Nikon Rumors: Z9II

If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).

I can not speak to the Red Video. I wish I knew about video but I know next to nothing. Any good reco for a course to get started?

For the Z9ii, 30 FPS raw and precapture, larger buffer to allow this to work, perhaps next generation CFexpress card speed, greater flexibility to assign functions to buttons, ... I think they will have a winner. Lot of us how have 2 Z9 will upgrade at least one.
 
unless it actually starts shipping before the holidays. remember the z9 shipped in limited quantities before the end of the year and i suspect they could do much better this time around since the z9ii won’t be a first of a kind, ground up design
In recent memory Nikon has had a couple of instances in which scarce new product availability didn’t help its reputation (Z9, 800 PF, 180-600). I can’t imagine Nikon doing that again on purpose.
 
I can not speak to the Red Video. I wish I knew about video but I know next to nothing. Any good reco for a course to get started?

For the Z9ii, 30 FPS raw and precapture, larger buffer to allow this to work, perhaps next generation CFexpress card speed, greater flexibility to assign functions to buttons, ... I think they will have a winner. Lot of us how have 2 Z9 will upgrade at least one.
I'd like to see it push for more than 30 FPS, otherwise it's just a A1ii that's behind by a year or two.

It already can do 60 FPS RAW at 38mp in 8.3k video, so I'm sure they can pull it off in some form, even if it's 38mp which I'm sure would be fine for many.

I'm sure most of it is the limitation of sustained card write speeds and the pipeline to the card, plus compression strategy's to fit in those constraints. The RED tech should help.
 
Nikon will need to remain competitive, so I wouldn't expect any less than the Current A1mkii or R5mkii in feature set. I am sure they will have pre capture and a even faster frame rate for those who need it. I'm sure the Exspeed 8 chipset will introduce even better AF with improved subject tracking algorithms. I'm certain CF express 4.0 will be fully supported to back those higher frame rates with bigger buffer. One question I have is, will they refine the body and controls/buttons in any way ? I have no doubt it will be a great camera
 
Nikon will need to remain competitive, so I wouldn't expect any less than the Current A1mkii or R5mkii in feature set. I am sure they will have pre capture and a even faster frame rate for those who need it. I'm sure the Exspeed 8 chipset will introduce even better AF with improved subject tracking algorithms. I'm certain CF express 4.0 will be fully supported to back those higher frame rates with bigger buffer. One question I have is, will they refine the body and controls/buttons in any way ? I have no doubt it will be a great camera
I'd expect them to exceed both by some amount. Why be the "catch up" brand when you can lead for a few years. If they tap into CF express 4.0 and leverage the HE/HE* or RED tech for compression without loss they can most likely push the speeds into new territory. I honestly wouldn't be surprised to see 60 FPS at at least 38mp just by taping into what the 8.3k 60 can already do on the Z8/9. They've got the best RAW compression of any of the brands right now, mix that with CFB 4.0 and you're pretty much there.

You can already pull RAW frame grabs at 60 FPS/38mp in 12 bit on the Z8/9 with davinci.

In the meanwhile add a 300 2.8 with built in TC or something of that nature, add in a 24-70 f2 or 28-70 f2, or 1.8.

I doubt they'll aim low.
 
I'd expect them to exceed both by some amount. Why be the "catch up" brand when you can lead for a few years. If they tap into CF express 4.0 and leverage the HE/HE* or RED tech for compression without loss they can most likely push the speeds into new territory. I honestly wouldn't be surprised to see 60 FPS at at least 38mp just by taping into what the 8.3k 60 can already do on the Z8/9. They've got the best RAW compression of any of the brands right now, mix that with CFB 4.0 and you're pretty much there.

You can already pull RAW frame grabs at 60 FPS/38mp in 12 bit on the Z8/9 with davinci.

In the meanwhile add a 300 2.8 with built in TC or something of that nature, add in a 24-70 f2 or 28-70 f2, or 1.8.

I doubt they'll aim low.
I have never tried pulling 38mpx frames from 8k/60 video with Davinci. That sounds like an interesting option that I will test soon. And I agree that Nikon has the best raw compression out there. The current RAW HE* files are superb
 
Nikon will need to remain competitive, so I wouldn't expect any less than the Current A1mkii or R5mkii in feature set. I am sure they will have pre capture and an even faster frame rate for those who need it. I'm sure the Exspeed 8 chipset will introduce even better AF with improved subject tracking algorithms. I'm certain CF express 4.0 will be fully supported to back those higher frame rates with bigger buffer. One question I have is, will they refine the body and controls/buttons in any way ? I have no doubt it will be a great camera
I mean, at this stage, shooting 8k60, barely saturates 50% of the max theoretical transfer speed of existing CFe 3 cards. Why would we think CFe 4 is required at this stage. Bragging rights, sure, but other than inflating the price of the body, as well as forcing the purchase of expensive CFe 4 cards, I can see many areas where Nikon could rather spend the R&D.

Just IMhO.

Edit: I’m a tech nerd btw, would love to see changes, but this is similar in a way to the PCIe 5 slots available on modern motherboards, where current games doesn't get close to saturation on the PCIe 4 interface. So we pay for hw features, where actual software to make use of it, is quite far into the future, and keep paying without back-end support ever catching up to hardware capabilities.

I very much suspect the Expeed 7 has a lot more oomph left in the tank as well. Money will most likely go into new sensor tech, and porting Red video stuff over, with hw/sw support.
 
I have never tried pulling 38mpx frames from 8k/60 video with Davinci. That sounds like an interesting option that I will test soon. And I agree that Nikon has the best raw compression out there. The current RAW HE* files are superb
Give it a try. I find the files are not the 14 bit RAW's we get for stills, but I think there's a lot of potential. You can edit them in Davinci for noise then export them as TIFF for lightroom or DXO. Davinci is quite impressive software.

The downside is holy gods does 8.3K60 NRAW burn through some storage. It would be amazing if Nikon let us pull frame grabs out of that feed in camera as NEF's.
 
I mean, at this stage, shooting 8k60, barely saturates 50% of the max theoretical transfer speed of existing CFe 3 cards. Why would we think CFe 4 is required at this stage. Bragging rights, sure, but other than inflating the price of the body, as well as forcing the purchase of expensive CFe 4 cards, I can see many areas where Nikon could rather spend the R&D.

Just IMhO.
"Whats it gonna take to earn this sale" "60, 120, whats the number". Haha. Wait until they make an offer you can't refuse, those CF cards will get swapped quick. Aren't the 4.0's less expensive, cards wise.

I know someone here's great with hardware. Whats the max theoretical you could push for framerate with HE/HE* file sizes and sustained write speeds to a 4.0 CFB card?

Is there a way Nikon could leverage the RAW compression they picked up with RED to enable some kind of high speed RAW file capture from video in camera? Could we see 8.3k 120 and the ability to frame grab 38mp NEF's in camera?

Whats the max the current Z8/9 sensor can handle? The V-Rapter can shoot 120 FPS 8k.
 
Last edited:
"Whats it gonna take to earn this sale" "60, 120, whats the number". Haha. Wait until they make an offer you can't refuse, those CF cards will get swapped quick. Aren't the 4.0's less expensive, cards wise.

I know someone here's great with hardware. Whats the max theoretical you could push for framerate with HE/HE* file sizes and sustained write speeds to a 4.0 CFB card?

Is there a way Nikon could leverage the RAW compression they picked up with RED to enable some kind of high speed RAW file capture from video in camera? Could we see 8.3k 120 and the ability to frame grab 38mp NEF's in camera?

Whats the max the current Z8/9 sensor can handle? The V-Rapter can shoot 120 FPS 8k.
I suspect the biggest challenge at these levels, will be power consumption and heat-dissipation. We’re already running into limitations on the Z8 wrt heat, so a big pro body would be a requirement for increased video-capture speed @8k.

We will have to go a long way before we will saturate the theoretical limit of the CFe 4 cards.
 
Last edited:
I suspect the biggest challenge at these levels, will be power consumption and heat-dissipation. We’re already running into limitations on the Z8 wrt heat, so a big pro body would be a requirement for increased video capture speed @8k.

We will have to go a long way before we will saturate the theoretical limit of the Cage 4 cards.
Yeah we're probably looking at full size Z9 body and battery's to make it work. Or a nuclear power backpack from ghostbusters.
 
I mean, at this stage, shooting 8k60, barely saturates 50% of the max theoretical transfer speed of existing CFe 3 cards. Why would we think CFe 4 is required at this stage. Bragging rights, sure, but other than inflating the price of the body, as well as forcing the purchase of expensive CFe 4 cards, I can see many areas where Nikon could rather spend the R&D.

Just IMhO.

Edit: I’m a tech nerd btw, would love to see changes, but this is similar in a way to the PCIe 5 slots available on modern motherboards, where current games doesn't get close to saturation on the PCIe 4 interface. So we pay for hw features, where actual software to make use of it, is quite far into the future, and keep paying without back-end support ever catching up to hardware capabilities.

I very much suspect the Expeed 7 has a lot more oomph left in the tank as well. Money will most likely go into new sensor tech, and porting Red video stuff over, with hw/sw support.
Agreed. I'm also a tech, systems engineer. here are some fun facts regarding Nikons video bit rates and actual bandwidth requirements. We can see that the CF express cards are hardly taxed with 8k video;

For 8K/60 recording, the minimum sustained write speed required by the CFexpress card is:

At least 162.5 MB/s for N-RAW.

At least 300 MB/s for ProRes RAW HQ.


A CFexpress Type B card (which the Nikon Z9 uses) typically offers sustained write speeds of 400–1500 MB/s, so high-end cards should handle this easily.
 
Whats the max theoretical you could push for framerate with HE/HE* file sizes and sustained write speeds to a 4.0 CFB card?

max theoretical for cfe-b v4 is 4000MB/s, 45MP he* is about 26MB per image, so that’s, uh 153 fps?

in reality tho, you aren’t going to be able to reach theoretical max.

so the v4 version of the 1.3tb delkin black says it can sustain writes at 2150 which gives you 82 fps.

so, less than one might expect.

given their strategy of not using cache, it totally makes sense for nikon to move to v4

Whats the max the current Z8/9 sensor can handle? The V-Rapter can shoot 120 FPS 8k.
well, we know it can do 120fps at least. of course we don’t know if that’s doing something like line skipping
 
I would think Expeed 7 is using a multiplexer to feed the different streams, not adding much processor-overhead in splitting the two data-streams that way. A 2nd processor will help with image-processing and improved AF, the former helping with raw-precapture.

I am thinking Expeed 7 still has a lot of available processing capacity, and that we won’t necessarily see a 45MP Z9II replacement soon - instead, Nikon will release something fast, with improved AF, for the media-crowd and sports-action guys, probably 24-30MP, and roll out new FW for existing Z9 users. Just IMHO of course.

Let’s see what the future holds, but naive to speculate anyway.

Mmmm…I was wrong. Looking a bit deeper into this, there are some high-speed memory and supporting pre-processors on the stacked circuitry on the sensor, supporting capturing two data-streams, and then feeding it directly and independently to the Expeed 7 processor, where (mostly) duplicate image-processors manipulate the data and export the processed streams separately to the EVF and the image-storage, i.e CFE card(s).

This is the cleanest way to do it in any event, with the lowest latency, but it speaks volumes to the capability of the Processor. Luckily, it appears there’s a fair bit of pre-processing done on the stacked circuitry, so I am still hopeful there is a fair bit of processing-horsepower available for RAW-precapture, at least where the Expeed 7 is concerned. Who knows what the pre-processors and memory on the stacked circuits can handle though. Of course, pre-capture buffer-size might also be a major limitation.
Untitled.jpg
You can only see EXIF info for this image if you are logged in.
 
Last edited:
there’s so much we don’t know, including how much we should read into their diagram.

for example are there really two feeds from the sensor? like actually two sets of wires. it would make more sense to have some buffer memory that both streams could read. but… maybe it might make sense to actually have two feeds if they’re doing something like line skipping or otherwise downscaling the image before feeding it to the live view stream which would make sense because it probably doesn’t need a whole image and you don’t want processing overhead for no benefit.

and… are they really separate, or just logical constructs?

i still suspect the pre-capture we have may be piggybacking the display feed and that might explain the lower res and jpg only.

i’d love to know more
 
there’s so much we don’t know, including how much we should read into their diagram.

for example are there really two feeds from the sensor? like actually two sets of wires. it would make more sense to have some buffer memory that both streams could read. but… maybe it might make sense to actually have two feeds if they’re doing something like line skipping or otherwise downscaling the image before feeding it to the live view stream which would make sense because it probably doesn’t need a whole image and you don’t want processing overhead for no benefit.

and… are they really separate, or just logical constructs?

i still suspect the pre-capture we have may be piggybacking the display feed and that might explain the lower res and jpg only.

i’d love to know more

What I could gather is that the stacked circuitry contains fast RAM, which most likely acts as the data-splitter, where it goes to pre-processors, which spits out the independent streams to the different cores on the Expeed 7. Anyway, nice to speculate about these things :)
 
i kinda doubt it. they've pushed this over-all layout down to cameras like the z6iii, i think they're more or less happy with it.
even if they remain in the same layout design, there could be a huge functionality leap if they will just “open up” the buttons customisation function options.
For example:
- The arrows multi selector is in prime location and could potentially have so many useful rolls in active shooting ( like what the Sony/Canon third wheel is doing for just one example) but currently it is just doing what the joystick is already doing.
- the zoom in/zoom out buttons are also very limited in theirs active shooting rolls currently. They just, well, zoom in and out… a huge wasted prime real estate!
- the nikon way of pressing a button and using a dial simultaneously to change stuff is sometimes unnecessary slow, they should follow what they did with the AF Cycle function - one push of a button to cycle between options- it is so much faster. They must enable it for Subject Detection asap

They really should look at what Sony (and Canon , I guess) are doing . In the A1 for example there fewer buttons but much more customisation options.
 
How much of this is real? We’ll see. I also expect some AF improvements (we’ve chewed over that at length on BCG Forums), some UI refinement, support for CFe 4.0 cards, and some improvement in pre-capture (perhaps HE* at 30 fps or more).
  • The new Z9II will not be a major update from the current Z9 – think about the Nikon D5->D6 update
  • It should be announced before the 2026 Olympics (February 2026)
  • It will have new RED video features
  • Same sensor
  • EVF from the Nikon Z6III (5760k dot)
  • Content credentials

https://nikonrumors.com/category/nikon-z9ii/
All the Z9II needs is X-Type AF points like the R1 and it'll be perfect. I had precapture on my E-M1X, never used it, I think it's kinda overrated.

I have a Z6III and I can't tell a difference in viewfinder fidelty compared to my Z9, and I have good, young eyes.
 
All the Z9II needs is X-Type AF points like the R1 and it'll be perfect. I had precapture on my E-M1X, never used it, I think it's kinda overrated.

I have a Z6III and I can't tell a difference in viewfinder fidelty compared to my Z9, and I have good, young eyes.
Same for me when I checked out the EVF on the Z6iii in store. Honestly not that much different. Nikon did a good job on the Z9 viewfinder and all the models that share it. The Z6iii was just a smidge clearer but not much. I think people get hung up on the dot counts on spec sheets.
 
unless it actually starts shipping before the holidays. remember the z9 shipped in limited quantities before the end of the year and i suspect they could do much better this time around since the z9ii won’t be a first of a kind, ground up design
The limits upon initial quantities were partially driven by COVID impacts, were they not? And who would have ever expected the demand for the Z9 to be what it was? That level of demand didn't help put and keep the Z9 on store shelves. It did, however, help Nikon's bottom line!
 
Back
Top