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 | Leica & Alternative Gear | Join Upload & Sell

1       2      
3
       4       end
  

Archive 2012 · X-Pro 1 tested by Pop Photo

  
 
aleksanderpolo
Offline
• • •
Upload & Sell: Off
p.3 #1 · p.3 #1 · X-Pro 1 tested by Pop Photo


Tariq Gibran wrote:
Sean Reid found that if you examine the X-Pro1 images in Lab mode, it's clear the A and B channels are being blurred. This results in the low noise but comes at the expense of color detail. It also results in so so looking B&W conversions (the loss of color detail kills B&W midtone richness). You can do something similar with any raw file in PP to reduce noise. The difference here is that Fuji does it in camera automatically (apparently).


Is it possible that the smoothing in A and B comes from adobe's interpretation of the data that also result in chroma smearing, rather than from Fuji's smoothing of raw? I have not tried to look at the files in Lab mode so I don't know the technical detail.



May 30, 2012 at 05:45 PM
Tariq Gibran
Offline
• • • • • •
Upload & Sell: Off
p.3 #2 · p.3 #2 · X-Pro 1 tested by Pop Photo


aleksanderpolo wrote:
Is it possible that the smoothing in A and B comes from adobe's interpretation of the data that also result in chroma smearing, rather than from Fuji's smoothing of raw? I have not tried to look at the files in Lab mode so I don't know the technical detail.


I don't think that's the case. The low noise at high ISO's is also there in the Fuji X-Pro1 jpegs. It must be taking place in camera - and on both jpegs and raws. For someone who shoot's high ISO and does not care to deal with PP, the in-camera noise processing appears to do a really great job. The negative of course is the loss of choice for the user to turn the Auto processing off.



May 30, 2012 at 05:55 PM
Tariq Gibran
Offline
• • • • • •
Upload & Sell: Off
p.3 #3 · p.3 #3 · X-Pro 1 tested by Pop Photo


Pixel Perfect wrote:
I've seen that phenomenon in plenty of POP photo tests. Not sure sure how they achieve those result either.


In this case, it's likely due to stronger in-camera noise processing kicking in.



May 30, 2012 at 05:58 PM
aleksanderpolo
Offline
• • •
Upload & Sell: Off
p.3 #4 · p.3 #4 · X-Pro 1 tested by Pop Photo


That would explain the loss of fabric texture in dpr and ir's test image

Um, I wonder if other camera makers will start follow suit and do noise reduction in A&B for the perceived advantage, people don't notice the loss of color detail/texture, but immediately see the noise performance. Hope not...



May 30, 2012 at 06:01 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.3 #5 · p.3 #5 · X-Pro 1 tested by Pop Photo


douglasf13 wrote:
Any idea what Fuji is doing exactly to achieve these results?


Well, it's fairly simple... It's not really Fujifilm "doing" anything, it's a byproduct of the type of interpolation you HAVE to use to get a usable image out of a X-Pro raw...

The images have low noise, due to two main reasons:
1) the "green" raw channel has more coverage but less positional symmetry than in a Bayer sensor - and it's fairly well balanced against CIE Y.
1a) as a biproduct, you get less high-frequency luma "miscalculations" in the interpolation, giving lower added raw converter noise

2) the fairly large B-B and R-R interdistances necessitates a fairly large chroma smoothing radius, unless you go to algorithms like POCS and such - which are very computationally heavy.

The camera doesn't "add" croma smoothing, it's there in the principle.

I would however think that the samples seen here exaggerates the bad parts of teh effect, much better results would be possible with an algorithm like POCS or anything else better adapted to uneven channel coverage.

The Fujifilm principle will in any case never get better than 50% of the Bayer chroma resolution, an inherent property of the scheme.
At higher ISOs this turns into an advantage in stead of a disadvantage.



May 30, 2012 at 06:09 PM
Tariq Gibran
Offline
• • • • • •
Upload & Sell: Off
p.3 #6 · p.3 #6 · X-Pro 1 tested by Pop Photo


theSuede wrote:
Well, it's fairly simple... It's not really Fujifilm "doing" anything, it's a byproduct of the type of interpolation you HAVE to use to get a usable image out of a X-Pro raw...

The images have low noise, due to two main reasons:
1) the "green" raw channel has more coverage but less positional symmetry than in a Bayer sensor - and it's fairly well balanced against CIE Y.
1a) as a biproduct, you get less high-frequency luma "miscalculations" in the interpolation, giving lower added raw converter noise

2) the fairly large B-B and R-R interdistances necessitates a fairly large chroma smoothing radius, unless
...Show more

So, you are saying the smoothing evident in the "A" and "B" high ISO Lab mode channels is inherently there due to the required interpolation for the CFA arrangement. No way to "turn it off" so to speak with a firmware update. Is this also causing the terrible water color artifacts with the subject matter I have shown at base ISO?




May 30, 2012 at 06:36 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.3 #7 · p.3 #7 · X-Pro 1 tested by Pop Photo


Yes, barring a more advanced interpolation algorithm. Since the layout is inherently asymmetrical, you would ideally have a multi-pass type of interpolation, with best-fit iterations. This is very resource heavy.


May 30, 2012 at 06:57 PM
rscheffler
Online
• • • • • •
Upload & Sell: Off
p.3 #8 · p.3 #8 · X-Pro 1 tested by Pop Photo


http://chromasoft.blogspot.ca/2012/05/lightroom-41-and-fuji-x-pro1-oh-dear.html


May 30, 2012 at 07:26 PM
kosmoskatten
Offline
• • • •
Upload & Sell: Off
p.3 #9 · p.3 #9 · X-Pro 1 tested by Pop Photo


And part three:
http://chromasoft.blogspot.ca/2012/05/demosaicing-fuji-x-pro1-part-3.html




May 31, 2012 at 02:21 AM
Osprey01
Offline
• •
Upload & Sell: Off
p.3 #10 · p.3 #10 · X-Pro 1 tested by Pop Photo


I believe it's the manufacturer's noise suppression algorithm kicking in.

Pixel Perfect wrote:
I've seen that phenomenon in plenty of POP photo tests. Not sure sure how they achieve those result either.




May 31, 2012 at 04:36 AM
Tariq Gibran
Offline
• • • • • •
Upload & Sell: Off
p.3 #11 · p.3 #11 · X-Pro 1 tested by Pop Photo


rscheffler wrote:
http://chromasoft.blogspot.ca/2012/05/lightroom-41-and-fuji-x-pro1-oh-dear.html


While the issue with foliage (watercolor effect) seems to be exaggerated in Adobe's raw converter, the problem is certainly still there with the Fuji/ Silkypix converter. In fact, the Adobe converter seems to give better green color differentiation and a bit more sharpness overall which may be why the effect shows up more.

For low ISO landscape, I think this quote from the above pretty much sums things up:

"Turns out that X-Trans sensor really wasn't a good idea."



May 31, 2012 at 07:19 AM
theophilus
Offline
• • •
Upload & Sell: Off
p.3 #12 · p.3 #12 · X-Pro 1 tested by Pop Photo


Tariq Gibran wrote:
While the issue with foliage (watercolor effect) seems to be exaggerated in Adobe's raw converter, the problem is certainly still there with the Fuji/ Silkypix converter. In fact, the Adobe converter seems to give better green color differentiation and a bit more sharpness overall which may be why the effect shows up more.

For low ISO landscape, I think this quote from the above pretty much sums things up:

"Turns out that X-Trans sensor really wasn't a good idea."



I observed the same watercolor effect this weekend on my X100 with a green wheat field.



May 31, 2012 at 09:39 AM
Tariq Gibran
Offline
• • • • • •
Upload & Sell: Off
p.3 #13 · p.3 #13 · X-Pro 1 tested by Pop Photo


theophilus wrote:
I observed the same watercolor effect this weekend on my X100 with a green wheat field.


I noticed it also with certain X100 files, which sort of debunks the idea that it's caused by the unique CFA arrangement of the X-Pro1 (since the X100 sensor uses the standard RGB arrangement). As mentioned by Henrik/kosmoskatten in the past, this effect shows up with other recent Fuji cameras so I do believe it has to have something to do with a specific choice Fuji has made with their processing in general and is not just X-Pro1 specific (Carsten mentioned the effect showing up in the files of the older Kodak SLR and I also remember seeing this watercolor effect in the files from the Mamiya ZD MF camera when it was first released as well).



May 31, 2012 at 10:06 AM
edge100
Offline
• •
[X]
p.3 #14 · p.3 #14 · X-Pro 1 tested by Pop Photo


Ok, instead of eating lunch at work, I did this little test.

Took a single Raw image (of my son Elliot) and did the default conversion in Silky 3.2.9.1 (the update from the other day), as well as RPP Beta 1521 (using VNG interpolation). I exported both as 16-bit TIFFs, imported to LR4.1, and then compared them to the same Raw processed in LR4.1 directly. Exported with standard screen sharpening from LR4.1

Here are the three @ 640px:

1. LR4.1 direct
http://farm8.staticflickr.com/7086/7309421162_afbbf61f04_z.jpg

2. Silky 3.2.9.1
http://farm9.staticflickr.com/8148/7309421098_6ef8f090c8_z.jpg

3. RPP Beta 1521
http://farm8.staticflickr.com/7241/7309420990_7ab80be85f_z.jpg

And the 100% crops of the same (pardon the very slightly different crops on each) :

1. LR4.1 direct
http://farm8.staticflickr.com/7221/7309454906_8d757db186_z.jpg

2. Silky 3.2.9.1
http://farm9.staticflickr.com/8162/7309454596_2d29cd94bf_z.jpg

3. RPP Beta 1521
http://farm8.staticflickr.com/7220/7309454836_e1b86b24e2_z.jpg

Aesthetically, I prefer the RPP file BY FAR over the other two. It looks like film to me. In terms of pixel-level detail, it's also the best of the three (though Silky isn't bad). But what really blows my mind are the colours; yes, they are all very slightly different with respect to white balance and overall colour balance. But take a look at the box of raisins in Elliot's hand; the LR and Silky versions (particularly the LR version) are almost pink, whereas the RPP version is essentially the "correct" colour. Also, check out the pink splotchiness on the top left of the LR crop (in Elliot's hair). I assure you, that's an artifact of LR's processing (and I've seen it in other Raws, too).

If I had to rank these, I'd say RPP > Silky >>LR.

C'mon, Adobe! This camera is capable of so much.



May 31, 2012 at 01:46 PM
douglasf13
Offline
• • • • •
Upload & Sell: Off
p.3 #15 · p.3 #15 · X-Pro 1 tested by Pop Photo


To be fair, I personally find RPP to be better than LR with just about any camera, although I still use LR for ease of workflow.


May 31, 2012 at 02:01 PM
alba63
Offline
• • •
Upload & Sell: Off
p.3 #16 · p.3 #16 · X-Pro 1 tested by Pop Photo


douglasf13 wrote:
To be fair, I personally find RPP to be better than LR with just about any camera, although I still use LR for ease of workflow.


Ditto! In order to use RPP, I have to transfer the images to an external HD, switch on my Mac mini (which I do 1-2x per week) and start RPP. For many (not all) cameras this little (not very beautiful and intuitive) program delivers the best, most film- like, most natural looking files. A shame to the big boys like Adobe who do not manage to replicate this level of quality.

But, as you, I often use LR for convenience.

Bernie



May 31, 2012 at 02:05 PM
edge100
Offline
• •
[X]
p.3 #17 · p.3 #17 · X-Pro 1 tested by Pop Photo


I'm going through some other Raws now and processing through RPP, and WOW! what a difference. That film-like look I've been trying to get from LR (with limited success) is coming through nicely from RPP.

It's a headache to go RAW -> RPP -> TIFF -> LR, but it's worth it if this is the result.



May 31, 2012 at 02:39 PM
boinkphoto
Offline

Upload & Sell: Off
p.3 #18 · p.3 #18 · X-Pro 1 tested by Pop Photo


Tariq Gibran wrote:
"Turns out that X-Trans sensor really wasn't a good idea."


I guess I disagree. I'm not saying it doesn't have flaws, but I bought it because I loved the output and owning it I continue to love the output.

Maybe I'm wrong, but I don't see sensor output as being generically interchangeable. I guess what I'm trying to say is if you have two cameras from two manufacturers that essentially equal in their specs and lenses and you take an image from one and compare it to the same image from another, one may asthetically appeal to me and the other may not. While people often say with PP they're all the same, I don't see that - NEX output isn't necessarily going to be able to look like Fuji output (nor for that matter do I want to adjust every photo so it does but...).

To me discussions of the merits of cameras tend to focus too closely on pixel peeping and rarely do people say, "I choose camera X because I like the look of the output better." To me that's the ultimate metric. Not paying attention to that seems to me to be like saying Kodak EBX and Fuji Velvia are "the same" - which clearly they are not (granted, with RAW and PP, there is more latitude - but not infinite).

That said, I'm not trying to imply that Fuji output is better - not at all. It just appeals to me more generally speaking than other brands I've looked at. Someone else might find NEX or G1 or whatever output appeals more to them, but my ultimate point is, the general "look" of an image shouldn't be dismissed.



May 31, 2012 at 03:02 PM
boinkphoto
Offline

Upload & Sell: Off
p.3 #19 · p.3 #19 · X-Pro 1 tested by Pop Photo


BTW - I should note that I understand some things may be deal breakers. Certainly if the "painterly" effect here interferes with producing fine art prints, and that's what you need the camera for, obviously it's a "no go".


May 31, 2012 at 03:07 PM
Tariq Gibran
Offline
• • • • • •
Upload & Sell: Off
p.3 #20 · p.3 #20 · X-Pro 1 tested by Pop Photo


boinkphoto wrote:
I guess I disagree. I'm not saying it doesn't have flaws, but I bought it because I loved the output and owning it I continue to love the output.

Maybe I'm wrong, but I don't see sensor output as being generically interchangeable. I guess what I'm trying to say is if you have two cameras from two manufacturers that essentially equal in their specs and lenses and you take an image from one and compare it to the same image from another, one may asthetically appeal to me and the other may not. While people often say with PP they're all
...Show more

Well, digital is quite different then film with regard to a specific "look". It's certainly more flexible and malleable, which is obvious from the different "looks" specific raw processors give out of the box just as mentioned above. A lot of this is simply down to different custom color profiles and curves. If we leave the lens out of the equation (which certainly does impart a unique fingerprint/ look), the fact is that these looks from different cameras can be emulated quite easily by those who know what they are doing. That's not to say there is not a benefit from getting exactly what you like ooc without PP and profiling. That's what Fuji is counting on in fact - attracting users who prefer the look they offer with minimal fuss.



May 31, 2012 at 04:44 PM
1       2      
3
       4       end




FM Forums | Leica & Alternative Gear | Join Upload & Sell

1       2      
3
       4       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.