Pistnbroke: You would have to try it. The issue is not as black and white as you suggest. Thom Hogan is talking about ultimate image sharpness. If that is what you want shoot the D850 sensor at f4 or f5.6. Thom is saying this is due to the high megapixels and therefore will be true for all sensors around 50mp. This is also true of almost all lenses. They are sharpest about two or three stops down from optimum and when you go to f8 you are trading ultimate sharpness for depth of field. You still will capture more detail with the D850 at f8 than you will with the D5 at f8 using the same lens. I generally shoot portraits at f4 or f5.6 and seldom at f8 unless there is a large group in multiple rows so it is not an issue for me. I don't think it would be an issue for your either if you compared a D810 image with a D850 image and you would have the near 24 mp DX crop you have always wanted. As to speed. Thom was shooting football action. Weddings are not moving that fast. You shouldn't be shooting long bursts of action. Also, it will depend upon whether you are shooting RAW or only jpeg. For jpeg I don't think it matters because you are moving so much less data through the buffer. I almost always set my bodies to send jpeg to the SD card and RAW to the XQD card. I could set it to use the SD card for overflow and than both RAW and jpeg would go to the faster XQD card. I don't think the speed issue would be a practical issue in your wedding business. But you would just have to rent one and try it to see if those issues occur for you in your use situation.
If you are shooting JPEGS the buffer is massive. I would not worry about that.
I am not sure exactly how many megapixels a diffraction limited lens at f/8.0s, but look at the Imatest results from PhotographyLife for the 58mm 1.4G, 105 1.4E and 400 2.8E. I just purchased these three lens so I am experimenting with them on the D850.
At f/8.0, they all achieve a score in the centre of around 2590 (note how the 105 beats the other two into the corners). At f/11.0, they are all around 2208.
Further note where these lenses are diffraction limited, meaning that the resolution of the lens seems to be limited by diffraction.
Tis seems to be f/8.0 for the 58. Resolution increases at f/5.6, but not by much so it does not seem to be due to diffraction.
For the 105, this seems to f/4.0 for the 105, which is quite amazing. The 105 is blowing me away. This might be the sharpest lens that Nikon makes shorter than the 400 2.8E - though I expect the new 200 2.0E and 300 2.8E to regain their titles. It is certainly the second sharpest in my bag besides the 400. Look at this lens at f/2.8 where it is at its sharpest, though just barely beating itself at f/4.0. At f/2.8, it is sharper then the 400 2.8E. However, this is not an apples to apples comparison because if you compare them both wide open, the 400 wins (but holy cow, by less then I expected - and the 105 at 1.8 is sharper in the centre than the 400 at 2.8).
For the 400, you will note that it tops out at 3,000 on Imatest. I wondered about this until I started talking to Imatest about buying their software and charts. I asked them if the charts were good enough for the 100 megapixels that the Phase 1 can achieve. The answer is no, they are not sharp enough, but they also said that their charts are not sharp enough for the D850 (they are working on a better chart). I suspect that PhotographyLife may have been working with a chart that topped out at 3,000 lines.
But to the point in Thom's blog, PhotographyLife was likely using a 36mp D800 sensor in his tests and the sensor seems to be capable of achieving an Imatest score of 3240. Since lens can just barely or not quite achieve this level of resolution at f/5.6, if you are shooting with a D850 or even a D800 at f/8.0 or less, then you are achieving less resolution than the camera is capable of "DUE TO THE LAWS OF PHYSICS".
There is a practical corollary here. If you want to maximize your resolution on a 36mp sensor, shoot at f/5.6 or wider and f/5.6 may not be quite enough. If you are shooting at f/8.0, you are likely only achieving around 24mp and shooting 12mp of mush.
With the D850, my new mantra is f/4.0 or wider on my good lenses unless I am willing to throw away resolution.
WestEndPhoto: Yes, when sharpness is the goal mid-value fstops will rule on the D850. Perhaps the new focus shift feature will come in as handy for landscapes as it is for macro. I could see three zones: near element, mid-distance element and infinity element. Perhaps we can learn to measure the distance to each element and then set the focus shift to each distance and shoot three exposures from a tripod. Maybe future versions of this body series will provide an menu item to automatically do that. Just focus a single spot on each distance and the camera automatically selects the best focus shift number of exposures and then merges all exposures in camera to create one image. You could shoot landscapes with great depth of field at f4. Now what about even higher megapixel sensors? If a 45mm sensor is less sharp beyond f4 or f5.6 what would happen with a 100mp FX sensor? Would it be less sharp beyond f2.8 or f4? See the issue here? At some point higher megapixel FX sensors will no longer make practical sense because most glass is not optimized for f2.8 or f4. Perhaps 50mp is as high as we should go in FX size?
Well my comment produced some usefull answers...Its horses for courses as they say . What would be the point of putting a 28-300 on a D850 and shooting large basic ...no point . I dont think I would have got such sensible and helpfull answers on the other N sites I am banned from for making "usefull comments"
A 28-300 on a D850? Surely, you wouldn't be getting the maximum resolution the sensor is capable of. However, that one lens on that body is highly likely not to be the only lens you use 100% of the time if you purchase a D850. There can be utility in using the body with other sharper lenses while also putting the 28-300 on it when you want to take only one lens with you. For example if you want "snapshots" of your time in a hostile environment where you don't want to be switching lenses but want that range of focal lengths available to you. Sometimes you may "trade off" ultimate image quality for the protection you achieve by not ever removing the lens from the body in a certain environment. Why shoot large basic jpeg? Look at the photos in Ken Rockwell's review of the D850. Ken tends to shoot basic or normal jpeg for family snapshots. Look at the crop of Katie's eyes and examine her eyelashes. That should certainly be "good enough" detail for anyone in any situation. If you set your Picture Profiles to your personal preference and if you get your exposures right and if you are shooting in situations where you will not need to bring up the shadows or reduce the highlights because you either have control over the light or the scene is even lit you don't need RAW files and you can let the camera do all the processing itself. I often shoot RAW to one card and jpeg to the other card. For studio portraits I very seldom need the RAW file because the slight changes I make in post processing are all well within the range of jpeg latitude. One final point. You, and the client, don't always want ultimate sharpness. For example, portraits of old people who don't want all their wrinkles and age spots to show. If you have a D850 and are shooting such a client you may be well advised to put a "resolution challenged" lens on that body!
@WestEndFoto: what f stops do you get if you apply that to the D750 and D7100?
My comment was somewhat broad so I am not sure which "that" you are referring to. I will assume that you are asking, "At which f-stop will a sensor will be out resolved by a lens, assuming that the lens is diffraction limited at all of its apertures?" In the real world, there may be some f/8.0 or f/5.6 lenses like this. But thinking about PhotographyLife's Imatest results for the 105 (I pick this because there are values higher than 3,000 unlike the 400) may be helpful. Look at the centre results:
Another data point is Imatest's statement to me that the test charts (and I have access to the same test charts as PhotographyLife) do not have enough resolution to test the D850 (when they figure that out, I will buy the test chart).
So my hypotheses are:
-that the best a 36mp sensor can do is roughly 3,200 and it is unable to exploit the perfect lens that I describe above at a wider aperture than say f/4.5 (I am interpolating the curve and making a guess on why there is not better performance than 3226 at f/4.0: lens performance, diffraction, Imatest performance are the candidates).
-since your D7100 has a higher pixel density, it should be able to do better ( exploit a wider f-stop), but if we are going to use Imatest to test it, we need a better chart than is currently available
Note: Given how diffraction works, pixel density is the determining factor, or rather a proxy for the determining factor, as the lens is the determining factor.
-since your D750 has a lower pixel density, this limit is somewhere between f/8.0 and f/11.0
I want to test the above hypotheses and as I have alluded to, I have been talking to Imatest about buying their software and test charts but am stuck because they don't have a chart that can properly test the D850 (or a Phase One for that matter). But given that I have the 105 and 400 I might be able to explore and test aspects of the above hypotheses. I have DSLR's with three pixel densities, namely the D5500, D800 and D850. If I can get my hands on a D750 or even a D600/610 I can round it out.
I just need to find the time. I have been very busy professionally and have just moved into the new house and a baby girl was due October 5th - so we are waiting around for that to happen. But this is an excellent project for the fall.
Looking at Photozone's imatest results and comparing the same lens on a Canon 21mp sensor vs 50mp, it's clear that at apertures I prefer for landscapes, f11 and smaller (higher number), there is some incremental gain in resolution with more MPs, but the resolution bell shaped curve is steeper at wide and stopped down beyond f8 with 50 MPs; meaning there is a higher degree of separation between the best and worse apertures. The question for me and FF is where is there no real absolute gain in resolution with greater MPs? At F11 anything above 50 MPs might not offer enough improvement. Not sure.
I wouldn't find myself using stacking often (most landscape situations have some movement), so we are back to legacy scheimflug techniques to maximize results with highest MP camera.
Revised 24, 45-50, and say 85-90mm t-s lenses are desired. Otherwise not sure there is enough resolution benefit to warrant a d850.
I love the discussions which are about technical stuff. The current cameras I have are still so far beyond my abilities, it will be a while before I go for the D850. Maybe when I sell the D4 and 400/2.8...
But, the ability to manage the capabilities of current equipment, do a decent job with composition, post processing, final presentation, these are what, IMO, make the shot.
No doubt, the D850 with its exceptional versatility makes this easier, but I have often found myself lost in the process, forgetting the final goal.... oh well.
Probably order mine for a Christmas present to myself... LOL
I decided when I bought the d7200 that I would use it for 3 years before going full frame. It's been 2 years and I have about 40,000 clicks on the shutter. The camera is fine and all but I shoot a lot of low light work and I'm thinking the full frame and better ISO ability of the 850 would help me. I also think it focuses faster than the d7200 which would help with my runway work.
I may rent one and see if it improves my work any.
I'm with you Msmoto, I'm also fascinated by all the tech talk (even though most of it is well over my head), but I try to stay 'focused' on subject matter before technically perfect shots. I'm sure I fail more than I succeed but the trying is the most enjoyable thing for me. Don't get me wrong I can see only too clearly that my D810 takes technically better shots than my D700, but I'm still in love with the 700 and really enjoy using it sometimes right alongside the 810 just with a different lens on the front. Of course the latest tech lets you do more in post as well, but I'm not heavily into all that stuff even though I know some of my shots would benefit from better PP by an 'expert'. I get my satisfaction from getting 'the' shot when I can and the rest of the time I'm happy capturing images that are pleasing to me (sometimes the wife!) when I view them a second, third, fourth time etc. I will chop the D700 in at some point in the not too distant future once I work out what I would be best replacing it with for the stuff I like to shoot. Initially I thought that might be a D500, but I'm not so sure I need the crop factor anymore. I'd love a D5 but just can't justify the outlay. Maybe I'll just wait for the D900!!!
I love the discussions which are about technical stuff. The current cameras I have are still so far beyond my abilities, it will be a while before I go for the D850. Maybe when I sell the D4 and 400/2.8...
But, the ability to manage the capabilities of current equipment, do a decent job with composition, post processing, final presentation, these are what, IMO, make the shot.
No doubt, the D850 with its exceptional versatility makes this easier, but I have often found myself lost in the process, forgetting the final goal.... oh well.
Probably order mine for a Christmas present to myself... LOL
Let me know if you are going to unload your 400mm F/2.8.
I decided when I bought the d7200 that I would use it for 3 years before going full frame. It's been 2 years and I have about 40,000 clicks on the shutter. The camera is fine and all but I shoot a lot of low light work and I'm thinking the full frame and better ISO ability of the 850 would help me. I also think it focuses faster than the d7200 which would help with my runway work.
I may rent one and see if it improves my work any.
You are right on target with your thinking. My base camera was the D7100 and I found its limitation at high ISO, DR, and focus speed often showed in low light. When I added the D750 I was blown away by image quality so I know adding the D850 to your tool belt will make a gigantic step forward. You will not be sorry.
D750 & D7100 | 24-70 F2.8 G AF-S ED, 70-200 F2.8 AF VR, TC-14E III, TC-1.7EII, 35 F2 AF D, 50mm F1.8G, 105mm G AF-S VR | Backup & Wife's Gear: D5500 & Sony HX50V | 18-140 AF-S ED VR DX, 55-300 AF-S G VR DX | |SB-800, Amaran Halo LED Ring light | MB-D16 grip| Gitzo GT3541 + RRS BH-55LR, Gitzo GM2942 + Sirui L-10 | RRS gear | Lowepro, ThinkTank, & Hoodman gear | BosStrap | Vello Freewave Plus wireless Remote, Leica Lens Cleaning Cloth |
Maybe when I sell the D4 and 400/2.8... Probably order mine for a Christmas present to myself... LOL
I keep thinking about sell my 400/2.8 but I'll never get another and when you need it you just need it. Seriously, I've found it to be fabulous at pano work. W/the new D850 it could be easier to focus stacking for an added depth of field. I'm going to hold pat with my current bodies for another year or two to see what shakes out.
Currently using D3s & mostly D810. I want to get back to 10 year body cycles. The D3s should be replaced in Dec 2019.
Most would argue that a 10 year body cycle in the current digital age is too long because there is just too much progress being made every few years. Probably skipping one iteration is a better plan. If you have a D3, skip D4 and get a D5 or if you have a D800, skip D810 and get D850 etc. But many people feel each iteration is worth the change for some new feature it offers which they want to use. For example, the silent shutter option alone in the D850 for wedding photographers may be worth changing from a D810. Progress seems to be happening faster and faster these days and significant new technologies appear frequently. Next year we will all lust after the "what you see is what you get" EVF in the new Nikon mirrorless bodies. As higher ISO image quality and EVFs and LED monolights all improve many portrait photographers will want to "see" their lighting set up directly before they press the shutter instead of previsualizing what the strobe will produce and checking it after they press the shutter. We will probably be at that stage in 2 to 3 years and may be there next year with a Nikon FX mirrorless body using the sensor in the D850 which I think is going to produce adequate image quality at ISO 2,000 with LED monolights. It was not like that back in the film days when a camera body was fine for 10 years and improvements in film stock were slow to appear.
I was thinking every second iteration would do for me, except I want a 2nd body and I wore out my D800 - so it is not much of a 2nd body.
Not sure if I will get a second D850 or wait for a D850s. And then if Nikon produces a D850 mirrorless, that could be an option too if the ergonomics are the same.
I now have a D800, a D810, and the fabulous D850 (I am trying to sell the D800). One thing that is starting to irk me is that the button layout is different between the D810 and the D850. I am really thinking that it would be nice with two cameras with somewhat identical layouts.
Unfortunately that means the D500, D5, or another D850 - and I don't want a DX camera, the D5 is too expensive, and then it's another D850. I guess I will just stick with my D810, but it still irritates me:)
I have a D500 and love it but the next one will be a D850 because of what it does that the D500 don't and the control layout is nearly the same. I will shoot them together and in a pinch one can back up the other. D5 is out of my price range also.
Comments
I am not sure exactly how many megapixels a diffraction limited lens at f/8.0s, but look at the Imatest results from PhotographyLife for the 58mm 1.4G, 105 1.4E and 400 2.8E. I just purchased these three lens so I am experimenting with them on the D850.
https://photographylife.com/reviews/nikon-58mm-f1-4g
https://photographylife.com/reviews/nikon-105mm-f1-4e
https://photographylife.com/reviews/nikon-400mm-f2-8e-vr
At f/8.0, they all achieve a score in the centre of around 2590 (note how the 105 beats the other two into the corners). At f/11.0, they are all around 2208.
Further note where these lenses are diffraction limited, meaning that the resolution of the lens seems to be limited by diffraction.
Tis seems to be f/8.0 for the 58. Resolution increases at f/5.6, but not by much so it does not seem to be due to diffraction.
For the 105, this seems to f/4.0 for the 105, which is quite amazing. The 105 is blowing me away. This might be the sharpest lens that Nikon makes shorter than the 400 2.8E - though I expect the new 200 2.0E and 300 2.8E to regain their titles. It is certainly the second sharpest in my bag besides the 400. Look at this lens at f/2.8 where it is at its sharpest, though just barely beating itself at f/4.0. At f/2.8, it is sharper then the 400 2.8E. However, this is not an apples to apples comparison because if you compare them both wide open, the 400 wins (but holy cow, by less then I expected - and the 105 at 1.8 is sharper in the centre than the 400 at 2.8).
For the 400, you will note that it tops out at 3,000 on Imatest. I wondered about this until I started talking to Imatest about buying their software and charts. I asked them if the charts were good enough for the 100 megapixels that the Phase 1 can achieve. The answer is no, they are not sharp enough, but they also said that their charts are not sharp enough for the D850 (they are working on a better chart). I suspect that PhotographyLife may have been working with a chart that topped out at 3,000 lines.
But to the point in Thom's blog, PhotographyLife was likely using a 36mp D800 sensor in his tests and the sensor seems to be capable of achieving an Imatest score of 3240. Since lens can just barely or not quite achieve this level of resolution at f/5.6, if you are shooting with a D850 or even a D800 at f/8.0 or less, then you are achieving less resolution than the camera is capable of "DUE TO THE LAWS OF PHYSICS".
With the D850, my new mantra is f/4.0 or wider on my good lenses unless I am willing to throw away resolution.
What would be the point of putting a 28-300 on a D850 and shooting large basic ...no point .
I dont think I would have got such sensible and helpfull answers on the other N sites I am banned from for making "usefull comments"
1.4: 2290
1.8: 2695
2.0: 2917
2.8: 3240
4.0: 3226
5.6: 2985
8.0: 2590
11:0 2203
16:0 1725
Another data point is Imatest's statement to me that the test charts (and I have access to the same test charts as PhotographyLife) do not have enough resolution to test the D850 (when they figure that out, I will buy the test chart).
So my hypotheses are:
-that the best a 36mp sensor can do is roughly 3,200 and it is unable to exploit the perfect lens that I describe above at a wider aperture than say f/4.5 (I am interpolating the curve and making a guess on why there is not better performance than 3226 at f/4.0: lens performance, diffraction, Imatest performance are the candidates).
-since your D7100 has a higher pixel density, it should be able to do better ( exploit a wider f-stop), but if we are going to use Imatest to test it, we need a better chart than is currently available
Note: Given how diffraction works, pixel density is the determining factor, or rather a proxy for the determining factor, as the lens is the determining factor.
-since your D750 has a lower pixel density, this limit is somewhere between f/8.0 and f/11.0
I want to test the above hypotheses and as I have alluded to, I have been talking to Imatest about buying their software and test charts but am stuck because they don't have a chart that can properly test the D850 (or a Phase One for that matter). But given that I have the 105 and 400 I might be able to explore and test aspects of the above hypotheses. I have DSLR's with three pixel densities, namely the D5500, D800 and D850. If I can get my hands on a D750 or even a D600/610 I can round it out.
I just need to find the time. I have been very busy professionally and have just moved into the new house and a baby girl was due October 5th - so we are waiting around for that to happen. But this is an excellent project for the fall.
I wouldn't find myself using stacking often (most landscape situations have some movement), so we are back to legacy scheimflug techniques to maximize results with highest MP camera.
Revised 24, 45-50, and say 85-90mm t-s lenses are desired. Otherwise not sure there is enough resolution benefit to warrant a d850.
But, the ability to manage the capabilities of current equipment, do a decent job with composition, post processing, final presentation, these are what, IMO, make the shot.
No doubt, the D850 with its exceptional versatility makes this easier, but I have often found myself lost in the process, forgetting the final goal.... oh well.
Probably order mine for a Christmas present to myself... LOL
I may rent one and see if it improves my work any.
Denver Shooter
|SB-800, Amaran Halo LED Ring light | MB-D16 grip| Gitzo GT3541 + RRS BH-55LR, Gitzo GM2942 + Sirui L-10 | RRS gear | Lowepro, ThinkTank, & Hoodman gear | BosStrap | Vello Freewave Plus wireless Remote, Leica Lens Cleaning Cloth |
Currently using D3s & mostly D810. I want to get back to 10 year body cycles. The D3s should be replaced in Dec 2019.
framer
Not sure if I will get a second D850 or wait for a D850s. And then if Nikon produces a D850 mirrorless, that could be an option too if the ergonomics are the same.
Unfortunately that means the D500, D5, or another D850 - and I don't want a DX camera, the D5 is too expensive, and then it's another D850. I guess I will just stick with my D810, but it still irritates me:)