There is something about the Z6 images compared to my d500

[JasonTheBirder]

I tell you one thing - If I could have a Z6 with the D500 focus snap and responsiveness, I would be a very very happy man, with or without the AI stuff. Sadly, I cannot afford a £4000 body.

Could the different be sensor or colour depth or something or dynamic range? I definitely notice that the highlights in the sun are better managed on the Z6 images, retaining more colour details - where as the D500 has more trouble, with more contrast. I wonder if this also contributes.

I shoot both cameras and I do get sharper shots with the Z6 more often. I think it's the shutter on the D500...it occasionally displays shutter shock (very occasionally).

 

[Ad12]

I got a Z6 with an adapted 100-400C and 150-600C. I loved it, but the AF for small birds and BiF was annoying me. I grabbed a d500, and it was night and day as far as focusing. I recall trying to shoot small dogs running head on towards me. The Z6 couldn't cope. The d500 had a 95% plus hit rate, really amazing. Since then, i've used the Z6 for travel, landscapes, night, and the D500 for all things wildlife.

Today I booked my first ever hide. Aside from my feelings of unease with the concepts of paid for hides that have food placed (I won't do one again despite the incredible images....), when I review the images, the Z6 photos are insanely wonderful, they pop so so well.

I had the D500 with 100-400, and the Z6 with the 150-600C. Same FOV, different bodies and capabilities.

WYSIWUG - I adore this for mirrorless, I just used the back screen. As you'd expect, the subject separation with the Z6 was so beautiful, much more so than the still wonderful D500 images. Also, when I raised shutter speed for movement, the Z6 as you'd expect, handles higher ISO's with ease, it doesn't even notice ISO 8000.

Perhaps it's the import settings in LR (Z6 imports with settings itself, D500 doesn't so I have to set presets myself), though I feel it's something else. The images from the Z6 are just popping. They are also sharper consistently, which makes me wonder if whilst the D500 images are sharp for sure, perhaps the focus adjustment stuff is not 100% nailed down. The D500 though is not out of focus - it's in focus, just less crisp to my eye.

I tell you one thing - If I could have a Z6 with the D500 focus snap and responsiveness, I would be a very very happy man, with or without the AI stuff. Sadly, I cannot afford a £4000 body.

Could the different be sensor or colour depth or something or dynamic range? I definitely notice that the highlights in the sun are better managed on the Z6 images, retaining more colour details - where as the D500 has more trouble, with more contrast. I wonder if this also contributes.

The Z6 is a full frame camera while the D500 is an APS-C camera and a larger sensor generally gives you less noise at high ISO and more dynamic range.

The difference you claim to experience in sharpness could be either that you're using different lenses or focus could be a bit off in your D500 images. It could also be that the Z6 has IBIS whereas the D500 doesn't so you might be getting some motion blur. Or are you perhaps shooting JPG? If so it could simply be because of different rendering and/or settings.

I had not considered ibis as a factor. Interesting.

 

[Ad12]

I tell you one thing - If I could have a Z6 with the D500 focus snap and responsiveness, I would be a very very happy man, with or without the AI stuff. Sadly, I cannot afford a £4000 body.

Could the different be sensor or colour depth or something or dynamic range? I definitely notice that the highlights in the sun are better managed on the Z6 images, retaining more colour details - where as the D500 has more trouble, with more contrast. I wonder if this also contributes.

I shoot both cameras and I do get sharper shots with the Z6 more often. I think it's the shutter on the D500...it occasionally displays shutter shock (very occasionally).

 

[Ad12]

The last two comments are interesting. I just looked on Lightroom for a little while. The Z6 shots are consistently very sharp for a stationary subject. The D500 has much much more variability, but can also yield very sharp results. Perhaps it is mirrorless focus accuracy, perhaps IBIS, perhaps the shutter. Hmmm. It does seem mirrorless offers greater potential for consistency with the mirror gone.

 

[JConrad]

I don't know what if anything will make the af and response characteristics of mirrorless compare on even keel with dslr's. There is a complete different situation going on in that comparison. But the image quality difference which seems like should not really be there, I have a personal crazy notion not accepted by most, that there is some AI involved in the processing of the images in the latest and greatest mirrorless that was not available in dslr's. It's almost like the mfg's are taking lessons from the cell phone mfg's. In my experience thus far with Nikon mirrorless, the image quality is just entirely too good and consistent across the entire range of lenses to be unassisted in some manner. But the response and quick handling of my dslr's still wins in many instances for me personally and in those situations my camera choice is always a dslr.

This isn't a mirrorless thing...it's a Nikon mirrorless thing until the Z8 and Z9. I have both a Canon R5 and R8, as well as a Z5, and I shot Sony for quite a while until 2021. The Z cameras (again, until the Z8 and Z9) just are not on par with the competition when it comes to autofocus, and it's true even in single shot. My Canon bodies will focus in super dim light with laser quickness and precision, along with AF tracking in action that easily bests any DSLR I've used. Meanwhile the Z5 is fairly pokey, especially in lower light, and just isn't even close to the recent Canon and Sony bodies. It is still accurate for stationary subjects, but it's really is quite a bit behind what other mirrorless cameras have been doing for a while.

Hopefully Nikon will filter that Z9 / Z8 AF down to future iterations of the other bodies. Canon and Sony are continuing to put pressure on that, as something like the R8 has an AF system that is simply incredible (and is even better than my R5, which is already outstanding).

I still like the camera a fair bit, and it’s very capable.

 

[Bob A L]

I don't really have any AF speed problems with my Z camera, the af is pretty quick. The problem for me is the evf lag which is part of any mirrorless.

 

[FingerPainter]

No, it's just simple physics. The larger senor results in a sharper image.

No, sharpness is an almost entirely lens sided characteristic and has very little to do with the sensor. Basically the only thing that affects sharpness on a sensor is whether it has an AA filter or not.

You are not talking about the same thing I am talking about. You are talking about the sharpness of the lens, I am talking about the sharpness of the whole image.

The D500 sensor doesn't have an AA filter while the Z6 does, so if anything you should be able to get slightly sharper photos with the D500.

You cannot. And if you tried it, you would see that you didn't.

If one gets less sharp photos with it it's either because a less sharp lens has been used or perhaps focus is a bit off. It could also be that the Z6 has IBIS and he's getting some motion blur in his D500 images.

Nope. You seem to fail to understand what the sharpness of an image is.

Sharpness is measured in line pairs over a distance. Sharpness of a lens is measured in line pairs per mm (lp/mm). Sharpness of an mage is measured in line pairs per picture height lp/ph).

Let's imagine using the very same hypothetical lens on a D500 and a D850.. I suggest the D850 because it has a very similar pixel pitch as the D500. Now our hypothetical lens can resolve, on an optical test bed, an average of 50 line pairs per millimetre over the height of the image circle. So how many line pairs are begin cast over the height of the sensor? Well, that depends on how high the sensor is. On the D500's DX sensor it will cast about 800 line pairs. That's because it is casting 50lp/mm over about 16mm. Over the D850 FX sensor, the very same lens casts about 1,200 line pairs.

So even before digitization, the image cast on the FX sensor has approximately 50% more line pairs per picture height. The digitization process will reduce the achieved resolution, but since the two sensors have similar pixel pitch the reduction on each sensor will be very similar.

Now, reality does get a little more complicated than this hypothetical example. One complication is that lenses tend to produce an image circle that gets decreasingly sharp as one approaches the edge of the circle from the middle. In the centre of the image circle, the lens will cast the same number of line pairs per millimetre on senors of either size. But at the edge of the DX image circle the lens will be casting a higher number of line pairs per millimetre than it will at the edge of the FX image circle.

The average image sharpness over the image circle will be about (Image height in millimetres x average of centre and edge sharpness. If C is centre sharpness of the lens in lp/mm, D is the sharpness at the DX edge and F is the sharpness at the FX edge, the number of line pairs cast over the FX sensor height will be better than 24mm x (C+F)/2 and the number of line pair cast over the DX image height will be better than 16mm x (C+D)/2.

Using basic algebra we can see what condition would have to be true for the DX image to be sharper (have more line pairs per picture height).

Simplifying, we want to see the relation of F and D where

12C + 12F < 8C + 8D

4C + 12F < 8D

C/2 + 3F/2 < D

So the DX image will only be sharper than the FX image when the DX edge sharpness in lp/mm is greater than the sum of half the centre sharpness plus three halves of the FX edge sharpness. There may be a few lenses where edge sharpness drops off this badly from DX edge to FX edge, but they are very rare.

In general, FX images from a given lens are significantly sharper than DX images from the very same lens (which obviously has the same sharpness itself regardless of which sensor you put it in front of).

Sensor size has a significant effect on image sharpness. Another sensor parameter that affects image sharpness is pixel count.

Imagine a 6-pixel sensor (not a 6MP sensor, a sensor with just 6 pixels). It should be obvious that the highest number of line pairs per image height it can digitize is one. Imagine the opposite case: a sensor with an infinite number of pixels. It should be obvious that the highest number of line pairs per image height it can digitize is the number of line pairs cast on the sensors by the lens. In between these two extremes, the maximum lp/ph achievable is a convolution of the pixel count and the lens' analog resolution.

So both the size and pixel count of the sensor influence how sharp the resulting image will be.

 

[FingerPainter]

The last two comments are interesting. I just looked on Lightroom for a little while. The Z6 shots are consistently very sharp for a stationary subject. The D500 has much much more variability, but can also yield very sharp results. Perhaps it is mirrorless focus accuracy, perhaps IBIS, perhaps the shutter. Hmmm. It does seem mirrorless offers greater potential for consistency with the mirror gone.

Beware of attributing to one factor (mirrorless vs mirror) what could be attributable to other factors (sensor size and pixel count).

To see if is being mirrorless that gives the Z6 the advantage, compere Z50 output to D500 output. These two cameras have the same sensor size and pixel count. I think if you make this comparison with an AF-S lens on an FTZ adapter shooting easy-to-focus-on subjects, you will see little difference in image quality.

I think most of the difference you see between Z6 images and D500 images is due to the difference in size of the sensors and a much smaller portion is due to the Z6 having a slightly higher pixel count.

 

[JayPhizzt]

No, it's just simple physics. The larger senor results in a sharper image.

No, sharpness is an almost entirely lens sided characteristic and has very little to do with the sensor. Basically the only thing that affects sharpness on a sensor is whether it has an AA filter or not.

You are not talking about the same thing I am talking about. You are talking about the sharpness of the lens, I am talking about the sharpness of the whole image.

The D500 sensor doesn't have an AA filter while the Z6 does, so if anything you should be able to get slightly sharper photos with the D500.

You cannot. And if you tried it, you would see that you didn't.

If one gets less sharp photos with it it's either because a less sharp lens has been used or perhaps focus is a bit off. It could also be that the Z6 has IBIS and he's getting some motion blur in his D500 images.

Nope. You seem to fail to understand what the sharpness of an image is.

Sharpness is measured in line pairs over a distance. Sharpness of a lens is measured in line pairs per mm (lp/mm). Sharpness of an mage is measured in line pairs per picture height lp/ph).

Let's imagine using the very same hypothetical lens on a D500 and a D850.. I suggest the D850 because it has a very similar pixel pitch as the D500. Now our hypothetical lens can resolve, on an optical test bed, an average of 50 line pairs per millimetre over the height of the image circle. So how many line pairs are begin cast over the height of the sensor? Well, that depends on how high the sensor is. On the D500's DX sensor it will cast about 800 line pairs. That's because it is casting 50lp/mm over about 16mm. Over the D850 FX sensor, the very same lens casts about 1,200 line pairs.

So even before digitization, the image cast on the FX sensor has approximately 50% more line pairs per picture height. The digitization process will reduce the achieved resolution, but since the two sensors have similar pixel pitch the reduction on each sensor will be very similar.

Now, reality does get a little more complicated than this hypothetical example. One complication is that lenses tend to produce an image circle that gets decreasingly sharp as one approaches the edge of the circle from the middle. In the centre of the image circle, the lens will cast the same number of line pairs per millimetre on senors of either size. But at the edge of the DX image circle the lens will be casting a higher number of line pairs per millimetre than it will at the edge of the FX image circle.

The average image sharpness over the image circle will be about (Image height in millimetres x average of centre and edge sharpness. If C is centre sharpness of the lens in lp/mm, D is the sharpness at the DX edge and F is the sharpness at the FX edge, the number of line pairs cast over the FX sensor height will be better than 24mm x (C+F)/2 and the number of line pair cast over the DX image height will be better than 16mm x (C+D)/2.

Using basic algebra we can see what condition would have to be true for the DX image to be sharper (have more line pairs per picture height).

Simplifying, we want to see the relation of F and D where

12C + 12F < 8C + 8D

4C + 12F < 8D

C/2 + 3F/2 < D

So the DX image will only be sharper than the FX image when the DX edge sharpness in lp/mm is greater than the sum of half the centre sharpness plus three halves of the FX edge sharpness. There may be a few lenses where edge sharpness drops off this badly from DX edge to FX edge, but they are very rare.

In general, FX images from a given lens are significantly sharper than DX images from the very same lens (which obviously has the same sharpness itself regardless of which sensor you put it in front of).

Sensor size has a significant effect on image sharpness. Another sensor parameter that affects image sharpness is pixel count.

Imagine a 6-pixel sensor (not a 6MP sensor, a sensor with just 6 pixels). It should be obvious that the highest number of line pairs per image height it can digitize is one. Imagine the opposite case: a sensor with an infinite number of pixels. It should be obvious that the highest number of line pairs per image height it can digitize is the number of line pairs cast on the sensors by the lens. In between these two extremes, the maximum lp/ph achievable is a convolution of the pixel count and the lens' analog resolution.

So both the size and pixel count of the sensor influence how sharp the resulting image will be.

Sure, but I only base my claim on what I've actually seen with my own eyes and I've never seen any difference in sharpness between sensors aside from like I said slightly sharper images from those that lack an AA filter.

 

[Jack Hogan]

[...] I've never seen any difference in sharpness between sensors aside from like I said slightly sharper images from those that lack an AA filter.

Have you ever shot with, say, a 1" sensor?

For birding it's hard to argue against pixels on the bird, assuming equivalent lenses. The smaller the pixels, usually the more of them on the bird, the sharper the resulting image.

AA or not AA is an interesting question. I've come full circle and now wish Nikon FF landscape cameras came with one. Once you start seeing aliasing you can't unsee it and then you see it everywhere, a bit like oversharpening halos introduced by some software. For birding it may be less critical because there are probably many more blurring factors at play, like camera/lens shake and subject motion.

Considering aliasing introduces another sensor variable to sharpness: effective pixel aperture size, see for instance the link below. Smaller pixel aperture means yet sharper but with more aliasing.

https://www.strollswithmydog.com/nikon-z7-insane-sharpness/

Jack

 

[areallygrumpyoldsod]

I don't know what if anything will make the af and response characteristics of mirrorless compare on even keel with dslr's.

Well please tryout one for a while and see.

There is a complete different situation going on in that comparison. But the image quality difference which seems like should not really be there, I have a personal crazy notion not accepted by most, that there is some AI involved in the processing of the images in the latest and greatest mirrorless that was not available in dslr's.

No AI in ANY camera sold today including Cell Phone cameras. What there are are algorithms populated by deep learning that have been tuned to identify certain subjects and to focus on key elements of them - eye, mid-bonnet, cockpit, etc.

The AI reference implies the camera does something to the image -- well not beyond optical correction and noise reduction and perhaps dome different minor adjustments in picture controls.

It's almost like the mfg's are taking lessons from the cell phone mfg's.

NOPE -- what one can do on a cell phone is load an app that messes with images - some of these apps do use the insane tools that appear like AI, but again are not.

In my experience thus far with Nikon mirrorless, the image quality is just entirely too good and consistent across the entire range of lenses to be unassisted in some manner. But the response and quick handling of my dslr's still wins in many instances for me personally and in those situations my camera choice is always a dslr.

I would hope that you get to try a Z8/Z9 "for a while" and see what you see. Then draw your own conclusions.

 

[areallygrumpyoldsod]

For birding it's hard to argue against pixels on the bird, assuming equivalent lenses. The smaller the pixels, usually the more of them on the bird, the sharper the resulting image.

Well not to infinity -- very real atmospheric effects (primarily heat haze but not just this), limitations on the resolving capabilities of lens (yes also real) AND ultimately diffraction do come into play -- and increasingly quickly with higher and higher pixel densities. AS does read noise.

Those striving for a 30+MP DX probody are in my view onto a losing wicket. Particularly if they expect it with any sensor in use today -- my personal view is a Z6iii with 26MP and 6.1k video is far more than the vast majority want or need, but of course the die hard BIF crew will howl until the end of time if this is "all they get". Remember folks such a camera will only be cheap if it is bought by a vast number of folk so compromise is your friend.

Until 2022 - I shot almost exclusively DSLRs since 2002 (my last were the D850, D5 and D500). I added first a Z7 and then a Z6ii, but was not excited about shooting until the Z9 was in my hands -- and it felt like the shackles had been cast off.

As great the D850, is and it is, or the Z6ii is, and it is as well, the results I am getting from the Z9 and now also the Z8 greatly superseded anything I achieved with these earlier models. WHY - it is simple more images with my subject in focus.

Back in 2017 I took a D5, D500 and pair of D810s to africa (all my D500/D810s had battery grips and EN-EL18 batteries on board) and YES - the crop factor of the D500 and the slightly higher fps gave great results -- but I have to say I noted at the time that I saw a "roughness" to the images from the D500 that simply was not present in those from the D810 and D5. (I only shot lossless raw, only use Camera profiles not adobe in LRC and ACR), but this was easily fixed with a wee bit of NR. However, I also found in low light the D810 appeared to significantly outperform the D500; and the D5 was well ahead of both. Images were just less ROUGH and easier to work when recovering dark shadows etc..

Even the AF in the top DSLR's had their own foibles - in ultra low light one routinely ONLY used the very central AF point and the camera took focus on whatever was under it. No possibility of focus recompose - just focus, shoot and crop to compose later.

AA or not AA is an interesting question. I've come full circle and now wish Nikon FF landscape cameras came with one. Once you start seeing aliasing you can't unsee it and then you see it everywhere, a bit like oversharpening halos introduced by some software. For birding it may be less critical because there are probably many more blurring factors at play, like camera/lens shake and subject motion.

The AA debate and moire is a long one - the D800E was the first 36MP Nikon to be offered alongside the same camera, with the same sensor, except that the its anti-aliasing filter was removed and here came Moire etc...

AND the D810 was launched without an AA filter. Shortly after Nikon added the D810A, where Nikon changed the IR-cut filter to pass the hydrogen alpha line; it's four times as sensitive at 656 nanometers as the regular D810. However because of this filter modification, the D810A gave poor color rendition for normal photography, so one still needed a regular DSLR for regular photography. [Folk are paying for these changes to be applied to more modern cameras, but they loose Phase detect AF as part of the conversion].

Considering aliasing introduces another sensor variable to sharpness: effective pixel aperture size, see for instance the link below. Smaller pixel aperture means yet sharper but with more aliasing.

True but...... OK it was a price many of us were willing to pay and fix in post.