@ haha same! Any time he releases a new video I have to watch it. I only get my Nikon news/gear reviews from Matt. I trust his views on it over other creators
In digital imaging, dynamic range is also expressed in decibels (dB), a logarithmic unit that describes the ratio of maximum to minimum measurable light levels.(1 stop ≈ 6.02 dB) 110 dB ≈ 18.27 stops DR
Mr. Irwin, this is a very technical video. I consider myself a geek and you went deeper then I would have. What I enjoyed was your way of explaining this technology. I felt like you were only presenting the basic information. Which made your video easy to watch. Even for me a geek. LOL 😂😂😂
The interesting thing about this sensor is that it's trading off its high frame rate to increase dynamic range by integrating exposures across sub frames through processing blocks on the lower chip. This can only be done with stacked sensors. They went from 110dB in a single 1000fps frame to 134dB with 16 subframes (lol 22 stops dynamic range).There will probably be some temporal aliasing artifacts with this technique but that should only matter in extreme action shots. I am absolutely excited for this technology and hope it ends up in a Z9II. PS. I used to design digital camera chips a long time ago.
Interesting, I do photogrammetry, and I primarily use micro four thirds or aps-c because I need more depth of field; I used to use a sony rx10 mark 4 because it got even more DoF, but I switched to MFT because it offered more versatility and control with interchangeable lenses. A high dynamic 1” sensor would be amazing if I could get Nikon level lenses. This is exciting, I hope they use it for more than just robots and cars (etc), I hope they re-launch their 1” sensor line!
Thank you for bringing up a very cool tech report from Nikon. But, keep in mind this is actually a report from 2022 (its header reads "Nikon Research Report Vol.4 2022") so I wonder how far further along this trail of tech development they are today? Would be really cool to hear what is happening with this tech 🙂
You unpacked a lot of info there Matt. Thanks. It seems to me another great thing about Nikon further developing expertise in visual sensor tech, in this growth industry that you have explained (car tech and robotic tech), means a lot of return-profit will likely sustain and expand R&D efforts for its core products. That's good for all of us.
I forgot to tell you Matt, that my son bought the Z 50 II, and is very happy with it and he did watch your video, that I told him to watch, on it and that pushed him to buy it.
Happy New Year Matt. Another wonderfully informative and studied video on Nikon’s progression. I have the Z9 and Z8 and I so much enjoy using them in tandem for my most important tasks. The Z7ii however is my walk around personal camera when I’m with the family and my dog. It is a great camera… small (but not too small), light (compared to the Z8/9) and it does everything I need for landscape and street photography. But there are times when unexpectedly, I need to switch gears and the UI or the camera’s response is not the same or at the level of either the Z8/9. My hope is that there will be a Z7iii with the responsive UI of its big brothers and some of the niceties they possess. I am quite happy with the resolution of all three cameras and since I photograph in raw, my computer storage setup is already nearing 80 terabytes, so I am very happy at the 45.7 mb range. Please Nikon, all I want in 2025 is the 35mm f/1.2 (I do so love “available darkness”photography) and the more responsive Z7iii. Thank you again Matt, for the beautiful photos and the informative videos.
After shooting my Z9 with a 180-600mm lens this week on birds in flight against a bright sky, I am thinking more dynamic range would be a very good thing. Bringing up the shadows on the birds at 340 ISO was extremely noisy.
You're getting it all wrong. The high fps is not intended for saving all those frames as images, but rather for exposing different areas and achieving a single, but very high dynamic range image. So you need the 1000fps for HDR at high fps (120+, even much more when dealing with computer vision)
I agree that these new sensors are likely going into the automotive application. So many requirements for ADAS, Advance Driver Assistance Systems, the need for high dynamic, fast sensors is the key to making collision avoidance, pedestrian detection and blind side alert function. I'm very impressed with the night vision my new vehicle has when I pull up the camera views.
Happy New Year. If I'm not mistaken there will be a joint booth next week at CES with Mitsubishi for some "car vision" tech. This sensor and its frame rate/DR is perfect for that application. Sure, it can trickle down to photography, but I'd very much love to see Nikon diversify in such a way that they can send more R&D money down the pipe to the imaging division. Automotive could prove to be extremely lucrative for Nikon. Glad someone on the UA-cam has finally figured out that Nikon and other brands aren't buying SKU'd/off the shelf sensors from Sony but sending their designs off to be manufactured by someone who can do it at scale. Will be an exciting year for photography! Cheers!
1” sensor pixel pitch 2.7um, sounds familiar, Nikon 1 v3 also 17mp and worked quite well, a sensor I’ve used on two occasions to capture surprisingly (for me) good images of a solar eclipse.
"What's special about Today's primes that they are extraordinary wide open" - that's true. I've the 35, 50, and 85 S primes (1.8, not the 1.4) and more than 70% of the time I'm shooting them wide open (that's also why I got the S/1.8s and not the 1.4s). Still I get amazing sharpness, and the only reason I stop them down sometimes is to have bigger DOF.
DR (dB) = 20log10(Nfull/Nnoise), DR (stops) = log2(Nfull/Nnoise). 20log10(2) = 6.02. So to convert stops to dB multiply stops by 6.02 or to turn dB into stops, divide dB by 6.02.
BTW, the dB measurement is the span of signal in the darkest to brightest areas, the larger the better. Recall that dB is logarithmic so conversion looks like this: Stops = log2(10^(dB/20)) Also note that this range does not specify the wave length but we can assume that it’s visible light they are talking about. Seeing a little below human range towards infrared would help with night vision though. Imagine being able to access that on the same sensor in a consumer camera with a toggle.
It is an interesting announcement. Even if it is really being developed for cars, the potential spin offs will inevitably find their way into photography. For cars, better fast sensor/guidance/warning and control systems are in high demand. Modern vehicles do bing and bong little warnings as you drive which is fine when they are correct but unhelpful if you are actually moving to avoid a pot hole or obstacle. Improving sensor accuracy is a key to improving processing driving assistance. However, if you have the technology for driving what else can you use it for? Autofocus anybody?
I'll take the dynamic range, but I don't need 1000fps. Actually, if Nikon could even bring Z9 sensor performance to a global shutter (or ultra fast rolling/global shutter hybrid) I'd call that a big win.
Thanks Matt and Happy to New Year to you and yours. So for me and many the current Z9 is almost perfect as it is for both stills and videos -- sure I want all the best of what is in the Z8 and Z6iii to be added in the next FW. AND yes there are a few long standing "improvements" that many have been asking for for some time -- ability to engage or disable Subject Detection, Multiple shooting recall hold per menu/shooting bank, some ability to export files from the camera onto an SSD (even if this has to have its own power), more video improvements (false colour).AND then the real winners if these are possible without physical changes -- 30FPS RAW and Pre-Capture in RAW - if not physically possible then these are must haves for a Z9ii. Then we have the Z9ii -- I am very happy with the current sensor in the Z9 -- as a predominantly wildlife shooter I would have preferred the base ISO to have been tuned to ISO 100 or 200 and a correspondingly cleaner/larger DR as higher ISO levels are used. I know of FEW sports shooters who do not also want even better low light performance -- I certainly do not want the "rubbish" performance the current Global Shutter hybrid camera delivers. Bigger/faster Buffer; Option for built in SSD; What about the NEW and even larger CF-Express Type C cards -- when will we see a body which uses these? Surely "better" (meaning higher res and higher refresh rate) EVF and LCD. I would like to see a 2nd USB-C data port and all such ports be USB-C 4 rated. A 1" sensor is much smaller than I expect in an FX body -- but the Global Shutter sensor in the $45k RED Raptor XL has a 40.96 mm x 21.60 mm (Diagonal: 46.31 mm) sensor, which is 3% larger area than a standard FX is 35.9 x 23.9 sized sensor. The Raptor sensor is 35.4MP which is 77% of the Z9 sensor's resolution. AND I doubt the Raptor's sensor has AF "points" built in -- but still a short walk from the current RED sensor. {I have little doubt this would take a while to develop for a Z9 model} Would shooters be willing to accept a Z9 variant with a 35.4MP global shutter sensor based on the RED design? What are the trade offs? [Even better DR and low light performance would be likely] - remember the Raptor XL offers 120 fps at 8K 17:9, 150 fps at 8K 2.4:1 and 2K (2.4:1) at 600 frames per second. We are still to see a new Nikon RED camera or a Nikon Hybrid that supports either bolt on or built in active cooling -- bolt on options have been tried with the Z6iii -- it would be far better to have these built by Nikon for a specific body (which would have to have a flippy screen)
110dB of dynamic range (log base 10 notation) is equivalent to just over 33 stops (log base 2 notation), but it really doesn’t translate like that unfortunately, because of things like the final ADC being 12 bit, and the eventual need for bayer filter arrays and debayering. it’s a high sensitivity/dynamic range sensor for sure, but the only way to know how it’ll compare to current sensors is to slap it in a camera and find out!
Thank you for that interesting article. It seems analogous to min-LED TV screens where small LED blocks are individually lit to give better contrast ratios. However, those TVs apparently can get things wrong such as rendering a small spot of light - the Nikon paper does not mention what happens if the lightness levels differs within one of those 16x16 exposure blocks. Just a warning that this technology may be good enough for industrial sensors but not for pixel peeping photographers :)
In regular chips, 65 nanometre is huge. The highest resolution steppers now are at 1 nanometre. Probably only the largest foundry (aka fab) (TSMC) has one machine at that resolution test printing (these print chips). And their finest operational machine will be something like 3 nanometre. In photography we need some area per photosite, or else we loose to much light sensitivity. In that sense the 65nm Nikon stepper (chip printer) is a great thing. Nikon (and Canon) have been leading this stepper machine market until about 2,000 when they lost their commercial edge in the regular chip market. In recent years, AFAIK Nikon has been R&Ding, looking to make steppers for this tiny scale market again. But the way I see it, no sensor will be made with 1 nm tech, or 3, or 7, or. Because light sensitivity. If that were possible, that would be a major revolutionary step change of several orders of magnitude. It would also facilitate maybe to have sub-photosites that together generate RGB instead of monochrome raw data. No, not T-cells, like Fuji. But demarcated regular shape sub-photosites. But that's a dream.
Yes I agree JP, photography will never utilise or need the lowest nanometer tech for light capturing. But lower nanometer chips may be used for on sensor processing, RAM, cpu's and more. And thus will still impact photography in general, just not the light collection part. :) Exciting times ahead.
Happy new year Matt. The most likely thing is that this sensor in FX format and with a higher number of pixels we are already enjoying it on the Z9 and Z8
I believe 110 db is approximately 18 stops of dynamic range. This is what a quick search came up with… “110 dB is equivalent to roughly 18 stops of dynamic range, as each "stop" in photography is considered a doubling of light intensity, which translates to approximately 6 dB in the logarithmic scale used for decibels”
Hi Matt - very interesting video! I wonder if part of the 1000 FPS is the ability to condense multiple shots into one as they do reference "computational" photography. Like on an Iphone, my understanding of how a phone creates very low light shots is through computational photography - essentially multiple exposures are taken very quickly and then combined to make a better single image - essentially its HDR but backed in and done very quickly. That said, this new tech does seem different in that it appears that the sensor itself is able to produce different exposures based on the different zones, so I get why one would think that the need to merge different frames is circumvented (and maybe it is), but I wonder if they are doing a little of both here. As you point out 1000 FPS seems a bit absurd and is definitely overkill, but if they are combining frames then this number would rapidly diminish - also 1 second is too long of a sample time to combine frames as there will be motion, but the camera could capture say 10 frames almost instantaneously in 1/100th of a second - then maybe with the combination of the zones and the available data of the 10 frames all of which an have slightly different exposure they may be able to produce a very impressive single frame with DR - like 20+ stops - who knows? All of this is speculation, but as you point out, whatever the case it is clear that things are progressing rapidly and that Nikon products have never been better. Indeed the future is very bright for this system - now we just need an inexpensive FF entry level camera like the Z5II to really get the newcomers to start with Nikon and boost that marketshare 😂. In all seriousness, I am sure something like that is on the near horizon. Thanks Matt! -PD
Hi Matt, happy New Year to you too! According to ChatGTP, Decibel can be converted, its 6.02 DB per stop (giving us 18.27 stops). No idea if correct, but perhaps an avenue you can pursue!
I can confirm that number. I think they can use Sub-pixel Architecture when there are 2 pixels with different size, so they can achieve multi-exposure in one shot.
My Z9 and Zf are just about perfect, but I'll grab a Z9II two years after it's released and easy to get (with firmware bugs fixed). That's how to upgrade. Always wait for the dust to settle.
I like the BNV alert label 🎉 I seem to remember that Canon also released some info on a 17mpix not long before the R1 announcement. I believe this was to reduce the disappointment of their flagship “only” having 24Mpix, compared to the larger MPix Sony and Nikon flagships. It’s possible Nikon is doing the same before a Z9ii with 20-24 MPix perhaps...
Hi Matt, Hope you have a fantastic 2025 ! It's funny when Nikon were considered late to the digital camera scene many, many people said Nikon are dead, they don't have the technical abilities to be competitive in such a market, those very same people were probably ignorant of the huge body of experience Nikon have in the specialist imaging field, Nikon have been world leaders in many different aspects of photography and imaging for decades, I think their future is very bright! Cheers.
The worry was real we are all happy nikon made it but it was only because of some really nice last models in camera world. in litho they (and canon) have been under great pressure from asml.
Cool !! Grok AI tells me 110 db is about 18.2 stops of HDR!!! I wouldn't mind a new Nikon1 camera with 18 stops and 17mp!! 134 db equals 22 stops !!! madness the future of camera tech is bright !
I've been saying for a few years now, Nikon is trying to remove itself from Sony. This type of sensor looks like it's going to before science and space. I can see the cars too. Knowing Nikon, they're not changing the Z9 sensor for several generations. If anything, they'll create a new camera series. If they were to go to the global shutter. It'd be called a Z1, Z100, etc.
I suspect. as you do, that this R & D is aimed at things like cars, robots, security etc etc. Especially cars. That said, things learned can be used for multiple applications. Personally, as far as cameras are concerned, I rarely shoot more than 10 Frames per second…most often far less. I would much much much rather a focus on dynamic range…give me an extra 2 stops of quality DR vs speed.
@boazsayar1193 there are quiet a good amount of one inch sensor cameras, i used o take my Panasonic one inch sensor, fixed lens camera to concerts where theres no bigger camera allowed and managed to get some decent results with it i believe it was the lx 15 or something like that
@@obedbrinkman I have Canon G3X, but it is getting old and I need a replacement. All cameras in the market now are pretty old. It is a great camera for travel. I take great photos.
The human eye can respond to pretty much any and all lighting conditions and adapt the amount of light entering through the pupil. It doesn’t make sense to refer to dynamic range for two reasons. 1. The eye picks up ALL light entering through the pupil and it’s absorbed via the retina. 2. The brain adapts light and defines contrast in an adaptive way. Ie, if you live in darker areas, you brain will adapt to differentiate between smaller difference of light. If you live in light areas, it will adapt to only differentiate larger differences. Human vision doesn’t make sense to define in the same way as a camera, though the concept is similar, a camera is not a dynamic organ like the eye/brain vision system.
Nikon making their own sensors will separate them form the field and no longer having to depend on Sony's basic sensor hardware manufacturing it for them. It's a great business move for their own bottom line as well as creating their own unique imaging ecosystem, plus as you detailed are open to pushing the technology into the future more efficiently for themselves and others who want to buy into it. Nikon is basically setting themselves up to be or better themselves compared to Sony... Competition drives innovation as they say. Additionally I think the acquisition of RED was a Huge step and commitment to this venture. I think Nikon felt they were being held back by Sony and RED's limitations and licensing that this was a way to free themselves to becoming independent and perhaps a leader in sensor technology for MANY different applications going forward.
I found this quite a bit "clickbaity" as technologies found in smaller sensors does not and cannot necessarily apply to full frame sensors, especially in the consumer camera space. And this one inch sensor is obviously for commercial use, not for consumer cameras.
Perhaps you did not watch the whole video. As I talk about a slower derivative with a larger pixel pitch, not this exact sensor and configuration. Or it may not ever make it to our mirrorless cameras. This development could be trickled down to countless variables. Which might only exist due to the prototyping Nikon engaged in. This is an iteration we know about. I wonder all the iterations we don’t know about.
In the Technology arena, it has progressed so much that businesses have to concentrate on developing their core competancy, instead of being a Jack of all Trade. Sony is just like TSMC a semicon Contract Manufacturer, providing sensor production services, in addition to servicing their own camera business. Obviously, Nikon develops their own sensors, manufactured by Sony.
Salut Matt, merci pour cette vidéo. L'IA pour la bande son traduite automatiquement dans la langue de l'auditeur est très... artificielle. A quand une IA qui peut reproduire fidèlement ta tonalité qui manque et deshumanise totalement ton propos. Peut on la désactiver pour revenir à un sous titrage classique qui nous permet en même temps d'apprendre l'anglais parlé en Australie ?
I am a bit tired about how image quality has gone lately , with privileging speed over image quality. It had been twelve years since the D800 and DR has barely improved. As a video shooter, i would love the dynamic range of an Alexa canera. If nikon wants to give a kick and regain market share, they should offer some distinctive capabilities and havr some independence from Sony.
Nikon cannot build sensors in anything near high volume. ASML builds lithography machines but it's not the leader in semiconductor manufacturing at any commercial node. Who is going to build the sensors for them? Finally, can they scale the technology? it's pretty difficult to do, in particular, FF sensors require multiple exposures. If they get a contract with a phone company for a specialized phone, they could use the money to become competitive, but it's unlikely. There is still hope, on the other hand: this can't be bad news. Automobil as he say is a good option too, as he say.
In no way do i want to say nikon is not putting effort in they know its probably a bad idea to keep depending on sony for their 'cake'. With now Red as part of nikon it seems very smart to evolve to have someone else fab their sensors but...... The claim that sony just bakes them is just well silly at best its a combination of both companies at the moment for the end result there is a reason why the stacked sensors in sony and nikon cameras are so close in testing. Do i feel nikon has gotten past its darkest point and would be smart to plan to have someone else as co-designers for their sensors for highend nikon and red cameras yes maybe the red partner for their sensors can help but its a risky step for nikon to already try that for the z9ii and not work with sony. To me it makes more sense to first expand/test in their film/red lineup before breaking with sony seems to early to me. The example for apple is also kind of proof of that what tsmc can make in their processes really effects apple a lot there is a reason why apple pre-book a lot of time and processes at asml->tsmc for that reason. And sorry nikon in 2024 is not like apple they can't force companies by promising orders of 5B-10B to be exclusive and demand what they want. I 100% agree with the direction nikon wants to go but its not that easy to break away from sony sensor tech just look at canon.
Out of curiosity what do you shoot on now? Fuji and Blad both use Sony sensors and are both known for outstanding image quality / colors. Not sure why one wouldn’t want a Sony made sensor honestly.
@@sequoi_ I shoot on Canon, and while I no longer believe in the "Canon Colors" secret sauce, I think having the entire industry depend on Sony is bad for innovation. I mainly do video and there are some very excting things coming from non-Sony sensors, like the new Blackmagic 12k basically outperforming the Venice 2.
@@avx111 for sure competition breeds innovation can’t argue that. I mostly shoot stills in RAW so the whole “color science” thing has never really been an issue for me. I could see shooting in log caring more about it though.
How to leave an obnoxious comment - not realize he’s talking about something that requires hard specs to then relay an opinion about.. he’s literally making sure you see it’s correct information then giving his take on it. Relax you can always play at a faster speed or scrub through it 👍
@@MattIrwinPhotography I did. Sony does the design AND manufacturing for Nikon's sensors. Your implication in the video was that Nikon may design their own sensor using this high-speed high-dynamic range stacked sensor tech. I find this unlikely as this is not a relationship the two companies have had before, Sony has its own competing stacked sensor technology, and Sony was already producing sensors with similar performance back in 2014.
65 nm is very doable in the 1990s. I would think it would not be difficult to do well below 20 these days. I have a feeling this is how they could achieve the advertised specs.
The lithography on an imagining sensor is not the same as that on CPUs. The smallest modern sensors are at 45nm and that isn't in our cameras, that is in scientific imaging. Your camera is probably 65nm or larger.
You might not need one in that long, but Nikon does. With the A1ii out, but more importantly the R5ii and R1 out, and the insane ai AF being put out by Sony and Canon, Nikon needs an update to make sure it stays competitive as a company. If you personally don't need a Z9ii, just don't buy one. But advancement is always nice
I've said the same in recent conversations. If the Z9 & Z8 were announced tomorrow, with all their latest features & capabilities as of now - these cameras would absolutely still be seen as revolutionary and class-leading. They would serve me well in any genre easily for 10 years (from today).
The camera industry have been lacking on DR compared to speed, it's really annoying as I'm not interested in speed. I'm waiting to buy a high DR nikon camera. I had a z7ii but the af was so bad I couldn't handle it.
@@meatbyproducts yea awful. Eye af was unusable in moderate light conditions, if you are happy with single spot af then it's fine, I would just go with dslr. For work we use Sony, af is 100x better. And that is on an old a7ii and a cheapo zv1!
@airjaff okay, i have had zero issues. I use all brands and take them into combat zones and crazy feild environments. None of the newer Nikon cameras have missed a shot for me nor any of the Soldiers I have worked with. Maybe it was user error.
Best way to start the day. Watching a Matt Irwin special on Nikon gear! Gotta love it!
Agreed. I have my coffee in hand, sitting on the couch with my dog and cat, and headphones on (while others sleep). Let’s do this.
@ haha same! Any time he releases a new video I have to watch it. I only get my Nikon news/gear reviews from Matt. I trust his views on it over other creators
In digital imaging, dynamic range is also expressed in decibels (dB), a logarithmic unit that describes the ratio of maximum to minimum measurable light levels.(1 stop ≈ 6.02 dB)
110 dB ≈ 18.27 stops DR
Amazing. Thank you.
@@MattIrwinPhotography you are welcome!!
Yeah, about 18 stops. Crazy.
Mr. Irwin, this is a very technical video.
I consider myself a geek and you went deeper then I would have. What I enjoyed was your way of explaining this technology. I felt like you were only presenting the basic information. Which made your video easy to watch. Even for me a geek. LOL 😂😂😂
Happy New Gear Matt and everyone!
The interesting thing about this sensor is that it's trading off its high frame rate to increase dynamic range by integrating exposures across sub frames through processing blocks on the lower chip. This can only be done with stacked sensors. They went from 110dB in a single 1000fps frame to 134dB with 16 subframes (lol 22 stops dynamic range).There will probably be some temporal aliasing artifacts with this technique but that should only matter in extreme action shots. I am absolutely excited for this technology and hope it ends up in a Z9II. PS. I used to design digital camera chips a long time ago.
Wow, which chips were you involved in creating?
@@MattIrwinPhotography Early CMOS image sensors at Intel back in the 90's, used in really cheap digital cameras, almost like toys.
Interesting, I do photogrammetry, and I primarily use micro four thirds or aps-c because I need more depth of field; I used to use a sony rx10 mark 4 because it got even more DoF, but I switched to MFT because it offered more versatility and control with interchangeable lenses. A high dynamic 1” sensor would be amazing if I could get Nikon level lenses. This is exciting, I hope they use it for more than just robots and cars (etc), I hope they re-launch their 1” sensor line!
Thank you for bringing up a very cool tech report from Nikon. But, keep in mind this is actually a report from 2022 (its header reads "Nikon Research Report Vol.4 2022") so I wonder how far further along this trail of tech development they are today? Would be really cool to hear what is happening with this tech 🙂
You unpacked a lot of info there Matt. Thanks. It seems to me another great thing about Nikon further developing expertise in visual sensor tech, in this growth industry that you have explained (car tech and robotic tech), means a lot of return-profit will likely sustain and expand R&D efforts for its core products. That's good for all of us.
For sure, pick up robotic and automotive markets, both massive growth industries for camera / sensor manufacturers.
I forgot to tell you Matt, that my son bought the Z 50 II, and is very happy with it and he did watch your video, that I told him to watch, on it and that pushed him to buy it.
Fantastic news Paul. Do you know what lens he purchased for it?
@@MattIrwinPhotography DX 16-50 VR. I got others with the Z 50 when he got that, 16-50, 50-200 both without VR.
Happy New Year Matt. Another wonderfully informative and studied video on Nikon’s progression. I have the Z9 and Z8 and I so much enjoy using them in tandem for my most important tasks. The Z7ii however is my walk around personal camera when I’m with the family and my dog. It is a great camera… small (but not too small), light (compared to the Z8/9) and it does everything I need for landscape and street photography. But there are times when unexpectedly, I need to switch gears and the UI or the camera’s response is not the same or at the level of either the Z8/9. My hope is that there will be a Z7iii with the responsive UI of its big brothers and some of the niceties they possess. I am quite happy with the resolution of all three cameras and since I photograph in raw, my computer storage setup is already nearing 80 terabytes, so I am very happy at the 45.7 mb range. Please Nikon, all I want in 2025 is the 35mm f/1.2 (I do so love “available darkness”photography) and the more responsive Z7iii. Thank you again Matt, for the beautiful photos and the informative videos.
After shooting my Z9 with a 180-600mm lens this week on birds in flight against a bright sky, I am thinking more dynamic range would be a very good thing. Bringing up the shadows on the birds at 340 ISO was extremely noisy.
You're getting it all wrong. The high fps is not intended for saving all those frames as images, but rather for exposing different areas and achieving a single, but very high dynamic range image. So you need the 1000fps for HDR at high fps (120+, even much more when dealing with computer vision)
that high? 1000fps?
Happy New Year Matt! 🥳 Your sensor / cake analogy is absolutely brilliant 👏🏻 Very interesting info here and looking forward to your 2025 content!
I agree that these new sensors are likely going into the automotive application. So many requirements for ADAS, Advance Driver Assistance Systems, the need for high dynamic, fast sensors is the key to making collision avoidance, pedestrian detection and blind side alert function. I'm very impressed with the night vision my new vehicle has when I pull up the camera views.
Happy New Year. If I'm not mistaken there will be a joint booth next week at CES with Mitsubishi for some "car vision" tech. This sensor and its frame rate/DR is perfect for that application. Sure, it can trickle down to photography, but I'd very much love to see Nikon diversify in such a way that they can send more R&D money down the pipe to the imaging division. Automotive could prove to be extremely lucrative for Nikon. Glad someone on the UA-cam has finally figured out that Nikon and other brands aren't buying SKU'd/off the shelf sensors from Sony but sending their designs off to be manufactured by someone who can do it at scale. Will be an exciting year for photography! Cheers!
1” sensor pixel pitch 2.7um, sounds familiar, Nikon 1 v3 also 17mp and worked quite well, a sensor I’ve used on two occasions to capture surprisingly (for me) good images of a solar eclipse.
I believe the 1 series cameras were using a 20mp Sony 1 inch sensor
"What's special about Today's primes that they are extraordinary wide open" - that's true. I've the 35, 50, and 85 S primes (1.8, not the 1.4) and more than 70% of the time I'm shooting them wide open (that's also why I got the S/1.8s and not the 1.4s). Still I get amazing sharpness, and the only reason I stop them down sometimes is to have bigger DOF.
I’m considering the 35mm S f/1.8 for Astro. Just how sharp is it in the corners on stars? I would try the 1.4 but it has NO manual focus button.
DR (dB) = 20log10(Nfull/Nnoise), DR (stops) = log2(Nfull/Nnoise). 20log10(2) = 6.02. So to convert stops to dB multiply stops by 6.02 or to turn dB into stops, divide dB by 6.02.
Yes, got my Nerd on Matt
Fascinating thanks for sharing
Happy New Year Matt!!
BTW, the dB measurement is the span of signal in the darkest to brightest areas, the larger the better. Recall that dB is logarithmic so conversion looks like this: Stops = log2(10^(dB/20))
Also note that this range does not specify the wave length but we can assume that it’s visible light they are talking about. Seeing a little below human range towards infrared would help with night vision though. Imagine being able to access that on the same sensor in a consumer camera with a toggle.
It is an interesting announcement. Even if it is really being developed for cars, the potential spin offs will inevitably find their way into photography. For cars, better fast sensor/guidance/warning and control systems are in high demand. Modern vehicles do bing and bong little warnings as you drive which is fine when they are correct but unhelpful if you are actually moving to avoid a pot hole or obstacle. Improving sensor accuracy is a key to improving processing driving assistance. However, if you have the technology for driving what else can you use it for? Autofocus anybody?
I'll take the dynamic range, but I don't need 1000fps. Actually, if Nikon could even bring Z9 sensor performance to a global shutter (or ultra fast rolling/global shutter hybrid) I'd call that a big win.
Thanks Matt and Happy to New Year to you and yours.
So for me and many the current Z9 is almost perfect as it is for both stills and videos -- sure I want all the best of what is in the Z8 and Z6iii to be added in the next FW. AND yes there are a few long standing "improvements" that many have been asking for for some time -- ability to engage or disable Subject Detection, Multiple shooting recall hold per menu/shooting bank, some ability to export files from the camera onto an SSD (even if this has to have its own power), more video improvements (false colour).AND then the real winners if these are possible without physical changes -- 30FPS RAW and Pre-Capture in RAW - if not physically possible then these are must haves for a Z9ii.
Then we have the Z9ii -- I am very happy with the current sensor in the Z9 -- as a predominantly wildlife shooter I would have preferred the base ISO to have been tuned to ISO 100 or 200 and a correspondingly cleaner/larger DR as higher ISO levels are used. I know of FEW sports shooters who do not also want even better low light performance -- I certainly do not want the "rubbish" performance the current Global Shutter hybrid camera delivers. Bigger/faster Buffer; Option for built in SSD; What about the NEW and even larger CF-Express Type C cards -- when will we see a body which uses these? Surely "better" (meaning higher res and higher refresh rate) EVF and LCD. I would like to see a 2nd USB-C data port and all such ports be USB-C 4 rated.
A 1" sensor is much smaller than I expect in an FX body -- but the Global Shutter sensor in the $45k RED Raptor XL has a 40.96 mm x 21.60 mm (Diagonal: 46.31 mm) sensor, which is 3% larger area than a standard FX is 35.9 x 23.9 sized sensor. The Raptor sensor is 35.4MP which is 77% of the Z9 sensor's resolution. AND I doubt the Raptor's sensor has AF "points" built in -- but still a short walk from the current RED sensor. {I have little doubt this would take a while to develop for a Z9 model}
Would shooters be willing to accept a Z9 variant with a 35.4MP global shutter sensor based on the RED design? What are the trade offs? [Even better DR and low light performance would be likely] - remember the Raptor XL offers 120 fps at 8K 17:9, 150 fps at 8K 2.4:1 and 2K (2.4:1) at 600 frames per second.
We are still to see a new Nikon RED camera or a Nikon Hybrid that supports either bolt on or built in active cooling -- bolt on options have been tried with the Z6iii -- it would be far better to have these built by Nikon for a specific body (which would have to have a flippy screen)
I loved Melbourne when I visited for work. There times when I felt like I was in Toronto, we have a lot of similar ‘isms and street names.
110dB of dynamic range (log base 10 notation) is equivalent to just over 33 stops (log base 2 notation), but it really doesn’t translate like that unfortunately, because of things like the final ADC being 12 bit, and the eventual need for bayer filter arrays and debayering. it’s a high sensitivity/dynamic range sensor for sure, but the only way to know how it’ll compare to current sensors is to slap it in a camera and find out!
Happy to you, Matt
Thank you for that interesting article. It seems analogous to min-LED TV screens where small LED blocks are individually lit to give better contrast ratios. However, those TVs apparently can get things wrong such as rendering a small spot of light - the Nikon paper does not mention what happens if the lightness levels differs within one of those 16x16 exposure blocks. Just a warning that this technology may be good enough for industrial sensors but not for pixel peeping photographers :)
In regular chips, 65 nanometre is huge. The highest resolution steppers now are at 1 nanometre. Probably only the largest foundry (aka fab) (TSMC) has one machine at that resolution test printing (these print chips). And their finest operational machine will be something like 3 nanometre.
In photography we need some area per photosite, or else we loose to much light sensitivity. In that sense the 65nm Nikon stepper (chip printer) is a great thing.
Nikon (and Canon) have been leading this stepper machine market until about 2,000 when they lost their commercial edge in the regular chip market. In recent years, AFAIK Nikon has been R&Ding, looking to make steppers for this tiny scale market again. But the way I see it, no sensor will be made with 1 nm tech, or 3, or 7, or. Because light sensitivity. If that were possible, that would be a major revolutionary step change of several orders of magnitude. It would also facilitate maybe to have sub-photosites that together generate RGB instead of monochrome raw data. No, not T-cells, like Fuji. But demarcated regular shape sub-photosites.
But that's a dream.
Yes I agree JP, photography will never utilise or need the lowest nanometer tech for light capturing. But lower nanometer chips may be used for on sensor processing, RAM, cpu's and more. And thus will still impact photography in general, just not the light collection part. :) Exciting times ahead.
This sensor was developed about 4 years ago. It is for industrial use. The actual images that it produces aren't very photographic.
Did you watch the whole video?
Happy new year Matt. The most likely thing is that this sensor in FX format and with a higher number of pixels we are already enjoying it on the Z9 and Z8
110dB = 18.33 stop of DR!! Extraordinary!!
Indeed! (Human eye 20-24)
I believe 110 db is approximately 18 stops of dynamic range.
This is what a quick search came up with…
“110 dB is equivalent to roughly 18 stops of dynamic range, as each "stop" in photography is considered a doubling of light intensity, which translates to approximately 6 dB in the logarithmic scale used for decibels”
Thanks so much for posting the info. Do you have a link. I’d love to read more about it. 😀
Hi Matt - very interesting video!
I wonder if part of the 1000 FPS is the ability to condense multiple shots into one as they do reference "computational" photography. Like on an Iphone, my understanding of how a phone creates very low light shots is through computational photography - essentially multiple exposures are taken very quickly and then combined to make a better single image - essentially its HDR but backed in and done very quickly.
That said, this new tech does seem different in that it appears that the sensor itself is able to produce different exposures based on the different zones, so I get why one would think that the need to merge different frames is circumvented (and maybe it is), but I wonder if they are doing a little of both here.
As you point out 1000 FPS seems a bit absurd and is definitely overkill, but if they are combining frames then this number would rapidly diminish - also 1 second is too long of a sample time to combine frames as there will be motion, but the camera could capture say 10 frames almost instantaneously in 1/100th of a second - then maybe with the combination of the zones and the available data of the 10 frames all of which an have slightly different exposure they may be able to produce a very impressive single frame with DR - like 20+ stops - who knows?
All of this is speculation, but as you point out, whatever the case it is clear that things are progressing rapidly and that Nikon products have never been better. Indeed the future is very bright for this system - now we just need an inexpensive FF entry level camera like the Z5II to really get the newcomers to start with Nikon and boost that marketshare 😂. In all seriousness, I am sure something like that is on the near horizon.
Thanks Matt!
-PD
Hi Matt, happy New Year to you too! According to ChatGTP, Decibel can be converted, its 6.02 DB per stop (giving us 18.27 stops). No idea if correct, but perhaps an avenue you can pursue!
6 dB generally represents a doubling of the voltage signal, which I assume is what they measure from the sensor so that could be true.
I can confirm that number. I think they can use Sub-pixel Architecture when there are 2 pixels with different size, so they can achieve multi-exposure in one shot.
according to chat GPT: A dynamic range of 110 dB corresponds to approximately 18.3 EV (stops).
My Z9 and Zf are just about perfect, but I'll grab a Z9II two years after it's released and easy to get (with firmware bugs fixed). That's how to upgrade. Always wait for the dust to settle.
+ all fault parts already called-out and replaced as Nikon tradition 😅😅
I'm a beauty and fashion photographer and have a z8. I would love a camera with a higher flash sync speed
I think the tech is awesome. Maybe they can incorporate it with MRMC/movie industry. Something similar to Freefly Ember system.
Zed Talks are back !
Big nerdy video is why I’m here. Bring it.
I am thinking this sensor could be useful in the medical field. Cleaner cell images. Whatever its developed for may we as photographers benefit also.
I like the BNV alert label 🎉 I seem to remember that Canon also released some info on a 17mpix not long before the R1 announcement. I believe this was to reduce the disappointment of their flagship “only” having 24Mpix, compared to the larger MPix Sony and Nikon flagships. It’s possible Nikon is doing the same before a Z9ii with 20-24 MPix perhaps...
Hi Matt, Hope you have a fantastic 2025 ! It's funny when Nikon were considered late to the digital camera scene many, many people said Nikon are dead, they don't have the technical abilities to be competitive in such a market, those very same people were probably ignorant of the huge body of experience Nikon have in the specialist imaging field, Nikon have been world leaders in many different aspects of photography and imaging for decades, I think their future is very bright! Cheers.
The worry was real we are all happy nikon made it but it was only because of some really nice last models in camera world. in litho they (and canon) have been under great pressure from asml.
What about military use? Expense is then less of a problem.
Cool !! Grok AI tells me 110 db is about 18.2 stops of HDR!!! I wouldn't mind a new Nikon1 camera with 18 stops and 17mp!! 134 db equals 22 stops !!! madness the future of camera tech is bright !
Interesting, but I’ll bet commercialization won’t be for some time after the Z9ii.
Happy New Year, Matt. Nikon is not resting on their laurels ... and they have plenty to rest on.
I've been saying for a few years now, Nikon is trying to remove itself from Sony. This type of sensor looks like it's going to before science and space. I can see the cars too.
Knowing Nikon, they're not changing the Z9 sensor for several generations. If anything, they'll create a new camera series.
If they were to go to the global shutter. It'd be called a Z1, Z100, etc.
By the way, Optimus already has two very good high dynamic rangecsensir, in the operator remotely controling it behind the curtain lol.
Have you considered broadcast TV cameras?
I suspect. as you do, that this R & D is aimed at things like cars, robots, security etc etc. Especially cars. That said, things learned can be used for multiple applications. Personally, as far as cameras are concerned, I rarely shoot more than 10 Frames per second…most often far less. I would much much much rather a focus on dynamic range…give me an extra 2 stops of quality DR vs speed.
one inch sensor? I say time for the Nikon1 J6 🤩
Or Sony RX10 rival. There were roomers some years ago (DL) but they didn't came to product.
@boazsayar1193 there are quiet a good amount of one inch sensor cameras, i used o take my Panasonic one inch sensor, fixed lens camera to concerts where theres no bigger camera allowed and managed to get some decent results with it
i believe it was the lx 15 or something like that
A rugged version that is waterproof etc.
@@obedbrinkman I have Canon G3X, but it is getting old and I need a replacement. All cameras in the market now are pretty old. It is a great camera for travel. I take great photos.
We're getting close to peak camera technology. Just glad to see Nikon innovating and putting the haters to bed.
The human eye can respond to pretty much any and all lighting conditions and adapt the amount of light entering through the pupil.
It doesn’t make sense to refer to dynamic range for two reasons.
1. The eye picks up ALL light entering through the pupil and it’s absorbed via the retina.
2. The brain adapts light and defines contrast in an adaptive way. Ie, if you live in darker areas, you brain will adapt to differentiate between smaller difference of light. If you live in light areas, it will adapt to only differentiate larger differences.
Human vision doesn’t make sense to define in the same way as a camera, though the concept is similar, a camera is not a dynamic organ like the eye/brain vision system.
NASA uses Nikon cameras, this sensor would be great for the next Mars rover.
100 mp at 20 fps with 12 stops of DR
100 mp at 10 fps with 15 stops in a medium format Nikon with Nikons autofocus
Whats not, 100mp at 120fps with 17+ stops of DR
Nikon making their own sensors will separate them form the field and no longer having to depend on Sony's basic sensor hardware manufacturing it for them. It's a great business move for their own bottom line as well as creating their own unique imaging ecosystem, plus as you detailed are open to pushing the technology into the future more efficiently for themselves and others who want to buy into it. Nikon is basically setting themselves up to be or better themselves compared to Sony... Competition drives innovation as they say. Additionally I think the acquisition of RED was a Huge step and commitment to this venture. I think Nikon felt they were being held back by Sony and RED's limitations and licensing that this was a way to free themselves to becoming independent and perhaps a leader in sensor technology for MANY different applications going forward.
I think it is already in the z9!! They will unlock it in the next firmware update!!
Photography has moved from a professional art to a "spray and pray" hobby.
I found this quite a bit "clickbaity" as technologies found in smaller sensors does not and cannot necessarily apply to full frame sensors, especially in the consumer camera space. And this one inch sensor is obviously for commercial use, not for consumer cameras.
Perhaps you did not watch the whole video. As I talk about a slower derivative with a larger pixel pitch, not this exact sensor and configuration. Or it may not ever make it to our mirrorless cameras. This development could be trickled down to countless variables. Which might only exist due to the prototyping Nikon engaged in. This is an iteration we know about. I wonder all the iterations we don’t know about.
O think d700 and z5 have rly cool sensors collor
id much rather 18mp and better D.R. not really arsed about the 1k fps thing though
oh i see why they're pushing for ultra high fps, thats how they get increased D.R. by stacking and auto-aligning with ai. clever.
In the Technology arena, it has progressed so much that businesses have to concentrate on developing their core competancy, instead of being a Jack of all Trade. Sony is just like TSMC a semicon Contract Manufacturer, providing sensor production services, in addition to servicing their own camera business. Obviously, Nikon develops their own sensors, manufactured by Sony.
110 dB corresponds to about 36-37 stops of dynamic range.
If Nikon makes a medium format camera ..... Does Nikon have a technical cooperation with Hasselblad?
Waiting for a 60 Mpix camera and a FTZ Screwdriver AF adapter. Else my D850 is good enouth
Salut Matt, merci pour cette vidéo. L'IA pour la bande son traduite automatiquement dans la langue de l'auditeur est très... artificielle. A quand une IA qui peut reproduire fidèlement ta tonalité qui manque et deshumanise totalement ton propos. Peut on la désactiver pour revenir à un sous titrage classique qui nous permet en même temps d'apprendre l'anglais parlé en Australie ?
Yes. I think it is settings at the bottom of the video. You can change back to my voice. 😀
Ok, ah, merci, j'ai trouvé, beaucoup mieux 😅
Nikon we want 60+ megapixels😢
I want a 100 mp sensor. I'm tired of Fugi having one but not Nikon.
110 dB is more than 18 stops, 6 dB = 1 stop
I am a bit tired about how image quality has gone lately , with privileging speed over image quality. It had been twelve years since the D800 and DR has barely improved. As a video shooter, i would love the dynamic range of an Alexa canera. If nikon wants to give a kick and regain market share, they should offer some distinctive capabilities and havr some independence from Sony.
Human eye has a dr of 20-24, about 1000 times more than any camera sensor.
Nikon cannot build sensors in anything near high volume. ASML builds lithography machines but it's not the leader in semiconductor manufacturing at any commercial node. Who is going to build the sensors for them? Finally, can they scale the technology? it's pretty difficult to do, in particular, FF sensors require multiple exposures. If they get a contract with a phone company for a specialized phone, they could use the money to become competitive, but it's unlikely. There is still hope, on the other hand: this can't be bad news. Automobil as he say is a good option too, as he say.
I take one frame every 5 minutes … hmm
In no way do i want to say nikon is not putting effort in they know its probably a bad idea to keep depending on sony for their 'cake'. With now Red as part of nikon it seems very smart to evolve to have someone else fab their sensors but...... The claim that sony just bakes them is just well silly at best its a combination of both companies at the moment for the end result there is a reason why the stacked sensors in sony and nikon cameras are so close in testing. Do i feel nikon has gotten past its darkest point and would be smart to plan to have someone else as co-designers for their sensors for highend nikon and red cameras yes maybe the red partner for their sensors can help but its a risky step for nikon to already try that for the z9ii and not work with sony. To me it makes more sense to first expand/test in their film/red lineup before breaking with sony seems to early to me. The example for apple is also kind of proof of that what tsmc can make in their processes really effects apple a lot there is a reason why apple pre-book a lot of time and processes at asml->tsmc for that reason. And sorry nikon in 2024 is not like apple they can't force companies by promising orders of 5B-10B to be exclusive and demand what they want. I 100% agree with the direction nikon wants to go but its not that easy to break away from sony sensor tech just look at canon.
At min light yourself as bright as the outside so you can cut the exposure that way
I love what Nikon is doing lately but I still can't bear to use a Sony sensor. They need to put RED sensors in their cameras ASAP.
Out of curiosity what do you shoot on now? Fuji and Blad both use Sony sensors and are both known for outstanding image quality / colors. Not sure why one wouldn’t want a Sony made sensor honestly.
@@sequoi_ I shoot on Canon, and while I no longer believe in the "Canon Colors" secret sauce, I think having the entire industry depend on Sony is bad for innovation. I mainly do video and there are some very excting things coming from non-Sony sensors, like the new Blackmagic 12k basically outperforming the Venice 2.
@@avx111 for sure competition breeds innovation can’t argue that. I mostly shoot stills in RAW so the whole “color science” thing has never really been an issue for me. I could see shooting in log caring more about it though.
How to drag out video time - read the specification on screen.
How to leave an obnoxious comment - not realize he’s talking about something that requires hard specs to then relay an opinion about.. he’s literally making sure you see it’s correct information then giving his take on it. Relax you can always play at a faster speed or scrub through it 👍
I'd just like to point out that Nikon does not make their own sensors. They use Sony sensors.
I suppose you did not what the whole video.
@@MattIrwinPhotography I did. Sony does the design AND manufacturing for Nikon's sensors. Your implication in the video was that Nikon may design their own sensor using this high-speed high-dynamic range stacked sensor tech. I find this unlikely as this is not a relationship the two companies have had before, Sony has its own competing stacked sensor technology, and Sony was already producing sensors with similar performance back in 2014.
65 nm is very doable in the 1990s. I would think it would not be difficult to do well below 20 these days. I have a feeling this is how they could achieve the advertised specs.
The lithography on an imagining sensor is not the same as that on CPUs. The smallest modern sensors are at 45nm and that isn't in our cameras, that is in scientific imaging. Your camera is probably 65nm or larger.
Title has wrong language
I don't see the point of the z9 II in the next 5 years.
You might not need one in that long, but Nikon does. With the A1ii out, but more importantly the R5ii and R1 out, and the insane ai AF being put out by Sony and Canon, Nikon needs an update to make sure it stays competitive as a company.
If you personally don't need a Z9ii, just don't buy one.
But advancement is always nice
I've said the same in recent conversations. If the Z9 & Z8 were announced tomorrow, with all their latest features & capabilities as of now - these cameras would absolutely still be seen as revolutionary and class-leading. They would serve me well in any genre easily for 10 years (from today).
The camera industry have been lacking on DR compared to speed, it's really annoying as I'm not interested in speed. I'm waiting to buy a high DR nikon camera. I had a z7ii but the af was so bad I couldn't handle it.
Certainly more DR advancements would be great to see.
The AF was bad?
@@meatbyproducts yea awful. Eye af was unusable in moderate light conditions, if you are happy with single spot af then it's fine, I would just go with dslr. For work we use Sony, af is 100x better. And that is on an old a7ii and a cheapo zv1!
@airjaff okay, i have had zero issues. I use all brands and take them into combat zones and crazy feild environments. None of the newer Nikon cameras have missed a shot for me nor any of the Soldiers I have worked with. Maybe it was user error.
@meatbyproducts not user error at all. Rude!
hi. Please put a film over your back window to cut the light . It's ugly seeing blown out windows like that
I think it looks great.
I'd love to see that sensor tech in a Z9 but I may have to collect it with my flying pig...🦘