Nikon Full Frame Sensor to Match Canon?

[Brad_]

<I>Nikon doesn't just "buy" sensors, they design them together with Sony.</I><P>

 

More like spec them. But in any case, until they have their own silicon foundry and can

manufacture their own wafers/sensors, with their own tweaks to the process, and be able to

develop and deploy their own IP, they will be at a severe cost (and probably performance)

disadvantage to canon, who has their own fab. There are a bunch of disadvantages to having

your own fab, though.

[andrew robertson]

There are a lot of good lenses that aren't extremely crappy on full frame digital, and for a

lot of people including myself a compromise will be gladly made. Until the lenses I

mentioned above are all out for DX, it clearly can't compete with 35mm (film or digital) for

many applications.

[._._z]

Don't forget that officially as of this past July, Sony is now partnered with Konica-

Minolta in the DSLR biz, and is said to be working on a full-frame sensors (or 1.1x-1.3x

if deadlines can't be met) to be used in K-M and Sony DSLRs this time next year.

 

Will Nikon get this chip from Sony if Sony becomes a direct competitor, especially

considering Sony's strong position in worldwide camera sales? Somewhat unlikely. More

likely Nikon will have to go shopping at Kodak or Fujiblad, or have to spend a lot of

money speccing out their own CMOS design for someone else to make. Tough road,

that.

[digitmstr]

>>They usually match Canon,<<

 

rather, they chase Canon...

[digitmstr]

>>that the corner quality of the images from the D2X is better<<

 

Ah! OF course, it's a cropped FOV so, ANY lens designed for FF will render better "edge" performance on a cropped sensor.

 

If we are citing professionals *opinions* just read the one by the makers or the stop animation movie recently released, Corpse Bride, I think it says it all. They could have used ANY digital imaging tool they wanted, they used a Canon 1Ds Mk2.

 

I seriously doubt that sensors have differences in performance from center to edge. If anything, it is other mechanical design issues that can render one design better that another one (i.e. mirrors, etc...).

[EricM]

I like Nikon skin tones.

[EricM]

love it!

[berg_na]

I guess if you prefer cartoonish colors then you should pick Canon...<p><img src="http://newpaper.asia1.com.sg/mnt/media/image/launched/2005-09-27/30reel_3.jpg">

[daniel_taylor]

It's funny to read the Nikon fan defenses of the DX format. On some level they have a point: as long as a DX sensor can deliver the image quality needed for a specific purpose, it doesn't make too much of a difference. I'm certainly not clinging to my 35mm equipment because it's "full frame". My DSLR beats it hands down.

 

But it's silly to ignore that a) DX is a mismatch for many existing lenses; and b) all things technology wise being equal, a FF sensor *will have* better S/N than a DX sensor of the same resolution.

 

The edge sharpness complaint has nothing to do with the sensor and everything to do with the lens. Shame on Nikon fans for trying to make this a "Canon sensor" issue. On the other hand, because an APS sensor sized image has to be enlarged more to reach a given print size, a FF sensor should show a slight resolution/MTF advantage even if both sensors have the same MP count, all other things being equal.

 

As for colors...there is no major difference in color rendition among any of the DSLRs. Photoshop color tweaks will change the color more than any inherent bias. "Orange tinted Canon photos" indeed...I've never seen any such thing in 12,000+ frames. You must need to calibrate your monitor if you're seeing orange tinted images from other photographers.

 

Having said all of that, sooner or later I imagine Nikon will offer a FF body. It will probably just be later since they don't have their own manufacturing for sensors.

[EricM]

"As for colors...there is no major difference in color rendition among any of the DSLRs. Photoshop color tweaks will change the color more than any inherent bias. "Orange tinted Canon photos" indeed...I've never seen any such thing in 12,000+ frames. You must need to calibrate your monitor if you're seeing orange tinted images from other photographers."

 

uhm, whoa. if you can't spot a canon jpg from a mile away...then...oh nevermind, my pizza's getting cold.

[berg_na]

<tt>a FF sensor *will have* better S/N than a DX sensor of the same resolution...</tt><br>Wrong. A 35mm-size sensor may or may not have better SNR than a DX-sized sensor, it all depends on the fill factor and the amplifier read noise.<p>

<tt>a FF sensor should show a slight resolution/MTF advantage even if both sensors have the same MP count...</tt><br>Not true. MTF is worse with larger pixels and unrelated to the sensor size.

[basscheffers]

Andrew, "biased" because it shows the Nikon to have better performance? :)

 

Anyways, I think they have the wide angle covered with 10.5/2.8 and 12-24/4. I don't understand your obsession with extremely fast wide angle and fish-eye lenses; I have personaly never seen an image taken with a lens like that which wasn't stopped way down to have front to back sharpness and what good is a f/1.4 lens if you always stop it down to f/8 or more?

 

Giampi: the sensor isn't different centre to edge. The problem is that unlike film, a sensor isn't flat and the photosites are down in a little well. At the edge, the light is at such a steep angle that not all of it reaches the the sensor becaus it doesn't make it over the wall. That is what causes the corner light fall-off and less resolution in the 1Ds.

[bengt_rehn]

You are right about MTF Berg Na, but you dont need that high MTF when you increas the sensor size. Compare with the state of the art 22 MP backs for medium format with its much larger pixels and medium format lenses with less MTF than 35mm lenses or DX-lenses. You will still get more resolution out of the larger system in reality.

[douglas_green1]

Bas: "Viewfinder size has, however, no influence on resolution or any other quality of the final image."

 

Really? I would contend that a viewfinder that is too small to use certainly degrades the final imagage quality. If you disagree, try composing without one.

[douglas_green1]

Obviously, I meant "image quality" not "imagage quality".

[andrew robertson]

From the final bit of the review from naturfotograf.com:

 

"The Canon combination is possibly just that little better to be declared the "best"."

 

The final comparison was the only really fair one in the test from my perspective.

Dpreview.com agrees that the 1Ds resolves more than the D2X as well.

[andrew robertson]

Bas, it's called shooting indoors in low light or at night. It's not an 'obsession', it's simply a

reality that the Nikon DX lineup, and the Canon EF-S line both have gaping holes in 'em.

[frank_skomial]

PC Magazine, October 4, 2005, provided resolution tests for D2X (1975 lines - 17-55/2.8DX lens), but Canon EOS-1Ds Mark II with unspecified lens (?) exceeded by unknown number max possible resolution above 2000 lines).

 

PC Magazine is no authority on photography, so they did not care to mention what lens they used on the 1DsMark II test. Seems for $7999 it must be sure bet that any lens would give high number of lines, that is higher than they have max limit of testing method (ISO or not ISO testing).

 

No one would even question that test result. Certainly pooling ISO method authority gives this test some sense of objectiveness. Who would doubt ISO?

 

The Canon test exceeded the 2000 lines resolution that was stated as maximum possible with their ISO conforming testing method, so the real resolution could be much higher ? who knows exactly? 2500, or 3000 lines ? - just a hint ?.

[oskar_ojala]

So when is Canon going to introduce a 6x4.5 sensor to match Mamiya and 'blad? Seems to me that they are falling behind ;-)

[basscheffers]

Frank, so they don't say by how much over 2000 lines and not which lens. Great test. You need about 1.25% more than 1975 lines to me "over 2000" and it also doesn't say where that was measured. If that was center, then obviously the Canon would be higher, as it has a higher resultion sensor. Duh! My point is that landscapes then to have detail in foliage in ever corner of the image, for example.

 

Douglas, you are being silly. The D2X does in fact have a viefinder and most reviewers will tell you it's pretty darn good and not much less than a FF viewfinder.

 

Oskar: good point. :) Also, have you ever noticed how these bigger photosites actualy have more noise at high ISO than DSLRs do? How can that be, BIGGER IS BETTER!

 

I don't own either of these cameras and I never will. But looking objectively at the tests I have seen, the D2X "matches" the 1DsII, it is too close to call which is best. If you have Canon glass, stick with it and get the 1Ds. If you have Nikon glass, you'd be a fool to jump ship.

 

In any case, still neither can match a drum scanned piece of Velvia, certainly not medium format. Well, not in glossy magazine prints anyway, they may for LightJets - I a have never seen them from these cameras - which generaly work a lot better for digital prints than a printing press does.

[daniel_taylor]

"uhm, whoa. if you can't spot a canon jpg from a mile away...then...oh nevermind, my pizza's getting cold."

 

You couldn't spot a "Canon jpg" from a foot away, and I would bet money on that. Care to submit to a test?

 

****

 

"a FF sensor *will have* better S/N than a DX sensor of the same resolution..."

"Wrong. A 35mm-size sensor may or may not have better SNR than a DX-sized sensor, it all depends on the fill factor and the amplifier read noise."

 

"ALL THINGS TECHNOLOGY WISE BEING EQUAL..." was the beginning of the quote you chopped up. Please don't misquote.

 

"a FF sensor should show a slight resolution/MTF advantage even if both sensors have the same MP count..."

"Not true. MTF is worse with larger pixels and unrelated to the sensor size."

 

Once again, lets try the FULL quote:

"On the other hand, because an APS sensor sized image has to be enlarged more to reach a given print size, a FF sensor should show a slight resolution/MTF advantage even if both sensors have the same MP count, all other things being equal."

 

This is true because ***LENS*** MTF drops as lpmm goes up. Say you want 12 MP for a 16x20 print. You have a FF 12 MP sensor, a DX 12 MP sensor, and the same lens. The FF image will have better MTF when enlarged.

 

http://www.photo.net/equipment/digital/sensorsize/

 

Please pay close attention to the paragraph that starts: "Line 9 of the table shows the MTF of a perfect lens operating at f8...".

 

****

 

"Giampi: the sensor isn't different centre to edge. The problem is that unlike film, a sensor isn't flat and the photosites are down in a little well. At the edge, the light is at such a steep angle that not all of it reaches the the sensor becaus it doesn't make it over the wall. That is what causes the corner light fall-off and less resolution in the 1Ds."

 

This is a widely quoted theory that is unsupported by real world results. http://www.luminous-landscape.com/reviews/cameras/1ds/1ds-field-3.shtml

 

Again, it's funny to read the Nikon-biased responses. Half-quotes, half-theories, and myths.

 

Having said that, if I had a Nikon lens collection I wouldn't be too disappointed in Nikon's offerings, and probably would not switch over the FF issue. But I imagine their are some Nikon photographers who want/need FF and are considering a switch, regardless of the "arguments" offered by Nikon faithful.

[douglas_green1]

Bas, EVERY nikon DSL has a viewfinder that is 1.5X smaller than it would be if a full frame sensor was used on it.

[berg_na]

<tt>Again, it's funny to read the Nikon-biased responses. Half-quotes, half-theories, and myths...</tt><br>

... and it's equally amusing to read the technically incorrect arguments of the other side...

[harvey_edelstein1]

The UK magazine Practical Photography tested the Canon FF vs the DX2 and the result was they rated the Canon higher. This magazine in the past rated the D70 the best dslr in its price range so I feel they are not biased towards Canon. The reason the Canon FF was rated better was because it was able to be set up at higher ISO and performed with less noise at 800 iso and higher.

 

A larger sensor area will be capable of gathering more energy from light striking it. This is translated into better electrical noise and color saturation characteristics. In turn less amplification and correction techniques are needed to handle noise so its easier to get clean sharp and better dynamic range in the photos.

[praveen murthy]

Canon 5D: $3000

Canon 24-105/4L: $1500

Lowepro to hold them: $100

 

Watcing Nikon users squirm and scream: priceless

 

There are some things money can't buy, for everything else, Bnetch.