Fuji XF50mm f1 (with Fuji X-T3) Review

Discussion in 'Photography and Cameras' started by purr1n, Jun 19, 2022.

  1. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Do any of you guys remember the phil.askey site before it became DPreview.com? And when Ken Rockwell actually wrote good informative reviews, going quite a bit into the nuances of operation? Sure Ken Rockwell was 100% Team Nikon back then, but that's what made him great. He offered a reference point where people could understand where he was coming from, even if you used Canon. Between those two sites, readers could get a pretty good picture of camera gear. Askey's site would do all the pixel peeping work. Ken could just call it as it is. Sadly those days are over. They are still good resources, but they have lost a bit of what made them special. Ken reviews everything and anything that comes out as soon as possible, sort of rapidly going through the same old motions. DPreview is mechanical and soulless. Things are made even worse because every single YouTube review of camera gear sounds like an extended informercial advertisement. Basically camera gear reviews in 2022 are like HF and ASR for audio. This is why I'm going to take a stab at it.

    I've always wanted a big wide aperture portrait lens. A few years back, I got to try out the Fuji XF56 f1.2 (full frame 85mm equivalent), but really wasn't happy enough to buy it. Well, I was happy with it's size, but it never got the background right - the bokeh was unimpressive. The XF56 wide open is f1.2, the depth-of-field is really more like that of an f1.8 or f2.0 lens because of the APS-C size sensor. But maybe it wasn't necessarily the bokeh. I rarely shoot totally wide open with super fast lenses (will say a bit more about this later). The XF56 just looked too perfect. Too sharp. The photos I took with it looked too antiseptic for what is supposed to be a portrait lens. (I blame the high capabilities of today's digital and the Internet's tendency to go "NWAVGUY/ASR/DXO" for this trend toward moar sharper, moar details, moar megapixels.) What I wanted was the Fuji APS-C equivalent of the Canon 85mm f1.2 for full frame. This Canon ain't exactly the sharpest lens in the world. Many have criticized it for this, but the photos that it takes are stunningly beautiful. These kinds of lenses are supposed to evoke emotion. I guess pick your poison: 8K Gonzo Porn or Bob Guccione / Penthouse.

    You are so full of shit! You are just a weekend photo hack! You don't know what you are talking about!

    Could be. But why then did Fuji later release the XF56 f1.2 APD (built-in apodization filter). And then the XF50 f1? I opted out of the APD after looking at the photos online. I didn't think there was a big enough difference from the regular XF56. I also got scared off from the APD lens because the filter turns off one of two modes of autofocus on the Fuji cameras. It's not like Fuji cameras can afford to lose any more capability with their autofocus, which still seems to be five years behind Sony and Canon. Finally, the APD lens also cuts down on the light. The APD just didn't seem right, considering that the price shot up to $1500, the same as the XF50mm f1. The 50mm APS-C comes out to 75mm FF equivalent. I actually prefer this to 85mm - a little less telephoto giving me more breathing room.

    The XF50mm f1 is actually interesting. It's not as sharp as most other lenses wide open. Most definitely not sharp, even if we compare it to even the typical F2.8 to f4-5.6 zoom lens. (This is a testament to the sharpness of today's lenses). However, we are comparing f1 to f2.8! The XF50mm f1 isn't something we can shoot wide open and get sharp photos. We have to work at it. We can forget about sharp photos or even in-focus photos if either the subject or photographer are even moving slightly. Let's get to autofocus first.

    I picked up a used X-T3 body for cheap. The X-T4 offered in-body stabilization which I really don't care about because I prefer to use faster lenses at sufficient shutter speeds rather than slow lenses at 1/10 sec shutter with stabilization. LOL. The X-T3 is also a smaller body where I can fit in my 18mm semi-pancake lens for a somewhat smaller package. The X-T4 isn't that much larger, but millimeters count. Trust me. It's also my understanding that the X-T3 autofocus is performance the same as the X-T4 autofocus after the firmware upgrade. I'd say that with most lenses, the X-T3 autofocus is about on par with the Sony A6400, a camera that came out four years ago. I know, it's kind of pathetic. Yeah, at least it the autofocus doesn't suck balls.

    The problem I'm running into is that the XF50 f1 is this huge ass lens that needs a lot of turns on the focus ring to get to where it needs to go. Focus speed with the X-T3 seems long and arduous. Coupled with wide aperture (super shallow depth-of-field) and coarse focus points (instead of Sony and Canons one-billion focus points), it's tough to nail focus in many situations. It's not the end of the world, but one needs to know the camera and know the lens, to "work it", to get the desired results. Think of it like The Mandalorian trying to wield the Darksaber. OK, it's not that bad, oh but maybe it is. It's not automagic. I'm not convinced this would be a good wedding lens yet, at least if you shot documentary style, because of these issues.

    For manual focus people, the focus ring is not mechanically coupled. It's a fly-by-wire system so that there's lag when we turn the focus ring (I can see @rhythmdevils shaking his head in amusement). Man, I miss the days where mid-end camera bodies like the Nikon F100 had these strong motors that could just grind the fuck outta the lenses and nail that focus. One of the lenses I loved with the Nikon was this 80-200mm f2.8 that depended upon the motor of the body. It was a bit cheaper, but with an F100, no problem with AF speed. (Thanks to Ken Rockwell for the recommendation and saving me some bucks). While we are talking about fly-by-wire, the same holds true for the aperture ring. It would normally be fine if the aperture ring had stronger detents, but it's loose. Almost every time I handle the camera with this lens on, I accidently move the aperture ring, stopping down a few notches. (I still use the aperture ring with Fujis). It's just a big honkin' lens and how the hand must support the lens from underneath, where the fingers end up, that causes these accidental ring slip issues.

    PXL_20220619_153217150.PORTRAIT.jpg

    With respect to handling, it's all sorts of fucked up. It's an acceptable compromise if I am going to bitch that the Canon EOS Rs are too bulky. Maybe the X-T4 with its more pronounced grip on the body would do better. I doubt it. The lens and body doesn't feel and balanced like a big Canon body with the 85mm f1.2. On the other hand, the XF50 f1 is still lighter than the Canon lens by 150grams, which is not insignificant. It's more comfortable holding modern assault rifles fully loaded (which are up to 7-8 lbs). Modern tools of death are surprisingly ergonomic. Go figure heh. The lens take a 77mm filter instead of something like a 62mm. The good thing is that I have 77mm filters around that I used with my old telephotos.

    Before I forget, the minimum focusing distance is 27-28". This has caused a few issues with not being able to obtain focus when trying to get close up, say to the cats.

    Given all these issues, why would I even consider this lens? Well, it's the magical creamy creme bokeh. One or two photos just don't do it justice. I think one really needs to handle this camera, look at what's out there, take the shot, and be amazed.
    DSCF0111 (Custom).JPG
    f1.3. 1/2200 sec. ISO160.

    I intended to take the above shot a f1.0, but accidently moved the aperture ring to f1.3 (again, handling issues). We can see the shallow depth of field since only one eye the the bridge of the nose is in good focus. Also note how the photo isn't tack sharp. I like the fact that it's not super sharp. In fact, photos taken at f2.0 still aren't sharp. It's at around f3.0 where things get sharp.

    Here's another one that really shows us the downsides of always shooting wide open at f1. I think it's important that camera gear reviewers post their fuck-ups so that people understand better the limitations. But they never seem to. I got one eye in focus when I wanted both. Maybe if I selected the focus point on the bridge of the nose I would have gotten the results that I wanted. Note only ISO1600 and 1/75 shutter, This taken at night with limited indoor light. This is what fast lenses get us. This is how good today's sensors are. High-speed black and white ISO1600 from the early 90s would look grainy as heck.
    DSCF0113 (Custom).JPG
    F1.0 1/75 sec. ISO1600

    This is a total fuck up. The focus point on was the cat's eye but actual focus fell to the neck. I blame the lens and Fuji's AF for this. I don't know why the camera system just screwed this one up. I did everything right. I'd say that maybe 33% of shots ended up like this. The X-T3 and XF50mm f1 would just go total fail. Note that these screw ups only seem to happen with the XF50 f1, and not any of my other lenses.
    DSCF0115 (Custom).JPG

    This one below came out OK. There's actually a lot of shit behind the cat on the floor. Soldering iron, xover parts, etc. Bless my wife and how she puts up with me. There's also the window. Bokeh to the rescue! Magical bokeh.
    DSCF0120 (Custom).JPG
    f1.0 1/75 ISO800

    The next post will demonstrate the true super power of the XF50mm f1. Hold on a bit.
     
    • Like Like x 8
    • Epic Epic x 1
    • List
    Last edited: Jun 19, 2022
  2. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The true super power of this lens is how it continues to pull of magical bokeh at f1.8 and 2.0! There's no way that I could have taken these shots at this distance wide open at f1.0. It would have been one petal or one ather in focus and everything else out of focus!

    Note that I used the Fuji Velvia emulation. No, I don't think it looks like Velvia slide positives. Well, maybe because Velvia is a slide film and slides just have a sense of slide film awesomeness when shown from a slide projector (or even one of those "Fisher Price" slide viewers)? Haha. Most of you guys are too young. Anyway, Velvia 50 was the high resolution film of that time. Modern digital kills it with respect to resolution, even though I took the shots in MED resolution and then scaled them down for web.

    Anyway, this is f1.8/2.0 and I really like the bokeh. We also also see that it's still not what I would call tack sharp. But I think this is the point. SINAD and MTF weirdos can go straight to hell.

    F1.8 1/1480 ISO80
    DSCF0143 (Custom).JPG
    f2 1/450 ISO80
    DSCF0146 (Custom).JPG

    Heh, I think I finally got why they are called dragonfruit. The flowers bloom at night, and then wither within hours.
     
    Last edited: Jun 19, 2022
  3. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Here's another screw up I blame on the X-T3 / XF50mm f1. I haven't figured it out yet, when it happens, why it happens. It just screws up suddenly no rhyme no reason. Still all a mystery. Still trying to wield this darksaber. It's upsetting because I hate chimping. But yet at the same time, I don't trust the lens/body, so I have to chimp. Grrrr.

    Out of focus - wut? Typical Fuji :p
    DSCF0096 (Large).JPG

    This came out OK. I cropped and then resized to web. ARCOS BW simulation.
    DSCF0095 (Large).JPG
     
    Last edited: Jun 19, 2022
  4. roshambo123

    roshambo123 Friend

    Friend
    Joined:
    May 26, 2018
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Los Angeles
    I've always thought apodization filters looked like ass. The bokeh they produce looks like fake photoshop gaussian blur.

    Velvia is also my favorite film for landscape and dissecting it has been a bit of a side hobby. I agree that Fuji's emulations do not look like Velvia 50. It's also worth pointing out that in-camera JPEG/TIF processed colors can be different than RAWs baked with Fuji's Velvia profile in ACR. I think in-camera is closer. But anyway,

    How to Velvia:

    Beastly contrast ratio
    Velvia is the densest film that's been widely available and has a DMAX of 3.8 which equates roughly a 10,000 : 1 contrast ratio. Outside of an Apple XDR and some of the newer QD-OLED's, you're not likely to run into many displays able to reproduce the kind of blacks found in a Velvia slide on a slide table. In general, most monitors are probably not better than 1:1000.

    Tone
    Velvia has an exposure dynamic range of about 7 stops which is challenging for newbies to learn how to meter but it greatly contributes to the look. While a negative film takes a huge 14 stop exposure range and rams it into a much less dense, lower contrast film, Velvia does the opposite, taking a much smaller range and stretching it across a very dense film. The end result is exaggeration, high contrast, and a harder tone.

    Velvia doesn't do medium orange
    Skin tones were always a disaster and sunsets always awesome for a reason. Digital almost always accurately renders orange clouds orange, while Velvia has a tendency push medium orange values into red, while slipping brighter orange and bright yellows into greens, plus compensating a bit by adding warmth to all highlights. As a result you get a much wilder sunset on Velvia than occurred in real life, as well as unexpected effects that lead to "holy shit" moments on the light table when you first get your film back.

    Blues have magenta
    Fuji's Velvia Emulation nails this bit. Velvia renders the whole sky with a lot more magenta which makes lower parts of the horizon that are typically more cyanic feel more like the upper atmosphere. End result is Velvia skies feel deeper to us than they are in reality, and often seem less hazy.

    Fierce greens that have blue
    Digital tends to render greens accurately with a lot of yellow while Velvia makes forests more verdant with deeper, purer greens and then adding a bit of blue that gives plants a healthy just watered look.

    Control of saturation
    Contrary to some opinions, ultra saturation isn't the key to Velvia. On digital, extremely saturated colors like bright red tend to clip easily and are difficult to reproduce naturally without photoshop work. To deal with this, Fuji created their 'Color Chrome Effect," which is a rather processing intensive in-camera option that progressively reduces values on saturated colors to keep them from being overexposed. Velvia does this naturally and renders saturated colors dark and deep, creating great tonal variation where digital has difficulty articulating nuances.

    Purple is pink
    Ken Rockwell pointed this out and it's true. Instead of rendering civil twilight as a blue-ish purple, Velvia shifts the entire range so it's a beautiful pink. It's completely inaccurate, and I've seen completely wrong flower colors on some Velvia slides I've taken, but as Ken said, Velvia looks how you want it to look, not how it is.
     
    • Like Like x 5
    • Epic Epic x 1
    • List
    Last edited: Jun 19, 2022
  5. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Here are some scans of Velvia 50 positives from a long long time ago. Ken Rockwell has actually been around this long. He used to inspire people to go out and try cool shit. I would have never shot on slide film but for him.

    pump.jpg
    trip1a-19x13.jpg
    trip2 19x13.jpg

    I'm still hoping Fuji can better their Velvia algorithm / color science. (Maybe it's simply not possible since color and light are so complex.) And while they are at it, make a Ken Rockwell emulation. :p

    ^ And oh, that's Wyoming. State is dirt poor. But so beautiful and majestic. I can't help but make an extended stop while passing through.
     
    Last edited: Jun 19, 2022
  6. YMO

    YMO Scatologically ribald obsessive

    Friend
    Joined:
    Apr 1, 2018
    Likes Received:
    8,907
    Trophy Points:
    113
    Location:
    DUVALLL-904
    Are Fuji's still solid for shooting people? I heard a while back that you get a Sony for its flexibility and mirrorless action, Nikons for its look, and Fuji's for people shots.
     
  7. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I guess so. I've heard the opposite with to Fuji and Nikon with respect to skin tones. I'd never had a problem with skin tone color science with any body from any brand with the exception of Canon in the 2000s that made Caucasian skin have a pink cast that was difficult or impossible to get rid of. The the color science evolves from each brand, so it's hard to pin down.

    If you are a JPG OOC shooter like me, every camera has a white balance shift. The Sony cameras tend to be a bit cold for my taste, so I shift the white balance a tiny bit to get a warmer color. I love the OOC colors from the Canon EOS R6 as is. The white balance is amazeballs on today's cameras. It's not like you have to be aware of light as much as with film cameras. Ultimately it comes down to personal preference.

    Every tool has its own peculiarities. Ultimately it's the man, not the machine. I may bitch about the autofocus XF50mm f1, but I'm confident I will learn to work around it to the point where I don't have to chimp.
     
    Last edited: Jun 19, 2022
  8. roshambo123

    roshambo123 Friend

    Friend
    Joined:
    May 26, 2018
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Los Angeles
    I don't believe they consider it a problem. I recall reading some article where Fuji stated the goal of the emulations was not to replicate the exact response of the films but to provide a characterization that was in-line with the general idea. Overall though, I agree with you, I'd like to see the emulation be closer.

    Added some pics:

    Mildly fair comparison between by GFX 50R Velvia emulation and actual Velvia, both taken in Hawaii. Minimal edits on the Velvia (it was a long exposure and has reciprocity failure color shifts) and none on the digital shot. I feel like the emulation looks out of box roughly what someone might post process Velvia film to look like.

    [​IMG]


    Random velvia slides for fun:
    [​IMG]
    [​IMG]
    [​IMG]
     
    • Like Like x 2
    • Epic Epic x 2
    • List
    Last edited: Jun 19, 2022
  9. rhythmdevils

    rhythmdevils MOT: rhythmdevils audio

    Friend
    Joined:
    Apr 15, 2020
    Likes Received:
    10,115
    Trophy Points:
    113
    Location:
    Bay Area, CA
    Home Page:
    Nice write up.

    I personally don't get the whole bokeh thing. I never shoot wider than about F4 but usually 5.6. Any wider than that and the subject is mostly out of focus. I guess it's just the kind of images I make, documentary mostly, and I want a healthy portion to be in focus. Too much blur just looks like "I just got a DSLR" pictures to me.

    I just got a Sony DSLR because I need a small one for now, the a7c. I got 2 35m lenses for it, the GM and the 2.8. I wound up keeping the 2.8 because I'll never use 1.4 and the image quality looked fine to me on the smaller lens. I don't pixel peep at all.
     
  10. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    It's important for stuff like documentary style weddings and to a lesser extent portraits where the photographer can't change the background. For example, a kid picking his nose a few feet behind the bride. Although that might make a great shot. Or taking photos of audio gear on a messy desk. Oh, and one final reason: bad light. Bad light was a huge problem with film and early digital. Today, everything does ISO256000, so less of a reason.

    Generally it's a cheap way out from proper composition. :D So yes, I agree with you. I'm seen plenty of audio meet photos where people shoot wide open, where only the volume knob on a DNA was in focus and everything else was out of focus. It's like dude, stop down to f4. You can still get rid of the distracting background at f4.
     
    Last edited: Jun 20, 2022
  11. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Fake Velvia just don't look right. Fuji needs to take thousands of actual processed Velvia, scan them, take photos of the sane scenes with their sensor, and then run all of it through machine learning to understand how to map the parameters. This is 2022. If they won't do it, someone else will.
     
    Last edited: Jun 20, 2022
  12. rhythmdevils

    rhythmdevils MOT: rhythmdevils audio

    Friend
    Joined:
    Apr 15, 2020
    Likes Received:
    10,115
    Trophy Points:
    113
    Location:
    Bay Area, CA
    Home Page:
    I'm always rolling my eyes at the FS pictures on Head-fi taken at F1.2. I can just feel the photographer thinking they're so fucking awesome because they have an expensive Sony and a GM lens. And the pictures suck ass.
     
    • Epic Epic x 2
    • Like Like x 1
    • List
  13. roshambo123

    roshambo123 Friend

    Friend
    Joined:
    May 26, 2018
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Los Angeles
    I've thought about this and I don't know that you can accomplish it with this approach. To me the chief problem is accurate A/D conversion.

    Drum scanned Velvia doesn't look like Velvia on a light table. A good drum scan gets all the information out but is going to look a bit flat, sort of in the way "Log" looks on some cinema cameras. So you're always applying post adjustments to get it to look like the slide on the light table, which is a bit subjective. Then there's the issue of every scanner has a profile and every scan file has a profile. Lots of opportunity for error and you can end up going to insane lengths attempting to preserve accuracy.

    Then there's the issue that Velvia is fixed white balance and RAW has no white balance until you bake it out. That process I believe has some RAW converter dependencies and you're looking at again, lots of chance for error. Maybe if you wrote your own RAW converter literally just for this you could overcome that.

    I view it as unnecessary though. To me the important thing is knowing what Velvia did that I liked and moving the picture toward that if I want to.

    For fun: Crops from random test between D600 and Velvia many years back. The pink shift is real (the flowers should be more purple-blue)

    [​IMG]
     
    Last edited: Jun 20, 2022
  14. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I secretly wanted to take photos of myself at f1.0 contemplating really hard looking at gear like that Rolf, Raff, Roff, Woof guy who took over InnerFidelity and ran in into the ground after Tyll left. Can another member fill in @rhythmdevils on Tyll's retirement and what happened afterward?
     
    Last edited: Jun 20, 2022
  15. Bina

    Bina MOT - Shanling

    Friend
    Joined:
    Sep 29, 2015
    Likes Received:
    403
    Trophy Points:
    63
    If somebody want to play with some extra bokeh at more affordable cost, I found the 7Artisan F0.95 quite fun choice for the money. I was running it on my Olympus Pen F, landing manual focus is nice challenge (even if I failed more than I would be happy with :D ) It's quite chubby and hefty lens, but a shorter one, so balanced ok.

    https://7artisans.store/collections...ony-e-nikon-z-fuji-fx-canon-eos-m-olympus-m43
     
  16. iFi audio

    iFi audio MOT iFi Audio

    Joined:
    Oct 3, 2017
    Likes Received:
    1,181
    Trophy Points:
    113
    Location:
    Many places.
    Home Page:
    50mm f/0.95 for $200, geez! Need to read about this one. Very happy with 60mm f/2.4 macro on my X-T2, but wouldn't mind a bright 50mm as long as it's sharp.
     
  17. purr1n

    purr1n Burned out

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    81,453
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    So to follow up on the AF issues: I was dumb.

    I used the Fuji X-T3 with wide/tracking AF mode and AF-C continuous autofocus, thinking it would track like a Sony (I'm not exactly doing sports shooting). It doesn't. Evidently, getting 50% of shots in focus is normal in this mode of operation. Grrrr, and all the Internet reviews never mention this.

    Single point and zone AF mode work fine though. Use it like a Canon AE-1 or Nikon F100, and the AF fine. Just don't try to use it like a modern high-tech camera from Sony. The cool features are still in beta.
     
    Last edited: Jun 23, 2022
  18. Martigane

    Martigane Rando

    Joined:
    Mar 5, 2019
    Likes Received:
    79
    Trophy Points:
    13
    Location:
    Belgium
    Last edited: Jun 25, 2022
  19. iFi audio

    iFi audio MOT iFi Audio

    Joined:
    Oct 3, 2017
    Likes Received:
    1,181
    Trophy Points:
    113
    Location:
    Many places.
    Home Page:
    Just took a looky and "nervous" is the right word to describe that Sigma's background. Not bad at all per se, but not the smoothest one out there either. Still, that's a sharp lens just where it needs to be.
     

Share This Page