How I (Finally) Chose the Fuji X-E1
The snake which cannot cast its skin has to die. As well the minds which are prevented from changing their opinions; they cease to be mind. –Friedrich Nietzsche
|X-E1 + XF14mmF28 R|
ISO 200 14mm ƒ/8.0 1/640s
Now, if you made it this far, consider me impressed. I would like to reiterate here that I don’t consider this to be any kind of condemnation of any other camera system. You will spend a large amount of time if you were to seek out a camera system that can not produce excellent images when applied with the appropriate level of knowledge and skill with that system. This is merely the system that hits the most key points of value for me personally. I hope that Fuji continues to advance the XF system in the same direction. I look forward to hybrid AF in future XF bodies and more wonderful lenses such as the upcoming XF 55-200mm ƒ/3.5-4.8, the XF 23mm ƒ/1.4 and the XF 56mm ƒ/1.2(!) coming over the course of this year.
Incidentally, all of the images in this post were JPEGs produced by the camera. The upload to Blogger may have had some impact on their appearance, but, that is always the case.
Resolution of X-Trans Derived Images
So http://www.dpreview.com/ updated their lab samples for the Fuji X-trans camera based reviews. The interesting thing I found was their explanation about why you can’t compare images from these cameras directly with others:
Using our standard processing, with Adobe Camera Raw’s noise reduction set to a minimum, the X-E1’s images look rather different to those from Bayer sensor cameras, and specifically show much lower chroma noise. This means that direct comparisons have to be treated with caution; in effect the demosaicing required by the X-Trans CMOS sensor includes a degree of chroma NR.Which, of course, makes perfect sense. The “semi-random” pattern in the Fuji X-trans would imply a need to apply an algorithm similar to noise reduction (if not identical) to smooth out colors properly. But that means that in order to compensate for this semi-random pattern—the pattern that’s used to allow the elimination of the anti-aliasing filter—detail-crushing noise reduction algorithms are applied. The result based on my own A) browsing of many, many, online samples, B) my own experiments with this camera and C) the lab samples on DPReview is that the images out of these cameras are less sharp, to a noticeable degree, when RAW converted outside the camera. The camera’s JPEG engine apparently applies a large amount of sharpening to compensate (along with harsh color and contrasts but that’s appealing to some people, Canon does the same thing to a lesser degree).
Ultimately, what this means is simple: The X-trans sensor does not produce as sharp of images as normal Bayer array sensors even with (weak) AA filters in place. This isn’t that big of a deal, it just means your effective resolution after processing is lowered, that you will probably want to reduce the estimated acceptable output size (i.e. if you thought 16mp was good for a 16x20 print, you might want to drop to 12x18 for X Pro 1/X-E1 cameras). And, of course, it is ultimately not meaningful for web-sized images.
It does make me wonder why they did not choose a 24mp base sensor as this would definitely compensate for the loss of overall resolution, and as Sony has shown, produces less total noise when normalized to the same typical print/output sizes.
It also helps explain why I spent 4 hours on Flickr never finding any sharp full sized images from these cameras. They just can’t produce them. The sharpening artifacts expose the lack of actual sharpness and it’s not the lenses (as far as anyone can tell), it’s this larger sensor pattern which makes perfect sense once you think about it a little.
So there you go. Shoot with the X-Trans sensors, they do offer slightly increased sensitivities at the same noise levels and reduced moiré without an anti-aliasing filter, but remember to compensate for this demosaicing when deciding on final output or sharpening when pushing for larger prints than the actual resolution would support and you’ll get the best results. I would say it’s more appropriate to think of these cameras as more like 12-14mp (it will depend on the colors in the subject matter) than 16mp.
One last comparison I would now like to make, but, can’t as I no longer have an X-trans camera in hand is the black and white processing in-camera. I wonder if the camera takes advantage of the RAW data to demosaic the image with less loss of detail when shooting B&W. It’s possible, especially in the large blocks of green. There are many 2x2 blocks of all green pixels which is probably a big contributor to detail loss in color demosaicing. Anyway, if I ever have another one of these cameras to play with then I’ll give this comparison a try.
Incognito Flash Lamps
Faced with another birthday party at Chuck E Cheese, a place my daughter loves but low ISO’s do not, I decided to get creative. I shot this album of images with a set of three Yongnuo YN-560 and YN-560 II’s with a diffuser cap/”omni bounce” inside of small lampshades generally placed along the table (and that can be seen in a few of the shots) or moved as needed for certain phases of the party.
I wanted to go for a look like dinner around a table with candles/lamps sort of low key, pleasant ambiance kind of thing. I also wanted to shoot at lower ISOs to get some real nice images of my kid as keepsakes. I don’t know if something like this would be viable/useful outside of a family event where I don’t care that I can see the bottoms of the lamps are actually flash bodies, but, I had thought that perhaps I could push the idea further in the future and hide the flashes in something that actually looks like a lamp. This was a prototype test run where at least I cared to get some decent shots.
Anyway, the flashes were all set at 1/64th power, zoomed out to 24mm, diffuser down, gelled 1/4 CTO, with Meike omnibounce-alike cap inside simple small white lampshades that I bought at Wal-mart for $3.50 each. The bounce caps (for bare-bulb-like effect) cost $5 each so the modifier cost was very low per flash. Camera was a Nikon D800 + Sigma 35mm ƒ/1.4 set to 100 ISO most of the time, occasionally up to 200 ISO, 1/60-1/100s shutter, f/1.4-2 all varied depending on my whim. Just one image from this album is not lit with the lights, it’ll probably be obvious which one.
Keep in mind that this was merely my own kid’s B-day party and not some event I shot for pay. I figured it might be interesting to the strobists anyway due to the unusual approach I took to dealing with a low/mixed light setting where the ceilings were too high and unpredictably colored to bounce (and I don’t really like the eye-shadow effect of bounced flash anyway.)
Review: Sigma 30mm F2.8 EX DN (NEX)
Sigma 30mm F2.8 EX DN (for NEX)
Let’s jump to the conclusion first: This lens is a very good buy at $200. This lens can not replace the Sony E 24mm ƒ/1.8 ZA, but, comes very close to its performance in many ways. It even bests it in size and sharpness.
I was able to get out and shoot a number of comparisons between the two lenses. I was motivated to buy and compare the 30mm to the 24mm because of some lab results from lensrentals.com. Essentially, they found the Sigma to be very sharp, even wide open. But, let’s start there. “Wide open” is relative. For a prime lens, ƒ/2.8 is not particularly large of an aperture. It’s about a stop larger than the kit lens at the same focal length. That can be the difference between 1600 or 800 ISO. It can be meaningful. It won’t, however, give you eye-grabbingly shallow depth of field for normally distant subjects.
It’s very sharp, especially at ƒ/5.6. Both lenses peak at ƒ/5.6. The Sigma’s sharpness is dulled somewhat by less than excellent contrast. It’s not bad, by any means, but it’s no Zeiss. It also doesn’t handle bright light sources nearly as well as the Zeiss, likely due to differences in coatings. In the dark, lights such as headlamps on cars generate huge halos and they are quite unattractive. However, in more normal lighting scenarios, you won’t notice these issues. I noticed it mostly on buildings with the evening sun shining bright on the sides. The Zeiss held the details far better in those areas than the Sigma. But make no mistake, under slightly more controlled lighting, the Sigma is definitely as sharp or sharper than the Zeiss at normal to long distances. The Zeiss does better close up, which brings me to…
Macro. Reproduction ratio. The Zeiss has about double the reproduction ratio as the Sigma. There is absolutely no contest here. I can shove the Zeiss several inches away from the face of a subject and show you the bacteria crawling around in the oil in the pores of their skin. (Okay, maybe that’s a slight exaggeration but it’s amazing. The Zeiss is very well corrected for close focus.) Even though the Sigma is a longer focal length, I can not fill the frame with as small of things. In fact, the “macro” capability of the Sigma isn’t even worth mentioning. It has nothing even close to macro capability. The Zeiss, with its sharp close range correction coupled with the 24mp sensor of the NEX-7 means I can shoot very close up on a whim, crop it, and you wouldn’t even know I didn’t have a macro lens with me. Versatility is where the Zeiss comes out clearly ahead.
The Zeiss has better bokeh in most every situation. Not by a huge margin, most of the time. Sometimes the Sigma has this split bokeh appearance, which has some Japanese
curse name because it’s notably ugly. It doesn’t always happen. Bokeh discs from bright lights are much worse on the Sigma than the Zeiss. They look like super zoom bokeh discs on the Sigma. They look like prime lens discs on the Zeiss.
In fact, that is a good summation of the Sigma, it performs like a great ƒ/2.8 zoom except it doesn’t zoom.
The Sigma focuses a little more slowly. I suspect this is due to its odd design. It seems to use magnets for focus, or the gearing is extremely loose without power because the entire focusing assembly slides freely when the lens is off the camera or the camera is off. A magnetically controlled focus assembly would be a cool way to keep size down.
The Sigma is smaller, significantly. However, it’s no pancake. It sticks out beyond the hand grip on the NEX-7. It is not flared towards the end like some pictures make it look, it’s a straight tube whose diameter is about half a millimeter less than the widest part of the metal lens mount on the NEX-7. Oh and the Sigma has a metal mount, not cheap plastic like the low end Sony alpha lenses.
As with all native E mount lenses, it focuses by wire. This is fine, once you get used to it. The faster you turn the ring, the fast the focus changes with a curved acceleration rate, which is nice once you get the feel for the rate. Specifically with the Sigma, however, there is a small problem. This is where the cheapness shows most clearly. If you apply even a small amount of pressure to the ring, which you are lightly to do out of habit of holding your camera steady, you will deform the ring slightly into the lens body and “brake” the ring against the internals of the lens. When this happens, it is difficult or impossible to turn the ring. You must hold the ring lightly and then it turns easily. This is something you can deal with consciously, but, may bite you in a stressful shooting moment.
Sigma thought it necessary to provide a thickly padded zipper case for the lens. However, they thought it unnecessary to provide any hood of any kind. The lens element is almost flush with the lens front. It’s small and slightly bulbous exactly like the Sony E 16mm ƒ/2.8. Both lenses are quite vulnerable to a scrape on the lens. The Sigma, however, seems fairly resistant to flare. I had no issues with flare and could not force it to occur when I tried. Perhaps on a day with fewer clouds and earlier when the sun is brighter I could get some blobs. I wouldn’t worry about that. I would worry a little about protecting the front element. But here’s the rub, if you buy and attach a filter for protection, the filter will undoubtedly flare like the cheap garbage most filters are made out of. Since this lens is so cheap and you’ll probably shoot at ƒ/2.8 - ƒ/4 99% of the time, I wouldn’t put anything on it. Just accept you might get a scratch or two on the front of your ridiculously cheap lens. No big deal. Finally, unlike Tokina lenses, Sigma’s lens caps both front and rear work well and fit correctly.
As mentioned, the focus ring has a flaw if squeezed. So don’t squeeze it. The mounting ring is metal, not cheap plastic. The rest of the lens is plastic. It doesn’t look expensive but it doesn’t look bad. It looks better than Sony’s cheap alpha lenses. Focus is internal, nothing spins or moves externally when using the lens. There is some noise in the lens even in manual focus. I suspect this is the focus motor maintaining the position of the focus group and/or aperture. I also suspect this is done electromagnetically (which would explain the nature of the noise. It sounds like a tiny speaker picking up interference and a speaker is controlled by magnets, so….)
This is pretty simple. Do you need a sharp lens for NEX-7, but, you can’t afford the Sony E 24mm ƒ/1.8 ZA? Buy this. Do you really like the field of view this lens offers? Buy this. The field of view of the 24 vs. the 30 is significantly different. Remember, the relationship between focal length and field of view is non-linear. The shorter the focal length, the bigger the angle difference for every millimeter. The difference between 24 and 30 is approximately 10 degrees. Personally, I find the 24 to be far more versatile. The 30 is definitely not “portrait” length so standing slightly closer to the subject or cropping gives the same angle of view. The 24 can be safely used at ƒ/2.2 for great sharpness to compensate in DOF terms. The 30 does not offer much that the 24 does not offer and as such is not a replacement, for any price comparison. If you can swing the $1100 USD for the 24mm, buy that. The Sigma 30mm is not smaller enough to consider it as a compact alternative.
At the end of the day, the Sigma 30mm F2.8 EX DN is a great lens, irrespective of price. It is the third best lens for the NEX system right now, in my opinion (the best is the 24mm, the second is the 50mm ƒ/1.8 OSS). It is the sharpest, in controlled lighting situations, if only by a tiny margin. It is lighter and smaller than the other two. It is one of only three E mount lenses that won’t make your NEX-7 sensor look like a snob. But, be warned, it’s a Sigma lens and I got a good copy. My copy is pretty much perfectly centered. I won the lottery. Almost every Sigma product is riddled with quality control problems. If you buy this lens and it isn’t as good as I say, return it and get a different copy. Do this up to 5 times. Yes, seriously, 5 times. Sorry, that’s the price to pay to get the price you pay for this lens.
Eventually I will post some images I made with this lens. You can find many examples around the web for doing your own pixel peeping comparisons, but, suffice it to say, if you have not had 5000+ exposures experience with the E 24mm, it will be very hard to make a reasonable comparison.
But hey, it’s $200 USD. (Well, $220 lately) so you won’t be taking much of a risk in buying this lens.
Buy this lens if you don’t have or plan to buy the E 24mm ƒ/1.8 ZA.
Until recently, all of the cameras I have used that had eye level viewfinders had those viewfinders located on the optical axis. In other words, they were in line with the lens and generally in a middle area of the camera body.
Now, I have the NEX-7 which has the viewfinder in the location more commonly associated with rangefinder cameras, on the left upper corner. So, I have been experimenting with different methods of looking through the viewfinder. I’ve discovered something surprising.
When looking through the viewfinder on a regular body, whether using my left or right eye, my alternate eye is obstructed by the camera body. Until recently, I had never given it any thought. However, using the rangefinder style viewfinder led me to try using my right eye more often, which is not the eye I most commonly used with past viewfinders and it has led to an interesting discovery.
Apparently, until now, I have not been closing my alternate eye. In fact, doing so is so foreign that I am having difficulty keeping my left eye closed when using my right eye in the viewfinder. I guess I will have to exercise this combination of open and closed eyes. Unlike shooting in a real rangefinder, one can not keep both eyes open using a rangefinder-style electronic viewfinder. It is like using a telescope with only one of your eyes while keeping both open.
Of course, I can use my left eye in the viewfinder and the camera blocks my right eye, but, that just doesn’t seem right for this rangefinder style of viewfinder.
I found this discovery surprising and amusing. It’s fascinating how an action can become so habitual that you lose awareness of precisely how you perform it.
to fall forever, then
It bounces. Twice.
Into the wall. Failure.
Glass shatters, sprays.
and teeth fly
diminished and then
You see a path.
Such short sighted thoughts
that is what led you here.
To think this is the end.
To forget the very
essence of living.
Every end is a beginning.
ness is the canvas
Alone at Sunset
outside city limits
asphalt becomes gravel
traffic signals blinking
bovine tails are swinging
beyond humble homesteads
sunlight begins fading
silence overwhelms me
grasses swaying gently
eagles flying nearby
floating over rivers
spying tiny rodents
sparrows in the bushes
shaking barren branches
swiftly gathers supplies
sunset’s cold approaching
water bubbling softly
flowing under bridges
shedding icy coating
harbors unseen nature
distant skyline’s glowing
imparts peaceful solace
showers golden half-light
precedes violet twilight
and here I find myself
watching with true wonder
wishing I could loiter
a part of this forever
secret power sought by all
glowing softly in the hall
in abundance they will come
treasure hunters one by one
seeking searching for the cup
never knowing down or up
vexing pathways adumbrate
foolish burglar’s grisly fate
endless stairways turn and twist
razing savvy to persist
gasping, starving, thin and frail
common ending to the tale
secret power sought by all
tempting heedless to their fall
passive baleful artifact
kills without a single act