Nexus D2x Housing

Diffraction is a softening of the image. The general limit for diffraction is when its effect becomes greater than the circle of confusion. The trouble is that it's hard to quantify but f/32 is generally considered "there". For DX sensors it's f/22. One full stop less. Most consider f/16 or f/22 the limit for full frame, so that's f/11 or f/16 for DX. Try shooting 2:1 macro underwater at f/11 and see what you get.

If I were to shoot the same nudi with a full frame sensor I'd use a half-stop smaller f-stop than you did but I'd get less relative diffraction, a sharper image plus more resolution. The effect is pronounced at high magnifications so the difference is slight with nudi-sized subjects. The reason I care is that I like shooting pygmies and I want large DOF.

Ok, here's a D70 pygmy shot. Strange enough @ f22. Do you have one to show me that you can point out the differences?? I really am not trying to make this a contest as I believe you. I'm just wanting to see it explained. Feel free to crop it to 100% or what ever shows it best.

I'll make myself popular by jumping in the middle and disagreeing with both of you.

Diffraction is blurring of the image due to making the light pass thru a tiny hole and is proportional to (the wavelength of light) * (distance between the image and the hole) / (width of the hole). It's a gradual blur and there's no clear onset. The way this and the lens parameters work out is that the distance range of this blurring as seen on the image sensor depends only on the aperture. If you plug in the numbers for red light (the longest visible wavelength) at around F/16 the blurring will have a range of ~ 6 microns which is the size of the pixels in the 6 MP 1.6x cropped cameras. Smaller apertures will decrease resolution. Actually, it's noticible on shots of test charts before that. That's why those are usually done at around F/8.

The 1DsII with 16MP on the full frame sensor has about THE SAME pixel size as the 1.6x 6MP sensor, and the resolution will become diffraction limited at about THE SAME aperture. However, it has 1.6x more pixels in each direction so will be able to capture more details. It's more complicated than that because if zoomed in to a higher magnification to get the same coverage, it will loose some depth of field, and if the aperture is reduced to get the same DOF, it will loose some details, but is overall still better.

It's a difference that's hard to demostrate on different pics shot under different conditions with different lenses. You'll need to shoot a test chart with the same lens to really show anything conclusive.

The D2x with the smaller pixel size will be limited by diffraction at a larger aperture. The extra pixels will not help much at F/16 and smaller.

Herb, I understand on paper these things make sense to you. These hard numbers have been posted but my eyes start to cross when I read about things such as circle of confusion. I'd very much just like to see two shots of similar subjects, especially Pygmies as conditions for shooting them are pretty universal in a very general sense, and have pointed out this diffraction difference. I think it's a given that these shots were done with the best possible result in mind. So Lens choice shouldn't be that big of a deal.

The FF cameras are multiples of the D70 in price. James bought an expensive Kodak only to have it fall on it's face. I'm sure he researched it, circle of confusion included, but when he looked at the final results from U/W, sold the camera. I'm just saying if there is a noticeable difference, I'd like to see it.

I too someday will be moving up and I'd like to understand this better. That's all I'm saying.

Rand

I'm just saying if there is a noticeable difference, I'd like to see it.

I too someday will be moving up and I'd like to understand this better. That's all I'm saying.

That's the first thing I have understood in this whole thread. I have gone the full circle of confusion.

I won't post a shot that contributes to a totally unscientific analysis. You have to shoot a single subject under controlled conditions with an identical camera and lens but with different apertures to judge this. You can find macro lens tests on the internet that show this or run the tests yourself. Diffraction effects are easily demonstrated and are the reason that lenses are typically claimed to be sharpest around f/8. There is a noticable difference.

Herb, acceptable levels of diffraction must be relative to the circle of confusion, not the pixel pitch. You don't say that higher resolution cameras have less depth of field after all and acceptable diffraction is just as dependent on viewing angle as critical focus is. In your definition you are simply stating what the limit is for maximum image quality and that unnecessarily penalizes higher resolution sensors. No one shooting film ever looked at it that way. It's not hard to find analysis on the web that matches your viewpoint but I've found it impossible to conclude anything from it. The 1DsII will not be diffraction limited at the same aperture as the 6MP Canon cropped sensor unless you deliberately choose different print sizes to make it so. Here is a link that discusses diffraction, circle of confusion, depth of field, etc.

Also, it is not physical aperture but effective aperture that determines depth of field. As a result, the full frame camera gets some DOF benefit at higher magnification due to bellows effect.

Rand, James can speak fo himself but I'm confident that he didn't abandon the Kodak because of any issues relating to this. The D70 is no match for the Kodak from an image quality perspective. Having owned the Kodak myself and used it underwater I can assure you that image quality was not a problem. The biggest problem shooting macro with the Kodak is being able to see your subject.

Whassup, guys? My comment was a little tongue-in-cheek, and I was a bit frustrated that I have to be so "straight" around here, lest I piss someone off.

But what I wrote is accurate for working pros. Nearly every underwater pro I've spoken to has switched to Canon or is debating switching to Canon. They are frustrated with Nikon being years late on having a professional body that produces good images. It's true that the D70 is an excellent camera, but a manufacturer's lowest-end digital SLR should not outperform their pro body! It's just ludicrous.

Anyway, I'm not defending myself or anything. A little levity always makes things better.

But, as DOF goes in macro, I'm seeing less as more.

 

Macro shots with everything in focus seem less interesting to me than those where DOF is controlled. I'm making a conscious effort to shoot wider. Trying to blur backgrounds, especially where the color contrast is not good but subject matter is.

 

Emphasizing certain elements while throwing the rest out. I'm not sure I would shoot f22 or higher even if the results were snap sharp.

 

I'd like to hear other viewpoints on this as I think it's a worthwhile discussion.

 

Rand

Would shooting the higher stop and then doing photoshop Blur accomplish the same effect for you? (I'm asking not stating this is how Rand should do it)

I prefer to try and get it right as much as possible out of the camera.

Hi Todd,

Might be a bit tough, as you'd just want to blur the background but keep the foreground sharp. Not sure how I'd do it.

FWIW, if anyone's getting this housing, there is a Nexus multiport and dome port for sale here in Houston for a VERY good price.

Cheers

James

James, congrats on being selected as the Inon artist of the month.

Woohoo! Thx Todd, I'll have to go check it out.

Cheers

James

I just got a pre-order price of 3460 USD from Japan through my local dealer. Thats nearly double the 2000 USD I paid for the D70 housing. Add that to the 180,000 Baht (4500 USd) I am being quoted for the D2x in Thailand I think my upgrade plans will have to be shelved for a long time to come.

But, as DOF goes in macro, I'm seeing less as more.

 

Macro shots with everything in focus seem less interesting to me than those where DOF is controlled. I'm making a conscious effort to shoot wider. Trying to blur backgrounds, especially where the color contrast is not good but subject matter is.

Yes, Rand, I agree. I specifically use a mask to sharpen so I am only sharpening the areas I want to draw the eye to.

But......

Would shooting the higher stop and then doing photoshop Blur accomplish the same effect for you? (I'm asking not stating this is how Rand should do it)

 

I prefer to try and get it right as much as possible out of the camera.

I agree with Todd also. Not so much the term "right". It is easier to reduce the DOF of areas you don't want sharp, than trying to sharpen areas that are not in focus. So, maximize DOF on the shot, smooth the areas you don't want in focus. Plus, the areas you soften don't have to be in the same plane.

My comment was a little tongue-in-cheek, and I was a bit frustrated that I have to be so "straight" around here, lest I piss someone off. Smile

Eric, Nikon zealots and apologists are getting a little sensitive these days. I suggest you try this for even more "fun"; Find a Leica forum and state that the 1DsMkII with a zoom lens has more resolving power than any Leica/lens combo. Then duck!

Got my hands on a late prototype D2X today. Very impressive and much lighter than I was expecting (but then I guess all cameras are light to us when they are not in housings).

The camera felt great in my hands, which I realise has no relevance for UW. As far as housing the camera goes - the body is identical to the D2H - which is very good news. This camera will slot into any D2H housing.

There were no on screen sample images. And no chance to take pictures on the day - which is the only real way to see what comes straight out of the camera. I wasn't allowed to stick a flash card in - although the card slot was not glued shut!

Despite this the Nikon Rep and Technical guy were bullish on the image quality. They were talking about out doing Canon 1DS Mk2 at low ISO. And being better than everyone expects at high ISO. But then they are Nikon Reps - and they are not going to say it is crap.

I had a good chat with the Nikon technical guy and he said one of the big advances on this camera is the colour handling and white balance control. He said that this was the big step - although he thought it wouldn't get the press as it is harder to quantify than resolution and noise improvements.

Regarding supply: Nikon UK are coordinating the orders - so cameras won't get sent out to the same photographer who has orders at several dealers! As of today, Japan are still sticking to the 25th Feb date to get cameras out. And intend to ship direct to dealers - with the first cameras designated to individuals who have pre-ordered them.

Alex

 

 

Despite this the Nikon Rep and Technical guy were bullish on the image quality. They were talking about out doing Canon 1DS Mk2 at low ISO.  .  

I'd love to see a Side X Side comparison.... :twisted:

Nikon zealots and apologists are getting a little sensitive these days. I suggest you try this for even more "fun"; Find a Leica forum and state that the 1DsMkII with a zoom lens has more resolving power than any Leica/lens combo. Then duck!

All brands have their zealots and apologists and Nikon's are no different. Try suggesting that an L series lens isn't better than its competition or that macs aren't more intuitive. People believe that buying into a brand makes them part of the team so they become cheerleaders. I don't think trolling helps though.

One thing that can be said for Eric's comment is that it's justifiable criticism. Whether it's the right place to say it or not I don't know but there's no denying that many agree with it. People would be crazy to think that the D2x can be a failure and Nikon not suffer seriously as a result. This needs to be a very good camera.