Lightroom colours tested vs in-camera

[aceflibble]

I typically test my new cameras and lenses with a simple product set-up and a ColorChecker Passport just to make sure everything is working properly, but this time around with the X-T10 I decided to also directly test the Fuji in-camera film simulations against Lightroom's rendering.

First, some notes on equipment and settings:

• Lights used were two InterFit daylight balanced continuous lights, which I frequently use for product photography. They're nearly perfectly balanced. (Tested with several shades of grey card, they consistently give a white balance of 5,540k needing a tint of just +2 towards magenta.)
• I tested the X-T10 with the 14mm, 35mm, 56mm and 60mm lenses, however I am only posting the results from the 56mm as I found this to be the most neutral. The JPG files of the 14mm and 35mm are given a lot more 'optimisation' by the camera and the 60mm has a slightly warm colour cast to it. The 56mm lens does still have some optimisation applied to it, though.
• Camera settings for all shots were ISO 200, 1/400th and f/4. Other settings were RAW+F, white balance set to Daylight, sharpness -1, noise reduction -2. Everything else left at default/0.
• Lightroom left everything at default/as shot/0, only reducing the default sharpening of 25 down to 15 and changing the colour profile to the appropriate film simulation.

Before you look at the images and try to guess which is which, I want to mention Fuji's white balance. Their 'daylight' white balance is, in fact, not a daylight white balance. A true daylight white balance should be 5,500k with no tint. (Some people argue daylight should be 5,600k.) Fuji's daylight balance, however, reads at 5,200k with a +7 tint towards magenta. This is clearly a cooler balance than the industry standard used by everybody else. I double-checked with my X100S and that gave the same 5,200k +7 result.
Not only that, but an actual perfect white balance of these Fuji  files comes out as 5,350k +28 (Lightroom auto) to 5,150k +35 (manually selecting a perfect balance). I double-checked this in SilkyPix and got the exact same results. I then tested again with a Canon camera and that read 5,570k and +5 in-file and 5,620k +6 corrected, which is much closer to what any daylight balance should be reading as.
This suggests that Fuji are running their colour much greener than it should be, either by having green read too strongly or magenta too weakly.
I will note that when I first bought a Fuji camera, the X100S, right away I felt the colour balances were all running a bit cool and I tweaked them all to run +1 towards yellow and red/magenta. Testing the X-T10 now, it seems I have been right to do this.
Of course, some people like Fuji film specifically because it has a cooler tone to it than the neutral-warm colours of Kodak, so this white balance bias may have been completely intentional by Fuji in order to replicate the cooler bias of their film stock. Even so, I think they've gone maybe a little too far with the green tint.


So, on to the actual images.

In order, we have Provia/Standard, Velvia/Vivid, Astia/Soft, Classic Chrome, Pro Negative High and Pro Negative Standard. I will not yet tell you whether the in-camera rendering or the Lightroom rendering is first or second in each pairing. That's our test, let's see if people can actually pick out which renderings they think are Fuij's or Adobe's. All I will say is that there are no 'trick' pairings, i.e. I didn't change the colour in any way, there are no mismatched pairs, etc. Each pair does contain one Fuji rendering and one Adobe rendering of the same film simulation.


 

And now the Monochromes. In order we have Monochrome, Monochrome + Yellow filter, Monochrome + Red filter and Monochrome + Green filter. No, I did not test the Sepia tone, because you have to be out of your mind to use Sepia.



 

My own observation is that the aqua, blue, purple and magenta colours barely change at all between Lightroom and in-camera, red only changes a noticable amount in one film simulation and yellow changes noticeably in two. The biggest offender is green, which is never matched well other than in Classic Chrome. The Classic Chrome simulation definitely is the one that Adobe and Fuji are the most closely-matched.
I also think it's interesting that mono and mono+Y are hardly different at all, with just a slight darkening of blue in one of the renderings of mono+Y, and that the green patch, which is the biggest problem for the colour renderings, is almost perfectly equal in all the monochrome renderings.
 
So, guess away. In each pairing, which do you think is Fuji and which do you think is Adobe? Make sure your monitor is calibrated properly! I'll give the answers in a few days.

[stillman3]

Thanks for doing that comparison, rather interesting.

>I'll give the answers in a few days.

There is an old joke that says

q) How do you keep an idiot in suspense?

A) I will tell you later.

Well this idiot is in suspense!!

I use Astia mostly so immediately looked there. Intrigued by the green difference. WHy do you think this occurs?

When you talk about the 35mm lens, you mention that the incamera rendering is optimised. How does this affect the results?

Cheers Eric

[aceflibble]

Well, since you're the only person to respond in a week  , I think I'll just say: all of the top lines are Fuji colours, all of the bottom ones in each pair are Adobe. Actually, the misalignment caused by the Fuji lens optimisation gives this away!

I think the heavy change in green is linked to the difference in white balance between Fuji and industry standards. Fuji film was known for having a cooler colour balance, so they set the white balance to run slightly blue. This probably makes green tones look unnaturally dull, so when the camera processes the .jpg files they have it increase the saturation and brightness a little and for some film simulations it also shifts green tones slightly back towards yellow. Conversely, Fuji doesn't seem to apply this shift to the yellow tones, which are simply left duller/bluer in most film simulations, or it could be that the corrections they apply to green bleed over into the yellow; either way, Fuji seems to be sacrificing yellow for the sake of green. (Classic Chrome being the exception, which has everything shifted slightly warm.)
Adobe Lightroom, and other programs, don't know the intentions behind Fuji's design decisions, so they don't over-correct the green. All the software knows is the colour balance is a little cooler than it perhaps should be, so greens are represented 'accurately' with a slightly drab and slightly bluer tone. (Velvia being the exception, which does have much more yellow in the greens; Adobe probably guessed that this would be used a lot for landscapes so they optimised the green tone themselves.) 

In other words, Fuji's style is partly done at capture (white balance) and partly after (processing the .jpg), whereas Lightroom is just aiming for an accurate representation of the capture (white balance).

For the lenses, Fuji applies profiles to even the raw files to fix colour fringing, distortion and colour casts. You can turn it off in most Fuji cameras, but once you've taken a picture with it on, it can't be removed from the file, even if you shoot raw. Adobe Lightroom warns you of this when you look at a Fuji image in the develop module. Every Fuji lens has some 'optimisation' applied to it, but some more than others. The 56mm has fairly little, just a tiny bit of distortion and fringing correction. The wider you go, the more heavily these things apply. The 35mm without optimisation has quite a lot of colour fringing, which in some cases can interfere with colour rendition (e.g. you photograph something purple and the camera reduces purple fringing, making the edges of your subject look drab) and wider than f/5.6 it shows some colour bleed, which is also corrected by Fuji's lens optimisation. The 60mm lens has less distortion and aberration fixes applied, but it has a warmer colour cast to it which is half-fixed by the software and makes it inappropriate for testing colour charts. The 14mm simply has so much distortion correction applied to it that everything is smeared. I've not yet tested the other Fuji lenses to see exactly what optimisation Fuji is doing.

[stillman3]

Thanks for the explanation. Boy, there is certainly a lot going on behind the scene with processing, that I was not aware of!

[flysurfer]

There is no digital distortion correction at all in Fujinon XF 14, 16, 23, 56 and 90mm primes.

[aceflibble]

This very test shows otherwise.   You can see each top row (all Fuji .jpg, optimisation on) is shifted compared to the bottom (all raw, optimisation off), with the shift originating from the centre. That's what I meant when I said the optimisation misalignment gives away which row of colour patches is from which source. The colour patches towards the left (red, orange) move more towards the left and the patches on the right side (purple, magenta) have shifted over more towards the right. The green patch is the only one truly centered in the original frame and it's aligned properly to the unaltered row. That is the result of correction of pincushion distortion; the image has been expanded outwards from the centre. In fact, if you look very closely, you can actually see on the raw rows (bottom of each pair) that the patches at each side are, very slightly, leaning inwards. (I.e. pincushion distortion.)

In other words, those very images explicitly show the distortion correction that the Fuji lens optimisation applies to the 56mm lens. If it didn't apply any correction for pincushion distortion the patches would align exactly with the raw files, or, even if I were to nudge the camera between pairs of shots, one row would be uniformly misaligned, not misaligned one way on one side of the image and in the other direction on the opposite side.

Following that, the 14mm and 35mm lenses both show barrel distortion when optimisation is turned off and focused at a medium or near distance. The 35mm's is fairly mild, the 14mm's is very heavy. I've not adequately tested the 23mm to tell and I do not yet have a 16mm, but I would assume those also show barrel distortion, given their focal lengths. I also do not have the 90mm yet—and do not intend to ever buy it—but given the 56mm displays some pincushioning, it should be safe to assume that the 90mm does as well. I don't own any of the zooms anymore, either, but I do remember the 18-55 showing both barrel and pincushion distortion when shot with optimisation off, at the wide and long ends respectively, and at the relevant focus distances.
Bear in mind that, to an APS-size surface receiving an image focused at this distance (not perfectly, but as close to 4' 6" as I could humanly manage), the only focal length which would be able to produce an image truly without any distortion whatsoever would be a 51.4mm lens, which does not exist in the Fuji line other than with zoom lenses. Though zoom lenses, at any focal length and focus distance, rarely exhibit a properly flat image anyway due to the nature of their construction.

So, there you have it. The 56mm lens does display pincushion distortion at this focus distance (and one would assume, further) and Fuji's lens optimisation is correcting for that. If the former wasn't true then Fuji would be breaking the laws of physics and if the latter was wrong then those pictures would not be misaligned as they are.

 

I will add that I was surprised to see pincushion distortion in these raw files, since I usually leave lens optimisation on and in any case, my natural assumption without doing the maths would be that a 56mm lens would show barrel distortion, not pincushion, but there you are. I tend to think of pincushion distortion only showing up >160mm at this sort of range, but that kind of thinking is just down to which lenses you have the most experience with, I suppose. The 56mm is not that far off the tipping point into barrel distortion, anyway. Shorten it by 5mm and it would be there. In fact, thinking about it now, it's actually surprising that the un-optimised lens has as little distortion as it does, and for a 'portrait' lens like this, pincushion at this distance is far more preferable than barrel!

[spivey]

"... Fuji's white balance. Their 'daylight' white balance is, in fact, not a daylight white balance. A true daylight white balance should be 5,500k with no tint. (Some people argue daylight should be 5,600k.) Fuji's daylight balance, however, reads at 5,200k with a +7 tint towards magenta. This is clearly a cooler balance than the industry standard used by everybody else. --- This suggests that Fuji are running their colour much greener than it should be"

 

Could you explain what this means for practical photography? If I want neutral white balance, should I bring a gray card with me everywhere and shoot manual white balance only? What about Fuji's auto white balance - does it also give greener colours?

[flysurfer]

This very test shows otherwise.   You can see each top row (all Fuji .jpg, optimisation on) is shifted compared to the bottom (all raw, optimisation off), with the shift originating from the centre. That's what I meant when I said the optimisation misalignment gives away which row of colour patches is from which source. The colour patches towards the left (red, orange) move more towards the left and the patches on the right side (purple, magenta) have shifted over more towards the right. The green patch is the only one truly centered in the original frame and it's aligned properly to the unaltered row. That is the result of correction of pincushion distortion; the image has been expanded outwards from the centre. In fact, if you look very closely, you can actually see on the raw rows (bottom of each pair) that the patches at each side are, very slightly, leaning inwards. (I.e. pincushion distortion.)

 

In other words, those very images explicitly show the distortion correction that the Fuji lens optimisation applies to the 56mm lens. If it didn't apply any correction for pincushion distortion the patches would align exactly with the raw files, or, even if I were to nudge the camera between pairs of shots, one row would be uniformly misaligned, not misaligned one way on one side of the image and in the other direction on the opposite side.

 

Following that, the 14mm and 35mm lenses both show barrel distortion when optimisation is turned off and focused at a medium or near distance. The 35mm's is fairly mild, the 14mm's is very heavy. I've not adequately tested the 23mm to tell and I do not yet have a 16mm, but I would assume those also show barrel distortion, given their focal lengths. I also do not have the 90mm yet—and do not intend to ever buy it—but given the 56mm displays some pincushioning, it should be safe to assume that the 90mm does as well. I don't own any of the zooms anymore, either, but I do remember the 18-55 showing both barrel and pincushion distortion when shot with optimisation off, at the wide and long ends respectively, and at the relevant focus distances.

Bear in mind that, to an APS-size surface receiving an image focused at this distance (not perfectly, but as close to 4' 6" as I could humanly manage), the only focal length which would be able to produce an image truly without any distortion whatsoever would be a 51.4mm lens, which does not exist in the Fuji line other than with zoom lenses. Though zoom lenses, at any focal length and focus distance, rarely exhibit a properly flat image anyway due to the nature of their construction.

 

So, there you have it. The 56mm lens does display pincushion distortion at this focus distance (and one would assume, further) and Fuji's lens optimisation is correcting for that. If the former wasn't true then Fuji would be breaking the laws of physics and if the latter was wrong then those pictures would not be misaligned as they are.

 

 

I will add that I was surprised to see pincushion distortion in these raw files, since I usually leave lens optimisation on and in any case, my natural assumption without doing the maths would be that a 56mm lens would show barrel distortion, not pincushion, but there you are. I tend to think of pincushion distortion only showing up >160mm at this sort of range, but that kind of thinking is just down to which lenses you have the most experience with, I suppose. The 56mm is not that far off the tipping point into barrel distortion, anyway. Shorten it by 5mm and it would be there. In fact, thinking about it now, it's actually surprising that the un-optimised lens has as little distortion as it does, and for a 'portrait' lens like this, pincushion at this distance is far more preferable than barrel!

 

Nope. Let me repeat it for everbody: There's not digital distortion correction in the lenses I mentioned. Easy to see in a RAW converter that offers to turn on/of RAW metadata based lens corrections.

 

LMO hasn't anything to do with distortion, anyway. LMO deals with corner sharpness and diffraction blur (and also a few other things, Fuji intends to broaden the LMO playing field).

[flysurfer]

Btw, X-T10 has a known firmware bug regarding image crop with certain lenses that will be fixed in a July update. I've already mentioned this in the German forum. 

 

That said, it's only supposed to affect lenses with digital distortion correction. 

 

(2) Angle difference between X-T10 and X-T1

The angle of jpeg image and RAW image converted in camera are different from ones taken by

X-T1 with certain lenses, which are XC16-50mm, XF18-55mm, XF10-24mm, XF16-55mm,

XF18-135mm, XF18mm and XF27mm.

 

 

In the end, the RAW converter makers decide which 16 MP of the 16.3 MP to use. Some even use all 16.3 MP, others output even less than 16 MP.

[aceflibble]

Okay, let me break this down for you because you are, categorically, wrong.

1. You can see in the very images in my original post that the raw files display pincushion distortion and the in-camera .jpg files display mirrored stretching originating from the centre.
2. The only things changed between shots was turning the lens optimiser on and off for .jpg and raw respectively, and changing film simulation every other shot. Focus distance was not changed, no additional processing or develop settings were changed. Lens optimisation on (jpg) and off (raw) only.
3. As I said in the original post, I did run the images through other software in order to double-check colour accuracy and the same discrepancy was shown in all.

To help you, here are three examples of the same test shot which explicitly illustrate the pincushion distortion of the raw file and the corrected Fuji .jpg.

Two full images, .jpg with lens optimisation on and then the raw with lens optimisation off:



 
Now the .jpg overlayed on top of the raw image, showing the difference in size withint he frame as caused by pincushion distortion correction pushing the image outwards originating from the centre:



 
And third, close crops of the top line of colour patches with a grid overlayed so you can see exactly how straight the jpg is compared to the inwards bow of the raw:



 
To deny that the lens optimisation is correcting for pincushion distortion would be to say either: Fuji's .jpg processing just naturally stretches everything out from the centre regardless of whether lens optimisation is on or off and by pure luck the 56mm has exactly the right amount of pincushion distortion to perfectly equal out the Fuji stretching; Fuji have broken the laws of physics and have created a 56mm lens which can generate an image with absolutely no degree of distortion at a focus distance of 4' 6" and Lightroom and Silkypix are so utterly confounded by this display of literal God-like dimensional manipulation that their software adds in distortion to the raw files just so those pesky Fuji files can adhere to the laws of our mortal realm; or you're wrong and just don't like to admit that you were mistaken about something related to Fuji.

It is factually not possible for a lens of this focal length and focused at that distance to not display some distortion, and the raw files with lens optimisation turned off show this, yet the Fuji .jpgs, with the lens optimisation turned on, do indeed show no distortion. More importantly than that, I have now posted a total of five images of the same scene taken with the same framing at the same distance with the same lens and the same focus which explicitly display the difference in distortion between the raw file with lens optimisation off and the jpg file with lens optimisation on. I will also note that I get the same results with my X-T1 as I do with the X-T10, so no, it's not a bug with the X-T10, or, if it is, then it's a bug in the X-T1 as well.

Your information is incorrect and you are mistaken. I have now proven this twice over with five illustrations. It's okay, you can just relay the correct information in the future. We all get things wrong sometimes. If you still believe that I am mistaken and my examples are inaccurate, please explain why, because I sincerely would like to hear your reasoning.
 

[aceflibble]

Could you explain what this means for practical photography? If I want neutral white balance, should I bring a gray card with me everywhere and shoot manual white balance only? What about Fuji's auto white balance - does it also give greener colours?

If you manually set a white balance by selecting a Kelvin colour temperature and use a grey card or other neutral reference then your white balance should come up perfectly neutral. If you select a white balance using the presets, such as Fine or Incandescent, when you check it with your grey reference you will find there is a slight blue and green tint. (The exception is the Underwater preset, which tints towards magenta.)

Auto white balance is tricky because even when photographing the same subject under the same light, it might change white balance between shots. It's hard to test consistently, but in my experience the Fuji auto white balance does tend to run a little cool for my taste, so I set the AWB to be +1 -1 (one point towards red and one towards yellow) to compensate.

 

Of course if you shoot raw then you can simply correct the white balance in whatever processing software you use. If you shoot jpg and have been happy with the colours you get until now, I wouldn't worry about it; cooler tones have always been Fuji's style. If you shoot jpg and really want to be sure you're getting a truly neutral white balance then yes, set your white balance manually using the Kelvin option and use a grey card or other grey reference whenever you are unsure of what Kelvin temperature to choose.

[flysurfer]

Okay, let me break this down for you because you are, categorically, wrong.

1. You can see in the very images in my original post that the raw files display pincushion distortion and the in-camera .jpg files display mirrored stretching originating from the centre.
2. The only things changed between shots was turning the lens optimiser on and off for .jpg and raw respectively, and changing film simulation every other shot. Focus distance was not changed, no additional processing or develop settings were changed. Lens optimisation on (jpg) and off (raw) only.
3. As I said in the original post, I did run the images through other software in order to double-check colour accuracy and the same discrepancy was shown in all.

To help you, here are three examples of the same test shot which explicitly illustrate the pincushion distortion of the raw file and the corrected Fuji .jpg.

 

Two full images, .jpg with lens optimisation on and then the raw with lens optimisation off:

 

 

 

Now the .jpg overlayed on top of the raw image, showing the difference in size withint he frame as caused by pincushion distortion correction pushing the image outwards originating from the centre:

 

 

 

And third, close crops of the top line of colour patches with a grid overlayed so you can see exactly how straight the jpg is compared to the inwards bow of the raw:

 

 

 

To deny that the lens optimisation is correcting for pincushion distortion would be to say either: Fuji's .jpg processing just naturally stretches everything out from the centre regardless of whether lens optimisation is on or off and by pure luck the 56mm has exactly the right amount of pincushion distortion to perfectly equal out the Fuji stretching; Fuji have broken the laws of physics and have created a 56mm lens which can generate an image with absolutely no degree of distortion at a focus distance of 4' 6" and Lightroom and Silkypix are so utterly confounded by this display of literal God-like dimensional manipulation that their software adds in distortion to the raw files just so those pesky Fuji files can adhere to the laws of our mortal realm; or you're wrong and just don't like to admit that you were mistaken about something related to Fuji.

 

It is factually not possible for a lens of this focal length and focused at that distance to not display some distortion, and the raw files with lens optimisation turned off show this, yet the Fuji .jpgs, with the lens optimisation turned on, do indeed show no distortion. More importantly than that, I have now posted a total of five images of the same scene taken with the same framing at the same distance with the same lens and the same focus which explicitly display the difference in distortion between the raw file with lens optimisation off and the jpg file with lens optimisation on. I will also note that I get the same results with my X-T1 as I do with the X-T10, so no, it's not a bug with the X-T10, or, if it is, then it's a bug in the X-T1 as well.

 

Your information is incorrect and you are mistaken. I have now proven this twice over with five illustrations. It's okay, you can just relay the correct information in the future. We all get things wrong sometimes. If you still believe that I am mistaken and my examples are inaccurate, please explain why, because I sincerely would like to hear your reasoning.

 

 

It seems that you have little understanding of what is actually going on. That's okay, but it's not helpful to distribute false information, because others users may take it seriously. I recommend that you read my book on the X-T1 (an X-T10 book will be out later this year), as it contains plenty of information with regards to RAW converters, lens correction metadata, the LMO and other topics. It may be helpful, although I'm not so sure in your particular case.

 

Obviously, all this is quite off topic here, as this thread is about film simulation color accuracy in Lightroom vs. the built-in RAW converter. I'd recommend trying the film presets in RFC EX 2 and especially Iridient Developer 3 (X-Trans APS-C sensor profiles v 3.3). IMHO, Brian's profiles are the most accurate (especially when used with the DR function), but I'm a bit prejudiced because I played a minor role in making them. There are also user-created Fuji film simulations for Capture One Pro. Sadly, Adobe's "Fuji-sanctioned" profiles do not cut it, but this is also due to Adobe's particular way of applying adaptive tone mapping.

[aceflibble]

Okay, 1) you don't need to quote a whole massive post, photos included, to write a reply that short, and 2) nothing you have said offers any explanation as to why my examples are wrong.

I'm not going to explain it again because I've already repeated myself. I gave examples and explained what was going on with the files. You said I was wrong without explanation as to how. I gave more examples, fully explaining the process behind taking the pictures, where the changes occur and why and how what you are claiming is not physically possible other than via the one differing factor, namely the lens optimisation feature. I have explained why the little you have put forward so far can not be factually accurate in any dimension known to the human race. You have, again, not actually explained why you think differently. You can't just say "you are wrong because you are wrong (by the way please buy my book)." You have to actually explain why you think you're right. I've explained why you're wrong. I've explained the science and I've explained the simple reality. I've provided multiple examples which explicitly show you are wrong.

Now, you can learn from this and correct your own information for the future or you can remain wilfully ignorant or you can prove me wrong, I don't mind which, but you need to do one of the three. You're not right just because you say you are right. You're right when you prove you are right, which I have done. If I am wrong, please show why and I will gladly amend my methods to reflect my mistakes.

[mjh]

And third, close crops of the top line of colour patches with a grid overlayed so you can see exactly how straight the jpg is compared to the inwards bow of the raw:

 

Are you sure that’s the image you wanted to show? Because, quite frankly, I don’t see the ‘bow’ you are referring to. And I’ve really tried, carefully scrutinising the image at 400%. There is a very slight (sub-pixel) inclination in both rows, that’s all.

[mjh]

Before you look at the images and try to guess which is which, I want to mention Fuji's white balance. Their 'daylight' white balance is, in fact, not a daylight white balance. A true daylight white balance should be 5,500k with no tint. (Some people argue daylight should be 5,600k.) Fuji's daylight balance, however, reads at 5,200k with a +7 tint towards magenta. This is clearly a cooler balance than the industry standard used by everybody else. I double-checked with my X100S and that gave the same 5,200k +7 result.

Not only that, but an actual perfect white balance of these Fuji  files comes out as 5,350k +28 (Lightroom auto) to 5,150k +35 (manually selecting a perfect balance). I double-checked this in SilkyPix and got the exact same results. I then tested again with a Canon camera and that read 5,570k and +5 in-file and 5,620k +6 corrected, which is much closer to what any daylight balance should be reading as.

This suggests that Fuji are running their colour much greener than it should be, either by having green read too strongly or magenta too weakly.

How would you know that Fuji’s daylight white balance was 5200 K? It is not like there was a colour temperature embedded in the raw file. A raw converter takes the white point data as specified in the meta data and applies some algorithm to translate that to a colour temperature value and a green/magenta bias, but as the algorithms differ, so will the values displayed by different raw converters.

 

Having said that, it is true that Fuji cameras will produce Fuji colours – they always have, just like Canon or Olympus cameras yield Canon or Olympus colours, respectively. Fuji has always tried to have their digital cameras mimick the characteristics of their film emulsions.

[aceflibble]

Are you sure that’s the image you wanted to show? Because, quite frankly, I don’t see the ‘bow’ you are referring to. And I’ve really tried, carefully scrutinising the image at 400%. There is a very slight (sub-pixel) inclination in both rows, that’s all.

Look at the grid line running along the top of the colour patches. On the Fuji file, the grid and patches line up perfectly. On the un-optimised raw, there is a solid black line—a gap between the patches and the grid—which is larger above the middle patches than the red and mageneta ont he edges. Your standard pincushion distortion.

 

 

 

How would you know that Fuji’s daylight white balance was 5200 K? It is not like there was a colour temperature embedded in the raw file. A raw converter takes the white point data as specified in the meta data and applies some algorithm to translate that to a colour temperature value and a green/magenta bias, but as the algorithms differ, so will the values displayed by different raw converters.

Which is exactly how I know Fuji are running their white balance cool. Every converter reads the Fuji white balance as 5200K. Yes, every converter is different, yet despite them all handling files slightly differently they all still agree that the Fuji daylight balance is running cold at 5200K. Additionally, you can even double-check this in-camera: take a picture with the white balance set to Fine/Daylight and then take another with the white balance set to 5200K. They come out exactly the same.

 

Which, yes, as I said, is likely intentional to copy the cooler colour tones of Fuji film, with Classic Chrome being warmed up to mimic Kodak's tones.

[nzswedespeed]

Wow thanks for the fantastic comparison! I'll have to have a closer look tomorrow. Cheers!

[jlmphotos]

Thank you for the comparison.  Good to know.  I usually am in Astia, and I shoot both JPEG and Raw.  I have my LR import preset set to Astia and quite honestly the jpeg + the raw look the same to me. 

Just an observation...

 

Thanks again.

[aceflibble]

Astia is very close. I think that and Classic Chrome are the two Adobe has matched best. It's only really pure green that the Adobe Astia and Fuji's Astia differ on, and it's pretty rare to find actual pure green in the world, so that's a simulation where you probably really are never going to tell the difference.

[Shadowside]

It seems that you have little understanding of what is actually going on. That's okay, but it's not helpful to distribute false information, because others users may take it seriously. I recommend that you read my book on the X-T1 (an X-T10 book will be out later this year), as it contains plenty of information with regards to RAW converters, lens correction metadata, the LMO and other topics. It may be helpful, although I'm not so sure in your particular case.

 

Obviously, all this is quite off topic here, as this thread is about film simulation color accuracy in Lightroom vs. the built-in RAW converter. I'd recommend trying the film presets in RFC EX 2 and especially Iridient Developer 3 (X-Trans APS-C sensor profiles v 3.3). IMHO, Brian's profiles are the most accurate (especially when used with the DR function), but I'm a bit prejudiced because I played a minor role in making them. There are also user-created Fuji film simulations for Capture One Pro. Sadly, Adobe's "Fuji-sanctioned" profiles do not cut it, but this is also due to Adobe's particular way of applying adaptive tone mapping.

This i s deeply offensive.... Ace has gone to great lengths to not only perform an objective test but, in true scientific form, has shared his results, methodologies and rationale as well. You (Surfer) however have countered with nothing short of curt posts with nothing to back your claims other than "You Should Read My Book" which is not even available yet! One does not require an entire book to counter Ace's data!

 

Opinions can either become fact, through scientific challenge, or be disproven by the same process. You points so far are nothing more than a marketing ploy and have no place in scientific discourse regardless of the subject matter. It's too bad really, as it may very well be that your book will be a worthy purchase, but I for one will not be buying it.

 

It is not constructive to be rude, condescending, or dismissive in order to illustrate a counter point or logical argument.... just be fair and back your claims in the same manner as your colleague. I possess a degree in sensitometry, and so far I aggree with Ace. This is not to say he is correct, just that he is the only one that has produced quantifiable data.

[Dave Rosser]

I came upon this thread while investigating my impression that my X-Pro1 gives cold green biased results in Capture One, Lightroom and Fast Raw Viewer when processing RAW files where the camera was set to auto white balance.  I take it others have seen this behaviour and it is just not me.  At the moment I am working on 600 odd RAW files taken with my X-Pro 1 set to auto white balance and every one has need a positive color temperature and tint correction to achieve correct color balance.  I do not see this problem nearly as severely with Nikon NEF and Sony ARW files.  Can anyone advise me if there is some way of applying something like a +300K color temperature +10 magenta tint correction to the auto color balance?

[om.ch]

I came upon this thread while investigating my impression that my X-Pro1 gives cold green biased results in Capture One, Lightroom and Fast Raw Viewer when processing RAW files where the camera was set to auto white balance.  I take it others have seen this behaviour and it is just not me.  At the moment I am working on 600 odd RAW files taken with my X-Pro 1 set to auto white balance and every one has need a positive color temperature and tint correction to achieve correct color balance.  I do not see this problem nearly as severely with Nikon NEF and Sony ARW files.  Can anyone advise me if there is some way of applying something like a +300K color temperature +10 magenta tint correction to the auto color balance?

 

 

In the menus, once you've set your WB you have the possibility of doing a WB shift, it's not number precise but I've managed to use it to correct my X-Pro1 and 2 colour bias.

[flysurfer]

Lightroom and in-camera WB and colors should always look the same as long as the same film simulations are applied. Of course, it's easy to adjust WB and colors "to taste" in Lightroom with the two sliders for color temperature and tint.

[Dave Rosser]

In the menus, once you've set your WB you have the possibility of doing a WB shift, it's not number precise but I've managed to use it to correct my X-Pro1 and 2 colour bias.

Thanks a lot.  I'd missed that possibility entirely - I should read the manual in more fully .  I will go away and experiment.

[Dave Rosser]

 Of course, it's easy to adjust WB and colors "to taste" in Lightroom with the two sliders for color temperature and tint.

I'm doing that but it does get a bit wearing after the 500th picture