Fuji dpreview: X20?

[Armanius]

That's a good point. I can't think of a DSLR with a histogram on its OVF either. Must be because I haven't touched a DSLR for years. I guess we all got spoiled with amazing OVF in the XPro1/X100/X100S.

 

[flysurfer]

That's a good point. I can't think of a DSLR with a histogram on its OVF either. Must be because I haven't touched a DSLR for years. I guess we all got spoiled with amazing OVF in the XPro1/X100/X100S.

In order to provide this feature, the X20 would need a larger and more complex hybrid viewfinder than the one in the X-Pro1, as it would have to be able to zoom. The X-Pro1 dual view (aka 2 magnification levels, no zoom) hybrid viewfinder is already so big that Fuji had to omit the diopter adjustment control, relying on screw-on diopters instead. This hybrid viewfinder is also the single most expensive part of the X-Pro1, it's easily responsible for several 100 dollars of its purchase price (that's why the X-E1 is smaller and can be sold so much cheaper).

A zooming hybrid viewfinder in a X20 would result in a camera that was bigger and even more expensive than an X-Pro1, even if it kept the small 2/3" sensor and compact lens. Obviously, nobody would buy such a ridiculous "compact" camera, they would laugh at Fuji for such foolishness.

So honestly, I'm at a loss about what's going on at DPR.

 

[carlb]

The one thing I miss in the X20 optical viewfinder is the electronic level like on the X-E1. I wonder if a clever liquid level could be incorporated into a future "X30" optical viewfinder ...

 

[otalp]

It is always a good idea, not to take reviews literally. No matter what subject you are dealing with. To me, DPR seems to be more on the reliable side, since they explain to some degree what they are doing. The review will not compensate for personal emotional preference, which of course is a significant aspect in a creative photographic process.

But, to be honest, from the sample images I have seen so far there I cannot tell a significant step in IQ due to the X-Trans sensor. Hopefully, that will change with the next firmware. Until then, I will have an eye on the decreasing prices of the X10 and have fun with my ancient G6. To me, the EXR concept is definitely more appealing than phase detect and enhanced OVF.

I am aware that I might be completely wrong.

 

[Armanius]

Looking at the RAW files, there's more detail on the X20 than X10. At least to my eyes. The surprising thing to me is that the X20 lags behind the likes of the XZ2, G15 and LX7 in spite of the larger sensor and no AA filter.

 

[flysurfer]

The surprising thing to me is that the X20 lags behind the likes of the XZ2, G15 and LX7 in spite of the larger sensor and no AA filter.

That would indeed be surprising, so I downloaded the ISO 100 studio sample RAWs of the 4 cameras you mention, normalized exposure and white balance in Aperture and set Apple Camera to maximize detail in each of the 4 shots. I have uploaded all 4 results in full size to my Flickr account, so everybody can have a look. It should be easy to identify the X20 version, as it supposed to lag behind quite visibly.

Here's camera 1:

Here's camera 2:

Here's camera 3:

Here's camera 4:
View attachment 12793

Of course, some people may think that I did a better job processing the X20 file than the other cameras' RAWs, because I'm either

• not trustworthy
• incompetent
• a fanboy
• paid by Fuji to spread lies
• any combination of the above

The good thing with this comparison is that it wouldn't matter, because everyone can prove my results wrong or otherwise inadequate any time by downloading the RAW files at DPR and processing them with Aperture, or any other RAW conversion software and conversion parameters of their choice.

By the way, since all shots were taken at f/4.5 and ISO 100, comparing shutter speeds and image brightness of the unadjusted RAWs is telling an interesting story about "inflated ISO", an no, it's not the X20 this time.

Btw, here's the same comparison with Lightroom 4.4:

Camera 1:

Camera 2:

Camera 3:

Camera 4:
View attachment 12797

 

[Armanius]

Of course, some people may think that I did a better job processing the X20 file than the other cameras' RAWs, because I'm either

• not trustworthy
• incompetent
• a fanboy
• paid by Fuji to spread lies
• any combination of the above

You forgot the special secret Fuji software that allows you to turn a 2/3" sensor images into full frame sensor images.

Can I get a copy of that software please? :biggrin_old:

 

[alessandro]

That would indeed be surprising, so I downloaded the ISO 100 studio sample RAWs of the 4 cameras you mention, normalized exposure and white balance in Aperture and set Apple Camera to maximize detail in each of the 4 shots

Aperture is clearly doing a better job than ACR/LR with the X20 file. The "winner" is the G15 to my eyes, but not by far. Only the LX7 "lags behind".

 

[flysurfer]

Can I get a copy of that software please? :biggrin_old:

Not yet, still working on the upgrade to turn it into a PhaseOne.

 

[flysurfer]

Aperture is clearly doing a better job than ACR/LR with the X20 file. The "winner" is the G15 to my eyes, but not by far. Only the LX7 "lags behind".

Yep, the G15 is looking very good, depending on what aspect of the image you look at. For example, the G15 shows plenty of color moiré in the head left of the little color balls.So do some of the other cameras, but not the X20.

 

[alessandro]

Yep, the G15 is looking very good, depending on what aspect of the image you look at. For example, the G15 shows plenty of color moiré in the head left of the little color balls.So do some of the other cameras, but not the X20.

Yep. I was looking at other parts of the image. Indeed it depends on where we're pixel peeping and what we're looking for. Strange: there are different parts where the "maze" pattern is more obvious for different cameras. Anyway, we agree that the level of detail is high enough. I think that the most important differences among (these) similarly targeted cameras don't belong to the sensors.

 

[hellwill]

Feeling foolish here, but I have no idea how to tell which camera goes with which test image. (I'm still in the market for a camera, sort of lusting after the x20 but need strong evidence before I commit the $$--and, apparently, the time to learn to deal with RAW files.) For what it's worth, I think I like the images from camera #3 best, though #4 competes closely on different criteria.

 

[flysurfer]

Feeling foolish here, but I have no idea how to tell which camera goes with which test image. For what it's worth, I think I like the images from camera #3 best, though #4 competes closely on different criteria.

Camera 3 is the X20, formerly known as "the camera that lags behind all others".

Camera 4 is the Canon G15, which seems to be a pretty close match, at least with this focal length for studio scenes at ISO 100 and if you don't mind the color moiré. By the way, camera 4 needed 1/60s to produce an ISO 100 image that is just as bright as the image from the X20 at 1/80s, so it looks like "Canon is inflating ISO".

However, Canon's ISO inflation is not even half as severe as in the LX7 (aka camera 1), which not only used 1/40s for the same scene but still needed a boost of about +0.4 EV to match the normalized brightness (no correction needed with the X20 RAW). So Panasonic appears to inflate ISO by a whopping 1.4 EV compared to the X20 (at least at ISO 100 with this scene and under this light). I am sure that seriouscompacts and other forums are filled with threads complaining about this, right?

The G15 is using the same 4:3 aspect ratio and 4000x3000 resolution as the X20. It's interesting that the Canon and Olympus are showing moiré despite their (reportedly both very weak) AA filters. Of course, a different scene could have produced a different result. There are situations where the X-Trans sensor will show moiré, too.

The Olympus (camera 2) overexposed the scene at 1/60s and had to be brought down about -1/2 EV to match the other samples. However, this is a problem of the test setup, as all shots were performed in manual exposure mode. Interestingly and unlike the 3 other cameras, the LX7 was not shot manually according to the EXIF data.

 

[hellwill]

This is starting to look like strong evidence. My wife will not be happy.

 

[Armanius]

This is starting to look like strong evidence. My wife will not be happy.

Not if you say, honey, I have a gift for you!