Image Sensor dynamics ~ 20 stops !!

[ahaslett]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

It's not true! It's much easier shooting RAW only, and less work overall. The only advantage of jpegs is instant availability. Exporting 1,000 keepers from a processor takes time and you need an assistant or to return to base. (I don't typically take more than 50-100 images per day, even in a photo rich environment).

Knowing one set of jpeg parameters and having the power, screen and keyboard access of a RAW processor makes things far easier than trying to adjust jpeg settings in camera.

You just have to get past the "jpeg+RAW" and "right in camera" mindsets, and understand how to expose to use the full DR of your body.

Also, you stop worrying about EVF and LCD IQ because you don't need to chimp, except maybe to check critical focus.

A

 

[tbcass]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

It's not true! It's much easier shooting RAW only, and less work overall.

I disagree. I shot JPEG for 19 years before I switched to RAW. RAW takes a bit more work and produces better results than JPEG no matter how much you change the in camera settings. That said I don't think JPEG users are lazy.

 

[ahaslett]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

It's not true! It's much easier shooting RAW only, and less work overall.

I disagree. I shot JPEG for 19 years before I switched to RAW. RAW takes a bit more work and produces better results than JPEG no matter how much you change the in camera settings. That said I don't think JPEG users are lazy.

Did you really adjust your jpeg settings for every shot? Would you be happy switching between cameras with very different jpeg engines? How did you use two different bodies at the same time, and match the look reasonably well without a lot of work?

I shoot a fair bit of landscape, so it only took me 10 years to realise that shooting jpegs was just frustating. I shot RAW+jpeg for a while before realising that the jpegs just cluttered up my workflow, and then that exposing for jpeg is a mistake (especially for landscape shooters).

Now I'm used to C1 and shooting RAW only, I just find jpeg shooting a lot of work. YMMV.

A

 

[Cherry_F]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

I'd say they use technology in order to save time, and dedicate themselves to whatever they consider more important.

I remember some scientist once said that there are only two personality traits that all animals (uncluding the human species) have in common: we don't like seing someone else hurt, and we are lazy.

Lionnesses for instance, go for hunting only when hungry. And spend much time playing with their cubs. I wish I were one !

___
Photography is so easy, that's what makes it highly difficult - Robert Delpire

 

[tbcass]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

It's not true! It's much easier shooting RAW only, and less work overall.

I disagree. I shot JPEG for 19 years before I switched to RAW. RAW takes a bit more work and produces better results than JPEG no matter how much you change the in camera settings. That said I don't think JPEG users are lazy.

Did you really adjust your jpeg settings for every shot? Would you be happy switching between cameras with very different jpeg engines? How did you use two different bodies at the same time, and match the look reasonably well without a lot of work?

That's a false equivalence. No I did not adjust for every shot other than the same things I adjust when shooting RAW and no JPEG shooters I know of have ever done what you are suggesting. That is why JPEG shooting is easier than RAW. JPEG shooters are happy with whatever the camera gives them. When they are not happy they shoot RAW. I will add that noise reduction while retaining detail is universally better with a good RAW developer than in camera noise suppression for JPEGs.

 

[ahaslett]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

It's not true! It's much easier shooting RAW only, and less work overall.

I disagree. I shot JPEG for 19 years before I switched to RAW. RAW takes a bit more work and produces better results than JPEG no matter how much you change the in camera settings. That said I don't think JPEG users are lazy.

Did you really adjust your jpeg settings for every shot? Would you be happy switching between cameras with very different jpeg engines? How did you use two different bodies at the same time, and match the look reasonably well without a lot of work?

That's a false equivalence. No I did not adjust for every shot other than the same things I adjust when shooting RAW and no JPEG shooters I know of have ever done what you are suggesting. That is why JPEG shooting is easier than RAW. JPEG shooters are happy with whatever the camera gives them. When they are not happy they shoot RAW. I will add that noise reduction while retaining detail is universally better with a good RAW developer than in camera noise suppression for JPEGs.

You can see then why I shoot RAW (all of the above).

A

 

[A74Me]

I sure can obtain the same results when processing raw files, but I'm not sure everyone is willing to do so. There was a popular saying in my trade : "software makes progress because its users are lazy".

Are you saying non RAW shooters are "lazy"?

It's not true! It's much easier shooting RAW only, and less work overall.

I disagree. I shot JPEG for 19 years before I switched to RAW. RAW takes a bit more work and produces better results than JPEG no matter how much you change the in camera settings. That said I don't think JPEG users are lazy.

Did you really adjust your jpeg settings for every shot? Would you be happy switching between cameras with very different jpeg engines? How did you use two different bodies at the same time, and match the look reasonably well without a lot of work?

That's a false equivalence. No I did not adjust for every shot other than the same things I adjust when shooting RAW and no JPEG shooters I know of have ever done what you are suggesting. That is why JPEG shooting is easier than RAW. JPEG shooters are happy with whatever the camera gives them. When they are not happy they shoot RAW. I will add that noise reduction while retaining detail is universally better with a good RAW developer than in camera noise suppression for JPEGs.

You can see then why I shoot RAW (all of the above).

A

for skin tones what cracks me up is i have my incamera jpegs tuned to perfection , so you take the raw file into ACR and hit "camera settings"

 

[Satyaa]

Image sensors with a dynamic range of 20 stops.

It’s known that Apple, Canon & Sony are working on producing consumer image sensors with a far higher dynamic range than the current sensor market.

The smartphone industry are the cash cows providing enough R&D cash to get Sony, for example, to have already on the market the consumer image sensor - the Sony LYT-828 which has a high dynamic range performance of over 100 dB.

These high dynamic range consumer image sensors are going to be the challenge to all camera manufactures.

OM Systems for example will have to depend heavily on either JIP providing cash to OM-S to be able to incorporate these sensors into their cameras, or JIP makes use of its newly acquired Toshiba company to do something. Toshiba already fabricates image sensors, so there is just a very small chance this is possible.

I think it highly unlikely that OM-S would ask Canon to supply it with their sensor.

Nikon will perhaps have Sony make a high dynamic sensor made to Nikon's own specifications.

OM-S has, I presume, a pretty go relationship with Sony Semiconductor Solutions Corporation. Well, let’s hope so.

I guess the question here is by how much will this type of smartphone affect the camera market?

I think your question above is partly in the context of the phone camera prices that you listed below (I could be wrong). In other words, how will these 'more capable' phone cameras further eat into dedicated/compact camera market?

I don't go to a phone camera for my photography hobby for multiple reasons. Let's leave that topic because it's been beaten to death.

My question is in the opposite direction. How can these sensors/technologies revive compact/bridge cameras?

In the past I have been interested in bridge cameras a lot. Panasonic FZ80 with 20-1200mm lens was particularly appealing. Nikon's P950/P1100 did not because the lenses were not the best. For IQ reasons, I did not go with sensors smaller than type 1.

I tried a couple of different type 1 sensor bridge cameras (still use one). They are excellent in good light but left lot to be desired in low light. They were all old sensor tech. I then settled with APSC/M43 as the sweet spot for minimum acceptable IQ for my needs.

Lot of tech available today was not in 2016. Most significant might be NR. This definitely lends new life to smaller sensors in low light situations. ISO invariance of the sensor combined with dual/triple gain will also help. Pixel-shift hi-res shooting is another that has proved useful and now included by most camera brands in some form.

Now coming to the article and tech that you referred to, if we start getting good sensors larger than 1/2.3", especially if the type 1 sensor size is revised with stacked, dual/triple gain and pixel-shift technologies, then can we expect a revival of bridge cameras?

That would be a reasonable and affordable option for hobbyists like me. Instead of handful of ILC bodies and a dozen lenses, I would be happy with two good bridge cameras. One could be a rangefinder style compact up to 120mm eq, and another could be a longer super zoom up to 1200mm eq.

While the larger expensive cameras will still be for serious amateurs and professionals, this could be the volume segment for everyone else. Camera brands lagging behind the leaders in larger sensor cameras can revolutionize this segment.

The cost of the Vivo X300 :

The Vivo X300 Pro begins at:
• ¥5,299 (~$744 / €643) for 12 GB + 256 GB
• ¥5,999 (~$843 / €728) for 16 GB + 512 GB
• ¥6,699 (~$941 / €813) for 16 GB + 1 TB
• ¥8,299 (~$1,166 / €1,008) for the X300 Pro Photographer Edition

HDR images produced using this new sensor LYT-828 are shown here :

https://www.provideocoalition.com/sony-lyt-828-a-new-50mp-sensor-for-smartphones/

Unboxing the Vivo X300 with Zeiss optics including a telescope : (This can be ignored if it’s already known.)

GSMArena has provided numerous images using the Vivo X300 :

https://www.gsmarena.com/vivo_x300_and_x300_pro_hands_on-review-2892p4.php#image8

--
See my profile (About me) for gear and my posting policy. My profile picture is of the first film camera I used in the early 80s, photo credit the internet.

 

[Ruby Rod]

My first digital was a Panasonic FZ-20. The EVF was horrible, but there are still posters on the wall here at work that I shot and that still look decent. Man, to have that super-zoom with a decent sensor and good EVF! It was a handy and capable camera, but we've sure come a long way since then.

 

[martinhb]

Stops quantify a ratio: Dynamic range is a ratio, and stops provide a log-base-2 (or f-stop) measurement of that ratio. A 10-stop dynamic range means the brightest possible pixel is 1,024 times brighter than the darkest possible pixel since 2 to the tenth power equals 1024.

Yes, but you're leaving out some critical elements. "Dynamic range" requires that you set specifics, like SNR for the floor and whether or not you allow non-linearity in the highlights. Most claims for expanded DR all revert to ignoring linearity and using a SNR of 1 as the floor. Try looking at an image with a SNR of 1 and you'll see that it's not practical.

What this means for photographers is crucial for scenes with high contrast, like a bright sky and dark foreground.

Almost certainly the final answer for image sensor "range" will be what you see in the patents from Nikon and others: a spillover count on the highlights. That, in essence, allows "infinite" dynamic range.

But while everyone seems to claim they want more dynamic range, exactly what are they outputting to? A display with a 1000:1 contrast ratio is 10 stops, and if you throw a 10-stop image at it, it wouldn't render exactly right due to the ways in which that 10 stops are generated. A print can handle maybe 10 stops, but is also non-linear at the extremes. Our eyes are about 10000:1 without pupillary adjustment, which is about what our cameras can capture currently.

Thanks Thom for fixing the .. uh, my flawed, absent SNR information. Much appreciated.

 

[monte12345]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

 

[tbcass]

Just something to ponder. Do we actually NEED 20 stops of dynamic range?

IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

I agree. It's just another way to sell cameras.

 

[WilyKit]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

Yes, we do. I would be really happy to be able to get high quality pictures of interiors with exterior in the background. Without "stacking" and other "computational photography" tricks that are suitable for scenes without movement. The cases for this are numerous.

 

[GarysInSoCal]

Nikon will perhaps have Sony make a high dynamic sensor made to Nikon's own specifications.

It happened 10 years ago... it's called the D810!... ;-)

https://www.dxomark.com/Cameras/Nikon/D810

 

[J A C S]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

Yes, we do. I would be really happy to be able to get high quality pictures of interiors with exterior in the background. Without "stacking" and other "computational photography" tricks that are suitable for scenes without movement. The cases for this are numerous.

Most of the noise in the (lifted) shadows with today's sensors are caused by low light, not by insufficient DR of the sensor.

 

[WilyKit]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

Yes, we do. I would be really happy to be able to get high quality pictures of interiors with exterior in the background. Without "stacking" and other "computational photography" tricks that are suitable for scenes without movement. The cases for this are numerous.

Most of the noise in the (lifted) shadows with today's sensors are caused by low light, not by insufficient DR of the sensor.

I'm talking about highlights.

 

[J A C S]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

Yes, we do. I would be really happy to be able to get high quality pictures of interiors with exterior in the background. Without "stacking" and other "computational photography" tricks that are suitable for scenes without movement. The cases for this are numerous.

Most of the noise in the (lifted) shadows with today's sensors are caused by low light, not by insufficient DR of the sensor.

I'm talking about highlights.

They have more or less a constant FWC per unit area. The new sensor in question claims a stop more in the highlights.

 

[WilyKit]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

Yes, we do. I would be really happy to be able to get high quality pictures of interiors with exterior in the background. Without "stacking" and other "computational photography" tricks that are suitable for scenes without movement. The cases for this are numerous.

Most of the noise in the (lifted) shadows with today's sensors are caused by low light, not by insufficient DR of the sensor.

I'm talking about highlights.

They have more or less a constant FWC per unit area. The new sensor in question claims a stop more in the highlights.

I don't know what "FWC" is, and I don't believe any claims about sensors' dynamic range (yeah even one stop better, not to mention "20 stops DR"), because companies are full of... So I'll wait and see.

 

[J A C S]

Just something to ponder. Do we actually NEED 20 stops of dynamic range? IMO it's just another Number that can be used to con us into purchasing new equipment we really do not have a need for.

Yes, we do. I would be really happy to be able to get high quality pictures of interiors with exterior in the background. Without "stacking" and other "computational photography" tricks that are suitable for scenes without movement. The cases for this are numerous.

Most of the noise in the (lifted) shadows with today's sensors are caused by low light, not by insufficient DR of the sensor.

I'm talking about highlights.

They have more or less a constant FWC per unit area. The new sensor in question claims a stop more in the highlights.

I don't know what "FWC" is, and I don't believe any claims about sensors' dynamic range (yeah even one stop better, not to mention "20 stops DR"), because companies are full of... So I'll wait and see.

Full Well Capacity - basically the number of electrons it takes for the sensor to saturate. More precisely, the cutoff which is a bit below the actual saturation.