Nikon D7100 preview updated with a look at effect of OLPF omission: Digital Photography Review

Am I the only one that is excited about the centre AF point that can autofocus at F8!!! This will let me use the new 80-400mm plus and 1.4X extender and the camera will still autofocus!

0 upvotes

miles green:

Except the point isn’t about a bayer pattern/matrix filter, the point is the Leica M9, Sigma Foveon cameras, Pentax K5 II, Nikon D800E, and now the Nikon D7100 do away with the optical low pass filters (yes, then Nikons may have them but the effects are then countered).

You’ve confused a discussion about cameras without OLPFs with the one about the newish Fuji Xtrans filter that replaces the bayer pattern filter on the newish Fuji X cameras.

Theses are different filters, doing different things. And one does not replace the other.

I’d imagine that Sigma Foveon cameras don’t really have a bayer patterning filter either, because each sensor photo cell point can record RGB and keep the data separate for later construction into a photo file.

0 upvotes

It's a pity the Sigma SD1 is not on the comparison tool. At base iso, it's phenomenal! Incredible clarity and no moire!

0 upvotes

Well you can check out the Leica M9 at base ISO.

0 upvotes

Found it!

http://www.dpreview.com/reviews/sigmasd1/19

The Leica M9 still has a bayer matrix... Gotta see it to believe it! (just make sure you look at the center of the frame, where the lens can keep up!)

0 upvotes

miles green:

Except the point isn’t about a bayer pattern/matrix filter, the point is the Leica M9, Sigma Foveon cameras, Pentax K5 II, Nikon D800E, and now the Nikon D7100 do away with the optical low pass filters (yes, then Nikons may have them but the effects are then countered).

You’ve confused a discussion about cameras without OLPFs with the one about the newish Fuji Xtrans filter that replaces the bayer pattern filter on the newish Fuji X cameras.

Theses are different filters, doing different things. And one does not replace the other.

I’d imagine that Sigma Foveon cameras don’t really have a bayer patterning filter either, because each sensor photo cell point can record RGB and keep the data separate for later construction into a photo file.

0 upvotes

I absolutely LOVE this body.

Pair it with a 35mm 1.8G or 50mm 1.8G and you will be amazed at the incredible detail, dynamic range, and color depth. It's absolutely beautiful -- you get so much bang for the buck.

Comment edited 5 minutes after posting

0 upvotes

I have both the D5200 and the D7100. For all practical purposes the image quality is the same. The only real difference that I've noticed between the two is that the D7100 is really starting to show the limitations of Nikon's lenses. The flagship DX lens, the 17-55 f/2.8G is almost unusable at f/2.8 because if the apparent softness. The D7100 retains detail, but the image is veiled in a soft glow similar to a diffusion filter.

What Nikon really needs to start working on is updating their pro lenses to match their sensors. My 14-24 just got back from NPS repair so I haven't had a chance to test it out, but so far the only two lenses I have that are holding up wide open are the Sigma 35mm f/1.4 and the new Sigma 17-70 f/2.8-4 C.

1 upvote

There are also the very good manual focus Zeiss lenses that mount on Nikon dslrs, no zooms though, but still much better optically.

But generally yes, Nikon should be improving the optical quality of its high end lenses; none of which come close to a good Samsung NX, Fuji X, or Olympus 4/3rds lens, let alone Zeiss or Schneider.

1 upvote

????
I've used the Nikon AF-S NIKKOR 300mm f/2.8G ED VR II, the Nikon AF-S NIKKOR 500mm f/4G ED VR II and the Nikon AF-S NIKKOR 600mm f/4G ED VR with D7100. Not only names and prices of these lenses are impressive, but certainly also the results with D7100.

Comment edited 24 seconds after posting

3 upvotes

I think you're missing the point here.

99% of the people that are going to be buying the D7100 have no need for fast exotic MF Zeiss primes or highly expensive super-teles.

As of right now there's no Nikon pro DX standard zoom lens that can handle the resolution of the D7100 competently. If the D7100 is Nikon's flagship DX camera they need to pair it up with a flagship standard DX lens that allows the camera to be used to its full potential.

Zeiss primes are great and all, but personally I'm not going to buy one for my Nikon. I'd buy one for my Leica.

2 upvotes

JDThomas:

Why not use a Zeiss on a Nikon, they're much better in low light high ISO situations than so called ED Nikon glass? And Zeiss lenses don't show the imperfections that Nikon lenses show when mounted on the D7100. I'm sure of both points.

As for mounting those M mount Zeiss lenses on a Leica, okay, but those are different lenses, and the reason for using a D7100 instead of a M9 or M8 is that the D7100 is less audible and much better at high ISO work than either Leica.

(I assume you're not talking about mounting Zeiss M mount glass on a 35mm film rangefinder (Leica, Contax, Minolta, Konica, etc)

Comment edited 5 minutes after posting

0 upvotes

My reasoning is simple. I use my Nikons for work. I prefer the speed of AF and the convenience of a zoom. This way I get things done quicker and I get paid quicker. I pay top dollar for my Nikon pro glass and I expect it to be on par with the cameras I'm pairing them with.

I use my Leica for personal photography and art. I don't mind taking my time and manual focusing. I don't shoot if high ISO noise is a concern. I'd prefer a nice Zeiss lens for the Leica over one for the Nikon any day.

In any case, I was specifically talking about Nikon needing a standard DX zoom that can handle the D7100. As I mentioned the Sigma 35mm f/1.4 is GREAT for the D7100 which really negates any need for me to buy a Zeiss MF lens for my D7100 anyway.

1 upvote

sorry to say that DX17-55/2.8 is no good lens.

0 upvotes

The 17-55 WAS a good lens when paired with earlier cameras. It's STILL a good lens when stopped down to f/5.6 on the D7100.

Heck, it's not really that bad when the image is downsized to about 12MP (but then why buy a 24MP camera if you have to downsize to get the best performance).

Really my only point is that Nikon needs to redesign the 17-55 f/2.8 to keep up with their cameras. And they likely are.

0 upvotes

24-85VR tack sharp on DX corner to corner.

0 upvotes

should get D800 and 24-70/2.8.

should giveup shooting with DX D7100 if is not tele resolution (small pixel pitch) or frame rate.

1 upvote

Reilly Diefenbach:

Lenses can be plenty sharp, but have poor colour performance; that's like a lot of Nikon ED lenses.

"Tack"? I've never thought tacks particularly sharp, maybe carpet tacks. Why not razor sharp? (Perhaps the term refers to something else other than actual tacks, anyhow it seems kind of silly to use it when describing lens performance.)

0 upvotes

@yabokkie:

I have a D800. I also have a D600, a D700, and a D5200 as well as the D7100 at the moment. I write camera guides. I have a pretty good grasp on what gear I "should" use. I'm writing a guide for the D7100 right now.

@Reilly Diefenbach:

The 24-85 isn't a standard DX lens, nor is it a pro lens. I have a 24-70 that is pretty sharp wide open, but that's not the issue. 24-whatever isn't a particularly useful range for DX.

1 upvote

think 17-55/2.8 is a pro-lens of good mechanical quality only.

Nikon gears can be categoried into two groups, pre-2007 are all sh!t (the company should be called Shikon) and a totally new company was born late that year.

the new Nikon is a superman 5 yo. it doesn't have a DX17-55/2.8. if they are planning their own product for DX it should be DX16-50/1.8 at about the same size and price as 24-70/2.8.

0 upvotes

No point in even continuing this discussion. You are completely derailing it with wacky talk.

16-50 f/1.8? Haha.

0 upvotes

I'm only saying that 17-55/2.8 is a bad lens at that high price and I think it's a good thing we put it and the old Nikon behind us.

what a beautiful world.

Comment edited 2 minutes after posting

0 upvotes

99% of people with D7100 users aren't going to buy pro zoom lenses either. Those are WAY expensive.

The 35mm and the 50mm 1.8G offer fantastic IQ at a cheap price.

1 upvote

Ok. My comment wasn't meant to start an argument about which lenses are sharper, what primes are better, etc...

ALL I was saying was that Nikon needs to update it's pro DX standard zoom to keep up with the resolution. The end.

0 upvotes

,

Comment edited 51 seconds after posting

3 upvotes

The colors of the D7100 seem punchier, aren't they?

0 upvotes

I know that. the main topic is the effects of no anti aliasing filter on high MP camera but what is surprising is the small amount who have commented on the rather obvious and surprising superiority of the pentax over all the others on all test modes . On the nikons I can see a difference between them but I admit they are small , even at 100%

1 upvote

Um, one has to account for the lenses used in the test too. It's not like the Pentax used Nikon lenses in the test shots.

My experience with the samples I've shot is that the Nikon D7100 easily beats the K5 II, but I was using a better lens on the Nikon and was particularly interested in high ISO work at which the Nikon D7100 and D5200 are better than the Pentax.

1 upvote

Remember the Pentax isn't working iwth 24MP. I'm sure a 10MP camera without the filter would look even more superior to one with it than the Pentaxes do. What you have to remembr is Nikon's isn't anywhere near as suceptible to moire as the Pentax is, due to the resolution.

1 upvote

@HowaboutRAW
Comparing the D7100 and K5iis at iso 6400 on this site, the K5iis looks well ahead (much cleaner) to my eyes.

Regarding AA-less cameras, I apreciate the increase of detail visible on large prints at at high screen manifications, but there is also a fair amount of false detail (moire banding) that is visible even at normal viewing.... I don't like it, so I'm not persuaded! AA-less sensors seem to work better the higher the resolution of the camera.

At the end of the resolution race, we'll all be using prime lens at f/5.6 or f/8....

0 upvotes

miles green:

Thank you I have my own raws from both; as I said. Now what I don't have are K5 II raws shot with a Zeiss.

But look the Pentax is a plenty good camera. Anyhow the problem with the Pentax, with Pentax lenses, is poor colour, when compared to a Zeiss lens.

This detail thing is mostly a distraction.

Comment edited 2 times, last edit 7 minutes after posting

0 upvotes

We are getting to the point where comparing higher MPs, AA vs AA-less, are all pointless and feeding hype and marketing when there are more important things to measure. The quality of shutters and mirror actuation and their influence on micro-blur (which is already of higher significance than these high MP AA vs AA-less tests). The quality and precision of AF sensors and *not* the number of them. The quality of lenses at different apertures and the quality of actually getting lenses within tolerance (so far we have very few tests on a few lens most of which seem to be super zooms).

9 upvotes

Speaking of all things 'AA,' how nice would it be if Nikon designed some upcoming bodies that would allow the option of using AA's, ala Pentax. It would turn my head and make me seriously think about them for my next system, to be sure.

Comment edited 29 seconds after posting

1 upvote

They do provide that option, via a battery holder you insert into the battery grip add-on.

0 upvotes

Indeed kimvette, that is a great feature that you mention, but I want to see it in the body as well, sans battery grip!

0 upvotes

"The Emperors New Clothes"
I've been told that it is superior therefore I do perceive its superiority.
It simply has to be better because it is promulgated by the pundits in this forum.
ha ha ha ... guys you made my day. Thanks.

2 upvotes

Sounds like somebody needs a new monitor. One wonders what the Nikon corp would have to do to impress some you would-be experts.

5 upvotes

You obviously don't understand what is going on here. The purpose of the anti-aliasing filter is to BLUR THE IMAGE at pixel scale to reduce moire, and at a high enough resolution that is not really a problem. If we have gotten to a point where we can remove that limitation on digital cameras, I'm all for it. Its not like anti-aliasing filters are analogous for photography, it was a limitation of early digital technology. It is better whether you like to recognize this or not. Just like a-spherical glass or improved coatings are a big step in optics, removing the aa filter is a big step in getting the most pure image from the camera and lens, and isn't that the point of all this gear talk?

0 upvotes

Couldn't there be a way to remove moire in pp? Like noise. Or even in camera like aberrations etc?

0 upvotes

Yes, there are ways to reduce moire, but a problem with that is that it's destructible process. If Moire is on a shot - you can't restore it to look as good as if moire would never be there.

2 upvotes

That said - for landscapes it's meaningless. For shooting cityscapes or fashion photography - any camera that produces visible moire is totally useless. So it very much depends on what kind of photography you do.

2 upvotes

it's impossible to remove moire in post. without LPF you are guaranteed to lose resolution and get noise whereever there is higher frequency projected on the sensor.

it is possible however to make the image look very sharp, which doesn't mean there is better resolution but opposite.

1 upvote

Adobe Camera Raw (and probably LR) is pretty damned good at removing moire. Along with that, the same tool can be used to reduce high ISO chorma splotchiness in areas with low detail (where it shows the most).

0 upvotes

For me, it is at the moment, a mystery, if a license or other authorization from the U.S. Government is a requirement, to export a Nikon D7100 (vacation).

B&H in regard to the 'Leica M9 Rangefinder Digital Camera Body (Black)' does state that this item is controlled on the U.S. Commerce Control List set forth in the Export Administration Regulations or the U.S. Munitions List set forth in the International Traffic in Arms Regulations, and may require a license or other authorization from the U.S. Government to be exported from the United States.

0 upvotes

Comparing the checker board at the top right, K5IIs>A77>D7100>D5200

Comment edited 34 seconds after posting

1 upvote

not really sure, but at first look yes. the 24mp images are just ... well, bigger.

0 upvotes

It's really hard to compare as dPreview AGAIN didn't focus all of the cameras at the same point. Click the Martini bottle - Pentax wins hands down. Click the Queen card - A77 wins hands down.

They really need to learn how to focus finally.

Comment edited 31 seconds after posting

5 upvotes

Interesting stuff, to my untrained eye the D7100 images look a lot better (sharper) than the D5200 - especially if you look at the moss on the trees and I like the way the D7100 renders colors ......

0 upvotes

if you shoot another image using the same camera/setting but at higher ISO (like 400 or 800), many may say that one looks sharper than the base ISO shot.

O-LPF-O does the same. all it can do is to introduce some noises and nothing else. the only difference is the source of noises. a high ISO (low exposure) shot gives you shot noise while O-LPF-O noises are high frequencies folded down (because it does not have LPF to filter them out).

0 upvotes

So for those of us trying to decide between the usefulness of the flip-out screen on the D5200 versus the slightly better resolution of the D7100, sounds like a wash maybe?

0 upvotes

Those are not the only differences though. The D7100 has a much better AF, Viewfinder, controls, top plate LCD, ect which will make a much bigger difference when using the camera than either the lack of OLPF or not or having a flip out screen or not. I would go to a store and try them both to make sure you can live with the dimmer, smaller, 95% coverage VF, only one control wheel and more menu centric controls in general of the D5200 before you deiced.

Comment edited 3 minutes after posting

0 upvotes

Judging by these shots the difference between the D5200 and D7100 is even smaller than the differnce between the D800 and the D800E. It isn't even noticeable in all shots. Of course the same is true for the D800E as well. I think it is just a marketing move to make the camera appeal to more serious users. What I am waiting to see is samples of fabrics like say wedding veil or macros of insects with the eyes in focus. That will be the real test for moire.

2 upvotes

The D800e uses a cancelling filter. I do not know how that works without looking at the patent or something, but the sensor isn't naked still. Id imagine the d7100 is the same. they say it is like an aliasiang filter that has a 2nd filter that cancels out the first one if that makes any sense (and it doesn't to me much). The k-5 IIs has no filter at all, but IMO I still don't like the moire. Maybe 24mp is better somehow in this regard even though it is only 30% more pixels. Maybe nikon has a good solution. I don't know. The shots I've seen from the 7100 really make me hope the next pentax is at least equal because they are amazingly detailed.

0 upvotes

So, what's the next marketing trick? 50 Mpx? Super RAW format or just the smallest DSLR with 100 colours available? Can't wait for new releases every 6 moths...

1 upvote

low res lovers will get their ultimate nuke: a single pixel sensor.

6 upvotes

50 mpx would not be a marketing trick. It would have noticeably higher IQ, even when down sampled, than even the D800 or any 24mp camera.

Comment edited 53 seconds after posting

3 upvotes

Josh152

Did you read the review of the D800 on this site? I'm not going to look at it again, but I believe that it said something like you need the top lenses and top technique (read: tripod at least) to squeeze out a tiny bit more resolution from the D800 than lower megapixel competing cameras.

As you increase pixel density, a tiny movement of a hand-held cameras, let's call it a one-pixel shake, that would have affected a single pixel now affects a swath of pixels. So image stabilization becomes paramount, without even beginning to talk about the necessary optics.

And how much resolution are you hoping to squeeze out of 50 megapixels? If you can already resolve every hair on your cat with 10 megapixels, and you can, then what exactly are you hoping to see with 50 megapixels?

2 upvotes

yabokkie

"low res lovers"? How about people with common sense.

I mean, if you buy a car that only goes 100 m.p.h., and you will never drive it over 60 m.p.h. anyway, because that's the law, are you a "low speed lover" just because you don't buy a car that goes 200 m.p.h., which you will never need?

It sounds like you're laughing at us, but maybe we're laughing at you too.

I advise you to buy a good 2 megapixel camera on ebay, they're everywhere cheap, and see what size prints you can get out of it. 8x10 is a given. You might be able to bigger. The need for more and more megapixels (16 isn't enough? are you kidding me?) is getting into compensation territory, like guys with big cars, if you get my drift.

Or maybe you just want a camera that goes up to 11.

Comment edited 46 seconds after posting

1 upvote

@bobbarber.

At the same viewing size the higher mp camera will always look as good or better with the same lens, technique ect. The differnce will only matter if you are viewing on a computer at 100% or making huge prints. In those cases you have to use the best lenses and technique possible to no matter how many megapixels the camera has.

The lower megapixel camera would merely be hiding your bad technique because it's lower megapixel sensor is limiting the resolution more than your bad technique is. I would much rather have my technique which can be improved be the limiting factor than the resolution of the camera which can't be changed without buying a new one.

3 upvotes

Josh,

I don't disagree with you, but it sounds like you are just chasing specs, instead of deciding what you need.

I HAVE made quite a few prints from a 2 megapixel camera, the first I owned, a Kodak, and they are as sharp as a D800 at 4x6. They are softer at 8x10, but they are still sharp at 8x10, especially at the right viewing distance.

Here's another thing I've tried. I shot my 16 megapixel GH2 at 8 megapixels (all cameras will downsize like this, just choose your size in the menu), then up-rezzed the 8 megapixel photos and compared. THERE WAS NO DIFFERENCE. Sometimes the 16 megapixel photo was slightly better, but sometimes the 8 megapixel photo was better! And the 8 megapixel photo was always smaller and easier to handle.

Yes, YMMV, but really, with this camera compared to previous cameras, I don't see the difference.

0 upvotes

I hate to say it to you all but like it or not, we are ramping towards an apex where diffraction will eat any further gains in resolution. Its a simple fact that the airy disks are only so small. These high density sensors are limiting at f9 or so. Not exactly a brick wall yet, but soon your resolving power will be at the limits of the lens more than anything. The D800 is already severely challenging for glass to fully resolve. Diffraction limits of f5.6 would be nearly useless on full frame. You would throw all that resolution away and be forced to have huge files with no more information in them. 35mm was a film size. Lenses only had to resolve maybe 20mp max due to the grain size in the film. Its already been noted that the gains in fine detail on the k-5 IIs are lost above f8, though overall the image stays sharper. That's on a 16mp aps-c camera, which scaled to FF equals the d800. The 7100 pictures look on the level of FF detail, so there are gains yet to be had, but clearly the MP war will become meaningless soon if you ask me.

0 upvotes

the optical resolution limit is the first radius of Airy disc. we need at least two 2x2 pixels or 4 photosites within this radius to read it out for Bayer sensors but that's dump calculation. practically we need only 3 (this number comes from lowest/comfortable/absolute resolution in DPR's camera tests).

we can make consumer grade lenses to reach perfect at f/2.8 but f/5.6 sounds good for lenses that most of us can afford. the radius at f/5.6 for 555nm wavelength is about 3.83um for 3 photosites, that's 1.26um pitch or over 500MP on a 35mm format sensor.

there could be 2000MP for perfect f/2.8 lenses, and it could be a mad 5000MP for "violet blue". but that's only calculation. I bet there will be little difference beyound 200MP.

but still I think we will get more and more pixels well beyond 5000MP, not for resolution but other image qualities. whether we should call them pixels is open to debate but to me, a perfect pixels is one that measures a single photon only.

Comment edited 5 minutes after posting

0 upvotes

atmospheric turbulence is another factor that limits resolution. pixel pitch will have to go 0.2um to make a standard 50mm lens affected by heat haze outdoor (it's about 1-2um for 500mm, depending on the weather, zenith angle, and distance to your subject).

0 upvotes

Are my eyes broken? When I look at the "globe" crops, it looks to me like the D7100 image is full of (non-color) moiré, particularly noticeable in areas without detail. Is that merely what the globe looks like, or is the globe printed with a half-tone and we're seeing moiré due to the absence of an OLPF?

1 upvote

Your eye are not broken, I don't know how the author could miss this.

3 upvotes

DPReview needs to give us a macro shot of that globe to put this to rest. If you look at the Pentax 645D and look at the globe it seems to indicate there *are* markings that represent waves or current, but then if you look at Africa you see the same waves on the actual land continent! If true I haven't been comparing real detail for a long time and just how one moire looks over another!

0 upvotes

It's halftone (4-colour process), and the patterns are also there (and only slightly less distinct) in the D5200 crops. Without using a significantly higher or lower resolution, or blurring to the point that the dots can't be resolved at all, moiré will happen on any sensor, OLPF or not, when two grids of slightly different sizes collide (in this case, the sensor and the halftone). That will be a constant in digital photography until a pseudorandom layout of sensels happens (and someone figures out how to extract colour information from it without paying a high sensitivity tax).

0 upvotes

Those crosshatch lines are some of the best things I've seen on that studio setup to judge optical resolution and high ISO settings. Only a few cameras can resolve those lines correctly, and only while set on RAW files, JPEG can't resolve the detail at all. The Nikon D800E shows the lines clearly on RAW, but on JPEG they disappear into a blurred mush, just like every other camera when set on JPEG. From best to worst, the D800E can resolve the lines clearly on RAW, the D800 too but slightly less clearly, the D7100 shows them but less clearly than the D800 and with lots of moire especially in the Arabian Sea, and the D600 misses most of them in the ocean except the ones near the Seychelles. The Canon 5D Mk3 does slightly better than the D600 but with moire, and the Canon 6D is prety close to the Nikon D600 but with more moire than the D600 or 5D Mk3. The Pentax 645D is very similar to the Nikon D800E, but the 645D can resolve the crosshatches around Tanzania that the D800E can't.

0 upvotes

As you bump up the ISO on all cameras they start losing resolution, and the crosshatch lines in the oceans eventually get blurred together with noise and disappear. The crop sensor cameras are much worse at that than full frame cameras. On the Nikon D800E by the time you get up to ISO 800 on RAW most of the crosshatch lines in the ocean are gone and are replaced by noise. On the Nikon D7100 by the time you get to ISO400 on RAW the lines are gone and replaced by noise. At ISO 3200 the D7100 is a grainy mess, but you can still see some lines on the D800E.

At ISO 1600 which is the highest setting on the Pentax 645D it is still showing lots of lines in the oceans, whereas on the D800E the lines are gone and replaced by noise and snow. The Pentax 645D can show the crosshatch lines clearly while set to JPEG, while the full frame 35mm and crop sensor cameras cant.

1 upvote

@photobfa: thank you for pointing out that the MF camera still wins even with an older CCD sensor. A lot of people are lost on that. Interesting how it still retains detail over the d800 even at 1600. I shall have to look myself soon now.

0 upvotes

Unless I'm missing seeing all the test images then this is like testing an underwater camera by taking photos of a cactus.

Buy a suit jacket with fine weave. Put the suit on a model.
Take photos at full, 3/4rd and portrait length.
Look for moire.
The original 1Ds, weak AA filter, images extremely sharp, chance of moire - high.
The 1DsIII, stronger AA filter, images slightly softer, chance of moire - low.
The low pass filter is a trade off but as far as I can see you are looking for the benefit of removing it without looking for the reason it was there in the first place.

Andrew

6 upvotes

moire is artificial noise in large scale for our eyes to recognize easily. more often is those noises do not construct large scale patterns. they are there but not easily seen by our eyes.

> images extremely sharp

color filter type sensors have the advantage to deliver far higher resolution than stacked ones. but that's not real reliable resolution. extremely sharp crisp image is not physically possible. if you see one, you either see noises caused by high frequencies folded down, or you see a heavily cooked image like found from Penta/Oly cameras.

1 upvote

I bought a refurb D3200 for about a 3rd of the cost of the D7100. I see no difference in the image quality between the D3200 and the D7100 from this comparison. Sure, it would be nice to have the better body with more direct access buttons, better build quality and the ability to use more lenses, but I only have g lenses, so it does not matter.

1 upvote

All modern cameras are very close in image quality. The main reasons to buy camera like D7100 are better controls, larger body, weather sealing, larger viewfinder, better autofocus system and so on.

3 upvotes

sergey has it right.

I think that manufacturers will have to start selling features as hard as they've been pushing ISO advantages, per-pixel sharpness, etc.

The IQ differences between modern cameras are insignificant for most people, especially when print and viewing sizes are taken into consideration.

The only reason that features are not pushed more now is that many people are listening to the IQ snake oil salesman.

Now, if you really do print billboards (And who needs high resolution for a billboard anyway? Look closely at one sometime. I've seen great billboards done with 5 megapixel point and shoot Fujis--I knew the photographer) and people really are judging your work with a magnifying glass...

I can't do it. I just can't finish the hypothetical. It's too stupid. Really, the IQ is there with all these cameras. Features are where it's at now.

Comment edited 2 times, last edit 1 minute after posting

4 upvotes

This update does not need a man -
costs too much

0 upvotes

Only Fuji with the x100s could take advantage of the removal of the low pass filter.

2 upvotes

Yes. Howerer, the resolution of XTrans is lower than one of bayer ;)

1 upvote

I'll stay with my D300s for a while longer. I don"t like the feel of the D7100 at all.
It's an ergonomic downgrade for me. The buffer is some kind of a sad joke. The D400 is probably just a fast fading dream. I'll have an eye on Canon, or Sony?

1 upvote

Take a serious look at Sony, IMHO you'll be positively impressed.

2 upvotes

D300/s owners who have upgraded state that even the AF is faster, and more precise in the dark. It's not just a sensor upgrade. Lots of other little things in there, too.

Comment edited 2 minutes after posting

1 upvote

I still have D200. I know where you are coming from. D300 is an awesome piece of equipment. That said, the sensor tech is lightyears ahead at this point. Unless you never shoot anything over ISO100, you should get a new camera.

Comment edited 34 seconds after posting

0 upvotes

Well, I got A77 and shot a wedding with my friend shooting D300s - in terms of usefulness A77 wins hands down. AF is noticeably better in low light, I could shoot under the angles he even didn't dream of, tiltable flash allows perfect, seamless shooting in vertical, 12 FPS is a great thing when you want to capture fast action (eg. catching wedding bouquet - I got 5 great shots, with everything composed perfectly he got 1 that was "ok"). Only thing we didn't compare was handling of the tough conditions, but as I washed my A77 under the shower few times (got weather-sealed lenses and weather-sealed flash gun) and it hit the ground few times over it's lifetime with no harm - I'm quite sure it'd be able to stand against D300s quite nicely :)
People complain about buffer size, but for me it was never an issue - I just bought the best Sandisk card they had and everything works fine with newest firmware.

2 upvotes

All this stuff - OLP vs no OLP etc - is just yet more marketing guff intended to differentiate products that are essentially functionally identical. It keeps the churn going, with people replacing perfectly serviceable products with marginally different "latest" versions. Of course the companies are only partly to blame; there's a constant demand expressed in sites like DPR for features and "improvements" to enable "better IQ" in snaps of dogs, cats, kids and cut flowers.

16 upvotes

Have a look at the Coolpix A gallery posted yesterday, or D800E images or even this D7100 test if you think no OLPF (not OLP) is strictly a marketing ploy.

As far as DRP Galleries and snaps, there is a bit of irony in your criticism of other peoples images.

2 upvotes

Yes, thank you DPR for helping instigate new improvements, innovations, experimentation and progress. Without it, everything stagnates, and stagnation becomes so... grimy and old.

So here's to marketing and innovation! With it, in just 200 years we went from pushing a plow to landing on the moon!

0 upvotes

Let's rather talk photography here. :)
If you start with the plow and the moon, I'd say that pushing a plow makes much more sense to me than landing on the moon...

0 upvotes

@fberns, you must have missed my thank you to DPR for publishing new innovations and experimentation here. In reference to this one about the low pass filter.

Now, if your reply was actually a desire to censor my supportive comments about innovation in general, that is certainly not talking photography, and I think you are splitting hairs pretty thin for your own self-interest. BTW, you are free to choose the plow, that's your choice. You might even post some pictures of your plowing.

0 upvotes

Why Not make cameras with a low pass filter. With the option to bypass. ie
Menu:
Low Pass filter:
On
Off

Comment edited 4 minutes after posting

1 upvote

Its a physical filter not some software emulation.

11 upvotes

A perfect example of 'thinking inside the envelope' brendon. You need to watch a video I saw yesterday, where a car door opens underneath the car, without changing the underbody clearance. I thought it was CGI'ed. It looks impossible. But it's not. The door disappears just as slick as rolling the window down.

0 upvotes

I bet that feature is so incredibly useful and cheap that we'll be seeing it in all of our cars within then next year or two, right?

1 upvote

Well, I don't want to stray too far off of Photography here, but dougster's desire for a switchable low-pass filter may not be impossible. It depends on usefulness and demand. Those van doors that slide along the body of the van have saved many lives eliminating the blind egress of kids from the side of a vehicle. Not swinging a car door out into traffic might actually be useful and desirable. And, if a switchable lowpass filter is useful, maybe dougster will invent one.

"This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us." -- Western Union internal memo, 1876

1 upvote

Following the mechanical principle of the hybrid viewfinder of Fuji X100/s and X-Pro 1, I think.

0 upvotes

low-pass filter can't be located in random place. It needs to be on top of the sensor, literally touching it. If it'd be flippable you'd encounter loads of issues with reliability (eg. one sand dust particle scratching whole sensor).

0 upvotes

With regard to D800 vs D800E this is hilarious !!
Nice evidence that D800E is the ultimate rip-off for would-be pros. Much reminds me of "Emperors New Clothes".

With stupidity the gods themselves struggle in vain.

3 upvotes

I think a test of the (very different pixel sized) D800vsD800E would be needed before you come to that conclusion.

2 upvotes

Almost all tests have revealed improvements with the E version in sharpness in optimal conditions. You buy this several percent mostly unobservable improvement with a large money difference, but when you do need the improvement it would be there.

4 upvotes

D800E image quality is absolutely fabulous. But better use the mirror prerelease even for handheld images (unless using LiveView) if highest resolution is required (and possible due to low iso).
24MP on DX is too much - the test results are no surprise.

0 upvotes

Your comment reminds me of "The Emperor doesn't know what he's talking about" as many test sites like DxOMark, LensRentals, et al have determined that the D800E is the highest resolving DSLR currently in existence.

2 upvotes

Hardly disputable, but what about Moiré?

1 upvote

what bout moire lucas? since you are hell bent on fabric etc, you do know that most medium format cameras do not have AA built in either, yet they are heavily used for indoor fashion photography. by the way there are a million tricks out there in the open on how to remove them - if they ever show up in 0.001% of your shots.

0 upvotes

@Devendra

First morie is not the only artifact that can be caused by lack of an AA filter.

High end indoor fashion shoots where Medium format is used are very controlled. A DSLR is designed for a more candid, faster paced, less controlled environment where you can't always make adjustments or shoot again to get rid of moire.

As far as removing it in post goes, when it is possible it is usually very time consuming if you want to do a good job. Fine if you are a pro retoucher paid by the hour working on high end fashion stuff where sometimes 8 hours is spent on the face alone, but a deal beaker for many armatures and pretty much all pro portrait/wedding photographers who's retouching is generally done as a cost of sale rather than as a for profit service.

0 upvotes

I just don't get this Mirror Up Tripod Thingy (MUTT) when talking about a D800 as it has slightly less pixel density than a D7000 or K-5 (APS-C crop = 15.3MP). The only source of extra blur could be rotational as the distance to the corners of the frame are much larger and could show angular movement based 'pixel blur'. I'm pretty sure my grip isn't as hard as a tripod and that any movement the mirror might cause would be absorbed by the flesh of my hands. A tripod would require the mirror up and shutter delay though. I have absolutely no issues with blur at the pixel level with my D800E's.

A 24MP APS-C camera is probably a different story though. I was hanging for a D7100 but I may just buy a K-5IIs or a K-3 when it arrives as all the studio tests I've seen point to the K-5IIs being superior to the D7100 for noise. (Please don't remind me that Pentax cooks the RAW files as I don't really care)

Comment edited 43 seconds after posting

1 upvote

I've noticed a huge difference in detail on a tripod on even my older k-7. Anything less than 1/200 has the potential for mirror movement and hand shake to rob fine detail. You may have a sharp looking shot, but even a pixel or two of blur is significant. At low shutter speeds handheld is no good unless you don't have the choice. I regularly shoot at 1/10s hand held, but high iso is already robbing detail there so its a wash. You should note that many serious photographers carry tripods regardless of their choice in camera.

0 upvotes

Penta's mirrors are not as good as Canon or Nikon which are nothing compared with Sony. sorry to say.

0 upvotes

Yabakko: pentax's mirrors arrent as good? You have evidence to support that? All slrs must deal with mirror slap. I find the k-5 to be well dampened.

0 upvotes

Disappointed there weren't any images looking for moire.

12 upvotes

Go to the studio comparison tool and compare it with the Pentax K-5 IIs. Look at the beetle and the Roman print. you'll see moire in the k-5 IIs and not in the D7100.

0 upvotes

> Amadou: The other FF cameras you mentioned have a lower pixel density so it's hard to make generalizations.

I find this interesting. If I did the math correctly, the pixel density of the 1.5 crop factor 24 mpx camera would be the equivalent to 54 mpx full frame (24 * 1.5 * 1.5).

That would seem to me to be making extreme demands of the lens, but at the center of the full frame capable lens. That should be the "sweet spot" of the lens, without the edges that tend to be softer.

And really, the high density pixel counts of the 10 to 14 mpx point-n-shoots is higher. DPR used to emphasize the mega-pixels per mm-squared in model summaries. IIRC, an older f.f. like the original Canon 5d would have a factor of 2.0 or so. The point-n-shoots would be > 40.

3 upvotes

Rumors are that this is the new FF sensor. If you expand the sensor to FF size, you get 56MP and that would be the new Nikon D4x.

On the same note, it will be difficult to find lenses that do the sensor justice. Image quality should go up but resolution will require new lenses like the Zeiss 55mm F1.4. These lenses need to perform wide open or close to it due to diffraction kicking in very early. (around f4 for a 56MP sensor, f5.6 for 36MP)

2 upvotes

MF lenses may be good for static landscape or studio shots.

the bottom line of diffraction is when the blue channel fails. but that's only pixel level. feel free to stop down further. as much as you want.

Comment edited 5 minutes after posting

0 upvotes

Sorry Yabokkie, I did not understand what your trying to say. Blue channel fail? So as long as it does not fail, you can stop down without any loss of contrast or sharpness? I think you might be thinking about the blue pill. If it lasts for more than 4 hours, please call your doctor.

Comment edited 20 seconds after posting

0 upvotes

I said that small pixel pitch is justified as long as the blue output is not flat. the signal may have very low contrast but that's the best we can do and should do.

also, you don't have to fully use the capability provided by small pixel pitch. like the max torque of your car. no matter how good it is, you are free to drive as you want becasue you are not the slave of your tools.

Comment edited 6 minutes after posting

2 upvotes

it's highly possible that pixels will go smaller far far beyond diffraction limit, not for resolution but other performances for better image quality. the output could be down sampled to generate a raw file.

0 upvotes

Base on Dpreview's studio comparison, the D7100 is still less sharp than NEX7

Comment edited 15 seconds after posting

1 upvote

well its not the same sensor as the nex7.

1 upvote

Perhaps Nikon should go back using the Sony sensor, as they do for other cameras...

1 upvote

I would like to see a more colorful world, though I do see some low effeciency behind it.

0 upvotes

perhaps Nikon was too conservative with the OLPF ommission sensor?

Others have shown better results without the AA filter.
btw, isn't AA filter or bayer filter an easier term than "OLPF ommission"

0 upvotes

no better results.

images with more noise (not shot noise here but "artificials") look appealing to some eyes, give them a wrong feeling of resolution.

but then photography isn't something "to tell the honest truth" and makers just want to sell what's appealing to the customers who have the right to pay the money for sh!t.

3 upvotes

The Bayer filter is not the same thing - it is still there, since the camera takes color pictures.

4 upvotes

LPF is an integral part of any color filter type sensor, which will not work properly without one.

but the negative effects may not that bad for smaller pixel sensors, and someone may even prefer those negative effects.

0 upvotes

Can someone explain please?

Just looking at the new "OLPF filter omission" section of this review (at page 6),DPReview are saying that in their testing there is no significant difference between the 24Mp with an Optical Low Pass Filter (OLPF) and the same/similar sensor without.

What I would like to understand is this:
(1) Is this because at a certain pixel count, or perhaps density, resolution is so high that removal of the OLPF doesn't really help much?
(2) And if so, what's the pixel threshold?
(3) Or could it be that there are benefits if shooting at certain distances? Perhaps the DPReview studio and outdoor tests weren't shot at distances that are like to show difference between sensors with and without an OLPF?
(4) So this leads me on to the question of cams like the Leica M, Nikon D800E, and all medium format cams, are there really no gains for these cams having omitted the OPLF? In other words, would the image detail look pretty much the same with these cams with an OPLF?

0 upvotes

You might need to re-read the review and look at the pictures. There is significant difference at f3.5 - the photos clearly show a difference. The OLPF /AA filter blurs the higher frequency detail. If the resolution is below the low pass filter frequency, then indeed it won't make the picture 'fuzzier' because it is already fuzzy. The 50/1.4 probably doesn't sharpen up until closed down to f/3.5, and at smaller apertures diffraction kicks in.

The high pixel density of the 24MP APS-C camera is pushing the limits of the system (more than the 36MP full-frame)

Comment edited 2 minutes after posting

1 upvote

We suspect that the answer lies along the lines of #1 and that 24MP on an APS-C gives a pixel density that may simply be beyond available optics.
The other FF cameras you mentioned have a lower pixel density so it's hard to make generalizations.

1 upvote

@f_stops
I don't need to re-read it. I saw the difference at f3.5 .......... but what DPReview pointed out was that it was only at this aperture that significant differences could be seen. Ergo, OVERALL,there is no SIGNIFICANT difference between the two 24Mp APS-C cams, in terms of perceived detail, when one does and the other cam doesn't have an OPLF.

@ Amadou Diallo
Thanks for the reply.

0 upvotes

(5) Is this because of those pictures were displayed on "retina" (high pixel density) display?

0 upvotes

> gives a pixel density that may simply be beyond available optics.

system = 1 / (1 / sensor + 1 / lens),

the sensor resolution will become far higher than lens. that if you have one dollar and are free to invest in either sensor or lens, put it in sensor will get you higher return because it's far easier and cheaper to improve the sensor.

because the pixels are getting smaller, any negative effects measured at pixel level are getting smaller, too. while the LPF is still expensive, anything non-semiconductor is expensive.

yet another incentive behind high resolution sensors.

Comment edited 4 minutes after posting

1 upvote

"gives a pixel density that may simply be beyond available optics." - you just need a lens that got it's maximum at f/5.6 or wider aperture. Same was said when A77 was released - that there's no glass to resolve that - comes out that there's no problem with buying one, only you really need to do your best to push the image quality to the maximum.

0 upvotes

Interesting results. It seems that omission of the OLPF really adds nothing tangible and is more of a "marketing" distinction than a true one.

1 upvote

Two ways to look at this from an IQ standpoint. A) it has no benefit or B) It has no downside. But you'd imagine that with one less part to manufacture and install, it does lower the production cost of the camera.

3 upvotes

cost saving with limited impact on quality.

the resolution is actually lower w/o LPF. you can get similar effect in post to make the images look crisp (standard for Penta and Oly) which will destroy a bit of real details.

it is Nikon's offical policy to "not mind the details". they have been cooking raws to take off some color noise for more than a decade.

Comment edited 3 times, last edit 14 minutes after posting

2 upvotes

More information is better. Always.

0 upvotes

nice ...!!

1 upvote


Review in progress: Pentax K-5 II and K-5 II S Apr 16, 2013	Nikon D800 review updated with D800E side-by-side testing Jun 11, 2012	Just Posted: Nikon D7100 in-depth review Apr 26, 2013	Nikon D7100 preview updated with studio scene and real world samples Mar 25, 2013

	Pentax Efina Jun 3, 2013
	Olympus PEN E-P5 May 10, 2013
	Samsung NX2000 May 1, 2013
	Sony Cyber-shot DSC-HX50V Apr 24, 2013
	Panasonic Lumix DMC-G6 Apr 24, 2013
	Panasonic Lumix DMC-LF1 Apr 24, 2013
	Ricoh GR Apr 17, 2013
	Samsung NX1100 Apr 11, 2013
	Panasonic Lumix DMC-GF6 Apr 9, 2013
	Fujifilm FinePix S8400W Mar 22, 2013
	Fujifilm FinePix XP200 Mar 22, 2013

Considering switch from Nikon, to replace D700 ...	150
Sobering Sales Numbers for Mirrorless	150
New Pentax K-50 leaked	74
Hmmm AF fine tuning : Bullsh#t	65
* Wed C&C (No Theme) Thread, Ed. 267, 13 06 05 *	59
Praise the Macro Lenses. Post your best macros ...	49
Poor digital camera sales continue....	47
How to turn my Canon PowerShot S3 IS into a ...	42
Still no signs of an Olympus f2.8 zoom?	39
Selecting Camera	36
Help me create the best 6D system on a $3500 ...	35
USPS... The worst mail servie in the world	34
electronic first curtain shutter - on or off?	34
K5(II) SD Card Door Coming Open	33
Mirrorless & DSLR Jan-April shipment & ...	32
Sigma 60/2.8... sigma be damned (for its ...	32
Done Deal - Switched to Fuji XE-1 from ...	30
Focal reducers, why complexity for the sake of ...	30
Replacing Olympus OM-D with Fuji XE-1 or Xpro-1	29
micro 4/3s , Owls IF	29

Your email address:
Your name:

Nikon D7100 preview updated with a look at effect of OLPF omission

Nikon D7100

Comments