The today sensor technology is reaching the limits of evolution. Color, noise, low light, dynamic range, resolution, fps, rolling shutter, everything has improved great.
So it seems only a new technology can make things better from now.
So the question is: will a new sensor technology come soon? Are the manufacturers developing a game changer sensor technology?
Yes. It is called AI :-)
And overall cameras production will drop and drop, as in the flood of AI stuff it'll be much less reason to buy camera.
“If we [make a] new camera, we can choose that kind of sensor. But it’s kind of an R&D matter. So it is not [certain] that we will use that sensor in the next camera or not.”
Well, as always they never tell you the whole story. It could be coming in a Panasonic camera this year and they would never tell you until it's official.
Dual layer sensor for low noise and better dynamic range:
https://www.sony-semicon.com/en/technology/mobile/2-layer-pixel.html
Old stuff, but good to know: dgo sensor - dual gain output:
https://www.canon-europe.com/pro/stories/dgo-sensor-explained/
Omnivision shows hdr sensor:
Interesting
Sony imx490 hdr sensor, 4 different exposures at same time, merge into hdr video:
Different manufacturers are showing hdr sensors but not for mirrorles camera yet...
the technology is ready to release a mirrorless camera with hdr sensor to solve the shadow/highlight problems and avoid using log profiles.
they just need to implement a triple native iso 800/3200/12800 or 640/2500/10000 and also implement a triple gain output, all at the same time to get 3 different exposures for each frame and then merge into a single frame inside camera.
dual native iso and triple native iso cameras already exists, and dual gain sensors already exists. hdr sensors for automotive and security cameras already exists.
so the question is: why are the manufactures avoiding to release a mirrorles or cinema camera with hdr sensor? maybe the log profile cameras are selling good and they want to make profit from it and just release the hdr sensor when the sales drop significantly...
there is a rumor about a S1H Mark ii with hdr sensor, but it seems to be a "i want this" rumor, instead of a real rumor:
https://thenewcamera.com/panasonic-s1h-mark-ii-specification/
Panasonic S1H II Rumored Specs
100dB DCG-HDR 22.5 megapixels BSI CMOS sensor
Triple native ISO 640/2500/10000
DR Boost on 3240p/80fps, 2160p/120fps, 1080p/240fps
Super-fast phase hybrid autofocus
ProRes RAW and ProRes RAW HQ internal recording
17stop Arri LogC4 gamma profile
Built-in dual fans for unlimited recording
60 fps burst shooting with AE/AF and pre-capture
Dynamic resolution function via subPDs with interlaced bayer CFA
Full Frame Open Gate /120fps (SSR: 5760×3840)
Full Frame 16:9 3240p/160fps (SSR: 5760×3240)
Full Frame 16:9 2160p/240fps (SSR: 5760×2160)
Full Frame 16:9 1080p/480fps (SSR: 5760×1080)
Super 35mm A6:5 3240p/160fps (SSR: 3840×3240)
Super 35mm 16:9 2160p/240fps (SSR: 3840×2160) Super 35mm 16:9 1080p/
480fps (SSR: 3840×1080)
HEVC 10bit 4:4:4 Intra/LongGOP, AVC 10bit 4:2:2 Intra/LongGOP codecs at all resolution
9.44M ultra-bright 40 00nits 120Hz EVF
Rumors from that website never turn out to be true!
Yes, this rumor is just people wishing...
so let's talk about wishes...
imagine a camera with triple native iso: 800 / 3200 / 12800. the first iso 800 can vary from 400 to 1600. the second iso 3200 can vary from 1600 to 6400. the third iso 12800 can vary from 6400 to 25600. all can vary in 1/3 fstop. each native iso deliver a frame, so you can get 3 simultaneous frames to merge in camera. and you can choose each iso to do the best combination for each scene. this would be a dream, the perfect camera, because you could adjust the middle exposure first and then how much to lift the shadows and how much to preserve the highlights...
if this camera become true someday it will be expensive, because this is all that pro people wants... and maybe the M43 will do this for the low budget people...
These tests shows the Alexa with 17 fstops dynamic range preserving the highlights much better than Sony:
https://www.redsharknews.com/how-can-the-alexa-35-achieve-17-stops-of-dynamic-range
https://www.arri.com/en/camera-systems/cameras/alexa-35/alexa-35-high-dynamic-range
DR Boost GH6 13 fstops, but we want more, 17 or 20 fstops...
the old sony a7sii and as7iii with 13 fstops since long time ago, if sony implement dual gain merge or dual layer sensor maybe it can rise to 15 fstops.
https://www.cined.com/sony-a7s-iii-lab-test-does-it-live-up-to-the-hype/
when you want more dynamic range the cameras price start to increase a lot, and even if you save money to pay more, you will never be completely satisfied, because the limit for average budget is 13 fstops dr today. so i decided to keep using my G95 with vlogl around 10 fstops dr.
Yes, sensor technology is truly a game changer. It’s transforming industries like healthcare, automotive, smart homes, and environmental monitoring. In healthcare, sensors enable remote patient care and better diagnostics. In cars, they’re key to self-driving technology. Smart homes use sensors for automation, making life easier and more energy-efficient. Wearable sensors help track fitness, while environmental sensors monitor pollution. With the growth of IoT, sensors are connecting everything, creating smarter systems that improve efficiency and safety. As the technology advances, its potential to drive innovation is enormous.
Honestly, a lot of the improvement seen above is based on properly mapping a wide DR image. Not feasible in batch processing today due to software limitations. Otherwise, the 12-14 stops today's cameras capture is enough to do most of what was shown.
8-year-old Sony A6500, single frame, exposed for the highlights, lens was the obvious limitation.
In this video recorded at sunrise/sunset it is possible to see the sky very bright and shadows very dark and the fire gets white overexposed when it is big in the scene, it is GH6 dr boost. So the conclusion is: dr boost with dual gain sensor can increase dynamic range from 11 to 13 fstops but it is not enough for all situations. If manufacturers implement dr boost with triple gain sensor it will increase to 15 fstops and this will be excelent for mirrorless cameras with the technology available today. 15 fstops in low budget cameras would be amazing, maybe it will happen someday.
i was trying to deal with a light situation in a room with a translucid white curtain in the window and a sofa under dim light from the window. if i set the correct camera exposure for the curtain the sofa gets 2 fstop underexposed, and if i expose correct for the sofa the curtain gets completely overexposed. i was thinking the problem was the camera dynamic range but i supose even the new cameras with 2 more fstop dynamic range would not solve the problem because the light difference would continue to exist. so the simple conclusion is: add light to the sofa. but low power led lights cannot solve the problem because they are not powerfull enough. so i did a research and i found a 400w led panel, 200w 5500k 200w 3200k 400w 4300k, so i hope this led light can solve the situation and i will try it with a soft white diffusion in front to avoid hard shadows. and if adding light cannot solve, also another simple solution is to avoid backlight situations when the camera cannot deal with the light difference between the background and the main subject, or just let the background overexpose and be happy. But there is a problem the 400w or 200w panels are 10kg weight, so the possible alternative is the yongnuo 72w 1,7kg. And then use the light source more close to the subject, and the dynamic range will be what is possible to do.
window backlight: dealing with dynamic range challenge
sometimes we start to believe in things that are difficult or almost impossible to do. magic lantern developed hdr video for up to 5 fstops difference in 5 exposures but in consecutive different frames which results in motion artifacts after merge. blackmagic introduced raw video and showed a scene in interior against a open window but not real time editing and time consuming in post. then cameras manufacturers start to put log profiles in camera to record flat image which have a difficult and time consuming grading in post. now we have dr boost and dgo sensors but only increase dynamic range a little bit somekind of two fstops more. all in an attempt to sell the idea of capturing the real life light to preserve the scene mood avoiding to add artificial light. and we see lots of advertising, reviews and tests that make us hungry to have it and to do it. But the old school solution continues to be the best practice: add light! Also there are big nd gel sheets for situations where it is possible, like cars... And so try to build a convincing light scene to simulate real life light.
the best light solution i found is the zhiyun g300, two units can equal to the sun at 1m distance
It looks like you're new here. If you want to get involved, click one of these buttons!