yayatosorus
Senior Member
I've tested the original 56mm with the X-H2s in dim light and the AF was very very quick. Still have to test AF performance on the 40MP bodies though, so can't comment on that.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
40MP DPR Test Scene comparison between the old and new 56mm f/1.2
- Thread starter Erik Baumgartner
- Start date
Rightsaidfred
Senior Member
- Messages
- 2,961
- Solutions
- 18
- Reaction score
- 2,996
First of all, many thanks to Erik.
However, I agree with Powerdoc. The fewest people will buy either 56/1.2 to mainly shoot at f/5.6. Most users will want to use such a lens wide open.
So we actually need to compare them at f/1.2.
Regards,
Martin
If DPR was having trouble focusing the original, and the challenge evidently continued with the new version, why wouldn't they have at least given the 50 f/2 an opportunity?
I'm not wanting to flame the 56mm which clearly has its charms, but sharpness at f/5.6 certainly isn't one its claims to fame. The 50mm - which is far sharper at f/5.6 and likely has less CA, as well - not only would serve the narrow chart shooting purpose better, but probably is easier to focus due to that sharpness!
I'm not wanting to flame the 56mm which clearly has its charms, but sharpness at f/5.6 certainly isn't one its claims to fame. The 50mm - which is far sharper at f/5.6 and likely has less CA, as well - not only would serve the narrow chart shooting purpose better, but probably is easier to focus due to that sharpness!
Erik Baumgartner
Forum Pro
- Messages
- 10,235
- Solutions
- 39
- Reaction score
- 14,241
If you actually tested it, possibly you got a bad copy of the 50mm. However, the two much-respected testing sites - Lentip and Opticallimits - both indicated that the 50mm was significantly sharper in the center; at the edges Lenstip had them even (only at f/5.6 but 56mm trailing at f/4) and Opticallimits had the 50mm well ahead at the edges. Overall, it wasn't particularly close.
I'm not saying that rendering is comparable (the 56mm is more impressive, of course), but for chart shooting your experience is not in line with the published, recognized data from two leading sources.
--
JNR
Last edited:
Pennyfan
Leading Member
I disagree, had the XF23 R and it wasn't sharp enough across the frame. Sold it, and been using Viltrox 23 until I got the new XF23 WR, which is a much sharper lens overall.
Same for XF56 WR, probe the sharpest lens I ever owned, a bit let down by the AF, but good enough for portrait photography...
The biggest improvements are sharpness at wide opened aperture (f1.2 / f1.4) and corner sharpness across all aperture. Those DPR tested them at f.5.6 thus you won't see significant different, but at f1.2/1.4, there are.
And don't tell me "just close your aperture"
, these lenses are designed for ultimate low light performance and background separation.If strictly comparing only f5.6, you can prob say there's no different between our kit lens XF18-55 and red badge XF16-55.
--
http://www.instagram.com/pennyfan.photography
Last edited:
JNR is right. For pure resolution and sharpness, the Fuji 50 f2 blew away the 56 f1.2 R and it the sweep spot the new 56 is equivalent according to DPR.
The 50 f2 is clinically sharp - maybe almost too sharp as it can impact its rendering.
bonneaufoto
Member
- Messages
- 30
- Reaction score
- 9
Hi!, Interesting if you look DPR Test Scene comparison between Canon R5 and Fuji X-T5 new 56mm f/1.2R WR. Fuji still handle better, more sharp, less moiré, if you look the scene black/white
Wow! This 40MP, beautiful.work from fujifilm team.
Canon R5-RF50mm F1.2 L
Fujifilm X-T5-XF56mmF1.2 R WR
Wow! This 40MP, beautiful.work from fujifilm team.
Canon R5-RF50mm F1.2 L
Fujifilm X-T5-XF56mmF1.2 R WR
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
Perhaps you have a bad copy of the 56mm. Fuji’s MTF charts show similar performances wide open, which obviously suggests an advantage to the 56mm. Other tests show the same thing.
How does a “too sharp” rendering manifest itself in photographs?
The higher frequency MTF suggests potential bokeh harshness 50mm - note the divergence in the meridional and sagittal curves.
Last edited:
Sorry but i´m not seeing better sharpness on the XF 56, not in the center text and not in the corners.Hi!, Interesting if you look DPR Test Scene comparison between Canon R5 and Fuji X-T5 new 56mm f/1.2R WR. Fuji still handle better, more sharp, less moiré, if you look the scene black/white
Wow! This 40MP, beautiful.work from fujifilm team.
Canon R5-RF50mm F1.2 L
Fujifilm X-T5-XF56mmF1.2 R WR
Yesterday i made this same comparison myself - downloaded the files, loaded them in C1, disabled sharpening on both and downsized the Canon file - and the XF 56 was still a little bit behind the RF lens.
Had to do this comparison because in the comments of this - https://www.dpreview.com/articles/7...ifilm-56mm-f1-2-wr-for-use-as-our-studio-lens - there was some guy hyperbolizing how disappointing the new XF56 was compared to the R7 and R5 plus RF 50 combo. He got me curious. Being that the Canon lens costs more than double the new XF 56 it´s not surprising it is (just slightly) better. But in no way i would call the new XF 56 optically disappointing. Never bothered to reply to his comment.
Dirk Dittert
Forum Enthusiast
That might be the case. I do find the sound of the motor driving auto focus annoying, though.
The reason to stick with the original lens is simple. If they changed then - one could not do a valid comparison between camera X using the the old lens and camera Y using the new. If one was interesting in comparing camera X to Y - such data would not be useful because they were not tested under the same conditions. Clearly DPR goes to great lengths in their test setup to insure all cameras are tested under the same conditions.
Since at the established f stop they use for testing previous both lenses are equivalent within their testing tolerance it makes no sense to change as that would call into question the validity.
They were testing the camera - not the camera and lens combination.
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
The photographs are not made with a high degree of rigor - they are suggestive of performance but shouldn’t be interpreted as dispositive in any way. For example, the GFX 100 images show banding that indicates the example they used had obsolete firmware.
Sadly, this is not the case, since cameras require lenses to make photographs. Obviously they are, in fact, testing camera and lens combinations. Because of this, and the fact that they use a single sample of the (variable quality) lenses, we cannot draw strong, independent inferences about the performance of the cameras alone, from the images. You seem to be overestimating the insights the studio scene images can provide.
Last edited:
This pretty much qualifies as changing the subject to an overall comparison of the two lenses, which is what I stated was not my goal in bringing up the 50mm option. Recall, I stated that the 56mm has more impressive rendering. You're disrespecting the thread topic. Chart shooting has nothing to do with which lens is better shot wide open, and as a 2D object, bokeh comparisons are irrelevant.Perhaps you have a bad copy of the 56mm. Fuji’s MTF charts show similar performances wide open, which obviously suggests an advantage to the 56mm. Other tests show the same thing.
How does a “too sharp” rendering manifest itself in photographs?
The higher frequency MTF suggests potential bokeh harshness 50mm - note the divergence in the meridional and sagittal curves.
Stick to the topic. The controlled comparisons of center and edge sharpness shot at f/5.6 (or f/4 as some prefer for higher quality optics) indicate that the 50mm is the superior choice for the particular, narrow task at hand.
Believe me, I'd be the first one to support rating lenses based more on rendering characteristics, as opposed to the over-emphasis on flat field sharpness. But for evaluating sensor qualities, the chart shot at optimal f-stop is the correct procedure.
--
JNR
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
I was responding to an assertion about the relative performance of the two lenses, so the source of your complaint lies elsewhere.This pretty much qualifies as changing the subject to an overall comparison of the two lenses,Perhaps you have a bad copy of the 56mm. Fuji’s MTF charts show similar performances wide open, which obviously suggests an advantage to the 56mm. Other tests show the same thing.
How does a “too sharp” rendering manifest itself in photographs?
The higher frequency MTF suggests potential bokeh harshness 50mm - note the divergence in the meridional and sagittal curves.
This seems like an over-reaction. I would suggest skipping this branch if it upsets you - take advantage of the threaded structure.
Why are bokeh comparisons irrelevant?
Which “controlled comparison”? The test here is hardly a controlled comparison. It is a somewhat casual exercise undertaken with single samples of lenses that are prone to sample variance. Other online tests indicate that the 56mm is slightly better in its ability to resolve. Again, these are based on single samples.
Hard to make objective standards for “rendering characteristics”. Subjective reviews are fine, but are subjective, obviously.
Over-emphasis is in the eye of the beholder.
If you wanted to carefully evaluate sensors, you would have to use the same lens for all of the tests. You could use adaptors and a single, tested-as-excellent example of a Rodenstock HR Digaron-S 100mm, for example, and isolate the sensor performance. That’s not practical here for various reasons, so we are left with a rather imperfect comparison based on single samples of lenses from the camera manufacturers, that may vary over time. There seems to be a tendency to impute more insight than is possible, given the practical constraints.
Last edited:
An MTF only tells you one thing. It tells you the lens transfer function at a fixed frequency in lp/mm as a function of distance from the center. From Fourier optics we know the max contrast is going to be in the center. So the MTF is scaled by the contrast at the center which is way the output plotted on the y-axis ranges from 0 to 1. While it measures the impact of many distortions caused by aberrations of the lens, for the most part the primary thing it measures is field curvature. A lens transforms a spherically far field wave radially wave into a collimated beam of (parallel rays) of light the MTF measures how way it does that at a distance from the center out.
What the MTF does not tell you is the center contrast at one frequency vs. another as 15 lp/mm vs. 45 lp/mm. That is how does the lens attenuate as a function of frequency. Although it seems that Fujifilms makes an attempt to supply it at least at two points. What the MTF does not measure is how the lens renders out of focus planes in front and/or behind the focal plane. It is produced with the illuminated chart perpendicular to the lens axis.However, in the real world a lens focuses light from a three dimensional space onto a plane. It is in actuality a projective transform when used in the real world - no the chart in the lab world. That all point sources of light on a line will end up at the same point on the sensor.
The design choices that produce the best MTF may not be the best choice for rendering of the the light point sources in out of focus areas and the transitional areas between the focal plane and the out of focus areas. Normally the rendering in these areas are lumped together under the term Bokeh. This rendering can result in smooth transitions, choppy transitions (usually called nervous Bokeh), or down right ugly such as onion rings, cat's eyes, or sharp circles around out of focus point sources.
The clinically sharp lens is only about the one infinitesimally thin plane perpendicular to the lens not the projective rendering of the volume of dispersed light in the cone defined by the field of view. The goal of designing a lens to get the best MTF involves trades that impact Bokeh or the volumetric rendering.
A pretty good description of rendering of point sources of the focal plane.
A descent lay discussion.
Ray tracing is the bread a butter tool for lens design today. Modern high speed computers allow very complex designs to be undertaken. However, geometric optics which is the basing of ray tracing is not adequate to accurate analyze how a lens renders light off the focal plane. However, it becomes a simple task using the fact that Fraunhofer diffraction shows that a lens is in fact Fourier transform and that allows one to perform analysis in the off focal plane volume of space where the bokeh lives and some say "character" lives.
The 50 f2 is a clinically sharp plainer designed lens. The plainer design emphasizes a flat field is based on the double Gaussian lens. The other classic Zeiss design was the Sonnar design.
These two designs are the starting point of most modern lenses.
So what clinically sharp gets you is highest resolution in the focal plane and the flattest field ( or sharpness in the corners) at the expense of the out rendering the out of focus areas. The beautiful rendering of the Sonnar lens design is the trade off of having some field curvature. Or as Carl Zeiss put it the eye is as important as mathematics in designing a camera lens. Interesting reading
Double Gauss and Sonnar lens designs: A complete comparison
A complete breakdown of the Double Gauss lens design and the Sonnar lens design. Includes history, technical examples, and performance evaluation of each.
www.pencilofrays.com
The double Gaussian Plainer lenses tend to be sharper while the Sonnar lenses produce more pleasing Bokeh and bette "character." Of course character is a subjective term based on how pleasing a person sees the rendering of the image.
In my view, the Fuji 50 f2 is a classic example of a highly corrected flat field design which is very sharp. It's a nice lens for I would not use it to produce a flattering portrait. It lacks in character compared to the 56 f1.2 and 50 f1.
Now others may have different objective views of the three lenses. You asked what I meant - so here it is. YMMV
You're right. This side trip isn't worth heading down at all, but your misunderstanding of aspects of my comments won't go unaddressed.
First, yes - both Erik and Truman made passing comments that enlarge the topic into the broader comparison - but your post entirely ignores the value of point I made, and is clearly aimed at debating the two lenses overall. Bad form.
Second, the "controlled comparisons" are the Lenstip and Opticallimits testing sites I referenced. Unless you've done something similar that negates those results, you really are just expressing an unfounded opinion. Try to at least to follow the point made before making the criticism.
If you're going to respond, please quote me in full, and don't personalize. Be civil.
First, yes - both Erik and Truman made passing comments that enlarge the topic into the broader comparison - but your post entirely ignores the value of point I made, and is clearly aimed at debating the two lenses overall. Bad form.
Second, the "controlled comparisons" are the Lenstip and Opticallimits testing sites I referenced. Unless you've done something similar that negates those results, you really are just expressing an unfounded opinion. Try to at least to follow the point made before making the criticism.
If you're going to respond, please quote me in full, and don't personalize. Be civil.
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
There are two lines in eqch graph - for meridional and sagittal lines, so we get insight into astigmatism, as evinced by a separation between the lines. This can have the effect of creating unpleasant bokeh, at least according to Roger Cicala.
Thank you for the references. I wonder about tradeoffs of out-of-focus rendering versus maximal sharpness in a, say, 17 element lens. I look forward to reading more.
If one solves the the propagation equations for the boundary conditions at an air/glass interface there is reflection. It is necessary so to maintain phase continuity across the boundary. A lens with N elements has 2 N air/glass boundaries. To some extent the internal reflection back into the glass and the reflection back into the air off the lens element. Modern coatings help but only on reflections off of the glass since the coating is on the outside of the element and none on the inside.
The heavy use of the MTF has generated a hunt for the holy grail - the elimination of field curvature. However, in doing so other effect creep in. I always read lens reviews with a large grain of salt. I don't really care about how a lens renders a chart. I care about how it renders the photographs I take and how I present the photographs. The 50 f2 is probably the flattest field lens I have and it is also the least like rendering. It has other positives, compact, fast and accurate to focus and works fine on some scenes. Although on other it leaves me longing.
Similar threads
Keyboard shortcuts
- f
- Forum