Popular Photography's take on X100s

[Ybor]

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

 

[cmc1]

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

 

[samhain]

Interesting...

 

[wchutt]

How odd.

The DPREVIEW data show the X100 raw S/N is lower than the raw S/N for both the X-Pro 1 and the XE 1.

For Popular Photography be correct, the 16 MP X100S Xtrans sensor would have significantly inferior performance to other two cameras using 16 MP XTrans sensors.

If the X100S color filter array lenses are still customized to match the angle of light exiing the lens, the sensor CFAs would not be identical for all the XTrans cameras. If anything, the more complicated X100S design would decrease vignetting at the frame edges which should increase the overall system S/N (less vignetting means more light and more light means mores signal). Another explanation would be the X100S electronics are inferior to the other three X APS-C cameras'. Why would Fuji do that? What would motivate them to use inferior electronics in the X100S?

Right now I don't believe the Popular Phoography results. They don't make sense.

 

[Jeff Charles]

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

Take a look at the DPR raw samples from the X100s and X100 at ISO 6400. The X100s image is definitely cleaner. The X100s photo did get about 0.6 stops more exposure*, but it's also a little brighter, so that more or less evens things out. Also, the X100s has more pixels, which helps reduce the visibility of noise when its images are printed or displayed.

* We don't know if the light levels in the DPR studio were the same for both sessions, which could explain the different exposures.

 

[Liviu Namolovan]

Are you sure?!

 

[chainmale]

Yes, according to Techradar as well, the X100 has better signal to noise ratios than the x100s at all sensitivity settings in RAW but not in JPEG. See it here -

http://www.techradar.com/us/reviews...as/fuji-x100s-1122175/review/5#articleContent

 

[John Carson]

Jeff Charles wrote:

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

Take a look at the DPR raw samples from the X100s and X100 at ISO 6400. The X100s image is definitely cleaner. The X100s photo did get about 0.6 stops more exposure*, but it's also a little brighter, so that more or less evens things out.

It more or less evens things out if you are testing for comparability of ISO ratings. It doesn't do anything to even things out if you are comparing noise levels. More light means less noise.

 

[pachura]

X100 has 12 MPix, X100s has 16.3. Of course it will have more noise. Plus, X100s has these on-sensor PDAF-pixels that could further degrade image quality...

 

[ljclark]

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw.

My comparison -- my X100 and X100s cameras, same place, same time: X100s has less noise at higher ISOs (images you can save as RAW -- up to 6400) than the X100.

Granted, that's a sampling of one (1) -- one of each body. You can see Liam's RAW conversions of two of my shots at the Fuji X Forum (you may need to log on to see the photos):

http://www.fujix-forum.com/index.ph...0s-x-trans-ii-sensor-great-or-horrible/page-6

 

[Jeff Charles]

John Carson wrote:

Jeff Charles wrote:

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

Take a look at the DPR raw samples from the X100s and X100 at ISO 6400. The X100s image is definitely cleaner. The X100s photo did get about 0.6 stops more exposure*, but it's also a little brighter, so that more or less evens things out.

It more or less evens things out if you are testing for comparability of ISO ratings. It doesn't do anything to even things out if you are comparing noise levels. More light means less noise.

You are right. I had it wrong. I agree that amount of light reaching the sensor, not ISO, is what determines noise.

It's a problem with DPR ISO testing (and Imaging Resource's too) that they adjust exposure to match output brightness, which tends to favor cameras with "overrated" ISOs.

 

[Jeff Charles]

ljclark wrote:

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw.

My comparison -- my X100 and X100s cameras, same place, same time: X100s has less noise at higher ISOs (images you can save as RAW -- up to 6400) than the X100.

Granted, that's a sampling of one (1) -- one of each body. You can see Liam's RAW conversions of two of my shots at the Fuji X Forum (you may need to log on to see the photos):

http://www.fujix-forum.com/index.ph...0s-x-trans-ii-sensor-great-or-horrible/page-6

Your samples that Liam converted do show a clear advantage for the X100s. The X100s image shows less noise, and in the X100 image, more detail is obscured by noise. Your X100s shot did get 1/3 stop more exposure, but I have a hard time believing that small difference would produce such a big difference in noise between the two photos. Liam did resize the images to the same output size, which gives an advantage to the X100s, but that's a real and beneficial consequence of the X100s's higher pixel count.

Why did both Techradar and Pop Photo find that X100s raws have more noise that X100 raws? Who knows? I am inclined not to worry about it.

 

[Ybor]

Jeff Charles wrote:

ljclark wrote:

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw.

My comparison -- my X100 and X100s cameras, same place, same time: X100s has less noise at higher ISOs (images you can save as RAW -- up to 6400) than the X100.

Granted, that's a sampling of one (1) -- one of each body. You can see Liam's RAW conversions of two of my shots at the Fuji X Forum (you may need to log on to see the photos):

http://www.fujix-forum.com/index.ph...0s-x-trans-ii-sensor-great-or-horrible/page-6

Your samples that Liam converted do show a clear advantage for the X100s. The X100s image shows less noise, and in the X100 image, more detail is obscured by noise. Your X100s shot did get 1/3 stop more exposure, but I have a hard time believing that small difference would produce such a big difference in noise between the two photos. Liam did resize the images to the same output size, which gives an advantage to the X100s, but that's a real and beneficial consequence of the X100s's higher pixel count.

Why did both Techradar and Pop Photo find that X100s raws have more noise that X100 raws? Who knows? I am inclined not to worry about it.

--
Jeff
"Please proceed..."

Well, I know that applying noise reduction eats away the resolution advantage of the 100s over the 100 (to some extent).

I should add that Popular Photography test results have flustered me in the past when they stood in contrast to some other reviews by generally respected sources. Then again, so has DPR.

Bottom line for me; noise is noise and the latest drive for pixel totals comes at a cost given present technologies.

--
Al

 

[Ybor]

cmc1 wrote:

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

 

[Sebit]

No need to worry. X100s is much better than X100 in real life (shooting RAW).

Starting from ISO 1600 the difference is very big. X100s holds colour and contrast much better, there is no banding from point lights, and all the way up to ISO 6400 there is no need for luminance NR in my opinion. I'm setting the chroma noise slider in LR from 2-3 (for ISO 1600) to 6 (ISO 6400). Default value is 25, but that's much to high.

If you would like to see my high ISO samples: http://www.dpreview.com/forums/post/51190251

 

[Jeff Charles]

Ybor wrote:

Bottom line for me; noise is noise and the latest drive for pixel totals comes at a cost given present technologies.

DxOMark says that Contrary to conventional wisdom, higher resolution actually compensates for noise

 

[headofdestiny]

Jeff Charles wrote:

Ybor wrote:

Bottom line for me; noise is noise and the latest drive for pixel totals comes at a cost given present technologies.

DxOMark says that Contrary to conventional wisdom, higher resolution actually compensates for noise

 

[Davidgilmour]

Ybor wrote:

I was disappointed as to the test outcome for the X100s in the noise department...that is, according to Popular Photography. More noise than X100 at every standard ISO in Raw. What surprised me most is its noise results looking far more like an m43 sensor (and not as good as the newest generation m43). The good news was 200-300 more lines of resolution than the predecessor. Maybe this will help offset some of the noise.

Hey, don't kill me, I have one on order, at least I still do as of the moment.

 

[tinpusher]

Davidgilmour wrote

Doesn't surprise me. x-100 colors and IQ is amazing. People who trade in their 100 for 100s are not better off...

I bought the 100s this afternoon to go with my 100.

I'm a cynic but even I'm impressed with the clean sharp images and speed of this new camera.

I'm shooting at ISO 6400 in dim lighting with all the colour of the X100 and very little noise.

The upgrade cost is high and not necessary but I can see what I'm getting here unlike the X10/X20 upgrade.

 

[mr moonlight]

One of the advantages of the X-trans sensor is the quality of noise. The noise is almost grain-like and practically free of chroma. I've also found that it retains colors better at higher ISO's. Weather it has more noise or not, the overall image quality is better. With my X100 I try to keep my ISO lower than 2000 if possible. With my Xpro1 I don't mind going all the way to 6400.