• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Handheld, long focal, comparison between 5D Mk2 and Pentax 645D

nicolas claris

OPF Co-founder/Administrator
As Asher as requested* here a handheld comparison, shooting flags on shore with long length with full frame DSLR and the MF Pentax 645 D, here we are, just hot shots…
Distance from camera to flags = 90 mètres / 296 feet.
All raw files edited in Lightroom 4.1 with nearly same settings. I tried to get them as close as possible to my taste, forgetting wich camera did shoot ; )

* Asher's requests are almost orders! ; )

A comparison between:
Canon EOS 5D Mark II - EF70-200mm f/2.8L IS II USM @ 200 mm - 1/1000s - ƒ8 - ISO 200
and
PENTAX 645D - smc PENTAX-FA*645 300mm F4 ED [IF] (24x36 = 236 mm) - 1/1000s - ƒ8 - ISO 200

_MG_5682_LR4_1000.jpg


_NCL8767_LR4_1000.jpg


CROPS at 100% (Canon 1st, Pentax below each crop)

_MG_5682_LR4_1000_a.jpg


_NCL8767_LR4_1000_a.jpg

 

nicolas claris

OPF Co-founder/Administrator
CROPS at 100% (Canon 1st, Pentax below each crop)

_MG_5682_LR4_1000_d.jpg


_NCL8767_LR4_1000_d.jpg


_MG_5682_LR4_1000_e.jpg


_NCL8767_LR4_1000_e.jpg

Of course the Pentax, at same aperture has less DOF, but look at that bokey! (and that's an even better advantage if you think about it)
 

Asher Kelman

OPF Owner/Editor-in-Chief
CROPS at 100% (Canon 1st, Pentax below each crop)

_MG_5682_LR4_1000_b.jpg


_NCL8767_LR4_1000_b.jpg


_MG_5682_LR4_1000_c.jpg


_NCL8767_LR4_1000_c.jpg



My dear friend Nicolas,

I'm so happy that you took up the challenge and help us with a real life comparison of the Pentax 645D and the Canon 5DII. The latter is a particularly good frame of reference, (for "yours truly", as I use one all the time. I notice the inside of the lamp housing in the Pentax is so well drawn in the Pentax and one not only sees the surface so much more clearly but also it conveys a sense of the material of the surface so it looks much more realistic and demanding of attention. In comparison, the rendering by the 5DII, although small is really just a swell done sketch and lacks the extra life.

That's my impression so far. Of course, one could use a Gretag MacBeth color card and correct each one to that for color, as they each deal with color differently. Still, I doubt that that correction would make any real difference! Perhaps the 5D III or the 1DX could stand a better chance!

Asher
 

nicolas claris

OPF Co-founder/Administrator
Thanks Asher
I don't have a 5D III nor the 1DX (and do not foresee to buy any of them…)

Yes the colors are different, so is the auto WB from the Canon and the Pentax.
I have noticed that the Pentax has, in daylight, a cooler WB than the Canon.
But you see this only when doing comparison or while shooting with both camera the same event.

Another point, I have forgotten to say that, for this comparison, the focus has been set on the orange flag.
 

Asher Kelman

OPF Owner/Editor-in-Chief
As Asher as requested* here a handheld comparison, shooting flags on shore with long length with full frame DSLR and the MF Pentax 645 D, here we are, just hot shots…
Distance from camera to flags = 90 mètres / 296 feet.
All raw files edited in Lightroom 4.1 with nearly same settings. I tried to get them as close as possible to my taste, forgetting wich camera did shoot ; )

* Asher's requests are almost orders! ; )

A comparison between:
Canon EOS 5D Mark II - EF70-200mm f/2.8L IS II USM @ 200 mm - 1/1000s - ƒ8 - ISO 200
and
PENTAX 645D - smc PENTAX-FA*645 300mm F4 ED [IF] (24x36 = 236 mm) - 1/1000s - ƒ8 - ISO 200

_MG_5682_LR4_1000.jpg


_NCL8767_LR4_1000.jpg



Nicolas,

It was so easy for me to challenge you but far more difficult to work on the result as they are all great! I'll tackle this first set. Hopefully other comments will follow!! :)

The most obvious differences relate to the larger format effect of the Pentax 645D at the relatively open aperture of f 8.0. Here, we have to choose what we value. For landscape images, we might really want to have the trees in the background very sharp with the form and colors perfectly drawn. So immediately, the Canon at 8.0 would do a far better job in the 35mm format. In fact, at any smaller aperture for the Canon lens, would lead to image degradation because of the tiny 5DII sensel size would show and define diffraction artifacts from the edges of the relatively tiny lens aperture.

OTOH, for isolating important subject matter, throwing the background out of focus with a pleasant blur is an advantage. So with the Pentax 645D, even at f8.0, the foreground flag is brought out so much more clearly against this softer background. That's an advantage for isolating a subject, especially in portrait. If one wants the same clarity of the background to appear, then one has to stop down considerably. I'd imagine that f22 would be about right. But then, one would have to push the ISO to 800 for the Pentax 645D which might mean noise reduction. However, likely as not, that would be accomplished with no degradation of the image.

For me, with a consideration of shooting portraits, the soft background of the Pentax 645D images is very pleasant. If one looks at the gaps in the leaves of the trees, there's a soft halo artifact. It's not jarring and doesn't degrade the photograph in any way, from my point of view. Still, one has to recognize that this edge effect is present. Could it be a jpg artifact? That I doubt too. Now one could describe that as painterly and desirable. Or perhaps it's a jog artifact, but that I doubt. It is almost as if the leaves are acting as a secondary aperture and we are getting artifacts.

Overall, at the chosen aperture of f8, the entire image, is obviously more satisfactory as a picture postcard from the 5D mark II, as everything is rendered well. However, that's not the subject of the test! Here it's just the flag and that's where the Pentax does an especially fine job, to be addressed in the following post.

Asher
 

Asher Kelman

OPF Owner/Editor-in-Chief
The appearance of the flags! This is where both cameras were focused!



Obviously both images are stunningly sharp and the colors are vivid and there's great contrast and detail. There are no obvious disappointments! Of course, the Pentax 645D is larger and naturally more impressive. But there's something else at work. The pictures have a different aura to them, don't you agree.

_MG_5682_LR4_1000_a.jpg


_NCL8767_LR4_1000_a.jpg



Nicolas Claris: CROPS at 100% (Canon 1st, Pentax below each crop)


Bart knows more than I do about the math and science of human perception of images, but here's my best guess. What the MF Pentax 300 MM FA lens is doing here is to provide a gradient of planes of focus, where the parts of the flag in front of and behind the sharpest plane are less sharp by a subtle but really significant amount, enough to signal to the brain difference in depth. So this might be what we can appreciate in sensing something different in the MF imaging of the flag. It seems to be more dimensional. I am surpassed to see this effect, and if it does indeed come from the subtle difference in focus, then it's still very different from the major jump in drawing the detailed flag as a whole and the soft background.


So does anyone else agree that the flag seems more real and has more presence in the Pentax picture? Or is this just my idiosyncratic impression? Anyway, perhaps we might also see this effect with the Canon 5DII using f2.8. If so, then we are indeed looking at an effect of the planes of focus getting softer either side of the shapes plane in the flag.

Asher
 

Michael Nagel

Active member
Asher,

The comparison is interesting, but the relative size of the sensors is IMHO a little bit unfair as you must account for difference in DoF and the (presumed) advantage in tonal nuances for the MF camera in this case.

The final outcome for Nicolas is a picture - according to his standards.
The question for him is AFAIK the usability of the camera and the available lenses.

The 645D has - for me - an edge compared to the Canon.

Best regards,
Michael
 

Jerome Marot

Well-known member
So does anyone else agree that the flag seems more real and has more presence in the Pentax picture?

Certainly, yes. But the resolution on the Pentax is higher, so we see more details. Maybe a better comparison would be between the 645D and the Nikon D800.
 

fahim mohammed

Well-known member
First, a big thank you to Nicolas for bringing this comparison for us.

The size ( physical dimensions and pixel ) of the Pentax helps it to show greater detail; in the flag and the lamp.
The texture details rendered in the lamp, as pointed out by Asher, is very apparent.

The shallower dof of the Pentax and the pleasing quality of the oof areas is an area I find most interesting.
The aesthetic possibilities available to Nicolas have just been increased considerably.

Asher has mentioned the ' planes of focus '. This is the feeling of depth that a viewer experiences on looking at prints from a MF as compared to ones from a 35mm full frame. Some call it a 3 dimensionality.
The Pentax is more geared to giving that feeling. Look at the flag pic from the Pentax.

Regards.
 

nicolas claris

OPF Co-founder/Administrator
Bonjour Asher, Michael and Fahim
Thank you for your inputs. I'll reply more in depth tonight while at home… ; )

For now, let me show you the first set of image post here Canon's and Pentax's) these are the frames showing the compete -not cropped- image.

I had previously (a month or 2 ago) some private chats with Bart about down sampling has it becomes difficult to get good results with such a down ratio… who could guess that 5 years ago, eh? : ) but I haven't find -yet- a good, easy and repeatable way of doing it.
So the ones shown on top of this thread where done from CS6 after some extra sharpening work (on both Canon's and Pentaxt's).
It shows artifacts and too much contrasts imo.

A liitle better are hereunder, direct exports from LR4:


_MG_5682_LR4_1000-2.jpg


_NCL8767_LR4_1000.jpg
 

Joachim Bolte

New member
I think you shouldn't have done ANY postprocessing, just to be sure that all you see it the output of the lens. That would be a more fair comparisson. The images now clearly show signs of oversharpening and noise reduction combined.

Could you repost them using neutral settings in your developing tool?
 

nicolas claris

OPF Co-founder/Administrator
I think you shouldn't have done ANY postprocessing, just to be sure that all you see it the output of the lens. That would be a more fair comparisson. The images now clearly show signs of oversharpening and noise reduction combined.

Could you repost them using neutral settings in your developing tool?
Joachim
First, thank you for stopping by.

I am not comparing lens (focal is not excactly the same (200/236) and the Canon's 200 is stabilized) but sensors.
IMHO:
Therefore there is no objective truth and no neutral settings as the 645D has no demoisacing filter,therefore it NEEDS to have some sharpening done, therefore again, to be fair, the Canon file has too be sharpened too.

I am trying to establish real life comparison, and as said in the OP I have tried to work as objectively as possible on each file. You may find that they are over sharpened, and I can agree with you. Or not… but imho again, this does not really change the feeling we get from these different cameras.

You should
 
I had previously (a month or 2 ago) some private chats with Bart about down sampling has it becomes difficult to get good results with such a down ratio… who could guess that 5 years ago, eh? : ) but I haven't find -yet- a good, easy and repeatable way of doing it.
So the ones shown on top of this thread where done from CS6 after some extra sharpening work (on both Canon's and Pentaxt's).
It shows artifacts and too much contrasts imo.

A liitle better are hereunder, direct exports from LR4

Hi Nicolas,

That's correct, Photoshop's default choice for downsampling, BiCubic Sharper (or Automatic), produces artifacts. Lightroom, since version 3, produces a much better downsampled image quality. After downsampling, a special sharpening can add a bit more punch to the image, but one should avoid creating new aliasing artifacts (stairstepping and halo).

Cheers,
Bart
 

Joachim Bolte

New member
...as the 645D has no demoisacing filter,therefore it NEEDS to have some sharpening done...

I suppose you mean an AA filter?

But then again, what purpose does it serve to compare two completely different sensors in two different configurations, with different glass in front of them? What will (CAN?) be your conclusion?

Also you are referring to some nice Bokeh effect, but I really can't judge that, because it is blown to pieces by the postprocessing.
 

Asher Kelman

OPF Owner/Editor-in-Chief
I suppose you mean an AA filter?

But then again, what purpose does it serve to compare two completely different sensors in two different configurations, with different glass in front of them? What will (CAN?) be your conclusion?

Also you are referring to some nice Bokeh effect, but I really can't judge that, because it is blown to pieces by the postprocessing.


Joachim,

In the end, it is what it is, window into the move up from 35mm DSLR to MF at a relatively modest price where AF lenses are plenty and cheap on the used market and where the finest lenses are also being made purpose built for this MF camera. So for me, for example, interested in giant portraits, the "presence" given by the Pentax 645D at the distance from the camera necessitated by the format, can be sensed already. This interests me a lot. It also tells me why such pictures in LF and ultra large format works so well.

Asher
 
That's correct, Photoshop's default choice for downsampling, BiCubic Sharper (or Automatic), produces artifacts. Lightroom, since version 3, produces a much better downsampled image quality. After downsampling, a special sharpening can add a bit more punch to the image, but one should avoid creating new aliasing artifacts (stairstepping and halo).

Continued, with some more details (pun intended).

I've just done a quick test and have come to the following conclusion, and recipe:

For an image that was exported with downsampling in Lightroom (I used LR 4.1) without additional sharpening in the Export process, and assuming correct sharpening at the full size, Lightroom will produce only mild aliasing artifacts (certainly when compared to Photoshop and others):

Rings_LR41_20pct_NoShrp.png

Depending on the subject, it may help to sharpen the downsampled result, but it should not create halos, and not introduce clipping. For halo free sharpening of LR downsampled images one can use the following filter in Photoshop (make sure your image is in 16-b/ch mode!):

Sharpen_LRdownsampling.png

To suppress clipping, it is best used in combination with a Luminosity Blend-if layer style for the sharpening layer with settings like this:

Non-clipped-sharpening.png

So, one downsamples the image on export with Lightroom, one creates a duplicate layer with a Luminosity Blend-if layer style in Photoshop, and one applies the above filter for sharpening to that duplicate layer which is positioned above the original in Photoshop. That still leaves the possibility to adjust the opacity of the sharpening layer, and or mask-out areas which create issues. It is very simple to combine all the Photoshop steps in a single action, which will remember all the blend and filter settings and executes them with one mouse click.

It's too bad we can't do it properly in one application, but such is today's reality.

Cheers,
Bart

P.S. I've inserted a note in the above recipe, that the Custom filter should be used in a 16-bit/channel image mode. I have noticed that at 8-b/ch, presumably due to rounding errors, the image becomes darker than it mathematically should.
 
Last edited:

nicolas claris

OPF Co-founder/Administrator
I suppose you mean an AA filter?

But then again, what purpose does it serve to compare two completely different sensors in two different configurations, with different glass in front of them? What will (CAN?) be your conclusion?

Also you are referring to some nice Bokeh effect, but I really can't judge that, because it is blown to pieces by the postprocessing.

Joachim
I draw and keep most of my conclusions for myself (one of the conclusions is that I switched from 1Ds3 to the Pentax 645, keeping the 5D2 as a back-up only).
Everyone can make his own advise, taking into account or not what I show here. They are many other posts on the net to help them too.
My purpose here, is to show real life shooting (normally I shoot to get my beef-steack at the end of the month) with some gear that many pros have. Canon or Pentax are the one I have. Many are wondering, as I did some months ago, if they should or not switch from one system to another one. Maybe and maybe only, I can help to make up their mind…

There are no lens that you can mount on a Canon AND on a Pentax. Sorry, I don't have more lens to compare with the exact same focal.

Yes I meant AA filter. Sorry for the confusion.

Now, after your so kindly made request, and because I don't want you to feel too much frustrated :

_MG_5682_LR4_1000_a2.jpg


_NCL8767_LR4_1000_a2.jpg


_MG_5682_LR4_1000_c2.jpg


_NCL8767_LR4_1000_c2.jpg
 

nicolas claris

OPF Co-founder/Administrator
Continued, with some more details (pun intended).

I've just done a quick test and have come to the following conclusion, and recipe:
………

It's too bad we can't do it properly in one application, but such is today's reality.

Cheers,
Bart

Thank you Bart, a very "positive" and instructive post. As usual ; )
 
Now, after your so kindly made request, and because I don't want you to feel too much frustrated :
[...]

Hi Nicolas,

Thanks for posting those, that helps for the analysis. Based on that, it looks like your Capture sharpening radius setting in Lightroom should be something close to 0.7 for these apertures . A more exact judgement, and/or tweaking of the amount and detail controls (and for different apertures) takes a bit more testing and is best based on a slanted edge test chart.

Cheers,
Bart
 

Asher Kelman

OPF Owner/Editor-in-Chief
A clear Demonstration of the Value of Bart-Nicolas Processing of Pentax 645D Files.


_NCL8767_LR4_1000_c2.jpg




_NCL8767_LR4_1000_c.jpg


The beauty of the processing speaks for itself. It's just like switching on a light and getting new spectacles, LOL!

Asher :)
 

Ivan Garcia

New member
This is an interesting change for you Nicolas.
What made you sway in Pentax direction as opposed to the obvious 1Dx upgrade path?
 

nicolas claris

OPF Co-founder/Administrator
Hi Ivan
Glad to hear from you!
When I made my decision to switch in December2011, no one could even know if and when the 1Dx will be available… plus 18.1 Megapixel is not enough for my huge prints.
Also, I have been somewhat knocked down by the MF quality when I had the HY6 from Sinar.
Smooth gradient etc… The Pentax 645D is far more easy to handle than the Hy6 (which anyway I cannot afford) and the files are supported by LR4 (not by C1, except their DNG version but many subtle settings in C1 aren't usable for thoses DNGs…)
 

nicolas claris

OPF Co-founder/Administrator


The beauty of the processing speaks for itself. It's just like switching on a light and getting new spectacles, LOL!

Asher :)


Asher
let's be honest with Bart!
The processed version here uses some action I made (with some pretty good help from Bart) about 2 years ago for the 1Ds3 files.
I do recognize that I am still too much "impregnated" by the habits for sharpening Canon files.
Her we can see some artifacts and they come from that habit, plus the fact that I'm a new comer to Lightroom.

As soon as I'll have some more time free, I'll dig seriously with Bart hints and help to get even better results, and I have no doubt that we can... and will let OPF know : )​
 

Asher Kelman

OPF Owner/Editor-in-Chief
Asher
let's be honest with Bart!
The processed version here uses some action I made (with some pretty good help from Bart) about 2 years ago for the 1Ds3 files.
I do recognize that I am still too much "impregnated" by the habits for sharpening Canon files.
Her we can see some artifacts and they come from that habit, plus the fact that I'm a new comer to Lightroom.

As soon as I'll have some more time free, I'll dig seriously with Bart hints and help to get even better results, and I have no doubt that we can... and will let OPF know : )

Hmmm, Nicolas!

What does Pentax recommend? They should have worked this out already! Is it silkypix?

Asher
 

nicolas claris

OPF Co-founder/Administrator
Hmmm, Nicolas!

What does Pentax recommend? They should have worked this out already! Is it silkypix?

Asher

They give a special copy of Silkypix but it is really too slow.
LR4 will do the job (BTW a new CR release 4.2 is downloadable there). It is just a question of mastering:
a - the software
b - the huge PEF file

: )
 
Asher
let's be honest with Bart!

Yes, please. ;) Although part of Asher's reaction was based on only that one parameter (Radius 0.7 for f/8) in the sharpening workflow that I approximated from your unsharpened sample, and it was already much closer to where we will ultimately get.

The processed version here uses some action I made (with some pretty good help from Bart) about 2 years ago for the 1Ds3 files.
I do recognize that I am still too much "impregnated" by the habits for sharpening Canon files.
Her we can see some artifacts and they come from that habit, plus the fact that I'm a new comer to Lightroom.

We shouldn't underestimate that, and I don't. Nicolas comes from a camera system designed for action, with a sensor that uses AA-filters and Microlenses, to a camera system without AA-lenses wich behaves a bit differently. On top of that, he is forced by the lack of 645D support in his favorite C1 Pro Raw converter to start using Lightroom/ACR as Raw converter, and that Raw converter just switched to a whole new Process version that requires unlearning old habits and learning new ones for everybody.

On top of that, I designed a method to improve the Capture sharpening quality, that even Adobe doesn't seem to intuitively understand (judging from the user interface design).

Let's call it a challenge.

As soon as I'll have some more time free, I'll dig seriously with Bart hints and help to get even better results, and I have no doubt that we can... and will let OPF know : )

Yes, we can and we will (even if I have to do some of the analysis steps myself). Take your time, take a shot of my resolution target at f/8, and send me the file. Ideally, shots at all apertures that can be used should be taken and analysed but that would take me some time. However, we can start with one (a target shot at f/8, ISO 100 or 200), and see how well people like its Capture sharpening. More importantly, how well Nicolas likes it, because the file should be a perfect foundation for artifact free enlargement to wall sized proportions.

Cheers,
Bart
 
Nicolas,

In case Bart doesn't come on line soon, this is a Photoshop setting and it is found in FILTERS>OTHER>CUSTOM, just one above where you find the High-Pass filter.

Hi Maggie, you just beat me to it.

Indeed, it's hidden in Photoshop's Filter menu under Other|Custom. It's one off those filters that normal people never need to use, but then I never said that I'm normal ;)

Cheers,
Bart
 
Top