Test images from the Touit 32/1.8 Fuji X mount

[awilder]

This isn't just a test of lens sharpness but more importantly how much of a difference the RAW file converter makes in delivered sharpness. Unlike the 35/1.4 Fujinon with true optical correction of distortion, the Touit utilizes a default digital correction in LR 4 to achieve this goal that results in slight sharpness loss across the entire image possibly because there is no lens profile to date for Touits in Fuji X mount, just the E mount version. Unfortunately, it can't be disabled. There are other RAW converters like Capture One that also let you set the distortion correction to "0".

With the Touit 32/1.8, there is a definite difference between opening up a RAW file in LR 4 verses AccuRaw (sorry, Mac app only). AccuRaw automatically gives a RAW image without the software correction for distortion. I also noticed a more neutral color balance in AccuRaw while LR 4 has a little more color saturation and possible slight magenta (?) color shift.

Here is a test shot with the 32/1.8 @ f/2.8. The first crop at 100% is the USAF resolution center target using LR 4 measuring about 64 lp/mm and the second crop is with AccuRaw that measures about 80-88 lp/mm. Frankly, I found the distortion level without correction to be minimally objectionable and would much prefer the greater sharpness.

 

[AdamJ]

I've never seen Adobe tie a default 'behind the scenes' distortion correction to a lens before, only on compact cameras or non-removable lenses. Although I am not surprised if this is true. Are you sure the worse resolution is due to this process, and not just the fact that Adobe's X-Trans demosaic is the softest of all the current RAW processors?

I've got some X-PRO1 / 35mm 1.4 images that are noticeably softer processed by Adobe than by CO, Aperture, or RPP...

 

[willie_901]

Adam,

I believe LR does this (apply manufacturer supplied corrections for raw images) for many Panasonic m4/3 lens.

 

[awilder]

I'm going by what others have said on the Fuji X forum and my own tests about the sharpness robbing effect of the LR digital distortion correction but was unaware of the LR's demosaic effect. Posted below are the full frame images comparing the Touit processed with AccuRaw (left) vs LR 4 (right). Note the difference in distortion, color saturation and slightly greater field without the distortion correction effect in AccuRaw. I did find the Fujinon 35/1.4 a little sharper measuring 72 lp/mm compared to 64 lp/mm on the Touit with the same chart using LR 4. It was only going to AccuRaw that I was able to extract the greater sharpness from the Touit at 80-88 lp/mm. I'll need to do further tests comparing Fujinon lens test shots between LR 4 and AccuRaw.

 

[Ron (Netherlands)]

This isn't just a test of lens sharpness but more importantly how much of a difference the RAW file converter makes in delivered sharpness. Unlike the 35/1.4 Fujinon with true optical correction of distortion, the Touit utilizes a default digital correction in LR 4 to achieve this goal that results in slight sharpness loss across the entire image possibly because there is no lens profile to date for Touits in Fuji X mount, just the E mount version. Unfortunately, it can't be disabled. There are other RAW converters like Capture One that also let you set the distortion correction to "0".

With the Touit 32/1.8, there is a definite difference between opening up a RAW file in LR 4 verses AccuRaw (sorry, Mac app only). AccuRaw automatically gives a RAW image without the software correction for distortion. I also noticed a more neutral color balance in AccuRaw while LR 4 has a little more color saturation and possible slight magenta (?) color shift.

Here is a test shot with the 32/1.8 @ f/2.8. The first crop at 100% is the USAF resolution center target using LR 4 measuring about 64 lp/mm and the second crop is with AccuRaw that measures about 80-88 lp/mm. Frankly, I found the distortion level without correction to be minimally objectionable and would much prefer the greater sharpness.

I'm going by what others have said on the Fuji X forum and my own tests about the sharpness robbing effect of the LR digital distortion correction but was unaware of the LR's demosaic effect. Posted below are the full frame images comparing the Touit processed with AccuRaw (left) vs LR 4 (right). Note the difference in distortion, color saturation and slightly greater field without the distortion correction effect in AccuRaw. I did find the Fujinon 35/1.4 a little sharper measuring 72 lp/mm compared to 64 lp/mm on the Touit with the same chart using LR 4. It was only going to AccuRaw that I was able to extract the greater sharpness from the Touit at 80-88 lp/mm. I'll need to do further tests comparing Fujinon lens test shots between LR 4 and AccuRaw.

Sorry but I can't make anything out of these very small thumbnales....

 

[awilder]

Click on the thumbnail to make bigger. The size unfortunately is limited to 600 pixels in the long dimension because of RFF upload limitations. The purpose of the images is simply to compare distortion, color saturation, a frame coverage for an identical file but opened up by different RAW converters.

 

[awilder]

It appears that LR4 Fuji X default settings for both sharpness and detail is 25. Changing both to 40 upon import improves resolution very close to that of AccuRaw so it seems to have less to do with LR's distortion compensation but rather the default settings for sharpness and detail.

 

[willie_901]

While I do not use AccuRaw, my experience is similar. Sharpening, clarity and even de-fringing parameters I use for Bayer raw are not appropriate for Xtrans raw. I also find it pays to carefully evaluate those parameters on an image-by-image basis for Xtrans raw. I rarely do this for Bayer raw unless I need to make a large print.