I'm creating this thread to hold all the random tests I'll be performing on my Z7. I plan to test areas I find important to me but I'm also open to requests, time-permitting.
To start things off, here's a video I recorded demonstrating the clarity and refresh rate of manual focusing at various magnification levels. It's also a demonstration of how sensitive the Z7's digital rangefinder is. This was done with yesterdays V1.01 firmware. Be sure to watch in 1080 HD:
Observations:
Excellent LV magnification clarity, best I've seen on a Nikon body
Very good refresh rate/lag, and noticeably improved vs V1.00 firmware
Digital rangefinder is very sensitive, almost too sensitive (conservative) on when it displays the center dot confirmation. Oddly I couldn't get the center dot confirmation at the second-level and last-level magnifications - I made sure to have the AF box on the area of text prior to switching to MF, not sure if it's sampling some other area for the rangefinder.
EVF is much sharper than LCD for focusing. For example I was able to successfully focus to the unmagnified view on the EVF but not on the LCD
For comparison here is the same test I just performed on the A7rIII:
Observations:
Very similar performance to the Z7 in terms of clarity and refresh rate
The EVF clarity between the two cameras was very close
Which lens you were using for focus confirmation test? Since I can’t use non native Nikon lens or even ZF.2 lens to get focus confirmation. I did a casual test with 24-70g and 58g there seems quite a bit play window if you do infinity test. I’d say it was not tighter than D850. Based on that I guess it is pure PDAF algorithm.
For close range like you did, the DOF is thin. Many Nikon body will give similar tight range, no?
Also I read from DPreview a while ago that Sony don’t have full resolution(means 3.7M) view on EVF during shooting. However if you use magnification at 10X, EVF will use its full 3.7M resolution. And in image review, it will show full resolution. I don’t have Sony anymore, can you confirm that?
Many thanks and look forward to see more work from you.
BTW, do we know more about banding?
snapsy wrote:
I'm creating this thread to hold all the random tests I'll be performing on my Z7. I plan to test areas I find important to me but I'm also open to requests, time-permitting.
To start things off, here's a video I recorded demonstrating the clarity and refresh rate of manual focusing at various magnification levels. It's also a demonstration of how sensitive the Z7's digital rangefinder is. This was done with yesterdays V1.01 firmware. Be sure to watch in 1080 HD:
Observations:
Excellent LV magnification clarity, best I've seen on a Nikon body
Very good refresh rate/lag, and noticeably improved vs V1.00 firmware
Digital rangefinder is very sensitive, almost too sensitive (conservative) on when it displays the center dot confirmation. Oddly I couldn't get the center dot confirmation at the second-level and last-level magnifications - I made sure to have the AF box on the area of text prior to switching to MF, not sure if it's sampling some other area for the rangefinder.
EVF is much sharper than LCD for focusing. For example I was able to successfully focus to the unmagnified view on the EVF but not on the LCD
For comparison here is the same test I just performed on the A7rIII:
Observations:
Very similar performance to the Z7 in terms of clarity and refresh rate
The EVF clarity between the two cameras was very close
Which lens you were using for focus confirmation test?
Hi Michael, this was with the Z 24-70 f/4
zhangyue wrote:
Since I can’t use non native Nikon lens or even ZF.2 lens to get focus confirmation.
Yep, I've read this as well. I don't have an FTZ adapter to check myself.
zhangyue wrote:
I did a casual test with 24-70g and 58g there seems quite a bit play window if you do infinity test. I’d say it was not tighter than D850. Based on that I guess it is pure PDAF algorithm.
Thanks, I'll do a check of the rangefinder @ infinity tomorrow and let you know how well it works on the Z 24-70.
zhangyue wrote:
For close range like you did, the DOF is thin. Many Nikon body will give similar tight range, no?
Agreed, the rangefinder might be more sensitive at closer distances due to the DOF. Will find out tomorrow.
zhangyue wrote:
Also I read from DPreview a while ago that Sony don’t have full resolution(means 3.7M) view on EVF during shooting. However if you use magnification at 10X, EVF will use its full 3.7M resolution. And in image review, it will show full resolution. I don’t have Sony anymore, can you confirm that?
I'm not sure if I'll be able to determine if Sony is using the full res for its unmagnified view via visual means. Comparing the A7rIII and Z7 side-by-side the EVFs looked comparable to me in terms of resolution at all magnification levels including unmagnified.
zhangyue wrote:
BTW, do we know more about banding?
We do. I've done a ton of experiments on the banding. You can read about them on the dpreview thread here (I'm 'horshack' there).
I ran into an issue that's likely academic in practice but thought I should share it in case anyone has a real-world scenario I can't think of where they'd run into it in actual shooting. The Z7's WB starts shifting slightly at -3EV underexposed and shifts significantly at around -6EV. Here's an animate GIF demonstrating the WB shifting from 6300 to 4100 between -5EV and -6EV underexposure:
RobCD wrote:
Hi Snapsy, great information. Do you have any comparisons to any other high quality standard zooms or is that on your to do list by chance ?
Thanks Rob. I don't have an FTZ adapter so I can't check any F-Mount lenses. I might compare it to some Sony lenses though if I get the time.
Tip/Trick: Normally the digital rangefinder is only visible in MF mode - when you switch to AF-S mode the digital rangefinder disappears. You can still override AF-S and focus manually by turning the focus ring but the rangefinder wont be available to aid in focus. I accidentally discovered a way to get it to show in AF-S mode. Do one of the following:
1. Press and hold AF-ON before turning the focus ring (downside is that it triggers AF first before you start manually focusing)
2. Or, if you don't want the AF to be triggered, start turning the focus ring first and while still turning it press and hold AF-ON button
Doing either will show both the digital rangefinder and the on-screen focus distance scale.
This trick also temporarily shows focus peaking highlights in AF-S if that's enabled as well.
A corollary to this is that the camera wont focus while the focus ring is being manipulated. This may seem obvious since the camera likely needs to disengage the AF to prevent fighting against the motor. However, due to the slight lag between the focus ring and EVF/LCD, the camera will still think you're manipulating the focus ring for a short time after releasing it, meaning if you press the AF-ON button soon after releasing the ring the camera won't actually focus. I found this yesterday when I was intentionally racking focus between multiple shots in AF-S mode and the camera would intermittently fail to focus after racking. The workaround is to wait a beat after releasing the ring before pressing AF-ON.
Taking a break from the indoor banding experiments today, I decided to try some real-world scenes throughout the day, starting from sunrise, to midday, to sunset. The banding proved extremely difficult to manifest under nearly all those conditions. It required very specific placement/size of the sun relative to a darker areas and even then the banding was usually intermittent. I wasn't able to induce outdoor banding at all if the sun wasn't in the frame. For example:
So in practice it might be less of an issue than feared. One suggestion I have is to take multiple shots of important scenes, since many of the scenarios that produce the banding do so only intermittently, so having multiple exposures gives you a chance you'll have an exposure without it.
thank you very much for taking the time to do this test and share it with us. I was trying to buy a Nikon Z7 and thanks to your article I have the answers to all my questions.
Ran into an issue yesterday using the electronic shutter with fast shutter speeds - a green band forms in the shadow region at the same horizontal plane as an overexposed highlight. Easily avoidable but thought I should document it anyway. It appears to be some type of sensor blooming issue but specific to the electronic shutter.
Here's what I see at 1/6400 on a +5EV push on Z7, comparing the electronic shutter to the fully mechanical one (EFCS tops out at 1/2000 so can't test it):
The Z7's green band gradually diminishes as I reduce the shutter speed (at the same absolute exposure, ie changing the aperture to offset shutter speed).
Here's the same scene and parameters on the A7rIII. You can see a tiny hint of the same phenomena:
Shot the Z7 all day two days ago and one thing I found annoying is a loud clicking noise when switching back and forth between shooting mode and image review mode. It only happens when IBIS is enabled, so I'm pretty certain the noise is related to Nikon engaging/dis-engaging the hard-lock feature of its IBIS mechanism. Not sure why they decided to lock the sensor down in review mode.
When image review is on it takes the camera approx 1.5 seconds to display the last shot on the LCD. To see if this was unique to playback or just a long image processing pipeline time I then turned image review off and watched how long it took for the camera to access the XQD card to write the last shot. By sight it appeared to be just under 1 second, which seems long for a Nikon body. For the camera to take that long it seems to be there must be some CPU involvement in the image processing pipeline, since Expeed wouldn't take that long. Btw the processing time seemed independent of image type - happens for both RAW and JPEG.
Starting with the Z7, Nikon is embedding lens correction profiles for Z lenses in the Z7 raws. ACR automatically applies those built-in corrections (screen shot), with no way to disable. However you can disable them by stripping out the profile form the raw before importing into ACR/LR, which can be done via exiftool -OpcodeList3= <rawfilename> (space after equal sign).
Here's an animated GIF I created showing what the 24-70 looks like at 24mm, 35mm, 50mm, and 70mm without and without the embedded profile, as processed by ACR:
snapsy wrote:
Starting with the Z7, Nikon is embedding lens correction profiles for Z lenses in the Z7 raws. ACR automatically applies those built-in corrections (screen shot), with no way to disable. However you can disable them by stripping out the profile form the raw before importing into ACR/LR, which can be done via exiftool -OpcodeList3= <rawfilename> (space after equal sign).
That's pretty cool, hope it works for the Z6 if it's the same.
Thanks for all of the time and resources you put into all of this testing and sharing your findings. I'm a "turn the key and hit the gearbox". I could never do all of what you've done with this and the other testing and research you've posted on here.
I too am enjoying your work. I have been away for a year or two but this body/system has me interested. I have not yet found a Sony that I just love to shoot. That of course is the Indian and not the arrow. I could put the time in but with Nikon, I started with.....and will probably end with one in my bag. I see this body as the start of their immediate future and honestly I did not expect their first effort to come this close to the best of breed.
I will be anxious to hear when you finish this up if the z is staying in your bag!
Thanks again!
Greg
Long-exposure comparison between the Z7 and A7rIII...
Using a Z 24-70 f/4 and FE 24-105 f/4, both @ 24mm f/4, I took two long exposures, one indoors @ 30 seconds and outdoors of stars @ 15 seconds. LENR disabled, WB matched, lens profiles disabled, all NR disabled, Adobe Color profile. The Nikon lens appears to have a brighter t-stop than the Sony, thus the brighter exposure.
Here's a 100% crop animation of the 30-second indoor exposures. Exposure adjusted +5EV in ACR:
Here's a 100% crop animation of the 15-second outdoor star exposure There's a ton of light pollution in my area so it's not the best for star shots. Exposure adjusted +3EV in ACR:
So I wondered, what will a 24-70Z image look like if I take a photo with the zoom ring set below 24mm, ie in the area considered "locked" (collapsed) by the camera. The only way to do this is to electrically disconnect the lens from the body by slightly turning it off the mount lock, otherwise you'll get the "Before taking photos, rotate the zoom ring to extend the lens" message.
First, here's an image at 24mm f/4, focused a few feet away. The lens correction profile has been deactivated for processing in ACR:
If I turn the lens below 24mm the image immediately goes OOF. So to get it in focus I set the focus to MFD prior to electrically disconnecting the lens. Then I turned the focal length ring as far below 24mm as I could and still get focus, which turned out to be about half an inch from the "24" label on the lens. Here's what the result looks like:
?