Nielk Mike wrote:
The difference between 26 and 40MP is relatively small and certainly not a step in the right direction on the part of Fuji. But: There seems to be little choice for Fuji given that they source their sensors from Sony. 26MP is more than enough for most every use case, and there is no lens out-resolving the 26MP sensor. All current Fuji primes including the older ones return sharp images on the 26MP sensor, though some need to be stopped down to f4 to get acceptable sharpness across.
But that is only one side of the story. Since I had to plunge into 40MP territory with the X-E5 (my other cameras are the X-T2, X-T3, X-Pro3 and X-E3) I decided that I can as well make use of the only advantage that I see with that sensor: Cropping. My basic lens set is consists of three lenses: Sigma 12mm, XF23f1.4WR and 56f1.2 WR. If I need small, I go with the 16/23/50 Fujicrons.
The 40MP sensor allows me to turn those 3 focal lengths into 7: 12/18 and 23/35 and 56/85/120. It also allows me to use 2x crop with the 12 and 23 if needed (if there is no time to switch lenses from, say, the 12 to 23. The images of the 12/23/56 are visibly sharper and very even across already at f1.4 when I crop compared to the older f1.4/f1.2 lenses.
Yeah but. The one thing more pixels does not cost you is any loss of lens resolution. One thing you gain and rarely mentioned is increased tonal gradation. Yes, you gain file size and lose some write speed, but storage is cheap and the 15 fps of the XH2 is more than fast enough for me. Noise is a wash; you get a bit more with more pixels, but wipe most of it out if you downsize to match the smaller sensor. You mentioned crop, which is a good thing. The other, again rarely considered advantage is a broader usable range of various aspect ratios; meaning square is better with more and so are any variety of pano crops.
Nielk Mike wrote:
The difference between 26 and 40MP is relatively small and certainly not a step in the right direction on the part of Fuji. But: There seems to be little choice for Fuji given that they source their sensors from Sony. 26MP is more than enough for most
every use case, and there is no lens out-resolving the 26MP sensor. All current Fuji primes including the older ones return sharp images on the 26MP sensor, though some need to be stopped down to f4 to get acceptable sharpness across.
MIke, you are not quite making sense here.
I can agree that the difference between 26MP and 40MP amounts to less than many people think, but in my experience it is, for example, greater than the difference between the resolving power of the f/2 and f/1.4 35mm lenses that seem significantly different to you in that regard.
I also disagree with your claim that “there is no lens out-resolving the 26MP sensor.” I’m pretty sure that several of mine have that capability.
It is true that “[a]ll current Fuji primes including the older ones return sharp images on the 26MP sensor,” but that doesn’t say anything about performance on the 40MP sensor. Unless “all current Fujifilm primes” have [b[exactly[/b the resolution that precisely matches but does not exceed that of 26MP sensor, at least some of them that “return sharp images on the 26MP sensor” have the potential to produce even more detailed images on a higher than 26MP sensor.
My experience with the 40MP sensor is that there are no performance downsides (unless image storage is a major issue for you, in which case I recommend a larger drive!), and that there is a potential upside. Going from 26MP to 40MP is not going to make a monumental difference in image quality (and I would not recommend that people with relatively recent 26MP cameras urgently upgrade to 40MP cameras), it will make some positive difference. With good lenses, careful focus, thoughtful aperture selection, attention to camera stability, skillful post-processing, and so on there is some room for even more image resolution, slighlty less pixelation at large reproduction sizes. slighlty smoother gradients, and smaller noise “grain” size… all with no downside.
In other words, it is a useful and meaningful bit of progress for these cameras — not enough that everyone with a 26MP sensor needs to upgrade, but enough that new buyers can appreciate it.
gdanmitchell wrote:
MIke, you are not quite making sense here.
I can agree that the difference between 26MP and 40MP amounts to less than many people think, but in my experience it is, for example, greater than the difference between the resolving power of the f/2 and f/1.4 35mm lenses that seem significantly different to you in that regard.
I also disagree with your claim that “there is no lens out-resolving the 26MP sensor.” I’m pretty sure that several of mine have that capability.
It is true that “[a]ll current Fuji primes including the older ones return sharp images on the 26MP sensor,” but that doesn’t say anything about performance on the 40MP sensor. Unless “all current Fujifilm primes” have [b[exactly[/b the resolution that precisely matches but does not exceed that of 26MP sensor, at least some of them that “return sharp images on the 26MP sensor” have the potential to produce even more detailed images on a higher than 26MP sensor.
Dan, back in the days of the NEX3, Sony had a nice little 16f2.8 pancake. It worked like a charm on that camera, sharpness was good. The sensor was 14MP. Then came the a6000 with 24MP - and the 16 f2.8 was widely see as almost unusable due to the sharpness fall of towards the outer edges, even stopped down. Was it noticeable when looking at an image taken with that lens down sampled to Full HD monitor resolution? Hardly. But if you zoomed (cropped) in with that 24MP sensor, it did show. The better lens (the Sigma 16f1.4) provided sharp detail.
Same today. The 35f1.4 is a very fine lens, I own it and don't intend to sell it, but it has limitations used with apertures below f4 and away from the center. It shows on a 40MP sensor at 100% and when cropping. If I compare it to the 35f2 which resolves more than the 35f1.4, I could see the difference, of course, also on a 26MP sensor. It is just more obvious using the 40MP sensor. The 40MP sensor allows me to crop 2x and still retain a 4K resolution, unlike the 26MP sensor. And yes, in order to make use of that additional resolution, one has to be using higher shutter speeds and to make sure that focus is perfectly where it is supposed to be (and not just within the DoF - ah, one of my favorite topics).
Did you have a chance to read the thread about resolution on dpreview that I linked to? No current lens in the Fuji line-up out-resolves 26MP - not even that Viltrox 75mm beast.
The cropping has become part of my photography. It is why I bought the Sony RX1RMKIII, and the X-E5. When I am out with only a small camera, the RX1 provides me with 3 (4) "lenses": 35/50/70 (and 100) with very respectable output even at 3x crop. That's all I need on my walks. As I said, the X-E5 and the Sigma12f1.4/Fuji 23f1.4WR/56f1.2WR provide a range of angles of view from an 18mm FF equiv. to 170mm - with an f1.4 stop in terms of light gathering ability. To make the best out of it, one needs lenses that resolve better than the older generations, and a sensor that has more pixels to crop from. So there is my use case for the 40MP sensor and the new f1.4 lenses.
I’m telling you, Mike, for all of those words… I use the 35mm f/1.4 on the XT5 and it performs beautifully, with excellent sharpness. And to get back to the main question, if one needs/desires f/1.4, then they should get the f/1.4 version of Fujifim’s 35mm prime. It is an excellent performer. If they don’t need f/2 and/or they prefer a somewhat smaller lens and a lower price, then the f/2 version is an excellent alternative.
Pretending that there is some major difference in their optical performance or that the 35mm f/1.4 is optically deficient compared to the f/2 lens or that it isn’t up to the resolution of the 40MP sensor is highly misleading.
Handheld, XT5, 35mm f/1.4 lens — full image and crop.
There is one more aspect Mike and the DPR article avoid, and that is the effectiveness of current digital sharpening technology. It is currently good enough to recapture what was most likely there, and the more pixels available, the better. Is it foolproof? Probably not, but I’ve never seen an unbelievable result when appropriately used. Does it work? Yup; and probably as well or even more accurately than AI.
Never had a problem with sharpness on the 35 f1.4, assuming the subject is in focus. It might even be sharper than the 35 f2, though the difference is negligible.
The 35 f2 is a little smaller, and it focuses faster and more quietly. It is also usually more affordable.
mdude85 wrote:
Never had a problem with sharpness on the 35 f1.4, assuming the subject is in focus. It might even be sharper than the 35 f2, though the difference is negligible.
The 35 f2 is a little smaller, and it focuses faster and more quietly. It is also usually more affordable.
That’s pretty much it.
One of the reasons I wanted to test it against the f/2 35mm lens was that I had read that it would focus faster than the f/1.4 lens. But using the two lenses side by side, I could not actually detect any functional difference in AF speed. It may be that that on the test bench there is some tiny difference, but if so it must be so small as to be unnoticeable.
I can’t speak to the quietness of the f/2 lens, since I didn’t notice any problematic level of AF noise from either lens so I did not compare them.
Again, the f/2 is a fine lens — optically excellent, small, and relatively inexpensive. I think it is a great choice for many people who want a 35mm prime and who don’t need the larger f/1.4 aperture.
Basically, Fujifilm has given us two excellent lenses to choose from, and we get to pick the one whose functionality (and price) best fits our needs.
(EDITED to remove my erroneous reference to a non-existent lens!)
mdude85 wrote:
There's a new 27 f2.8, but not a new 35 f2.8...
I need to stop making that mistake! I know there’s no 35mm f/2.8… but that’s the second time I’ve done that. (I’ve altered the post to fix that error.)
It's a long time since we had a 1.4 vs 2 thread, used to be common. It always amazes me that so many find the 1.4, for all its other virtues, as sharp or sharper than the f/2. When I bought my Xpro2 all those years ago I looked at all the tests, every one said that the f/2 was sharper at every aperture they had in common, and I've not seen any since that measured anything else. I don't think the difference is huge, but every testercould measure it.
So I bought the f/2, have used it a lot over about 8 years, never regretted it. In particular it's usable across the frame at 2 and 2.8 whereas the 1.4 in tests is quite soft at the edges till well stopped down.
If you need sharp at 1.4 buy the 33mm, you want character and soft edges for portraits at 1.4, buy the original 35/1.4
gyoung143 wrote:
It's a long time since we had a 1.4 vs 2 thread, used to be common. It always amazes me that so many find the 1.4, for all its other virtues, as sharp or sharper than the f/2. When I bought my Xpro2 all those years ago I looked at all the tests, every one said that the f/2 was sharper at every aperture they had in common, and I've not seen any since that measured anything else. I don't think the difference is huge, but every testercould measure it.
So I bought the f/2, have used it a lot over about 8 years, never regretted it. In particular it's usable across the frame at 2 and 2.8 whereas the 1.4 in tests is quite soft at the edges till well stopped down.
If you need sharp at 1.4 buy the 33mm, you want character and soft edges for portraits at 1.4, buy the original 35/1.4
Depends a lot on how critical frame edge sharpness is to one's actual images; and then there may also be broader sample variation than I'm aware of -- and perhaps Dan and I have and had outstanding samples... As re importance, in my case, softer edges are often a pretty big plus as sharpness tends to pull the eye to it, similar to a vignette. In the case of my copy of the 35/1.4, that edge softness was additionally pretty subtle at f1.4, and virtually non-existent at f2.8 on my 40mp cameras.
Other than astro, I'm not really sure what the use case is for a 35mm f1.4 that is sharp at the edges at f1.4, but here are a few shot at f1.4. Web display doesn't tell much but these look plenty sharp to me at home on a 28inch 4k screen.
a few more shot at f1.4, with at least some parts of the subject in the focus plane at the edge of the frame. The horse trough is interesting. While this was a discard as I find the out of focus foreground distracting, the bottom right corner of the frame is in focus. While I wouldn't conclude it is super sharp, I don't think it is bad and certainly not something that would make me choose another slower lens.
About the Fuji 35mm f/1.4 lens, people here say that you could photograph an overturned trash bin with it and it would still look good. The 35mm f/2 lens has a slightly more boring, more modern rendering.
There is a lot of "who needs sharpness away from the center wide open", or "a little less resolution doesn't hurt the image, and "the character of the lens is key".
A good start would be to accept that the 35f1.4 according to the reviews I have seen and my own tests resolves less than the f2 version.
Once one accepts the difference, one than has to decide whether the one stop higher light gathering ability of the f1.4, or its "character" is a good reason to buy a bigger and more expensive lens.
No one questions that the f1.4 is a good lens capable of producing sharp and good looking images. 35mm images often have the center of attention at the middle of the frame anyway. And without much cropping, sharpness will not been issue, even at f1.4 in the center.
But no sample images in the world can negate the difference in sharpness performance between the f1.4 and the f2.
Nielk Mike wrote:
There is a lot of "who needs sharpness away from the center wide open", or "a little less resolution doesn't hurt the image, and "the character of the lens is key".
A good start would be to accept that the 35f1.4 according to the reviews I have seen and my own tests resolves less than the f2 version.
Once one accepts the difference, one than has to decide whether the one stop higher light gathering ability of the f1.4, or its "character" is a good reason to buy a bigger and more expensive lens.
No one questions that the f1.4 is a good lens capable of producing sharp and good looking images. 35mm images often have the center of attention at the middle of the frame anyway. And without much cropping, sharpness will not been issue, even at f1.4 in the center.
But no sample images in the world can negate the difference in sharpness performance between the f1.4 and the f2. ...Show more →
Perhaps you could post images showing the sharpness difference? I have both lenses and can't see the difference. It may be there, but the question is whether the difference is significant enough to be a differentiator. There are other aspects that differentiate the lenses. For my needs at least, sharpness does not.
Geoff D F wrote:
Perhaps you could post images showing the sharpness difference? I have both lenses and can't see the difference. It may be there, but the question is whether the difference is significant enough to be a differentiator. There are other aspects that differentiate the lenses. For my needs at least, sharpness does not.
Don't have membership to post images - so you have to take my word for it: Even the Sigma 18-50 is sharper at 35f2.8 than the 35f1.4 at f2.8. About 2/3rds away from the center, resolution begins to degrade. It is fine from f4. None of that matters if you use the image at down sampled to view on a PC monitor or TV screen (4K), or print and view at appropriate distance. But one you crop in, or view at 100%, it is more than obvious. So it is not a lens I would use on the 40MP sensor with the intention to crop up to 2x. And you don't need my sample images, anyway. Just review the MTF charts at Lenstip or Optical Limits for the two lenses. Comparable at the center, significantly different mid-frame and outer part. My own test confirm that performance.
Nielk Mike wrote:
Don't have membership to post images - so you have to take my word for it: Even the Sigma 18-50 is sharper at 35f2.8 than the 35f1.4 at f2.8. About 2/3rds away from the center, resolution begins to degrade. It is fine from f4. None of that matters if you use the image at down sampled to view on a PC monitor or TV screen (4K), or print and view at appropriate distance. But one you crop in, or view at 100%, it is more than obvious. So it is not a lens I would use on the 40MP sensor with the intention to crop up to 2x. And you don't need my sample images, anyway. Just review the MTF charts at Lenstip or Optical Limits for the two lenses. Comparable at the center, significantly different mid-frame and outer part. My own test confirm that performance. ...Show more →
You don't need a membership to post. If you post on flikr you can link from there. That's what I do.
I tried the Sigma 18-50 after hearing so much about it. I wasn't impressed at all. It's one of the few lenses I have returned. Maybe I got a bad copy, but it was no better than my Fuji 18-55.
If you are really exacting about performance why choose a lens to crop in 2x in the first place? Why not just use a different focal length? It seems odd to me to worry about minute sharpness differences and then say it is because you want to crop in 2x or view at 100 percent.
Why am I not using a prime without cropping? Well, I do, but I think I also explained in detail why it is the best use of a 40MP sensor to crop - thereby being able to carry only two or three lenses to cover a FF equiv. field of view from 18mm to 150mm. The Sigma 12 mm is fast and sharp, easily usable at 1.5 an 2x crop. That's 18, 28, and 35mm. The 23f1.4 WR has outstanding performance across. It is on my camera most of the time now. 35, 50 and 70mm -easy. And then for short to medium tele the 50f2 or 56f1.2WR: 75, 100 and 150mm. The 2 crop returns 4K images. The better the lens resolves, the better the result when cropping.
