The X-Pro 2 on the runway - a mini-review

[Darthfeeble]

I am sending mine back, We're not sure on the other, I had hoped for some insight to other post processing info. I know the Phase One program is supposed to be good. A little expensive.

 

[gavinlg]

I am sending mine back, We're not sure on the other, I had hoped for some insight to other post processing info. I know the Phase One program is supposed to be good. A little expensive.

I'm using capture one after using LR for years because it doesn't work well on my new windows laptop. They are much of a muchness.

A fuji file shouldn't be noisey in the skies at all. I can shoot at iso6400 with a previous generation x-t1 and it has about the same noise in the sky as an iso1600 canon 5d or sony a7 file. If you want to post some examples I'd be happy to have a look and tell you if it's normal?

As far as the mushy landscapes go, lightrooms sharpening for x-trans files is pretty average. The trick is - don't sharpen your files. They don't need it. You'll find they look far more natural.

 

[Darthfeeble]

Thank you. I did a little playing around and while I agree that the Pro1 is less noisy, exactly what you say about sharpening is the case. The 2 seems to not like LR sharpening at all, though Topaz detail makes it happy but not much is needed. I found that for the big expanses of sky, the best thing seems to be to just brush in some blur. I'll be keeping it now.

 

[willie_901]

Got my wife an X-T2 and liked what I saw so I bought a used X Pro2. We both went on an outdoor trip and I find that I am unhappy with the image quality in my Pro2. Very noisey particularly in the sky and the the mushy foliage blur that was such an issue with the first Xtrans sensor. Even the T2 has it. Both processed in Lightroom. Thoughts? Suggestions?

The X-Pro 2 and X-T2 should not render with such dramatic differences. Their data-stream technologies are almost identical.

Skies are always noisy... even when one uses base ISO with a perfect exposure. The noise comes from quantum noise (a.k.a photon noise or shot noise). This is true for all digital cameras. But the noise should only be obvious in extreme crops.

I suspect a more rigorous comparison is in order. Shoot the same subject at base ISO (200) with optimum and identical shutter times and apertures. Use the same DOF. When comparing green foliage, wait until the air is still. Use raw files or at least take care that all in-camera JPEG rendering parameters are identical and DR =100.

For shadow regions the X-T2 should be a bit cleaner than the X-Pro 2. For a sky the noise should be essentially identical. Here's a comparison of the in-camera electronic noise levels (read noise) vs ISO for both cameras.

For the X-Series cameras (and many other newer models from other brands), as light levels decrease the main noise contribution transitions from read noise to quantum noise. This is not the case for older digital camera technologies.

 

[Darthfeeble]

I confess to pixel peeping. One thing that surprised me is the Classic Chrome jpeg setting. I could see myself giving up RAW altogether. I won't though. Obviously my expectations were a bit unreasonable. All that said, my trusty X100 offers the cleanest files of any camera have owned. I can get shots at ISO 6400 that while noise is present, it is very manageable.

One point you made is to have the DR at 100%, does increasing that create some of my problem? I believe on that day I had it on the Auto setting and I was shooting at some pretty small apertures, thus it might have been more.

 

[adamjbonn]

I confess to pixel peeping. One thing that surprised me is the Classic Chrome jpeg setting. I could see myself giving up RAW altogether. I won't though. Obviously my expectations were a bit unreasonable. All that said, my trusty X100 offers the cleanest files of any camera have owned. I can get shots at ISO 6400 that while noise is present, it is very manageable.

One point you made is to have the DR at 100%, does increasing that create some of my problem? I believe on that day I had it on the Auto setting and I was shooting at some pretty small apertures, thus it might have been more.

Yes the DR modes can add noise to the RAWs, there's some very detailed info on this out there, so I won't rehash... but the DR modes basically underexpose the RAW 1 (DR200) or 2 (DR400) stops

It gets a little convoluted at this point, as it depends which photo editor you're using (as they handle the fact the DR mode has been used in different ways, some ignore it completely others automatically implement their own approximation of it)

But bottom line - underexposed RAW + lift exposure in post = more noise

LR/ACR is (in)famous for doing a poorer job with Fuji X-Trans. Many people have dropped $30 into the Iridient RAF to DNG convertor, which resolves the problem.

As far as I'm concerned, that ACR does a mediocre job with X-Trans is 100% an Adobe issue, pretty much every other RAW convertor out there does a better job (than ACR) and most people rate Iridient as the best.... and that's a one man band.

 

[Darthfeeble]

Thank you, I tend to underexpose for highlights as well as use the DR settings. I'll work on that to see what happens.

 

[ptpdprinter]

The OP seemed a little surprised that the Fuji did fine on the runway and backstage, even suggesting that he was taking a risk using it rather that his usual Canon gear. I don't see any reason why the Fuji, and a half a dozen other cameras, wouldn't do fine. What's with all the drama?

 

[honozooloo]

The OP seemed a little surprised that the Fuji did fine on the runway and backstage, even suggesting that he was taking a risk using it rather that his usual Canon gear. I don't see any reason why the Fuji, and a half a dozen other cameras, wouldn't do fine. What's with all the drama?

When you're assigned with shooting every single look during a weeks' worth of fashion shows without missing a single one, you tend to stick to the gear you know best. Unpredictable lighting and challenging conditions means you try to control what you can, in most cases even if you don't know what to expect, using a familiar set of gear may be one of the few factors you can actually get a handle on.

"I think it will work in these conditions" is not "I know it will work in these conditions" - hence the acknowledgement of the inherent risk in trying something challenging with new, untested gear, and also my surprise in finding out that the Fuji was up to snuff. I certainly didn't buy the X-Pro 2 for its tracking abilities, but finding out that they worked as well as they do was a pleasant surprise...I've owned the X-Pro 1, X-T1, XE-2, and would not have expected any of those cameras to handle these conditions nearly as well. Just in my experience.

 

[ptpdprinter]

I've owned the X-Pro 1, X-T1, XE-2, and would not have expected any of those cameras to handle these conditions nearly as well. Just in my experience.

I understand going with what you are familiar with, but that goes to your familiarity with the gear, not its inherent capabilities. What specifically is it about the Fuji cameras that led you to the expectation that the Fuji cameras wouldn't handle the assignment nearly as well as your Canon gear?

 

[willie_901]

...

One point you made is to have the DR at 100%, does increasing that create some of my problem? I believe on that day I had it on the Auto setting and I was shooting at some pretty small apertures, thus it might have been more.

When DR is >100 exposure (shutter time and aperture) is never maximized. The camera intentionally underexposes. Then, during in-camera JPEG rendering, the shadow regions are selectively pushed. Setting DR > 100 all but eliminates the risk of loosing highlight information caused by overexposing the sensor when the shutter is open and, or clipping the analog-to-digital converter after the shutter closes.

So, when DR is > 100 the images' signal-to-noise ratios are always compromised. The noise remains relatively constant, but the intentional automated underexposure reduced the signal levels. So DR > 100 reduces image quality in order to protect the photographer from blowing highlight regions (usually skies).

If you have DR on auto and use JPEGs, a photo from one X-Series camera could appear noisier than one made with a different X-Series body because of a difference in the DR parameter. The one with a higher DR has less signal because it is underexposed.

I use raw with LR CC. DR is always =100. I can try different the FUJIFILM rendering Camera Calibration Profiles.

I do protect myself from loosing highlights. I almost always automatically bracket aperture three raw exposures. I typically use +1/3, 0 and -1/3 stops. In very bright conditions I sometimes switch to -2/3, 0 , +2/3 aperture steps. In post-production I select the image with the optimum exposure (only unneeded highlights are lost). I delete the other two. I do this to maximize the raw data signal-to-noise ratio. When appropriate I selectively push shadow region brightness.

 

[Darthfeeble]

Again, many thanks for all the help here. What a great forum! If I understand the whole in camera adjustment thing, the DR setting should only affect the jpeg. My isssue was with the RAW output. None the less, I shoot a third of a stop under most of the time and sometimes even more. I'm trying shooting at 0 under to see what goes. The other compensation is to use the brush tool to reduce noise where it is most visible and as suggested, no sharpening. That has had the most affect toward improving my opinion of the image quality. The one factor that was most influential for buying the camera was the increased detail. I rather blindly continued the same workflow as the XPro1 and it's apparent that was a mistake.

 

[honozooloo]

I understand going with what you are familiar with, but that goes to your familiarity with the gear, not its inherent capabilities. What specifically is it about the Fuji cameras that led you to the expectation that the Fuji cameras wouldn't handle the assignment nearly as well as your Canon gear?

In the case of the X-T1, X-Pro 1 (lol) and X-E2, neither had buffers big or fast-clearing enough to handle rapid fire bursts of RAW files the way I needed them to. High ISO performance wasn't quite there either. The tracking AF on all three cameras was not bad (most notably the X-T1), but they were not as fast as the X-Pro 2's, or any number of Canon DSLRs. The responsiveness the older X-cameras and AF accuracy in low light was also not quite fast enough for me either - Canon performed just a little better before the X-Pro 2.

This is NOT a critique of the previous X-Cameras, which I would expect to be more than good enough for most. Hell they're more than good enough for me 95% of the time! But the older cameras weren't quite enough for my needs under demanding conditions, like fashion week. They were 'close' but not quite there.

 

[ptpdprinter]

In the case of the X-T1, X-Pro 1 (lol) and X-E2, neither had buffers big or fast-clearing enough to handle rapid fire bursts of RAW files the way I needed them to. High ISO performance wasn't quite there either. The tracking AF on all three cameras was not bad (most notably the X-T1), but they were not as fast as the X-Pro 2's, or any number of Canon DSLRs. The responsiveness the older X-cameras and AF accuracy in low light was also not quite fast enough for me either - Canon performed just a little better before the X-Pro 2.

Makes you wonder how photographers were able to cover runway/backstage back in the film days.

 

[back alley]

all the photographers had the same 'advantages'...just like today...

 

[adamjbonn]

Again, many thanks for all the help here. What a great forum! If I understand the whole in camera adjustment thing, the DR setting should only affect the jpeg. My isssue was with the RAW output. None the less, I shoot a third of a stop under most of the time and sometimes even more. I'm trying shooting at 0 under to see what goes. The other compensation is to use the brush tool to reduce noise where it is most visible and as suggested, no sharpening. That has had the most affect toward improving my opinion of the image quality. The one factor that was most influential for buying the camera was the increased detail. I rather blindly continued the same workflow as the XPro1 and it's apparent that was a mistake.

No - the DR expansion modes very much affect RAW files.

Firstly lets take a quick look at how digital ISO works

ISO above native (200 for the Fuji) works by UNDEREXPOSING the raw file, then applyng a global exposure push digitally, 1 stop underexposure and 1 stop push for iso 400 2 for 800 etc

This is basically the same as using the exposure slider in LR (or whatever)

When we import the raw in to our editing app, the ISO and corresponding exposure push are in the meta data of the raw, and the app acts on the info

The DR expansion modes work by under exposing the raw file 1 or 2 stops (as per ISO 400 or 800)

Then FOR THE SOCOC JPEG the shadows and midtones are lifted as the jpeg is created

This is basically the same as using the shadow slider in LR

But the RAW file doesn't get the global exposure meta data (you want to retain highlights, not apply a global positive exposure value, which is the antithesis of highlight preservation), so you end up with a RAW file that's 1 or 2 stops under exposed (depending on DR200 or 400 mode)

What's the easiest way to make a noisy file? Underexpose it, which is what the DR modes do

The largest amount of DR is available at base ISO

You'll get a cleaner image shooting @ iso200 and exposing for critical highlights and doing the shadow recovery yourself in post, than you will using DR400 and throwing away data before you even take the RAF from th SD card

 

[infrequent]

Basically if you shoot jpgs and prefer the SOOC route, you are better off using Auto DR. If you shoot RAW only, depending on your converter the metadata recalls the DR setting and either uses it or ignores it. As a sidenote, the DR setting will also override the Auto ISO setting or whatever maximum threshold you might have. For a scene that has a really high dynamic range, shooting RAW+Fine (hit Q or program a Fn button) and/or exposure bracketing is recommended.

Did I get the gist of it?

 

[lcpr]

This is NOT a critique of the previous X-Cameras, which I would expect to be more than good enough for most. Hell they're more than good enough for me 95% of the time! But the older cameras weren't quite enough for my needs under demanding conditions, like fashion week. They were 'close' but not quite there.

I agree, I've used the XP1 at London Fashion Week before (a mix of Front of House/Backstage work) and while I could get some great images out of it I always felt that I had to finesse things a little more compared to my 5DIII/D800E. Some images in the link below:

http://www.lloydramos.com/foh/

http://www.lloydramos.com/pringlescotland/

http://www.lloydramos.com/katie-eary-aw14/

 

[willie_901]

No - the DR expansion modes very much affect RAW files.

....

What's the easiest way to make a noisy file? Underexpose it, which is what the DR modes do

The largest amount of DR is available at base ISO

You'll get a cleaner image shooting @ iso200 and exposing for critical highlights and doing the shadow recovery yourself in post, than you will using DR400 and throwing away data before you even take the RAF from th SD card

Absolutely. The sole purpose of DR > 100 is to protect photographers from blowing highlights. FUJIFILM's product groups' embrace an in loco parentis philosophy.

 

[honozooloo]

I agree, I've used the XP1 at London Fashion Week before (a mix of Front of House/Backstage work) and while I could get some great images out of it I always felt that I had to finesse things a little more compared to my 5DIII/D800E. Some images in the link below:

http://www.lloydramos.com/foh/

http://www.lloydramos.com/pringlescotland/

http://www.lloydramos.com/katie-eary-aw14/

Killer work Lloyd! I agree - nothing against the X-Pro 1 as I LOVE mine and still use it...but it had its trade offs.