Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Fuji Forum | Join Upload & Sell

1      
2
       3       4       end
  

Bang for your buck body - Travel focus

  
 
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #1 · p.2 #1 · Bang for your buck body - Travel focus


RoamingScott wrote:
You get FAR better results sharpening X-Trans files with unsharp mask in PS than any combination of LR sharpening settings. I would personally not use Fuji X-Trans if I wanted to exclusively use LR for post, it just doesnít play well with the files.

Capture One works significantly better, but for all my trying, I canít get along with that program.



We're going to have to disagree on this. Though see a note at the end.

I used to use the old "unsharp mask" approach extensively, back when that was often one half of the sharpening process and it was done in Photoshop.

But since more robust and flexible sharpening tools have been added to ACR I no longer use unsharp mask, and I haven't used it for years.

There are analogs to it in ACR and Lightroom: for example the "clarity" setting, which used in combination with texture (and at times, "dehaze") is very powerful and flexible. The combination of sharpening controls over radius, amount, details, and the ability to mask this is also quite useful.

I encourage anyone who still thinks that they are seeing the so-called "worms" issue to try the technique of reducing "details." It works quite well. If you think about how the details control works and about what actually happens in the "worms" scenario, this will make sense to you.

About that note: In one respect we might be in partial agreement. After minimizing or eliminating the use of the details adjustment in ACR, I use slightly different sharpening settings in Photoshop than I use with, for example, my Canon files. I might use a slightly larger radius for example. Of course, that is just an example of something pretty normal when it comes to sharpening, namely that we tailor the precise setting and methods to the particular files we are using.

Dan



Jun 17, 2022 at 12:22 AM
RoamingScott
Offline
• • • • •
Upload & Sell: Off
p.2 #2 · p.2 #2 · Bang for your buck body - Travel focus


Iíve tried your method and every variation of it. Itís not adequate in my opinion.


Jun 17, 2022 at 12:42 AM
molson
Offline
• • • • • •
Upload & Sell: On
p.2 #3 · p.2 #3 · Bang for your buck body - Travel focus


https://imageswest.blogspot.com/2022/06/fujifilm-x-trans-artifacts.html


Jun 17, 2022 at 07:55 AM
Makten
Offline
• • • • •
Upload & Sell: On
p.2 #4 · p.2 #4 · Bang for your buck body - Travel focus


molson wrote:
https://imageswest.blogspot.com/2022/06/fujifilm-x-trans-artifacts.html


The C1 example looks pretty bad too, IMO. Smeared, as if there is a ton of NR applied.



Jun 17, 2022 at 08:59 AM
molson
Offline
• • • • • •
Upload & Sell: On
p.2 #5 · p.2 #5 · Bang for your buck body - Travel focus


Makten wrote:
The C1 example looks pretty bad too, IMO. Smeared, as if there is a ton of NR applied.


That's probably because it's shown at 200%, so there are a lot of interpolation artifacts that are not there if you view it at actual pixel resolution. There is no noise reduction applied (unless C1 applies it automatically, which would be pretty strange for an image shot at base ISO).

The point is, the C1 software doesn't create the unrepairable "Adobe worms"...



Jun 17, 2022 at 09:30 AM
Makten
Offline
• • • • •
Upload & Sell: On
p.2 #6 · p.2 #6 · Bang for your buck body - Travel focus


molson wrote:
That's probably because it's shown at 200%, so there are a lot of interpolation artifacts that are not there if you view it at actual pixel resolution. There is no noise reduction applied (unless C1 applies it automatically, which would be pretty strange for an image shot at base ISO).

The point is, the C1 software doesn't create the unrepairable "Adobe worms"...


Exactly what "200%" means unfortunately isn't as clear as it used to be, since high resolution monitors don't always show 1 pixel as 1 pixel. But since I can see how large the crop is in relation to the full image, I still think it doesn't look very good. No "worms", but still blotchy and a bit strange compared to (edit: sharp) Bayer images.



Jun 17, 2022 at 10:55 AM
molson
Offline
• • • • • •
Upload & Sell: On
p.2 #7 · p.2 #7 · Bang for your buck body - Travel focus


Makten wrote:
Exactly what "200%" means unfortunately isn't as clear as it used to be, since high resolution monitors don't always show 1 pixel as 1 pixel. But since I can see how large the crop is in relation to the full image, I still think it doesn't look very good. No "worms", but still blotchy and a bit strange compared to (edit: sharp) Bayer images.


I assume "200%" means 200% of the original pixels in each axis, i.e. 4x the total number of pixels compared to the original image, so your interpolation algorithm has to create three new pixels for every one pixel in the original image.

I's a moot point, since the C1 images look great when you look at the actual pixels or lower magnification.



Jun 17, 2022 at 11:51 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #8 · p.2 #8 · Bang for your buck body - Travel focus


Makten wrote:
Exactly what "200%" means unfortunately isn't as clear as it used to be, since high resolution monitors don't always show 1 pixel as 1 pixel. But since I can see how large the crop is in relation to the full image, I still think it doesn't look very good. No "worms", but still blotchy and a bit strange compared to (edit: sharp) Bayer images.


Two points.

First, your point about the meaning of "200% magnification" is right on. We used to sort of understand that this meant one image pixel per screen pixel, and back then most monitors were displaying in the range of 70-something to 90-something pixels per inch. If I created a 100% magnification screenshot on a 72 ppi monitor and you looked at it on a 90 ppi monitor we weren't seeing the same thing, but it wasn't too far off.

But today screen resolution is generally over 200 ppi and sometimes a lot higher. Using the old notion of 100% magnification, a 100% crop would not be that much larger than a print made with a resolution of 300, which isn't uncommon at all. But back in the days of 72 ppi monitors, the equivalent print size would have been three or four times wider! Basically, we were looking a lot "closer" at the image.

Given this issue and the variations in actual screen resolutions, I think it is a good idea to add one piece of information when sharing so-called 100% magnification crops, namely the actual size (in inches or centimeters) of the image on your screen. If we don't do this, I might take something that is 2" x 2" at 100% magnification on my screen, share it, and you might be viewing it as a 1" x 1" or a 4" x 4" image or something entirely different.

So, while looking at images at high magnification on screen is very useful, it is hardly objective unless we are clear on what the magnification factor actually means these days.

I still hold that the ideal (at least for those of us who print) is to judge printed versions of images at some particular image size and resolution.

All of this raises another really important issue, and one that too often gets ignored in photography forums. Basically, any image from any digital capture system will break down in various ways as the magnification increases ó for example, it might pixelate, it might become "plasticky" (by means of compensation for pixelation), various tiny distortions and artifacts eventually become more visible. This is unavoidable and the predictable result of looking at digital images at huge magnifications. (It is useful, but it can also mislead.)

So, again, the real question isn't "what error can I find as the image breaks down at giant magnifications?" It is, "how does the image look when reproduced at sizes that I will use?"

The so-called "worms" issue is a great case in point. First of all, the problem used to be more of an issue than it is today. I used to encounter it with some regularity in certain kinds of images with Fujifilm files when I first used the system a decade ago. Today I almost never see it.

It is a very rare problem, and the situation in which it is more likely is somewhat predictable. It usually (virtually always) involves the green color channel, and it happens in situations where there is very fine detail (on the order of a single pixel) that is very slightly out of focus and the subject's details are random. The classic example is very fine detail in distant vegetation.

I know for certain that it some significant subset of the rare images in which this occurs I can eliminate it by using sharpening techniques suitable to x-trans files, most notably by avoiding the use or overuse of the details slider and compensating later with other sharpening adjustments if necessary. It is pretty straightforward once you figure it out, and it works in almost all cases.

What about the cases where it doesn't work perfectly? If you find one of these and look at an extremely high magnification on the screen you may still detect some remnant of the "worms" effect. (I'm convinced that it is a side-effect of the x-trans color filter array, and particularly of its lower frequency of repetition across the sensor and how this plays out in the green channel.)

But, how about making a print of one of these very rare examples where you can still find some "worms" at very high resolution on the screen, or. even how about looking at the image at typical screen display sizes?

After a decade of using this system, I have yet to run into one image in which this was a problem even at large real-world print and display sizes.

To summarize: The "worms" effect may may appear on x-trans sensor images in extremely rare cases when certain features are present. The effect can typically be eliminated by using lower (or zero) details settings and other Fujifilm-specific adjustments to sharpening. In the even rare situations when it cannot be fully eliminated from 100% and larger screen views, it virtually never is visible in a large print or other display.

Finally, regarding your point about Bayer images, we're almost on the same page (at least in the same chapter of the book!) here. My view is that both Bayer and x-trans sensor have pluses and minuses. For example, we might accept that Fujifilm claim that x-trans reduces the (very rare) occurrence of aliasing effects while it may also increase the (also rare) potential for needing to deal with "worms." Bayer may be (very slightly?) more likely to create aliasing at certain frequencies (though I virtually never see it), but it also might do a more accurate job of accurately rendering pixel colors in final files.

That's a lot of words to explain why I think Fujifilm should just move all of their cameras to Bayer. They've already done so with no downsides in the miniMF systems.

Dan




Jun 17, 2022 at 01:11 PM
fjablo
Offline
• • •
Upload & Sell: Off
p.2 #9 · p.2 #9 · Bang for your buck body - Travel focus


russellnadler wrote:
After a long...I mean LONG hiatus from photography, I'm hoping to rekindle things with an upcoming month long backpacking trip to Nepal. I'll have my S22 Ultra in my pocket, but I was hoping to supplement with a Fuji setup. Not trying to break the bank, but I'm think TBD body with the 18-55 kit lens. I used to have the X-E2 and loved it.

Anyone have a recommendation on what can be had on the cheap used? Don't need the latest features - still only, mostly landscape and street.


I've seen you also posted on the Nikon board and since I shoot Fuji & Nikon AND am also planning to do a backpacking trip in Nepal next year (if things go as planned), I thought I'd share my thoughts.

Currently planning to bring the following:
- Z7: because I have it, but Z5 or Z6 would do as well (Z5 offers piece of mind due to dual card slots, but Z6 has the slightly better sensor and much nicer rear LCD)
- 24-200mm: a compromise, but I like having a longer lens in the mountains and Nepal can be quite dusty so not too keen on changing lenses frequently. Of course 24-120 or 24-70 f4 would also be an option.
- 40mm f2: for low light & when I want a smaller kit and because it weighs almost nothing
- small tripod & CPL

I've decided against the Fuji because to me that system is great with primes, but doesn't offer zooms that can compete with the Nikon Zs. And I don't think frequently changing lenses would be an enjoyable experience (though I have no problem with it on city trips etc). Lastly, a comparable Fuji setup wouldn't be that much lighter either.

So while I like the Fuji system a lot, I think it's currently not the best option for that kind of trip. On the other hand a friend of mine did a Nepal trip 3 years ago with my old X-T1 and a XC 16-50mm and got some really cool shots as well - so all these gear considerations are a bit overrated anyways



Jun 18, 2022 at 04:01 AM
Makten
Offline
• • • • •
Upload & Sell: On
p.2 #10 · p.2 #10 · Bang for your buck body - Travel focus


molson wrote:
I assume "200%" means 200% of the original pixels in each axis, i.e. 4x the total number of pixels compared to the original image, so your interpolation algorithm has to create three new pixels for every one pixel in the original image.

I's a moot point, since the C1 images look great when you look at the actual pixels or lower magnification.


What I mean is that your 200% crop won't look the same on every viewers screen. High resolution screens will interpolate too, which makes it even worse in some cases.
It also depends on the software you use when making a screenshot (or crop). Some interpolate, some don't (upscaling doesn't require interpolation, you can just quadruple one pixel into four identical pixels).

Theory aside, I simply don't like the look of that crop, even if it's much better than the wormy ones.



Jun 18, 2022 at 05:52 AM
 


Search in Used Dept. 

AmbientMike
Offline
• • • • •
Upload & Sell: Off
p.2 #11 · p.2 #11 · Bang for your buck body - Travel focus



russellnadler wrote:
After a long...I mean LONG hiatus from photography, I'm hoping to rekindle things with an upcoming month long backpacking trip to Nepal. I'll have my S22 Ultra in my pocket, but I was hoping to supplement with a Fuji setup. Not trying to break the bank, but I'm think TBD body with the 18-55 kit lens. I used to have the X-E2 and loved it.

Anyone have a recommendation on what can be had on the cheap used? Don't need the latest features - still only, mostly landscape and street.


Here's a similar thread I started 2 years ago. Should still be useful mostly 24mp recommendations

https://www.fredmiranda.com/forum/topic/1642545



Jun 18, 2022 at 12:07 PM
zConcept
Offline

Upload & Sell: Off
p.2 #12 · p.2 #12 · Bang for your buck body - Travel focus


I would go with a xe4 with 27mm and pick up the fuji 14/2.8 used. I've seen them for 400 Canadian. Maybe then later get the fuji 35/1.4. This lens lives on my xe4 and my copy is actually super sharp.


Jun 19, 2022 at 11:59 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #13 · p.2 #13 · Bang for your buck body - Travel focus


zConcept wrote:
I would go with a xe4 with 27mm and pick up the fuji 14/2.8 used. I've seen them for 400 Canadian. Maybe then later get the fuji 35/1.4. This lens lives on my xe4 and my copy is actually super sharp.


That makes a nice little trio. The 35mm f/1.4 isn't too big, but you can get even smaller (and less expensive) with the 35mm f/2 as an alternative, also.



Jun 19, 2022 at 11:40 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #14 · p.2 #14 · Bang for your buck body - Travel focus


RoamingScott wrote:
You get FAR better results sharpening X-Trans files with unsharp mask in PS than any combination of LR sharpening settings. I would personally not use Fuji X-Trans if I wanted to exclusively use LR for post, it just doesnít play well with the files.

Capture One works significantly better, but for all my trying, I canít get along with that program.



300% magnification crop, XPro2 file, post processed with ACR and Photoshop. Roughly equivalent to a section of a 60" wide print. (But uninterpolated here.)

Seems fine to me...







Full image:








Jun 19, 2022 at 11:47 PM
Makten
Offline
• • • • •
Upload & Sell: On
p.2 #15 · p.2 #15 · Bang for your buck body - Travel focus


gdanmitchell wrote:
Seems fine to me...


Looks wormy to me. Especially considering the low contrast of the structure in the wall.
"Fine"? Probably. Still would not want that sensor when there are ones that don't do worms.



Jun 20, 2022 at 04:52 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #16 · p.2 #16 · Bang for your buck body - Travel focus


Makten wrote:
Looks wormy to me. Especially considering the low contrast of the structure in the wall.
"Fine"? Probably. Still would not want that sensor when there are ones that don't do worms.


You really, really, really want to believe this, don't you? ;-)

300% magnification.

The size on your screen should be equivalent to what you'd see in a print that is five feet wide.

From a 24MP APS-C sensor camera.

What you imagine to be "worms" is the normal random texture of a stucco-type wall surface...

... at 300% (!) magnification.

You do you. ;-)

But don't expect anyone to give a lot of credence to your pronouncements at this point.



Jun 20, 2022 at 09:28 AM
Makten
Offline
• • • • •
Upload & Sell: On
p.2 #17 · p.2 #17 · Bang for your buck body - Travel focus


gdanmitchell wrote:
You really, really, really want to believe this, don't you? ;-)

300% magnification.

The size on your screen should be equivalent to what you'd see in a print that is five feet wide.

From a 24MP APS-C sensor camera.

What you imagine to be "worms" is the normal random texture of a stucco-type wall surface...

... at 300% (!) magnification.

You do you. ;-)

But don't expect anyone to give a lot of credence to your pronouncements at this point.


Funny how the texture of the wall is visible even in the metal part and buttons to the left then.
As usual; you not seeing something doesn't mean that someone who does, "believes". It's clearly visible.

Magnification is irrelevant. A bayer sensor doesn't to that at all. Ever. Why would I choose a sensor that produces results that I don't like?



Jun 20, 2022 at 09:45 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #18 · p.2 #18 · Bang for your buck body - Travel focus


RoamingScott wrote:
You get FAR better results sharpening X-Trans files with unsharp mask in PS than any combination of LR sharpening settings.


I want to come back to this for a moment.

I wonder if you actually understand what unsharp masking is and how it is used as part of a sharpening workflow.

This is a little hard to describe in a text-only manner, but unsharp masking (USM) derived from an old film technique in which a somewhat unfocused and inverted version of the negative (thus a positive) was produced and layered with the original negative for printing. (This out-of-focus layer is the "unsharp mask," with mask here "masking" the amount of sharpening.)

The result was that contrast was somewhat increased in areas of existing sharp luminosity contrasts. By controlling the amount of "unfocus" the printer could control the (to use the Photoshop terminology) "radius" of the effect.

Generally USM was used produce what has been called local contrast enhancement near edges, but it would usually have a very large radius of many (perhaps 25-50 or even more) pixels in digital post. The effect is very similar to what the "clarity" control does in Adobe software, though I understand that Adobe applies some added "smarts" to clarity beyond pure USM sharping.

In Photoshop terms, what we used to refer to as USM was "low amount, very large radius" sharpening.

It also is not a thing that is remotely related to the so-called "worms" issue that is being discussed here.

That would be connected to what might be called "details sharpening," which uses high amount and small radius settings ó the opposite of what we typically think (though?) of as USM. Rather than spreading a subtle contrast enhancement over larger areas around high contrast areas, this "detail" sharpening constrains the sharpening to very small dimensions (often less than a pixel) and much larger "amounts" ó e.g. "more sharpening right at edges."

There are a ton of misconceptions floating around about how sharpening works and about the various methods to use when sharpening. Fortunately it is easy to do the experiments yourself and learn how these things work. A couple of good experiments are a) to work with an image of alternative pure black and white areas and/or b) work with areas of pure gray alternating with white and/or black. As you play with the various sharpening settings and look at the image at a high magnification the effects of each type of adjustment become pretty darned clear.

Try it. This is a super useful exercise.

Dan

Edited on Jun 21, 2022 at 09:44 AM · View previous versions



Jun 20, 2022 at 09:45 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.2 #19 · p.2 #19 · Bang for your buck body - Travel focus


Makten wrote:
Funny how the texture of the wall is visible even in the metal part and buttons to the left then.
As usual; you not seeing something doesn't mean that someone who does, "believes". It's clearly visible.


That's pure nonsense.

ANY image at 300% ó x-trans, Bayer, film ó†is going to show _some_ artifacts at 300% magnification.

The texture of the wall is not remotely "visible in the metal part and buttons." In fact, even at 300% magnification, it is arguably possible to distinguish the text on the buttons in this handheld shot.

Confirmation bias is a strange thing, and it is fascinating to watch it play our here. I guarantee that if I posted an example like this in a thread extolling the excellent sharpness of Lens X or Camera Z, where the fans of Lens X or Camera Z hang out, they would be using this very example to demonstrate how well their favorite lens/camera performed. But here, where the image is used to challenge nonsensical and hyperbolic negative claims, it "proves" the opposite to the true unbelievers.

Ah, photography forums.

I wonder about some of you folks.

Meanwhile, I'll continue to make excellent and quite large prints from these files while y'all moan about phantoms.

Have fun. ;-)



Jun 20, 2022 at 09:53 AM
Makten
Offline
• • • • •
Upload & Sell: On
p.2 #20 · p.2 #20 · Bang for your buck body - Travel focus


gdanmitchell wrote:
That's pure nonsense.

ANY image at 300% ó x-trans, Bayer, film ó†is going to show _some_ artifacts at 300% magnification.

The texture of the wall is not remotely "visible in the metal part and buttons." In fact, even at 300% magnification, it is arguably possible to distinguish the text on the buttons in this handheld shot.

Confirmation bias is a strange thing, and it is fascinating to watch it play our here. I guarantee that if I posted an example like this in a thread extolling the excellent sharpness of Lens X or Camera Z, where the fans of Lens
...Show more

You are projecting your own inabilities on others, just like you always do. Please don't quote me again, I'm not interested in your excuses.

Edit: For other readers, here's a 200% crop with various structures, from a bayer sensor... Artifacts? I can't see them.








Edited on Jun 20, 2022 at 10:50 AM · View previous versions



Jun 20, 2022 at 10:35 AM
1      
2
       3       4       end






FM Forums | Fuji Forum | Join Upload & Sell

1      
2
       3       4       end
    
 

You are not logged in. Login or Register

Username      Reset password