The Sony rumors seem to be true

Kerry Pierce wrote:
rhlpetrus wrote:

ISO can't be much better than what we see in the latest sensors, the limits are being reached already.
They managed to tweak about another stop out of the d3 sensor, to make the d3s. IMO, they could have done something similar with the d7k sensor, if they really wanted to.
They managed to tweak out quite a bit over the D7k sensor with D7100. Well, not Sony, but Toshiba. Though, I would certainly expect Sony to improve too, seeing that their sensors evolved with every new generation. No reason why that should stop now.

Still, there are real limits at play, as Renato hinted. You're not gonna see huge improvements every time you want it, and it's not a matter of engineers not wanting it to happen.
 
Last edited:
coudet wrote:

Still, there are real limits at play, as Renato hinted. You're not gonna see huge improvements every time you want it, and it's not a matter of engineers not wanting it to happen.
Yes!
 
coudet wrote:
Kerry Pierce wrote:

I'm in full agreement with jkjond, in that, I simply can't imagine that a 16-18mp camera would be lacking for my sports/action requirements.
16mp wouldn't be lacking in some sports situations, but it'd be lacking at everything else. Make the camera more versatile and more people will be interested in it.
I would suggest that any 16mp DX camera that was dedicated to sports/action shooting would sell much better than their current action camera offering.

The notion that the d400 has to be a camera that appeals to everyone, is most likely the reason why we don't have one at all. It makes no sense that every DX camera must be high MP. There are already 3 24mp DX cameras for which one can choose to use for portrait/landscape/etc, ie general purpose cameras. There isn't a high speed, high performance camera. The FX market is broken down with both a high performance and general purpose cameras. There's no logical reason that DX should be different.

Kerry
 
coudet wrote:
Kerry Pierce wrote:
rhlpetrus wrote:

ISO can't be much better than what we see in the latest sensors, the limits are being reached already.
They managed to tweak about another stop out of the d3 sensor, to make the d3s. IMO, they could have done something similar with the d7k sensor, if they really wanted to.
They managed to tweak out quite a bit over the D7k sensor with D7100. Well, not Sony, but Toshiba. Though, I would certainly expect Sony to improve too, seeing that their sensors evolved with every new generation. No reason why that should stop now.
No, no reason to stop at all.
Still, there are real limits at play, as Renato hinted. You're not gonna see huge improvements every time you want it, and it's not a matter of engineers not wanting it to happen.
History is replete with naysayers about the limits of science. When they hit the limits, maybe that is the time to be satisfied, but certainly not before.

Nobody said it was a matter of the engineers not wanting to do it. The engineers most likely do what they're told to do by the management.

Kerry
 
Kerry Pierce wrote:

Nobody said it was a matter of the engineers not wanting to do it. The engineers most likely do what they're told to do by the management.
Sorry Kerry but managers can't sucessfullt tell engineers to break the laws of physics.
 
Phil_L wrote:
Kerry Pierce wrote:

Nobody said it was a matter of the engineers not wanting to do it. The engineers most likely do what they're told to do by the management.
Sorry Kerry but managers can't sucessfullt tell engineers to break the laws of physics.
Nobody is asking them to do that. What I said was that they certainly improved the d3s sensor, over the d3 and that, IMO, they could have done something similar with the d7k sensor.

Depending on who you ask, the d4 sensor is a slight, to rather significant, noise performance improvement over the d3s, even with the smaller pixels that the extra 4mps bring. To me, that means that technology is still advancing and that it is reasonable to expect that they could certainly improve on an older sensor design like the d7k.

I haven't seen anything to indicate that the d7k sensor represents the absolute limit of noise performance. For me, even a slight improvement would be welcome. If you have information to prove that the d7k sensor could not be improved, I'd like to see the citations.

Kerry

--
my gallery of so-so photos
http://www.pbase.com/kerrypierce/root
 
Last edited:
n057 wrote:
Tidewater wrote:

My personal interest in 32mp would frankly be for cropping. As every bird photographer knows, there is never enough reach. My D200 has 10mp. Having 32mp should enable use of my light and portable 70-300 as a 600mm+.

Correct me if I am wrong.
That is, assuming the new camera has very clean high-ISO. The 70-300 at 300mm is at f/5.6, while Nikon's 600 mm is f/4. Combine that extra stop with the possible need for high shutter speed and that will bump you into hi-ISO, which will show up in your cropped image.

JC
Some cameras, some lenses, some computers
I seem to be missing your point. I shoot with a 10mp, Nikon D200 camera now with a slow lens and a low iso camera. Probably many of us do. Thats just the way it is. I manage those parameters pretty well and shoot a lot of BIFS. (See gallery) But I would still like to have 3x as many pixels thus giving me more reach through cropping , without having to use a heavy lens. Seems like a great feature to me. The dx format makes the 300 into a 450 and the 400 becomes 600 and I am often at full 100% crop, so I could crop less or have even longer reach than that. Yes if I used a 2.8 or an f4 I could get another stop or two of shutter speed as well. For me photography has to be fun so I even find the wonderful 80-400 to be somewhat burdensome. So I guess the guys over in nature and wildlife or sports and action would be more excited by the 32MP. BTW you can get a 2 terabyte HD for about $129 for external storage and back-up.
 
Tidewater wrote:
n057 wrote:
Tidewater wrote:

My personal interest in 32mp would frankly be for cropping. As every bird photographer knows, there is never enough reach. My D200 has 10mp. Having 32mp should enable use of my light and portable 70-300 as a 600mm+.

Correct me if I am wrong.
That is, assuming the new camera has very clean high-ISO. The 70-300 at 300mm is at f/5.6, while Nikon's 600 mm is f/4. Combine that extra stop with the possible need for high shutter speed and that will bump you into hi-ISO, which will show up in your cropped image.

JC
Some cameras, some lenses, some computers
I seem to be missing your point. I shoot with a 10mp, Nikon D200 camera now with
I understand. I have two of those D200, which I still use because I do not have a D400 yet :-)


Sorry if I was not clear enough. If you look at, say, a D800, and compare it to a D7000, the DX mode of the D800 gives practically the same image as an uncropped D7000 image, including its noise characteristics. If you take an FX sized image of the D800 and crop it down to D7000 size, you get an identical image as you would if taken with a D7000 - including its noise characteristics.

In other words, the crop that you get with the hypothetical 32 MP will have an increased noise level too.
a slow lens and a low iso camera. Probably many of us do. Thats just the way it is. I manage those parameters pretty well and shoot a lot of BIFS. (See gallery) But I would still like to have 3x as many pixels thus giving me more reach through cropping , without having to use a heavy lens. Seems like a great feature to me. The dx format makes the 300 into a 450 and the 400 becomes 600 and I am often at full 100% crop, so I could crop less or have even longer reach than that. Yes if I used a 2.8 or an f4 I could get another stop or two of shutter speed as well. For me photography has to be fun so I even find the wonderful 80-400 to be somewhat burdensome. So I guess the guys over in nature and wildlife or sports and action would be more excited by the 32MP. BTW you can get a 2 terabyte HD for about $129 for external storage and back-up.
I have 4 x 2 TB drives in my main computer, and two RAID NAS boxes with 2 x 2 TB drives each on my lan, for backup :-)


JC
Some cameras, some lenses, some computers
 
Kerry Pierce wrote:
For me, even a slight improvement would be welcome.
Me too. But Nikon chose not to??

I think we are into an area of marginal diminishing improvements that cost a lot of engineering money and don't wow the customers.

Binning is an interesting idea, selectable levels of binning all the more so and a possible work around to the problem!??

Nikon have done it before, on the D1 afaik.
If you have information to prove that the d7k sensor could not be improved, I'd like to see the citations.
Now I didn't say that! :-D
 
n057 wrote:
another stop or two of shutter speed as well. For me photography has to be fun so I even find the wonderful 80-400 to be somewhat burdensome. So I guess the guys over in nature and wildlife or sports and
I have 4 x 2 TB drives in my main computer, and two RAID NAS boxes with 2 x 2 TB drives each on my lan, for backup :-)

JC
Some cameras, some lenses, some computers
Wow.... ;-) Makes my lil computer with 1 X 1 TB & 1 X 500 GiG drive & couple 250 Gig external backup drives seem minute... ;-)

However, what is the recommended 'final' file version to keep , JPEG at what resolution ?

Some of you also keep NEF's and I agree for important photos but is it necessary for all?

Finally, why do TIFF's explode so much compared to NEF's? I would prefer to keep TIFF's if I was keeping a second format besides the final JPEG, because it would have my initial RAW conversion.

But the TIFF's are large...even from the D300.

Thanks!

--
Amateur photographer. Enjoy.....believe in yourself..
 
Last edited:
However I have just gone for a D800 to replace my D300S as my main camera.
 
Kerry Pierce wrote:
rhlpetrus wrote:
Kerry Pierce wrote:

Well, I guess that 32mp would certainly be the cat's meow, *IF* high ISO and low ISO noise performance was better than the d7k/d7100, preferably significantly better.
ISO can't be much better than what we see in the latest sensors, the limits are being reached already.
They managed to tweak about another stop out of the d3 sensor, to make the d3s. IMO, they could have done something similar with the d7k sensor, if they really wanted to.
At full res I bet it'll look worse, reduced maybe about same. DR/Color performance more relevant than noise IMO, as ISO moves up colors tend to shift badly. People worry about perception of noise, but that's not the main point IMO.
Depends on what you're shooting. For sports, I don't worry about color fidelity. Try shooting a hockey game that is lighted with those goofy lights that change colors with the fluctuation of the AC power. Florescent lights are bad, see this link, but there are others, such as Mercury Vapor lights, which are far worse.

How about having red, white and blue tints in the same photo, like this one in Mercury Vapor lighting? It's very common in the indoor arenas that I see. :-)

70031866.jpg

I've been doing a lot of editing this week, on a maternity gig that I recently shot with the d3s and d800. On my PC, LR4 often chokes on those big files, making any serious editing a real pain. I can't imagine trying to edit them under pressure of a short time frame.
D800 is a camera for low intensity shooting, like landscapes or studio. Nikon still lacks a compact FF with the same characteristics as the D700 and updated sensor.
Yes. I'd really love to see another pairing like I have with the d300/d700. I like my d800, but it's not a sports/action camera.

Kerry

--
my gallery of so-so photos
http://www.pbase.com/kerrypierce/root




My lord, that looks hideous with those color shifts. Ruins an otherwise good photo.
 
Phil_L wrote:
Kerry Pierce wrote:

For me, even a slight improvement would be welcome.
Me too. But Nikon chose not to??
I was referring to them putting a tweaked version of that sensor into a d400. That is something that they could have easily done, perhaps concurrently with the development of the d7k.

We could be using that sensor/camera today, griping about not seeing a d500, but at least we would be using a very good performing sensor, with MPs that are close to what the Canon boys already have.

That is something that Nikon has apparently forgotten/ignored and usually is never mentioned here. The 7d users already have 18mp cameras that they use for sports/action/wildlife. Downsized to 12mp, I'd assume that they'd have better noise performance than we have with the d300. Plus they have plenty of cropping room or pixels on target, when the noise isn't bad. Those are significant pluses, leaving the 7d users well ahead of the game, IMO, while we still use the old 12mp sensor.
I think we are into an area of marginal diminishing improvements that cost a lot of engineering money and don't wow the customers.
That may well be. I have read that we are relatively close to the theoretical limits of performance from the current DX sensors. But, AFAIK, that doesn't account for things like binning or other technological tricks that may be employed.

From the d3 to d3s improvements, I simply extrapolated the probability that most, if not all, sensors are not fully exploited for performance, ie noise, speed, color fidelity, etc, in the first iteration. Perhaps they could squeeze out another 1/2 to 2/3 stop of noise performance. If so, I wouldn't call that marginal.
Binning is an interesting idea, selectable levels of binning all the more so and a possible work around to the problem!??

Nikon have done it before, on the D1 afaik.
Yes, I believe that the d1 was actually a 10mp sensor. I don't know how they'd go about doing that today, but I assume that they could improve noise performance by binning.
If you have information to prove that the d7k sensor could not be improved, I'd like to see the citations.
Now I didn't say that! :-D
Okay. :-)

Kerry
 
ranalli wrote:
Kerry Pierce wrote:
How about having red, white and blue tints in the same photo, like this one in Mercury Vapor lighting? It's very common in the indoor arenas that I see. :-)
70031866.jpg
My lord, that looks hideous with those color shifts. Ruins an otherwise good photo.
Yes, of course you're right, but that isn't the only problem. The exposure in the red area is about 1/2 to 2/3 a stop darker than the brightest area of the ice. It makes full ice shots like this impossible to take. You must shoot lower shutter speeds and everything as isolated as possible, to get the exposure and the colors somewhat close to correct and to a point that you can fix them in post. You can easily see the difference in the 2 players that are at mid-ice, on the fringes of the white/red toned area, to the closest players in the full red tone.

Kerry

--
my gallery of so-so photos
 
Bajerunner wrote:
n057 wrote:
another stop or two of shutter speed as well. For me photography has to be fun so I even find the wonderful 80-400 to be somewhat burdensome. So I guess the guys over in nature and wildlife or sports and
I have 4 x 2 TB drives in my main computer, and two RAID NAS boxes with 2 x 2 TB drives each on my lan, for backup :-)

JC
Some cameras, some lenses, some computers
Wow.... ;-) Makes my lil computer with 1 X 1 TB & 1 X 500 GiG drive & couple 250 Gig external backup drives seem minute... ;-)

However, what is the recommended 'final' file version to keep , JPEG at what resolution ?

Some of you also keep NEF's and I agree for important photos but is it necessary for all?

Finally, why do TIFF's explode so much compared to NEF's? I would prefer to keep TIFF's if I was keeping a second format besides the final JPEG, because it would have my initial RAW conversion.

But the TIFF's are large...even from the D300.

Thanks!
 
Kerry Pierce wrote:
coudet wrote:
Kerry Pierce wrote:

I'm in full agreement with jkjond, in that, I simply can't imagine that a 16-18mp camera would be lacking for my sports/action requirements.
16mp wouldn't be lacking in some sports situations, but it'd be lacking at everything else. Make the camera more versatile and more people will be interested in it.
I would suggest that any 16mp DX camera that was dedicated to sports/action shooting would sell much better than their current action camera offering.

The notion that the d400 has to be a camera that appeals to everyone, is most likely the reason why we don't have one at all. It makes no sense that every DX camera must be high MP. There are already 3 24mp DX cameras for which one can choose to use for portrait/landscape/etc, ie general purpose cameras. There isn't a high speed, high performance camera. The FX market is broken down with both a high performance and general purpose cameras. There's no logical reason that DX should be different.

Kerry
 
JimPearce wrote:

Nikon could do 32 MP at 9 fps. Hard to imagine a mirror moving much faster at this price point. Unfortunately, it seems that once again my attempts to understand what the future holds have come up short. In particular, I now think that a 32 MP DX sensor could deliver up to 10% more linear resolution than a 24 MP sensor does with the best lenses. I really don't want to have to store the files from a 32 MP sensor.
 
n057 wrote:
Tidewater wrote:

My personal interest in 32mp would frankly be for cropping. As every bird photographer knows, there is never enough reach. My D200 has 10mp. Having 32mp should enable use of my light and portable 70-300 as a 600mm+.

Correct me if I am wrong.
That is, assuming the new camera has very clean high-ISO. The 70-300 at 300mm is at f/5.6, while Nikon's 600 mm is f/4. Combine that extra stop with the possible need for high shutter speed and that will bump you into hi-ISO, which will show up in your cropped image.

JC
Some cameras, some lenses, some computers
not trying to be harsh..how much cropping do you plan on doing?..you may need to invest in longer lenses..fair to say that most photogs here dont crop heavily..just asking
 
Kerry Pierce wrote:
Phil_L wrote:
Kerry Pierce wrote:

Nobody said it was a matter of the engineers not wanting to do it. The engineers most likely do what they're told to do by the management.
Sorry Kerry but managers can't sucessfullt tell engineers to break the laws of physics.
Nobody is asking them to do that. What I said was that they certainly improved the d3s sensor, over the d3 and that, IMO, they could have done something similar with the d7k sensor.

Depending on who you ask, the d4 sensor is a slight, to rather significant, noise performance improvement over the d3s, even with the smaller pixels that the extra 4mps bring. To me, that means that technology is still advancing and that it is reasonable to expect that they could certainly improve on an older sensor design like the d7k.

I haven't seen anything to indicate that the d7k sensor represents the absolute limit of noise performance. For me, even a slight improvement would be welcome. If you have information to prove that the d7k sensor could not be improved, I'd like to see the citations.

Kerry

--
my gallery of so-so photos
http://www.pbase.com/kerrypierce/root
Kerry,

I've not held D4 in my hands, but if DxO is something to go by, D4 mainly has some advantage over D3s at the low ISO, which is probably not where your interest lies. D7K also came out a year after D3s, has pretty much the same characteristics as the D800 in DX mode, and could be assumed to incorporate all the ISO improvements D3s had received earlier. That was the last significant high ISO step we've seen so far: D3s over D3/700 and D7000 over D300, and in both cases we gained about 2/3 of a stop, give or take. I'm not sure we can realistically expect much more with the existing technologies, which might explain the shift from ISO race to pixel race, apart from the fact that someone at Nikon probably realized that increasing sensor resolution is a great device for making people want to buy new lenses.
 
Last edited:

Keyboard shortcuts

Back
Top