Pages: 1 2 3 [4] 5 6 ... 15   Go Down

Author Topic: "The Nikon D3x offers the finest image quality in a DSLR the world has yet seen"  (Read 131298 times)

dwdallam

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2044
    • http://www.dwdallam.com

Quote from: Slough
Outourcing chip fabrication is normal. In fact having two source protects you against e.g. a fire in one fab.



Pure speculation on your part. We know that Sony had SOME input in the design. That's all we know.




You think Canon design and manufacture everything in their camera? They don't.



See earlier.



Pure speculation on your part.

What makes you think that Nikon haven't had key input to the design of the chip? After all, the D3 chip is 'pure' Nikon. People assume that Sony have Nikon by the nadgers. But it could be more of a symbiosis, Nikon chip design expertise, Sony fabrication and electronics expertise. Pure speculation on my part. But just as (un)likely as yours.

That's the problem with the inter-porn-net-web. It's full of people making statements with 100% certainty but which cannot be substantiated. Remember the Nikon MX camera that was a dead cert? Even the PC-E lenses were used as evidence by the experts.  

It's similar to but not the A900 chip, but it IS manufactured by Sony:

http://www.robgalbraith.com/bins/multi_pag...cid=7-9318-9761


"While it's acknowledged that Nikon uses image sensors principally designed by Sony in many of its digital SLRs past and present, the D3X's sensor was described in a recent briefing by Nikon USA's Silverman as an "original Nikon design" that does not and will not appear in cameras from other digital SLR manufacturers. In addition, Nikon issued the following response to the questions they received from us and others about the D3X imager's roots:

The Nikon D3X’s 24.5-megapixel FX-format (35.9 x 24.0mm) CMOS sensor was developed expressly for the D3X in accordance with Nikon’s stringent engineering requirements and performance standards, with final production executed by Sony.  Featuring refined low-noise characteristics, 12 and 14 bit output, Live View capability and more, the D3X’s unique sensor design was carefully blueprinted to perform in perfect concert with proprietary Nikon technologies including EXPEED Image Processing and the Scene Recognition System. Meticulous efforts allowed the sensor to become one of the many essential components and technologies which contribute to the D3X’s superior image fidelity."
Logged

dwdallam

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2044
    • http://www.dwdallam.com

Quote from: Slough
No, they haven't. They have said it is unique to the D3x. And they have indicated involvement with Sony. And they have said that it is fabricated by Sony. Anything more is supposition.


Yep, it's MADE in Sony plant. It was designed by Nikon. But from my other post you can see it is very similar to the Sony chip. According to my other post showing you the link, Nikon has been using Sony as a chip maker for some time. That's not a bad thing. Sony has decades in the sensor business, and I'm sure their plants are up to task. It's just a fact: Sony manufactures Nikon chips and Canon manufactures Canon chips. Neither method is better or worse, necessarily.
Logged

Slough

  • Guest

dwdallam: Thanks, I knew that information from Nikon was somewhere.
Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547

Nikon has cleverly worded the D3x sensor scenario in that simply changing the AA and CFA would make the sensor "exclusive." Lets be a little realistic here. What are the odds that a D3x sensor that is exactly the same size as A900's, has the same unusual on chip A/Ds as the A900, and is made by the same company as the A900, is not the same silicon?  Heck, even the actual pics of the sensors look identical. LOL.  Now clearly the A900 and D3x have different output due to filter differences and processing, and each has it's strengths and weaknesses, but they are clearly closely related. I could get into the idea of Nikon resampling the camera's 12bit output to get 14bits, but I'll leave that to the tech guys
Logged

NikosR

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 622
    • http://

Who cares?
Logged
Nikos

Slough

  • Guest

Quote from: NikosR
Who cares?

Exactly.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

There may be some significant differences between Nikon's and Sony's sensors even if they use the same chip. The color filter array may be somewhat different. It's pausible that Nikon and Sony would prefer different spectral transmission characteristics. The other obvious difference is that Sony has on chip AD-converters (12 bit) and Nikon uses a few auxilary 14 bit AD converters AFAIK.

Erik

Quote from: douglasf13
Nikon has cleverly worded the D3x sensor scenario in that simply changing the AA and CFA would make the sensor "exclusive." Lets be a little realistic here. What are the odds that a D3x sensor that is exactly the same size as A900's, has the same unusual on chip A/Ds as the A900, and is made by the same company as the A900, is not the same silicon?  Heck, even the actual pics of the sensors look identical. LOL.  Now clearly the A900 and D3x have different output due to filter differences and processing, and each has it's strengths and weaknesses, but they are clearly closely related. I could get into the idea of Nikon resampling the camera's 12bit output to get 14bits, but I'll leave that to the tech guys
Logged
Erik Kaffehr
 

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547

Quote from: ErikKaffehr
Hi,

There may be some significant differences between Nikon's and Sony's sensors even if they use the same chip. The color filter array may be somewhat different. It's pausible that Nikon and Sony would prefer different spectral transmission characteristics. The other obvious difference is that Sony has on chip AD-converters (12 bit) and Nikon uses a few auxilary 14 bit AD converters AFAIK.

Erik

  I agree that there are still differences outside of the sensor itself, and that's why I mentioned the color filter in my post (CFA.)  According to the Nikon website, the D3x A/D converters are on-chip, which is unique to Sony sensors.  The real question is how they are attaining 14bits, which looks to be a matter of resampling (and thus slowing down fps,) but it's not known.  

  As far as the "who cares?" post, I agree.  I'm not sure why so many Nikon users are reluctant to believe the D3x chip is a Sony.  Sony has been in digital imaging as long as anyone, and is certainly capable.
« Last Edit: January 07, 2009, 04:14:05 pm by douglasf13 »
Logged

Dennishh

  • Full Member
  • ***
  • Offline Offline
  • Posts: 160

What makes me very happy is that my 1dsMk3 is going to stay current for at least another year or maybe two years. The quality is just amazing and there is no reason on earth to upgrade. In this economy it makes perfect sense to stay put. $8000.00 price point is a thing of the past and might be one that 2 1/4 systems will have to take. Canon and Nikon will have to come up with new technology, larger sensors or a lower price point to make me jump to a new body.
Dennis
Logged

inissila

  • Newbie
  • *
  • Offline Offline
  • Posts: 36

Quote from: dwdallam
I also think it's a fallacy--and I'm not saying you made this mistake--to say that the "camera" is superior. It's not the camera, it's the electronics (CPU) in it  + the lens used + post processing, which is the software after the fact. If the review states that JPG quality is "better" and here is why, x, y, z (e.g., w/o any post processing) then that is saying something.

A part of it may be due to the antialiasing filter - which is hardware. If the same or better algorithms than what Nikon use in their camera and Capture NX2 are not available elsewhere, does it really matter whether the quality is achieved in hardware or software - in either case it isn't available with other hardware. And even in the future, if Adobe catches up to Nikon's algorithms for raw conversion etc. they won't be tailor made to Nikon's optics (lenses, microlenses, and AA filter) so you have to spend time finding the right settings. I've never been able to obtain NX2 quality out of D3 NEF files shot at high ISO using ACR ... you can get lower noise but at far greater compromise to detail and vice versa. And finding the parameters which result in natural results using ACR is a royal pain whereas NX2 does it out of the box with minimal tweaking. This is because it's designed for Nikon's cameras and their optics, whereas ACR is generic.

Any interesting comparison of noise or detail needs to include both parameters as you can always trade off one for the other by a choice of parameters. And it should be made with the best raw converter available - which is the manufacturer's code usually. It really shows best when shooting and processing portrait images - as they have some fine detail, some relatively smooth areas, and the subject is 3D. The natural appearance of shadows and highlights, and the overall appearance of the image IMO comes out better with NX2. The general consensus seems to agree with me though some claim that it's just a question of parameters. IMO it is not. Software should be optimized to the hardware to provide optimum results. This is a generic principle of imaging systems.
Logged

Slough

  • Guest

Quote from: douglasf13
As far as the "who cares?" post, I agree.  I'm not sure why so many Nikon users are reluctant to believe the D3x chip is a Sony.  Sony has been in digital imaging as long as anyone, and is certainly capable.

Because you seem to be making assertions without proof. If you want to say "I think the chips excluding the cover-plate and AA filter are the same" fair enough. But if you say "They are the same" and provide only hand waving arguments, well, I'll refrain from drawing a conclusion.

Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547

Quote from: Slough
Because you seem to be making assertions without proof. If you want to say "I think the chips excluding the cover-plate and AA filter are the same" fair enough. But if you say "They are the same" and provide only hand waving arguments, well, I'll refrain from drawing a conclusion.

  Maybe I worded things incorrectly, because I was typing quickly on my danged cell phone.  Sorry   What I'm saying is exactly that, and I don't mean to imply more.  I think the chips of the D3x and A900 are the same, with different CFA, AA, and an unknown special sauce for 14 bits, which leads to a different output than the A900.  That being said, I wouldn't say one is superior to the other.  A few dual users, like Iliah Borg, have shown that each camera has it's own IQ positives and negatives, depending on the situation, and I guess that's an appropriate thing to mention, considering the title of this thread.


Logged

jani

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1624
    • Øyet

Quote from: douglasf13
Maybe I worded things incorrectly, because I was typing quickly on my danged cell phone.
Your wording was just fine, fairly precise. I think Slough was reading more into your statements than what was really there.
Logged
Jan

Fine_Art

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1172

Quote from: douglasf13
Nikon has cleverly worded the D3x sensor scenario in that simply changing the AA and CFA would make the sensor "exclusive." Lets be a little realistic here. What are the odds that a D3x sensor that is exactly the same size as A900's, has the same unusual on chip A/Ds as the A900, and is made by the same company as the A900, is not the same silicon?  Heck, even the actual pics of the sensors look identical. LOL.  Now clearly the A900 and D3x have different output due to filter differences and processing, and each has it's strengths and weaknesses, but they are clearly closely related. I could get into the idea of Nikon resampling the camera's 12bit output to get 14bits, but I'll leave that to the tech guys

Nikon did the same there as they did with the D300 vs the A700. Same chip with different hardware attached for 14 bit, custom "Gapless microlens", etc.

A900 & D3x
http://www.photoclubalpha.com/2008/04/16/n...el-sensor-leak/
http://www.sony.net/SonyInfo/News/Press/20...010E/index.html

A700 & D300
http://www.sony.net/SonyInfo/News/Press/20...072E/index.html

http://www.sony.net/SonyInfo/IR/financial/...conductor/2007/

Logged

Fine_Art

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1172

Nikon press statement, yes its a sony.

http://www.robgalbraith.com/bins/content_p...cid=7-9319-9802

look at these pictures of the chips side by side (scroll down the reviews).

http://www.imaging-resource.com/PRODS/D3X/D3XA.HTM

http://www.imaging-resource.com/PRODS/AA900/AA900A.HTM

They are identical.
« Last Edit: January 08, 2009, 02:57:12 am by Fine_Art »
Logged

dwdallam

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2044
    • http://www.dwdallam.com

If someone or some test said something like, "The D3X is capable of capturing detail that the 1DS3 cannot produce, no matter what post processing you use"--now you have my attention. That is saying something substantial, although the extra detail the DX3 does produce may be tick terds in any real test, such as large printing. But it would still be saying something not so general or ethereal.

I wonder why one poster thinks the DS3 will be around for another year and maybe two? I'll bet it's gone with spring when Canon announces their new 1DS4 model at 25+MPs. There is no way Canon is going sit around and have the top end line suck second to Nikon, not even in pixels. Canon hasn't even worked on the 1DS3 anymore as far as firmware or anything else, plus the 1D model was from what I've read, a disappointment.  Canon has some catching up to do.

And look what we have here:

AMAZON:
Canon EOS 1Ds Mark III 21.1MP Digital SLR Camera (Body Only)
List Price:    $7,999.00
Price:    $6,549.88
You Save:    $1,449.12 (18%)

Nice, I just lost 1500.00 dollars in less than a year. I actually lost more than that, because if that price hadn't dropped so radically, I could have probably gotten 6500 for it today, if the Nikon hadn't forced the price down.

This is good though. The next Canon flag ship may open at 8000, if it equals or betters the D3X. But it won't stay that way for long because people will say, hmmm, Nikon, good, Canon good, Nikon 6000, Canon 8000, Nikon good. Same thing that Canon is trying to do right now--keep people from moving or initially going to Nikon by dropping their prices 1500.00US You buy the 1DS3 and you have enough to buy a 70-200MM lens too. Nice. Or, Canon 1DS3 with Canon 70-200mmL IS 8000. NICE deal!

The way, and this is all speculation on my part, but somewhat educated in this game, I see it is that right about the time Canon was getting ready to release the 1D3 and 1DS3, they had this vision that Nikon was about to steal their prize. So they hurried to market the 1DS and 1DS3, and immediately started working on the updates for those two cameras. They probably found out about one year before they were going to release the 1DS and DS3 that Nikon was planning to knock their blocks off with a 24MP FF camera that would out resolve and have less noise and higher ISO. Canon got lucky--they got both the 1DS3 and the 5DMKII out before the release of the D3X. They got very lucky indeed, even though the 5D2 would probably gone over well anyway, since Nikon has nothing to compete with in that area, especially price wise. Canon was very savvy.

I think we'll see the release of a 30sih MP 1DS4 this spring. Can you imagine that? I mean have you thought about that? Canon not only reclaims the high end FF market share--given that the quality is up to task--but also begins to REALLY bring in the medium format people over to FF. So Canon not only gets back it's top position from FF people, but the MF people start buying the new Canon too. At that point, unless MF has something that you just gotta have, MF is in trouble big time, and Canon is there to lick up the profit. I jsut wish Canon's felt as good in my hand as the Nikons. Am I the only one who "feels" that advantage from Nikon?
Logged

dwdallam

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2044
    • http://www.dwdallam.com

Quote from: Fine_Art
Nikon press statement, yes its a sony.

http://www.robgalbraith.com/bins/content_p...cid=7-9319-9802

look at these pictures of the chips side by side (scroll down the reviews).

http://www.imaging-resource.com/PRODS/D3X/D3XA.HTM

http://www.imaging-resource.com/PRODS/AA900/AA900A.HTM

They are identical.


It's not really a Sony from what I have been able to dig up. It's designed by Nikon. The specs are from Nikon., The pixel pitch is different than the Sony A900 chip and slightly different size too. Pretty much they just said, "Hey, we want one like this, can you make it, good, deal? It's still Nikon technology, just made in a Sony fab plant.
Logged

NikosR

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 622
    • http://

Quote from: dwdallam
Same thing that Canon is trying to do right now--keep people from moving or initially going to Nikon by dropping their prices 1500.00US

How many times must this be said to register? Canon (or Nikon) do not do most of the price reduction. Dealers and to a lesser extent the local distributors (i.e. the market) is doing it by lowering their margins. At least this is what industry watchers are maintaining. Does it make a difference? Well, yes in the context of these fairly useless discussions cum speculations, no if you're making your research as a prospective buyer.


Both Canon and Nikon have repeatedly shown they don't want a price war. You can only judge manufacturer pricing by RRP prices at introduction. Anything other than that and you end up making judgement errors since you discount the effect of the elasticity of margins down the supply chain. Also, as of now, because of the very volatile exchange rate markets, the best comparative indicator is to look at the Yen price and not the prices in $, Euro or other currency markets.
« Last Edit: January 08, 2009, 04:42:10 am by NikosR »
Logged
Nikos

dwdallam

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2044
    • http://www.dwdallam.com

Quote from: NikosR
How many times must this be said to register? Canon (or Nikon) do not do most of the price reduction. Dealers and to a lesser extent the local distributors (i.e. the market) is doing it by lowering their margins. At least this is what industry watchers are maintaining. Does it make a difference? Well, yes in the context of these fairly useless discussions cum speculations, no if you're making your research as a prospective buyer.


Both Canon and Nikon have repeatedly shown they don't want a price war. You can only judge manufacturer pricing by RRP prices at introduction. Anything other than that and you end up making judgement errors since you discount the effect of the elasticity of margins down the supply chain. Also, as of now, because of the very volatile exchange rate markets, the best comparative indicator is to look at the Yen price and not the prices in $, Euro or other currency markets.


So what %age are they making then on the 1DS3? 1500 price drop from 8000 assuming a 30% profit margin is pretty drastic. That's almost 19% profit they just lost.
Logged

Slough

  • Guest

Quote from: douglasf13
Maybe I worded things incorrectly, because I was typing quickly on my danged cell phone.  Sorry   What I'm saying is exactly that, and I don't mean to imply more.  I think the chips of the D3x and A900 are the same, with different CFA, AA, and an unknown special sauce for 14 bits, which leads to a different output than the A900.  That being said, I wouldn't say one is superior to the other.  A few dual users, like Iliah Borg, have shown that each camera has it's own IQ positives and negatives, depending on the situation, and I guess that's an appropriate thing to mention, considering the title of this thread.

No problem, and you might be right. We don't know.

One way that manufacturers sometimes work is to have the same hardware for multiple products, but only the top end product uses all of the features. It's often cheaper due to economies of scale to produce twice as much of the nominally more expensive to make hardware, than making two different pieces of hardware with relatively small differences. So it could be that the chip is pretty much the same (subject to the AA filter, microlenses etc. as you mention) but that there are features used by Nikon alone due to licensing i.e. Nikon allows Sony access to some of its design expertise, but not all. Pure speculation on my part of course. And largely sterile, since all we care about is the price and the functionality.

And MR has shown that in terms of IQ alone the two cameras (A900 and D3x) are pretty damned close in IQ. But that is not what differentiates them. As to the post that started this thread, well, even as a Nikon user I feel it has over-egged the pudding somewhat.
Logged
Pages: 1 2 3 [4] 5 6 ... 15   Go Up