Pentax 645D vs. Nikon D3X (print quality in A2)

[ErikKaffehr]

Hi,

I made a quick comparison of the Pentax 645D vs. the Nikon D3X, based on images from "Imaging Resource". The question I asked was essentially if we can see a significant difference in A2-size prints. The method used was the following:

Downloaded two similar test shots from Imaging Review, both in RAW

Made three different "developments". To begin with I white balanced all the pictures on the color checker card second gray field from left. After Develop->Setting->MatchTotalExposures was done. Identical crops were than established.

Both images were sharpened with Lightrooms Narrow Edge Sharpen. The Nikon image was additionally sharpened using the deconvolution option ("Detail slider" fully to the right" and a bit extreme setting).

The images were exported for printing on glossy paper at 480 PPI in A2 format. Normal sharpening for output was used.

The images were than checked out at actual pixel in Photoshop, see the enclosed screen dumps. I looked at the central part and at the scale at the right. Images at top are Nikon and bottom Pentax 645. The deconvoluted image is on the right. This is extreme pixel peeping! When looking at the images at this absurd magnification the Nikon image has a lot of jaggies and artifacts on the scale.

I also made A4 prints from approximately half of the image. This is quite good approximation of an A2 print. The differences are actually quite small in print. I guess they would be hard to tell apart. The scales on the Pentax image are darker.

To sum it up, I have little doubt that the Pentax images are better but I'm also quite confident that the differences are very hard to see in an A2 size print.

Addition: 2010-12-14

Now I also made 1/4 crops, corresponding to A0 prints. In this size the Pentax image is smoother but the deconvoluted Nikon image still holds up. Disturbing moiré is now visible on the Samuel Smith label.

Best regards
Erik

[Ray]

Hi Erik,
That seems like a very thorough analysis and comparison. Thanks.

Do I take it you will not be buying a Pentax 645D?  

My own view is, big cameras with high pixel counts tend to need large printers.

My printer is the 24" wide Epson 7600. I consider a 12mp image from a full frame 35mm sensor sufficient for a 23"x35" print, lacking only in the finest detail that could be visible with one's reading glasses from a distance of 300-400mm.

A 24mp image for that size would certainly be ample, even for close viewing, in my opinion.

[sojournerphoto]

Erik

I only want to comment on your good taste in drinking one of my local beers. Sam', as it is affectionately known around here, has a long and honourable tradition.

Mike

[ErikKaffehr]

Hi,

Sorry, the image is coming from Imaging Resource, so those are the people having an excellent taste!

Best regards
Erik

Erik

I only want to comment on your good taste in drinking one of my local beers. Sam', as it is affectionately known around here, has a long and honourable tradition.

Mike

[ErikKaffehr]

Hi,

The Pentax is certainly a camera I'd loved to test. In my view it's competitive with the Nikon D3X on price, especially if you have some decent lenses collecting dust in a wardrobe. I don't happen to have a ton of Pentax 645 glass, what I happen to have are 5-6 Pentax 67 lenses. They are usable on the Pentax 645, with an adapter.

Nevertheless, you are probably right. I'm not really a Pentax 645D purchaser.

Best regards
Erik

Hi Erik,
That seems like a very thorough analysis and comparison. Thanks.

Do I take it you will not be buying a Pentax 645D?  

My own view is, big cameras with high pixel counts tend to need large printers.

My printer is the 24" wide Epson 7600. I consider a 12mp image from a full frame 35mm sensor sufficient for a 23"x35" print, lacking only in the finest detail that could be visible with one's reading glasses from a distance of 300-400mm.

A 24mp image for that size would certainly be ample, even for close viewing, in my opinion.

[JR]

If you need to print really big the 645D is probably a good choice. But, as you have demonstrated, one does not need the most expensive equipment for A2/A3 prints. I guess professionals/working photographers already know that but I doubt people in general know that. Good analysis.

[uaiomex]

MR "suggested" a while ago that at A3 size even a good P&S competed with a P1 digital back (if I recall well). In the darkroom days my "standard" print size was 16X20". With digital, my standard is now 24X30-36". I'm sure my prints would benefit with the use of bigger than FF sensors. But the big question remains: Up to what extent? Is it worth it?. I wish I could have access to live comparisons among prints. Unfortunately I don't so I have to rely on posters comments, general perception and the fact that FF sensors are good enough for me. So, I'm happy with FF cameras. Some sort of "if I would know better"" case I think. 
Eduardo

[ErikKaffehr]

Hi,

I have discussed my A2 prints with a friend and colleague and he could clearly tell P645D apart from Nikon D3X regardless of sharpening. I think he was checking fine print on beer bottle.

Best regards
Erik

MR "suggested" a while ago that at A3 size even a good P&S competed with a P1 digital back (if I recall well). In the darkroom days my "standard" print size was 16X20". With digital, my standard is now 24X30-36". I'm sure my prints would benefit with the use of bigger than FF sensors. But the big question remains: Up to what extent? Is it worth it?. I wish I could have access to live comparisons among prints. Unfortunately I don't so I have to rely on posters comments, general perception and the fact that FF sensors are good enough for me. So, I'm happy with FF cameras. Some sort of "if I would know better"" case I think. 
Eduardo

[BernardLanguillier]

Hi,

I have discussed my A2 prints with a friend and colleague and he could clearly tell P645D apart from Nikon D3X regardless of sharpening. I think he was checking fine print on beer bottle.

Considering that current printers are said to deliver their best possible print quality at 480 dpi, it shouldn't come as a surprise.

Now I would like to be shown images where the gap between the 2 is large enough to make or break a print.

I don't doubt that there are subjects for which more resolution helps, but for that kind of subjects what we need is typically 400 megapixels and not 40. At the risk of awaking a sleeping daemon, stitching remains the only credible option in such cases.

Cheers,
Bernard

[ErikKaffehr]

Hi,

I plan to write an article with my findings and publish scans of the printed images.

I'd say that the theoretical advantage of doubling MegaPixels would correspond to an A-size in enlargement. That is a 40 MP camera would be able to make a perfect A1 print if an 24MP camera would be able to make a perfect A2 print. My friend observed directly that there were tonal differences between my samples. He says that you can always mistake contrast for sharpness and so on. He looked at fine text on the beer bottle and could identify the P645D easily.

Best regards
Erik

Considering that current printers are said to deliver their best possible print quality at 480 dpi, it shouldn't come as a surprise.

Now I would like to be shown images where the gap between the 2 is large enough to make or break a print.

I don't doubt that there are subjects for which more resolution helps, but for that kind of subjects what we need is typically 400 megapixels and not 40. At the risk of awaking a sleeping daemon, stitching remains the only credible option in such cases.

Cheers,
Bernard

[NikoJorj]

I don't doubt that there are subjects for which more resolution helps, but for that kind of subjects what we need is typically 400 megapixels and not 40.

Indeed!
http://theonlinephotographer.typepad.com/the_online_photographer/2009/02/why-80-megapixels-just-wont-be-enough.html

[Rob C]

Hi,

I plan to write an article with my findings and publish scans of the printed images.

I'd say that the theoretical advantage of doubling MegaPixels would correspond to an A-size in enlargement. That is a 40 MP camera would be able to make a perfect A1 print if an 24MP camera would be able to make a perfect A2 print. My friend observed directly that there were tonal differences between my samples. He says that you can always mistake contrast for sharpness and so on. He looked at fine text on the beer bottle and could identify the P645D easily.

Best regards
Erik

Yep, Eric, and the more empty bottles the more certain he became! It's always like that: you start cool and sceptical and end up effusive and generous. Or looking for fights at a time when others seek sex!

That old beer test was disproven years ago.

Rob C