Pentax 645D, or 645Z: CCD vs. CMOS

[SeattleDucks]

Hi All,

Some around the web have commented about CCD sensors having a different look than CMOS, saying the CCD gives a more analogue film-like rendering. For Pentax 645D owners, do you find this to be true? Are there differences you see in the output of the 645D vs. other modern Sony CMOS sensor based cams such as the D800, that could be attributed to some quality of the CCD sensor in the Pentax?

I'm aware of the modern CMOS advantages in terms of lower noise at higher than base ISO's, increased dynamic range, and better performance in long exposures. But my main concern right now is determining if the CCD in the 645D actually has some advantage in it's rendering for a landscape shooter despite CMOS' advantages. If so I may be inclined to purchase a 645D rather than the Z I have preordered.

Perhaps Phase One and Hasselblad [edit: and Leica] users may have thoughts to share on this as well.

Thanks in advance,
Ross

 

[gazwas]

I imagine its a very difficult question to answer for anyone with anything other than an oppinion at the moment as the camera has not been released yet. I personally feel the advantages of CMOS more than make up for the (subjective) difference in look over CCD.

As processing techniques/engines evolve I imagine the difference will be less apparent.

 

[fotografz]

The only similar experience I have had is when Leica went to CMOS for the M(240) compared to CCD in the M9.

Personally, after two weeks of testing and experimenting with post processing approaches, I subjectively didn't like the new camera's image qualities (look and feel) and for the first time in 35+ years did not buy a new M despite improved operational characteristics.

Whether that will hold true for MFD in general, or Pentax specifically, is anyone's guess right now. One thing is certain, any of these cameras will deliver stunning detail … then it comes down to personal aesthetic preferences weighted against operational needs.

BTW, I will be facing the same decision soon enough when Leica goes CMOS for their S camera … I currently shoot with a S2-P that uses a CCD sensor, which I love because the files have a certain naturalness to them, something others have also commented on. That may be an attribute of the lenses for all I know, but I can't determine that for sure until I can try a CMOS version.

When it comes to MFD, I have a pretty firm policy of "try before you buy".

- Marc

 

[Jay Emm]

...I will be facing the same decision soon enough when Leica goes CMOS for their S camera … I currently shoot with a S2-P that uses a CCD sensor, which I love because the files have a certain naturalness to them, something others have also commented on. That may be an attribute of the lenses for all I know, but I can't determine that for sure until I can try a CMOS version...

Ditto for me re-Hasselblad and the new CMOS model.

This post comments on Leica CCD vs CMOS in the M-Series, and it seems the CMOS ticks enough boxers to make it a wise choice. But I get Marc's point - as for me the CCD in the Hasselblad delivers wonderful color - and is frequently commented on my customers (and yes, lens and processing play a part too). Anyway, more food for thought here Leica M (typ 240) Field Test and Review

 

[D&A]

The only similar experience I have had is when Leica went to CMOS for the M(240) compared to CCD in the M9.

Personally, after two weeks of testing and experimenting with post processing approaches, I subjectively didn't like the new camera's image qualities (look and feel) and for the first time in 35+ years did not buy a new M despite improved operational characteristics.

Whether that will hold true for MFD in general, or Pentax specifically, is anyone's guess right now. One thing is certain, any of these cameras will deliver stunning detail … then it comes down to personal aesthetic preferences weighted against operational needs.

BTW, I will be facing the same decision soon enough when Leica goes CMOS for their S camera … I currently shoot with a S2-P that uses a CCD sensor, which I love because the files have a certain naturalness to them, something others have also commented on. That may be an attribute of the lenses for all I know, but I can't determine that for sure until I can try a CMOS version.

When it comes to MFD, I have a pretty firm policy of "try before you buy".

- Marc

I completely share Marc's opinion regarding the differences in output between the CCD based M9 and CMOS based M240. Sure on can fiddle and develop a workflow that emulates many of the aspects of the M9 look in M240 files, but it's not uniformly consistent and there are some types of images/lighting that simply come close but not the same. Again this is very subjective but that's been my experience working with similar and closely matched pairs of files between the two cameras.

The M240 is a lovely camera and I could use many of it's advantages over the M9....but the look of the image and output comes first before anything else. This is especially true when the shear number of files I need to adjust and tweak is substantial/ I don't want to spend days tweaking endlessly images, to just to get a large percentage to get close or somewhat near the look I want. The M9 RAW files have a look I love and so does the 645D.

I have no basis to say this yet and it may be completely wrong, but I have a sneaking feeling that a similar situation to a degree will hold true between the CCD based 645D and CMOS 645Z. That's why I'm treading slowly before considering selling my 645D for the 645Z. Time will tell....it always does.

Dave (D&A)

 

[Ed Hurst]

I fully agree Dave - can't wait to get a proper body of 645Z files to consider in order to make this decision. But Pentax has certainly done a good of job of making the decision a tough one with their improvements to the body's functioning. Here's hoping the file properties don't let the thing down.

 

[D&A]

I fully agree Dave - can't wait to get a proper body of 645Z files to consider in order to make this decision. But Pentax has certainly done a good of job of making the decision a tough one with their improvements to the body's functioning. Here's hoping the file properties don't let the thing down.

Ed, analogous to Pentax making things difficult with all the added features of the 645Z, Leica did exactly the same with the M240, maybe even more, since operationally, the differences between the M9 and M240 were striking. At the end of the day, there were a fair number of individuals who decided to stay with the M9 or simply after moving to the M240, went right back.

Dave (D&A)

 

[Paul2660]

This subject was addressed in detail over at LuLa, and you might want to take a few minutes and read through this post. Some pretty heavy hitters over there, offering both pro and cons on color, CCS vs CMOS.

The Future of CCD Sensors

Paul

 

[wentbackward]

I'll be contrary to other opinions here. I believe the differences between CMOS and CCD are there, but other things affect the image MUCH more, such as the filters used, and pixel site size. The whole optical pipeline is different when comparing say a D800 to anything else. Here are my insights:

Optical pipeline
To play the High ISO game, I believe Nikon uses far weaker colour filters and more mathematical methods than Canon or Sony, hence I personally feel Nikon let us down with the D800, since they didn't address the weakness in their system (colour, especially red->oranges->yellow in bright conditions). I did some analysis of this (using various nikon files vs A77 and A99) and recall it was around 2/3rds or more of a stop extra speed Nikon vs Canon/Sony with a degradation in colour, especially skin tones, especially around the hair line of brunettes and fawns.

Filtration
Many people preferred the M8 to M9 (Due to filtration changes), now argue M9 vs M[240] (ccd to cmos and different filtering), I think that it's more accidental quirkyness of a very special system than anything specific to CCDs. Due to the differences from M8 to M9 with essentially the same silicon and precisely the same optical pipeline beyond the sensor filter, I argue filtration is far more important than CCD/CMOS.

I also think with the early M's we were intrepid explorers, the latest M is far more complete. This resulted in a different mindset where we were prepared to work around the IQ quirks in the early M's, coming to endear their unusual properties.

Sensor site
Going from a Mamiya 33 (Aptus 7II) to a Phase One IQ160, I felt disappointed for several months, lets say even deep regret about upgrading. In part due to the superior Mamiya/Leaf profiles, but also the larger pixel site sensor just rendered skin tones and also natural greens to brows, with such incredible beauty. This is well documented on many forums and there are people who seek out these larger pixel site sensors still today. I therefore argue that the ability to capture a larger area of light has significant affect on the sensor.

- Paul

 

[Shashin]

Since you have no history with either camera, it should make no difference which you choose as you are going to have to work with each to understand the images they will produce. Neither are going to end up making bad images. The only cravat I would make is if you are shooting portraits in the studio and then I would test the cameras side by side.

The discussion around CMOS and CCD is rather hard with no real data. The only thing that I have heard, and here again with no real confirmation, is that to get higher ISOs the color filter array is made weaker. But yet CMOS sensors make great images--I have an RX-1 and D800 along with my 645D. I do prefer the 645D, but that could be simply because I use it more. It could also be Pentax profiling and colors rather than Sony or Nikon.

Another hypotheses could be the effect of greater dynamic range. The greater the dynamic range, the flatter the image. I have heard lots of people say the CMOS sensors are "flat." That can be fixed with a tone curve, but how many photographers are wiling to give up highlight and shadow detail? The current aesthetic favors greater dynamic range.

The only CCD/CMOS color test I have seen is between the D800 and Hasselblad. I don't know how much is the sensor and how much the camera profiles, but you can scroll down the page to see a color checker comparison:

Nikon D800E v.s Hasselblad H4D40: the end of medium format superiority? Round two «Photigy: Online Studio Photography Lessons

If I did not have a 645D and was shopping for a Pentax MFD, I would probably go for the 645Z. There are a number of improvements that are good: ISO, live view, faster preview times, etc. However, owning a 645D, I feel no great need to change. What I get out of the camera is great--I print large on 44" roll paper.

One thing that no one has brought up about the 645Z is that Pentax only give lens profiles with the FA and DFA series lenses. They do not profile the A series lenses. For most of the A lenses, it might not be a big deal, but if you are thinking about the A35mm, then it might be worth it to wait until someone has tested that lens with the 645Z. CMOS might be a little more sensitive to lens cast and such. At least the Phase One test of the IQ250 on tech cameras seem to indicate that.

 

[D&A]

This subject was addressed in detail over at LuLa, and you might want to take a few minutes and read through this post. Some pretty heavy hitters over there, offering both pro and cons on color, CCS vs CMOS.

The Future of CCD Sensors

Paul

Paul thanks. I've been following that discussion but hadn't contributed a post till now.

When I say I prefer CCD based cameras to CMOS ones in general, I'm not saying it's simply the output I find more attractive and preferable and is solely due to the actual CCD chip. It could be the filter array or some other factor in the supporting cast to the sensor, but the bottom line is the look and output I prefer has been pretty consistent in cameras containing a CCD sensor.

Dave (D&A)

 

[GrahamWelland]

Sensor site
Going from a Mamiya 33 (Aptus 7II) to a Phase One IQ160, I felt disappointed for several months, lets say even deep regret about upgrading. In part due to the superior Mamiya/Leaf profiles, but also the larger pixel site sensor just rendered skin tones and also natural greens to brows, with such incredible beauty. This is well documented on many forums and there are people who seek out these larger pixel site sensors still today. I therefore argue that the ability to capture a larger area of light has significant affect on the sensor.

- Paul

I have to agree with this. I still think that my Aptus 65M files right from the back have a nicer colour subtlety than my IQ260 files until I've worked them. :chug:

 

[SeattleDucks]

Thanks everyone for the thoughtful discussion. Some good points brought up.

My background is many years of landscape shooting with 4x5, 6x7, and 6x4.5 film, and I began making the transition to digital several years ago with various DSLRs, currently the D800. The Nikon has plenty of resolution and loads of dynamic range, but I am not thrilled with the files even after much work in LR5. The color often does not look quite right to me, and I sense a bit of lifelessness in the overall rendering.

Looking at 645D images from around the web I have often seen a quality in the images that feels closer to the organic look of my film drum scans. But it is very hard to quantify exactly what it is in the files that's grabbing me.

Though the new 645Z is not released yet, we do have current Sony CMOS sensors out there that may give a general feel for what to expect in the rendering of the Z. As mentioned, I'm not thrilled with the D800 files overall look, nor have I been drawn to the IQ250 sample images I examined. And yet images like these from the 645D have drawn me in with a sense of inner glow, local contrast, and a natural look that, again, reminds me of the high quality film scans I've worked with:

645D review images: Google Translate

Unfortunately, the samples posted by Pentax from the Z are not helpful at all to this investigation. Most of them are quite awful.

I apologize if my attempts at describing what I *think* I'm seeing are not specific nor scientific enough. I do often doubt my own eyes on this topic, wondering how much my brain is being influenced by things I've read, so I appreciate the discussion here with others to help flesh these things out.

Ross

 

[D&A]

Ross,

No need to be sorry or apologize for describing what you see or feel when looking at certain images. Photography is a visual art form and although there may not be rigid scientific explanations of why you derive a certain feeling or reaction to certain images as those from a CCD based camera ( or any kind of camera or lens for that matter), you are not alone.

As often said, it's an art form like most that is very subjective and there is no right or wrong....simply what you like or don't like. Now as for the best tasting ice cream flavors, that's a whole different matter .

Dave (D&A)

 

[Shashin]

Well, it you had to pin me down, I would say my 645D is on the yellow-green side whereas my Sony camera are yellow-red, but it is a lot more complicated than simply color balance--kind of like trying to nail down why Kodak and Fuji film are different. I cannot tell you why, but I do like the 645D files very much. But if you want me to demonstrate a difference and a consistent difference, honestly, I would be quite hard pressed to give you something. However, if you are seeing it, who am I to call you crazy?

 

[tsjanik]

Ross:

I bought the 645D in 2010 and intended to continue using a 67 film camera as well. I have been so pleased with the files from the 645D that I have developed only two 220 rolls from the 67 since that time. I've owned only a few digital cameras, but the files I've liked best have come from CCD sensors. Will the 645Z be the same? We all wonder.

Tom

 

[Pics2]

The discussion around CMOS and CCD is rather hard with no real data. The only thing that I have heard, and here again with no real confirmation, is that to get higher ISOs the color filter array is made weaker. But yet CMOS sensors make great images--I have an RX-1 and D800 along with my 645D. I do prefer the 645D, but that could be simply because I use it more. It could also be Pentax profiling and colors rather than Sony or Nikon.

Shashin (and Tom), I'm thinking about switching from Nikon D800E to Pentax 645D for my studio still life work. I want to go for larger sensor and CCD.
The only thing that still stops me from doing so is that I'm used to Capture One software and the way it handles Nikon files. How do you process Pentax files, what's the best RAW converter and will I miss C1?
Thanks! (Sorry, it's a bit off topic)

 

[wentbackward]

You'll be switching from small format to medium format, that's more significant than you can imagine. Firstly the workflow is slower (studio still life should present little difference). DoF will wonderfully creative in some ways, but a royal PITA in others.

One thing is that going for the Pentax for still lifes maybe somewhat limiting in the sense that a modular camera system would give you more options to upgrade (e.g. to get a Linhof techno or similar).

 

[peterv]

On page 5 of the aforementioned LuLa thread, someone posted a comparison between the Hasselblad 60MP CCD and the new 50MP Sony CMOS:

The Future of CCD Sensors

What I see on a pixel peeping level is a litlle less noise and a little more dynamic range in the CMOS. Looking at the complete images side by side, I like the CCD versions better. Maybe it's the slighty smaller DR, wich gives the CCD version more apparent local contrast and a bit more 'bite'.

I'm in the same boat as Marc and others, shooting with an S2-P and not in a hurry to move to CMOS.

 

[Pics2]

You'll be switching from small format to medium format, that's more significant than you can imagine. Firstly the workflow is slower (studio still life should present little difference). DoF will wonderfully creative in some ways, but a royal PITA in others.

One thing is that going for the Pentax for still lifes maybe somewhat limiting in the sense that a modular camera system would give you more options to upgrade (e.g. to get a Linhof techno or similar).

Thanks! I have experience with MF - Phase One owner. But, for my current workflow Pentax 645D looks fine to me (I don't need movements, digital back versatility etc). It's not expensive, no need for complicated dealer contracts, and it's a decent jump from DSLR.
It's, like I said, RAW convertion workflow that worries me. I'm used to C1 - it's colors, rendition etc.