score
- 1,894
One of the things I like to do when purchasing a new camera is to shoot a “studio” shot that will allow me to evaluate the performance of the camera and lenses at various apertures and focal lengths. It’s a pixel peeping exercise, but with the clear objective of letting me know what apertures are best at a given focal length, what’s going to produce results I’m happy with.
In addition, with the Fuji x-trans sensor there is the additional complication of how the various software packages handle the de-mosaic process. Is detail equivalent? Moire suppression? Noise levels? Saturation/color balance? There usually aren’t major surprises here with regard to different software packages, but because the x-trans sensor is different in its architecture, the results had the potential to be quite different.
To that end, I shot exposures of the same scene using Fujifilm 14mm (21mm equiv), 18mm (27mm equiv), 35mm (53mm equiv), and 60mm (90mm equiv) lenses. The camera was tripod mounted, and a cable release and 2s self timer were used throughout. White balance was performed manually and was not adjusted in software. All images are from processed raw files. Exposure adjustments were performed to set the black point and to allow identical (or nearly identical) exposures. No other adjustments in the raw converters were made.
As for the raw converters themselves… I already own Lightroom 4.4 and Aperture 3.4.5 (latest raw converters for each product as of 6/24/13). I have not tried Silkypix which shipped with the camera since, unless I am seriously dissatisfied with the other two packages, it is very unlikely I would change my workflow. I standardized on Lightroom about three years ago. All images were shot at base ISO of 200 (which I would rate at more like 160). The camera was moved to provide approximately the same field of view at all focal lengths.
Here is the full image as rendered in Lightroom
Lighroom Full Scene
And here is the same image as rendered by Aperture
Aperture Full Scene
First, the de-mosaic differences… Both Aperture and Lightroom showed very high levels of detail in the images—detail that compared well with my Leica M9 shot at base ISO of 160 with a 50mm Summilux Asph at f/4. The Leica had a slight edge in what I will call “three dimensionality” that the Fuji couldn’t match, but it was only visible at 100% magnification and was quite subtle. Both packages were winners in terms of detail—I didn’t feel like there was anything left on the table by either package—they were drawing out as much detail as I think a 16mp sensor with no anti-alias filter can provide. Again, the Leica provided a good reference here.
There were some obvious differences between Aperture and Lightroom, though. Color balance was slightly different between the two programs with Aperture favoring the greens and Lightroom adding some additional warmth which came across in the yellows and greens, but, strangely, not in the reds. At least with this scene, I prefer Lightroom’s interpretation, but it’s hard to say which is technically more accurate.
Aperture is significantly more heavy handed with regard to edge sharpening, to the point that details feel somewhat false when viewed at 100%. There is plenty of “pop” in the image, and it would likely print well, but screen appearance is slightly over-sharpened. This is true even with sharpening turned down significantly from the default value. I found I was happiest with this scene when sharpening was turned off completely in the default adjustments screen, but left at 100% in the Raw Fine Tuning screen. Even with these settings, though, Lightroom provided the more natural appearing scene to my eye.
Here is a 200% crop from the center of the image for Aperture and Lightroom. Both look pretty good considering the magnification. To my eye, Aperture appears more vibrant, slightly over-sharpened, and less natural. Remember, this is 200%!
200% Crops from Aperture and Lightroom
I prefer the way Lightroom renders the image. It isn’t a dramatic difference, and I’m sure excellent prints could be made from either presentation, but now that Adobe has improved the de-mosaic algorithm for x-trans sensors I feel they have a slight edge. Personally, I’d be happy with either and see no reason to try Silkypix.
What about the lenses?
Here are center crops and corner crops from all four lenses shot wide open and then at optimum aperture...
Center crops wide open
Corner crops wide open
Center crops at f/5.6
Corner crops at f/5.6
Summary of findings on the lens IQ:
14mm—Sharp in the center at all apertures. Sharp in the corners from f/4 on up. Best at f/5.6. Diffraction effects are visible by f/8 and significant by f/11. Use between f/4 and f/8 whenever possible, at f/11 if you need the extra depth of field.
18mm—Soft in the middle at f/2 with resolution improving by f/2.8 and quite good by f/4. Never reaches the standard set by the 14mm in terms of acutance, though this can be compensated for in Lightroom by careful selection of “clarity” and “contrast” values. Soft in the corners at all apertures.
35mm—Sharp in the middle even wide open at f/1.4 becoming exceptionally sharp by f/4. However, field curvature is quite significant, leading to soft edges and corners at f/1.4. The corners don’t really get good until f/4. Diffraction starts to become visible at f/8.
60mm—Sharp in the middle at all apertures with maximum resolution and contrast at f/5/6. Diffraction is visible by f/11. Corners are a little soft at f/2.4 but improve significantly by f/2.8. By f/4, corners are nearly as sharp as the center of the field. Saturation appears slightly lower in this lens for some reason—easy to adjust if required.
- Jared Willson
In addition, with the Fuji x-trans sensor there is the additional complication of how the various software packages handle the de-mosaic process. Is detail equivalent? Moire suppression? Noise levels? Saturation/color balance? There usually aren’t major surprises here with regard to different software packages, but because the x-trans sensor is different in its architecture, the results had the potential to be quite different.
To that end, I shot exposures of the same scene using Fujifilm 14mm (21mm equiv), 18mm (27mm equiv), 35mm (53mm equiv), and 60mm (90mm equiv) lenses. The camera was tripod mounted, and a cable release and 2s self timer were used throughout. White balance was performed manually and was not adjusted in software. All images are from processed raw files. Exposure adjustments were performed to set the black point and to allow identical (or nearly identical) exposures. No other adjustments in the raw converters were made.
As for the raw converters themselves… I already own Lightroom 4.4 and Aperture 3.4.5 (latest raw converters for each product as of 6/24/13). I have not tried Silkypix which shipped with the camera since, unless I am seriously dissatisfied with the other two packages, it is very unlikely I would change my workflow. I standardized on Lightroom about three years ago. All images were shot at base ISO of 200 (which I would rate at more like 160). The camera was moved to provide approximately the same field of view at all focal lengths.
Here is the full image as rendered in Lightroom
Lighroom Full Scene
And here is the same image as rendered by Aperture
Aperture Full Scene
First, the de-mosaic differences… Both Aperture and Lightroom showed very high levels of detail in the images—detail that compared well with my Leica M9 shot at base ISO of 160 with a 50mm Summilux Asph at f/4. The Leica had a slight edge in what I will call “three dimensionality” that the Fuji couldn’t match, but it was only visible at 100% magnification and was quite subtle. Both packages were winners in terms of detail—I didn’t feel like there was anything left on the table by either package—they were drawing out as much detail as I think a 16mp sensor with no anti-alias filter can provide. Again, the Leica provided a good reference here.
There were some obvious differences between Aperture and Lightroom, though. Color balance was slightly different between the two programs with Aperture favoring the greens and Lightroom adding some additional warmth which came across in the yellows and greens, but, strangely, not in the reds. At least with this scene, I prefer Lightroom’s interpretation, but it’s hard to say which is technically more accurate.
Aperture is significantly more heavy handed with regard to edge sharpening, to the point that details feel somewhat false when viewed at 100%. There is plenty of “pop” in the image, and it would likely print well, but screen appearance is slightly over-sharpened. This is true even with sharpening turned down significantly from the default value. I found I was happiest with this scene when sharpening was turned off completely in the default adjustments screen, but left at 100% in the Raw Fine Tuning screen. Even with these settings, though, Lightroom provided the more natural appearing scene to my eye.
Here is a 200% crop from the center of the image for Aperture and Lightroom. Both look pretty good considering the magnification. To my eye, Aperture appears more vibrant, slightly over-sharpened, and less natural. Remember, this is 200%!
200% Crops from Aperture and Lightroom
I prefer the way Lightroom renders the image. It isn’t a dramatic difference, and I’m sure excellent prints could be made from either presentation, but now that Adobe has improved the de-mosaic algorithm for x-trans sensors I feel they have a slight edge. Personally, I’d be happy with either and see no reason to try Silkypix.
What about the lenses?
Here are center crops and corner crops from all four lenses shot wide open and then at optimum aperture...
Center crops wide open
Corner crops wide open
Center crops at f/5.6
Corner crops at f/5.6
Summary of findings on the lens IQ:
14mm—Sharp in the center at all apertures. Sharp in the corners from f/4 on up. Best at f/5.6. Diffraction effects are visible by f/8 and significant by f/11. Use between f/4 and f/8 whenever possible, at f/11 if you need the extra depth of field.
18mm—Soft in the middle at f/2 with resolution improving by f/2.8 and quite good by f/4. Never reaches the standard set by the 14mm in terms of acutance, though this can be compensated for in Lightroom by careful selection of “clarity” and “contrast” values. Soft in the corners at all apertures.
35mm—Sharp in the middle even wide open at f/1.4 becoming exceptionally sharp by f/4. However, field curvature is quite significant, leading to soft edges and corners at f/1.4. The corners don’t really get good until f/4. Diffraction starts to become visible at f/8.
60mm—Sharp in the middle at all apertures with maximum resolution and contrast at f/5/6. Diffraction is visible by f/11. Corners are a little soft at f/2.4 but improve significantly by f/2.8. By f/4, corners are nearly as sharp as the center of the field. Saturation appears slightly lower in this lens for some reason—easy to adjust if required.
- Jared Willson
