Industry News...Nikon Z9

Considering all the reports of financial issues, I am actually happy to see Nikon release the Z9. On paper, this is an amazing camera, and easily you can see the R1 specs matching it in a year or two. While I've already seen some people emphasizing how much better this is than the R3, to me...we are talking different shades of excellence. If you really need AF 120/sec vs 60/sec, here you go. If you really think you need 45 MP over 24 MP, here you go. If you want 30 fps vs 20 fps, R3. If you want smaller files/less system bandwidth (which Canon continues to say is a market driver), R3.

No doubt both will take many excellent images.

I see it as a strong convergence statement in video.

I am sure the videofiles can explain to me what is missing from a $10k $20k $30k cine camera, but shootong raw what else do you need???

Global shutter perhaps, depends on how minimized rolling shutter might be. Film had rolling issues

DPreview in testing was estimating that the sensor read rate of the Z9 was ~1/270th second (seems odd, but that is what he said). For the R3, it was ~1/180th. And, Nikon dropped the mechanical shutter entirely. So, in their flagship, they seem to seen reading a sensor in ~1/270th of a second is fast enough. Canon did come out with a video recently where Rudy Winston said the R3 was designed to be primarily electronic shutter.

I wonder if we are going to see a global shutter. I've been thinking this, and I wonder how much a global shutter is more of a concept and less of a reality. Ultimately, pulling information from millions of pixels instantaneously is a challenge, but then they have to be managed, kept in order, and recorded as a single image. To do it, you'd almost have to have RAM like memory associated with each pixel that could keep up with 20-30 fps. I am sure possible, but if it was easy, I think we would have seen it in the R3/Z9. Instead, we see faster and faster rolling shutters. Nikon seems to think 1/270th of a second is fast enough a mechanical shutter isn't even needed. I wonder if it is less expensive to get to say 1/500th of a second with rolling shutter read speeds and we never see a global shutter in this level of camera.

Back to the Z9, I mostly watched DPReview's preview in its entirety, but looks like limited buffer 40 frames at 20 FPS for full RAW, or 2 seconds. They have faster fps and larger buffers, but at increasing levels of image compression. It is also heavy like a 1Dx and Dx body. No weight savings.

Its funny, as I am seeing a lot of comparisons to the R3, but I have to admit I very much prefer the R5's buffer depth. Anyways, at this level any more, shades of excellence.

That's similar to the max shutter rate when using a flash, so similar to full open/full closed shutter timing... but faster shutter speeds tend to start the closing cycle before the opening cycle is done, leaving only a partial sensor exposed. Logically, this partial exposure window scrolling over the sensor over time would have the same rolling shutter effect as the limited read speed, since the overall timing is roughly the same.

I think.

I agree with Brant. Dont know of the flash synch which clearly is linked to read speed and very well could be linked to effective electronic max.

There are already very high speed video cameras, lower res than 30/45 mpx, there may be many choke points thar need data/hardware engineering.

At 20k fps rolling shutter isnt an issue in most cases :)... O assume.

Which brings me to my next point. I saw the update of dpp for raw vid. Will be doing my 1st SHORT test on 4k raw this weekend. What I used refet to as my mini mainframe may not be up to much of the task. What cpu memeory needs these imaging monsters are going to require?

I have no idea what the speed is on the Z9. When I mentioned the flash speed above, I meant historically for Canon it's been 180-250, with 250 on the higher end bodies, and I was also under the impression all manufacturer's flash speeds were similar. Canon HSS repeatedly flashes to account for the partially open shutter.

My point was since the shutter movement has such a speed limit, and a object can move across the sensor while differing portions of the sensor are exposed, then a sensor read speed that reaches that level of performance should have the same level of rolling shutter/jelly effect as an actual moving shutter... assuming the sensor values are wiped with the same sort of scan time, approximating the other half of the shutter.

We're reached the point where the sensor read is as fast as a shutter can move.

...and now the rumor is that a global shutter is not coming to the R1.

I was not the source for the rumor. :p

Rumors are rumors, but if the read rate of rolling shutter is fast enough, I get it.

This is not a hard number, but I've heard the read rates of DSLRs like my 5DIV were ~1/20th to 1/30th of a second. Supposedly, the R5 is ~1/60th of a second and the R3 ~1/180th of a second with the Z9 at ~1/270th of a second. Its not hard to see that these are getting much faster that "the bad days" of rolling shutter. I think I've seen issues with the R5 and electronic shutter perhaps once. Not saying I couldn't create an issue if I tried, but I was using Electronic shutter on a few bird shots a week ago and noticed a funny looking tree in the background. But, wasn't even that bad. Otherwise, I haven't seen any rolling shutter issues. I will say, I do think AF is better in Electronic shutter. No split second that the shutter blinds the sensor.

Definitely makes me less enthusiastic about the R3.

Seems to me the Z9 sacrifices nothing over the R3 at a similar price point.

Kinda makes the R3 irelavent unless you are tied to the Canon ecosystem.

Especially with the A1, it just double downs on the fact that resolution is a selling force today for flagship cameras.

Competition is good. Canon, Nikon, and Sony all push each other and the consumers benefits. The spec sheet of the Z9 will push Canon.

Dave

This is a bit of a chicken or the egg debate, but I agree in that Nikon and Canon, in particular, have always had very similar flagship cameras. Is that because they looked at each other's specs and wanted to compete head to head or is it because they had the same read the market? I suspect a bit of both.

I still think there are two scenarios that could unfold: 1) R1 and R3 are both going to exist moving forward, so two camera models to target two markets, which seems to be the general assumption at this point; or, 2) the R3 was a placeholder as Canon developed the tech for the R1 and there will be one "flagship" model moving forward.

Assuming that there are two "Flagship" models, the R3 and R1, moving forward, I saw some Pros conversing and their general sentiment was that the R3 would be the de facto heir to the 1Dx line and the "Pro's" flagship. Their comments make sense to me, but if this scenario plays out, camera manufacturers may now see enthusiasts with large amounts of disposable income that want the "best" as a large part of the market, thus they get the "R1" with the high MPs, etc.

Not saying that some working professionals won't go with the R1, almost certainly many will for one reason or another. But there are many reasons why 24 MPs could be enough. Quick example, but you can control your working distance in a studio, thus there really is a limited benefit of cropping. My first two cameras were less than 24 MP. The 5DIV was 30 MP. Assuming 24 MP is good enough, the R3 is significantly lighter than the Z9 (for example), gives you full RAW at 30 fps, has lower bandwidth/storage issues, eye-controlled AF, etc. Of course, no one has seen a production model of the R3 or Z9, so we need to wait and see about other features.

There was a post by Craig/CRguy in response to Neuro indicating that the R1 would be "more than double" Neuro's prediction of 30 MPs. Personally, I can see the R1 having too many MPs for my purposes. With the R5, I needed to upgrade my home computer and I can see the differences in performance between my 1 yr old laptop (Ryzen 4700) and my new desktop (Ryzen 5900). I can see the benefit of more MPs for birds, but that is really it and even then....45 is exceptional. Everything else, 45 MPs is too many, IMO. So, I very well might go R3. It would depend on other features.

I'm guessing that the idea of a global shutter is going to go away, in place of faster read speeds on conventional electronic shutters. From what little that I know, a global shutter cannot be done without adding noise to the system.

Regarding the Z9, I'm surprised to see the price come in lower than the R3. That is quite a bold move on Nikon's part, and I think that Canon is going to have to drop the price of the R3 at least when the R1 is announced, if not sooner.

I think the justification stage for the R3 is over, we can talk all day what pro's might need. Why someone would be cool with less resolution ect.

But at 6k and as a current flagship those excuses fall apart. And when the R1 comes, it's gonna push the price to the moon.

I agree with Jonathan that the value is not there anymore and I proves my point in previous posts that low resolution is a bad marketing move for Canon at this price point.

My take is Nikon left nothing in R&D and put it all out.

A market share move. Capture as much as they can.

Why leave it in R&D when fighting for your existence?

Trying to chip into the vid market. Etc.

I agree with Jonathan, global shutter is the realm of the hyper cine cameras. With computing power they might be correcting it in post....

Z9 looks like an R5 on steroids!!!!

A grab on market share during a rare Canon mount change, when pros might consider replacing a lot of glass with the new versions as they release, along with a body change. Smart move.

Could only be more optimal if they also released a EOS to Nikon mount converter to let them keep using their existing glass, but I think Canon would cry foul if they reverse engineered the communications to do that well.