Pages: 1 ... 6 7 [8] 9 10 ... 29   Go Down

Author Topic: Nikon D850: Announcement of an Announcement of Development  (Read 1000280 times)

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13985
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #140 on: August 01, 2017, 01:12:11 pm »

We'll know soon enough.

As I said, the point isn't the D5x, it is the exustence of a plan for FX DSLR above the D850.

Cheers,
Bernard

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #141 on: August 01, 2017, 01:15:27 pm »

Increase the size of the capacitor and you've just increased your low-ISO DR.

Yes. I mentioned that earlier in the thread. You can increase the capacitance by making the plates bigger, by getting them closer together, or by increasing the dielectric constant. Or, going with your 3D idea, stacking capacitors vertically. I'd go with a mirror-image design if I were going to do that.

Jim
« Last Edit: August 01, 2017, 01:19:29 pm by Jim Kasson »
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #142 on: August 01, 2017, 01:16:40 pm »

This would probably require destructive testing, but does removing the filter stack from all three sensors equalise their sensitivity?

I'm not going to try that experiment. I doubt it.

Jim

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #143 on: August 01, 2017, 01:39:42 pm »

We'll know soon enough.

As I said, the point isn't the D5x, it is the exustence of a plan for FX DSLR above the D850.

Cheers,
Bernard

So, D850 vs A7r3 and D900 vs A9r?

Price point would be critical in the relative appeal of the four cameras.

A7r3 vs D850 would be an interesting matchup at the same price point, with the A7r3 likely having a significantly better sensor (maybe 72MP vs 46MP) but the D850 having a faster frame rate and (possibly) better AF (unless the A7r3 is more like the A9 than the A7r2 - it's hard to say, since the A7r2 is more like a prototype than a true part of a mature product lineup).

A9r vs D900/D5x is less certain. How good can they make the A7r3 without encroaching on the A9r? Make it not good enough and it basically becomes a landscape/studio one-trick wonder, as the A7r2 is regarded now, with limited appeal beyond this base compared to the 5Ds2 and D850. Make it too good and the A9r becomes pointless. And where would Nikon source a competitive sensor? Certainly Sony won't sell them theirs if they plan to put it in their own A9r. Perhaps a lesser Sony sensor - 55-60MP?

Certainly,  there is room for a top-of-the-line, no-holds-barred high-resolution body. It would be different,  rather than necessarily better, than the action bodies - for instance, bandwidth may be similar,  for 72MP at 6.7fps instead of 24MP at 20fps - with AF and other features being similar,  and a similar price point. Similarly, there is room for an entry-level, high-resolution/DR one-trick wonder, in the same vein as the 5D2, for those who need the sensor and not much else.

In that case, if there is a D5x or D900 in addition to the D850, perhaps the A9r would be the intended competition for both of them, with the A7r3 being an entry-level 'bare sensor' not intended to compete with these bodies at all. Or perhaps the A7 lineup will take a step up (like the 5D line with the 5D3), matching the D850 level of AF and other non-sensor performance, with a third, entry-level A5 line being introduced to compete with the 6D and D600 lines (with the high-resolution 'bare sensor' belonging there). That would certainly introduce symmetry into the lineup of the three major players. But is symmetry really necessary? And can Sony really support - or does it really need - a third line of full-frame mirrorleaa bodiea?

And where does all this leave Canon? The 5Ds2 is a near-certainty. Sensor-wise, it will likely be more A9r/A7r3 than D810 - they've proven their ability to manufacture high-DR sensors, and they're hardly likely to go backwards on resolution - bit the other aspects of performance are far leas certain. Does Canon have room for a 1Dxs in addition to the 5Ds line?
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #144 on: August 01, 2017, 01:48:18 pm »

I'm not going to try that experiment. I doubt it.

Jim

Then what else could be responsible for the reduced sensitivity, yet equal ultimate DR (and same half-a-stop lag in DR-vs-ISO performance) in a sensor whose core design appears to be otherwise identical to the other two, and is likely the same basic sensor?

The D810 is no less noisy and has no more DR at ISO 64 compared with the D800e and A7r at ISO 100. It simply needs a lower ISO to achieve the same output quality. Usually, you'd regard that as a step back. What other aspect of performance could be improved by Nikon making that sacrifice? Did they use a stronger colour filter for better colour accuracy and/or gamut? Would be interesting to know.
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #145 on: August 01, 2017, 01:51:58 pm »

Then what else could be responsible for the reduced sensitivity, yet equal ultimate DR (and same half-a-stop lag in DR-vs-ISO performance) in a sensor whose core design appears to be otherwise identical to the other two, and is likely the same basic sensor?

The D810 is no less noisy and has no more DR at ISO 64 compared with the D800e and A7r at ISO 100. It simply needs a lower ISO to achieve the same output quality. Usually, you'd regard that as a step back. What other aspect of performance could be improved by Nikon making that sacrifice? Did they use a stronger colour filter for better colour accuracy and/or gamut? Would be interesting to know.

I don't buy that the D810 and D800 sensors are the same. I've seen no evidence of that. I don't see how they could be, since the D810 has EFS and teh D800 does not. Are you saying that the D800 sensor has all the circuitry for EFCS and just doesn't use it?

Jim

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13985
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #146 on: August 01, 2017, 01:53:20 pm »

I don't think the a7rIII will be 70mp.

Sony is going to keep that sensor for the a9r and those who can afford it will have to pay 5,000+ US$ to get one.

Cheers,
Bernard

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #147 on: August 01, 2017, 01:59:59 pm »

I don't buy that the D810 and D800 sensors are the same. I've seen no evidence of that. I don't see how they could be, since the D810 has EFS and teh D800 does not. Are you saying that the D800 sensor has all the circuitry for EFCS and just doesn't use it?

Jim

Possibly an upgraded version of the same basic design. Same thing at the core, with different bits and pieces tacked on.

A bit like how a Lexus and Toyota are the same basic car - same chassis, same engine. But different bits and pieces attached to it. In the D810's casr, this may include additional circuitry, which doesn't actually change the underlying basic sensor design.
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #148 on: August 01, 2017, 02:07:42 pm »

Possibly an upgraded version of the same basic design. Same thing at the core, with different bits and pieces tacked on.

A bit like how a Lexus and Toyota are the same basic car - same chassis, same engine. But different bits and pieces attached to it. In the D810's casr, this may include additional circuitry, which doesn't actually change the underlying basic sensor design.

How do you know that? Putting a rolling reset in affects a lot.

Jim

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #149 on: August 01, 2017, 02:08:28 pm »

I don't think the a7rIII will be 70mp.

Sony is going to keep that sensor for the a9r and those who can afford it will have to pay 5,000+ US$ to get one.

Cheers,
Bernard

Without a significant sensor advantage, the A7r3 would need either a huge AF and general performance boost over the A7r2, or a much cheaper price point than the D850 and 5Ds2, to be competitive with the Canon and Nikon. No-one's going to buy an A7r3 with D850-level sensor performance and much poorer other features. Sensors are Sony's strength, while competing on price has never been one of its strengths...

It may be easier to use the same sensor on both cameras and make AF and general performance the distinguishing feature, rather than using different sensors on each one and having to match the D850's speed and AF with the A7 line.
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #150 on: August 01, 2017, 02:10:13 pm »

Then what else could be responsible for the reduced sensitivity, yet equal ultimate DR (and same half-a-stop lag in DR-vs-ISO performance) in a sensor whose core design appears to be otherwise identical to the other two, and is likely the same basic sensor?

The D810 is no less noisy and has no more DR at ISO 64 compared with the D800e and A7r at ISO 100. It simply needs a lower ISO to achieve the same output quality. Usually, you'd regard that as a step back. What other aspect of performance could be improved by Nikon making that sacrifice? Did they use a stronger colour filter for better colour accuracy and/or gamut? Would be interesting to know.

D810 Claff PDR at base SISO is 11.6 stops
D800 Claff PDR at base ISO is 11.42 stops.

Jim

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #151 on: August 01, 2017, 02:13:14 pm »

How do you know that? Putting a rolling reset in affects a lot.

Jim

It seems more than mere coincidence that a sensor from the same maker, with the same resolution, used in a camera that's essentially a mid-generation upgrade of another using a similar sensor, happens to have a near-identical DR-to-ISO curve, except shitted half a stop to the left. How many completely-different 36MP sensors with near-identical performance would Sony have designed and put into production?
Logged

Michael Erlewine

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1033
    • MacroStop.com
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #152 on: August 01, 2017, 02:17:19 pm »

I've measured the sensitivity of all three cameras in a rough sense since I routinely do comparison shots and note the differences in raw values for the exposures. The D810 sensor at base ISO is indeed about two-thirds of a stop slower than the a7RII at base ISO. The GFX sensor is slower than the a7RII at the same ISO.

Jim

Thanks Jim. I am beginning to understand. As mentioned, my eyes tell me that the C810 is a better low ISO camera than my Sony A7RII... and while I'm not sure, my eyes tell me that neither the GFX or the X1D can match the D810 in low ISO, although (IMO) the X1D is better at handling highlights than the D810.
Logged
MichaelErlewine.smugmug.com. Founder MacroStop.com, MichaelErlewine.com, YouTube.com/user/merlewine

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #153 on: August 01, 2017, 02:21:06 pm »

D810 Claff PDR at base SISO is 11.6 stops
D800 Claff PDR at base ISO is 11.42 stops.

Jim

D800e is 11.45, A7r is 11.71.

In addition, we know that the A7r and D810 (the two higher values) lack AA filters, while the D800 and D800e have ine (although the D800e has a second filter to reverse the first one). Perhaps this has a small impact on base DR.

These values - and the similar way the sensors behave at higher ISOs - all seem close enough to be due to statistical variation between individual sensors or measurements, rather than being indicative of the three sensors being of different underlying designs. Particularly since the D810 value sits between the D800 and A7r values - two cameras which show even more similarities in sensor performance - rather than off to one side of them.
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #154 on: August 01, 2017, 02:25:04 pm »

It seems more than mere coincidence that a sensor from the same maker, with the same resolution, used in a camera that's essentially a mid-generation upgrade of another using a similar sensor, happens to have a near-identical DR-to-ISO curve, except shitted half a stop to the left. How many completely-different 36MP sensors with near-identical performance would Sony have designed and put into production?

The a7R and the D800 have similar sensors.

http://photonstophotos.net/Charts/PDR.htm#Nikon%20D800,Sony%20ILCE-7R

Now, take a look at the FWCs of the a7R vs the D810:

http://blog.kasson.com/the-last-word/d810-vs-a7r-sensor-performance/

The FWC of the D810 is much larger. They are not the same sensor cell design.

Jim

JKoerner007

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 262
  • "A picture's worth a thousand words."
    • John Koerner Photography
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #155 on: August 01, 2017, 02:51:14 pm »

Possibly an upgraded version of the same basic design. Same thing at the core, with different bits and pieces tacked on.

A bit like how a Lexus and Toyota are the same basic car - same chassis, same engine. But different bits and pieces attached to it. In the D810's casr, this may include additional circuitry, which doesn't actually change the underlying basic sensor design.

More pure speculation.

Act as if you know the first time, then 'speculate' when someone calls you on it.

The truth is, you have no idea.

I will address your other tirade (on a new thread topic I create) when time permits.

Let's keep this to the D850.

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #156 on: August 01, 2017, 08:36:31 pm »

More pure speculation.

Act as if you know the first time, then 'speculate' when someone calls you on it.

The truth is, you have no idea.

I will address your other tirade (on a new thread topic I create) when time permits.

Let's keep this to the D850.

And what exactly does your post add? He's brought out evidence and charts I hadn't seen before. You're just here to fling personal insults and attack the person rather than the argument.

Interesting how you always maintain that others have 'no idea', without once spelling exactly how it is that you do have an idea, or how your thoughts are any more valid. I don't recall you even once laying out the reasoning behind your conjecture, beyond simply saying things like 'It's wrong', 'You're clueless' and 'You're dumb' without any attempt to demonstrate how or why there is an error, or why your reasoning holds any more water. That's not debate - it's just trolling.

In the absence of hard data, you lay out reasoned arguments, together with your reasoning and the evidence behind it (in the absence of hard data, that will be circumstantial). Others can refute this with hard data, or, in the absence of hard data, with their own reasoned arguments. That is how academic debate works. It is not 'speculation', any more than reasoned but untested scientific hypotheses are theories. Or was Einstein merely speculating about relativity a century ago, when he had no hard data and no means of testing any of his theories, despite having a logical reason for thinking what he did?
« Last Edit: August 01, 2017, 09:34:52 pm by shadowblade »
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #157 on: August 01, 2017, 09:21:29 pm »

The a7R and the D800 have similar sensors.

http://photonstophotos.net/Charts/PDR.htm#Nikon%20D800,Sony%20ILCE-7R

Now, take a look at the FWCs of the a7R vs the D810:

http://blog.kasson.com/the-last-word/d810-vs-a7r-sensor-performance/

The FWC of the D810 is much larger. They are not the same sensor cell design.

Jim

How much of a modification to the sensor is necessary, though, to achieve a higher FWC? Since it's a matter of capacitance, rather than necessarily the light-collecting area itself, wouldn't a redesign of the capacitance area alone - a thicker capacitor layer, or higher-capacitance capacitors, as you mentioned earlier, for example - achieve this, without requiring a full redesign of the sensor?

Also, I note that your PDR values differ from those provided by Claff. He gave a value of 11.71 for the A7r at ISO 100 (base ISO) and 11.60 for the D810 at ISO 64, dropping to around 11.02 at ISO 100 (I say 'around' because I don't have the exact number in front of me).

Or are you saying that the base ISO of the D810 is actually lower than 64 (with ISO 64 also being some multiple of base ISO) and Nikon actually crippled the sensor by not allowing access to even lower native ISOs, with commensurate higher DR?

Do you have any results for the Pentax K-1? If so, which version of the 36MP sensor does it use? If it uses the D810 sensor rather than the D800/A7r sensor, could it be the case that the D810 (and possibly K-1) merely represent an updated version of the 36MP sensor, perhaps with the sensitivity difference accounted for by a different filter stack in front? The K-1 has a PDR of 11.43 at its base ISO of 100, and is also consistently a third of a stop ahead of the D810 at any given ISO, suggesting greater sensitivity.
Logged

Jim Kasson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2370
    • The Last Word
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #158 on: August 01, 2017, 11:38:18 pm »

How much of a modification to the sensor is necessary, though, to achieve a higher FWC? Since it's a matter of capacitance, rather than necessarily the light-collecting area itself, wouldn't a redesign of the capacitance area alone - a thicker capacitor layer, or higher-capacitance capacitors, as you mentioned earlier, for example - achieve this, without requiring a full redesign of the sensor?

I'm not a sensor designer so I can't say how different, but there is sufficient evidence here -- the PDR, FWC, EFCS -- to indicate that the deisgns are different.

Also, I note that your PDR values differ from those provided by Claff. He gave a value of 11.71 for the A7r at ISO 100 (base ISO) and 11.60 for the D810 at ISO 64, dropping to around 11.02 at ISO 100 (I say 'around' because I don't have the exact number in front of me).

Bill and I use different methods and usually different serial numbered cameras.

Or are you saying that the base ISO of the D810 is actually lower than 64 (with ISO 64 also being some multiple of base ISO) and Nikon actually crippled the sensor by not allowing access to even lower native ISOs, with commensurate higher DR?

I'm not saying that at all.

Do you have any results for the Pentax K-1? If so, which version of the 36MP sensor does it use? If it uses the D810 sensor rather than the D800/A7r sensor, could it be the case that the D810 (and possibly K-1) merely represent an updated version of the 36MP sensor, perhaps with the sensitivity difference accounted for by a different filter stack in front? The K-1 has a PDR of 11.43 at its base ISO of 100, and is also consistently a third of a stop ahead of the D810 at any given ISO, suggesting greater sensitivity.

I have no Pentax results at all.

Jim

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13985
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: Nikon D850: Announcement of an Announcement of Development
« Reply #159 on: August 02, 2017, 01:14:14 am »

Without a significant sensor advantage, the A7r3 would need either a huge AF and general performance boost over the A7r2, or a much cheaper price point than the D850 and 5Ds2, to be competitive with the Canon and Nikon. No-one's going to buy an A7r3 with D850-level sensor performance and much poorer other features. Sensors are Sony's strength, while competing on price has never been one of its strengths...

Look at the Canon 6DMkII. It offers very little over the 6D and even less over competition.

An even better data point is the a7rII vs a7r.

And finally, the main concern of Sony will be to sell as many a9r as possible, so the differentiationn they will want to protect is btw a7rIII and a9r.

Anyway you look at it, it makes more sense to Sony to keep the a7rIII an alrounder camera which means small increases of resolution.

And I believe that the D850 is targetting the same positioning in the Nikon line up, which means a D900 is coming too.

Cheers,
Bernard
Pages: 1 ... 6 7 [8] 9 10 ... 29   Go Up