story_teller wrote:
The Zf looked intriguing to me as well. I wanted better AF than my Z6, but after holding the Zf for a few minutes, the body was too thin for my hands. Glad I waited for the Z6III. It’s a great replacement for my Z6 and very similar to my Z8, so it will serve as a backup and for low light shooting (events, etc.).
I actually don't have any Nikon gear at the moment and I'm only interested in it because it may solve my issues with critically focusing my M lenses on my Leica and the hassle of focus peaking and magnifying on other cameras. Curious to see if Canon introduces something similar on their new R5II. They already have nice MF implementation, but only for RF and EF mount lenses. Hopefully, they match Nikon, so my Canon gear can fulfill multiple roles.
I just discovered the Touch Fn feature, which was first introduced on the Zf. You can configure what a tap on a specific area of the rear LCD does when looking through the viewfinder. I configured my Z6 III to magnify to 1:1 on a tap in the upper-right corner, which is a natural location for the hand holding the camera. Very useful feature.
In my ISO test I'm using a fixed exposure (aperture, shutter speed) and only varying the ISO between raw images. At ISO 100 the scene is -7EV underexposed. I pull all the raw images into ACR/PS and then normalize the output brightness to be the same for all images. For ISO 100 that means +7EV exposure in ACR/PS, for ISO 125 +6.66EV, for ISO 160 +6.33EV, etc...
By using a fixed exposure I'm isolating any noise difference to ISO gain alone. Most people who do a noise ramp vary the exposure and ISO together, so that they reduce the exposure (ie, faster shutter speed or small aperture) as the ISO increases. That matches how people actually shoot in low-light conditions but by varying the exposure they're incorporating not just ISO gain noise differences but differences in shot noise.
Gerald's ISO 6400 comment is for N-Log, which uses a sensor ISO that's -3EV below the notional one, which means N-Log ISO 800 is actually sensor ISO 100...and N-Log ISO 6400 would be ISO 800. You can read more about how ISO is managed for N-Log in my thread here:
Thanks for clearing up what your ISO ramp test is showing, snapsy, I suspected you were providing Camera Stills RAW. FWIW, there's a lot of (understandable) confusion between Stills/Linear RAW or Video/Logarithmic RAW, um, "out there" and labeling/informing what is being presented might help clear some of that fog.
Re: your "only varying the ISO" methodology, it's my preferred presentation method when reviewer folks decide to provide an ISO ramp comparison, it provides a clearly-visible understanding of what the camera's gain control is actually doing...then, if the presenter wishes to show a "shot-matching" ISO ramp, then one can better, er, internalize what their camera is actually doing to create that shot-matched exposure. (Others MMV)
Thanks for the DPR link, I'll reserve my more technical posts over there so as to not side-rail this thread.
johnvanr wrote:
I actually don't have any Nikon gear at the moment and I'm only interested in it because it may solve my issues with critically focusing my M lenses on my Leica and the hassle of focus peaking and magnifying on other cameras. Curious to see if Canon introduces something similar on their new R5II. They already have nice MF implementation, but only for RF and EF mount lenses. Hopefully, they match Nikon, so my Canon gear can fulfill multiple roles.
I know you mentioned sensor stack thickness before, and I'm sure you know that M lenses don't really adapt perfectly to Z. Is "critical focus" important enough to lose IQ over?
johnvanr wrote:
I actually don't have any Nikon gear at the moment and I'm only interested in it because it may solve my issues with critically focusing my M lenses on my Leica and the hassle of focus peaking and magnifying on other cameras. Curious to see if Canon introduces something similar on their new R5II. They already have nice MF implementation, but only for RF and EF mount lenses. Hopefully, they match Nikon, so my Canon gear can fulfill multiple roles.
The MF eye/subject tracking is outstanding with my Zf and I can only imagine it's better with the Z6III. I bought my Zf and a Voigtlander 15 4.5 and 40 1.2 and it's been so much fun to use. I has cured my lust for a M10/11 body.
RoamingScott wrote:
I know you mentioned sensor stack thickness before, and I'm sure you know that M lenses don't really adapt perfectly to Z. Is "critical focus" important enough to lose IQ over?
The IQ depends on the lens and seems mostly an issue with vignetting, especially with wide angle lenses. I'm not really a wide angle shooter, but in any case focusing with a wide angle isn’t really that challenging anyway. It’s when you’re using 50mm or longer wide open that the issues start. And at those focal lenghts, sensor thickness seems to be less of an issue.
In any case, at the moment I’m just curious, esp. now that my M11 will be out of use until hopefully the Leica folks in Wetzlar can bring it back to life next month. I actually miss manual focus now in some circumstances, but not enough to deal with focus peaking and the cumbersome implementation of magnification on just about any camera except for the Leica SL line.
For those that have the Z6iii in hand, can you check and see what the maximum burst rate you can select without getting the "slideshow" in the EVF, i.e. what's the fastest fps that still gives you blackout. I would assume AF-C works all the way up to 14fps in mech shutter, but if that's wrong, please correct me.
The manual is clear as mud as usual I'm assuming H* (Continuous H Extended) is slideshow and cannot accurately meter each shot based on their language but good to know for sure.
RoamingScott wrote:
For those that have the Z6iii in hand, can you check and see what the maximum burst rate you can select without getting the "slideshow" in the EVF, i.e. what's the fastest fps that still gives you blackout. I would assume AF-C works all the way up to 14fps in mech shutter, but if that's wrong, please correct me.
The manual is clear as mud as usual I'm assuming H* (Continuous H Extended) is slideshow and cannot accurately meter each shot based on their language but good to know for sure.
Steve Perry covered this well in his early preview - starts at 9:00
1bwana1 wrote:
If only the Nikon people could discuss Nikon products without comparing them to other brands. As is usual on the Nikon Forums it is the Nikon people who turn the conversation towards other brands.
In this thread it was our friend bernardl who is a frequent practitioner of the art.
snapsy wrote:
Steve Perry covered this well in his early preview - starts at 9:00
Good info, thanks. Would be interested to see faster action/BIF stuff, or panning shots where straight lines exist in the background because I don't believe the latency situation is as bullish as he makes it sound.
RoamingScott wrote:
Good info, thanks. Would be interested to see faster action/BIF stuff, or panning shots where straight lines exist in the background because I don't believe the latency situation is as bullish as he makes it sound.
I'll try to get something for you in the next day or so. It's on my list of things to check and document.
RoamingScott wrote:
Good info, thanks. Would be interested to see faster action/BIF stuff, or panning shots where straight lines exist in the background because I don't believe the latency situation is as bullish as he makes it sound.
Btw the latency will still be there because it's a function of the latency of Expeed processing each image. That latency is about 50ms, which equates to the 20fps electronic shooting rate. Someone on DPR asked by the Z6 III has a blackout in single-shot mode and not Continuous H. Here's the explanation I just gave him:
When you are idle in Live View the camera is continuously sampling the sensor and routing that data into Expeed to convert it into the visual feed you see. To accomplish that Expeed has to perform a pipeline of image processing steps on that data, including demosaicing, WB scaling, color matrix transforms, picture control application, noise reduction, etc... Basically everything the camera has to do for actual photos/video it has to do for the Live View feed.
When you take a photo the camera interrupts the LV feed in order to read the sensor for the exposure, then it has to route the sensor data into the same Expeed pipelines in order to generate the raw/jpeg file that is written to the card. That pipeline takes approximately 50ms on Expeed 7, during which Expeed is unavailable to return to processing the continuous sensor sampling for the LV feed.
The Z8/Z9 solve this issue by having parallel pipelines in Expeed, so that it can process the continuous LV feed at the same time it processes the sensor data any photos you take.
The reason the blackout disappears at Continuous H+ on the Z6 III is because the camera replaces the usual LV feed with a slideshow of the exposures being taken and processed by Expeed for the photos you're taking.
That's how every camera handles extended shooting as far as I know and have experienced. The difference is that it's never before been usable for fast action tracking in the way Steve mentioned in the video. It's now usable because that slideshow is actually approaching watching a 24fps live feed video instead of a series of stills that happened in the past. For most slow to mid moving action, this seems ideal.
I would think all the same things that trip up non-stacked electronic sensor shots (baseball bats/balls, soccer balls in flight, panning background, etc) would still trip this sensor up, just less so.
RoamingScott wrote:
That's how every camera handles extended shooting as far as I know and have experienced. The difference is that it's never before been usable for fast action tracking in the way Steve mentioned in the video. It's now usable because that slideshow is actually approaching watching a 24fps live feed video instead of a series of stills that happened in the past. For most slow to mid moving action, this seems ideal.
I would think all the same things that trip up non-stacked electronic sensor shots (baseball bats/balls, soccer balls in flight, panning background, etc) would still trip this sensor up, just less so....Show more →
Agreed. The latency for processing photos on Expeed 7 is 50ms, so I expect the individual frames of the LV slideshow at 20fps to exhibit around 70ms of latency, which is 50ms from Expeed + 14ms for the sensor readout (1/70), plus a little extra to account for various overhead.
snapsy wrote:
Agreed. The latency for processing photos on Expeed 7 is 50ms, so I expect the individual frames of the LV slideshow at 20fps to exhibit around 70ms of latency, which is 50ms from Expeed + 14ms for the sensor readout (1/70), plus a little extra to account for various overhead.
How did you measure these 50ms?
How is the Expeed 7 able to capture jpg at 60 fps, which is actually more compute heavy than to shoot raw, if that is the case?
I had some time to use the Z6 III and do some basic comparisons with my Z8 and Zf.
The Z6 III's ergonomics are much better than a Zf, which is expected. It actually handles quite well with the big lenses, including the 1.2 primes. Still, because of how big Nikon's 1.2 primes are, a Z8 and Z9 are still the best choices for that since they balance the big lenses better.
Here is another consideration - Why purchase the Z6 III or Zf instead of the Z8? The 24 megapixel sensor is not just a lot better at high ISO, but the colors are more vibrant and painterly as well. This is not something that the specs will reveal to you. Even if you tweak the Z8 files in post production, the results are not identical. Z8 gives you clinically sharp images with flatter colors in comparison. I shoot portraits, fashion, and weddings, and yet there are many instances where I feel the Z8 is too much camera.
I could have easily bought two Z8's for the cost of these cameras, so the differences matter to me. Will have to do more testing between the Zf and the Z6 III.
I've encountered an intermittent issue with the Z6 III's WiFi whenever the HDMI port is in use. Either the camera will fail to connect to the WiFi network, or it'll get intermittent errors after connected while transferring images. It seems to be aggravated by certain HDMI cables, which implies a potential shielding sensitivity on the camera between the HDMI port and its WiFi antenna. That said, none of these cables produces the problem with my Z8.
Here's a quick demonstration, first showing the camera failing to connect to the network (error 1F), then later failing to transfer an image while connected (error 37). Nikon's wireless troubleshooting document for the Z9 says the following about these errors:
Error 1F: Solution - Turn the camera off and then on again.
Error 37: Check firewall settings, Check PASV mode settings
Is this with your $12 Chinese Amazon HDMI monitor? It’s possible it’s not well shielded from EMF.
snapsy wrote:
I've encountered an intermittent issue with the Z6 III's WiFi whenever the HDMI port is in use. Either the camera will fail to connect to the WiFi network, or it'll get intermittent errors after connected while transferring images. It seems to be aggravated by certain HDMI cables, which implies a potential shielding sensitivity on the camera between the HDMI port and its WiFi antenna. That said, none of these cables produces the problem with my Z8.
Here's a quick demonstration, first showing the camera failing to connect to the network (error 1F), then later failing to transfer an image while connected (error 37). Nikon's wireless troubleshooting document for the Z9 says the following about these errors:
Error 1F: Solution - Turn the camera off and then on again.
Error 37: Check firewall settings, Check PASV mode settings
I'm assuming H* (Continuous H Extended) is slideshow and cannot accurately meter each shot based on their language but good to know for sure.
Jun 27, 2024 at 08:41 PM