The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

If CCD rendering can be achieved with CMOS where are the examples?

Guy Mancuso

Administrator, Instructor
Agree. Things have changed in CMOS vs CCD. Never before was it the same format but now with the same MF I'm afraid all our old well it's just better trust me comments are more like it's basically the same. Our old 35mm versus the world thinking is still there but if within the same format they are just so close it may not matter. Plus we pick up some functionality. I'm still old school thinking in be a expert in post as that's what really counts.
 

iiiNelson

Well-known member
Well I can see a distinct difference in saturation and brightness. Neither look are offensive to me. I like both. The 645Z clearly has more DR but the 645D has punchier colors.

My take at least regarding CCD/CMOS when I shot my M9-P next to an A7/A7R were similar to what you see with the 645D/645Z. It actually was a bit shocking to me that these CMOS was closer in the "CCD look" of the M9 than the CMOSIS chip Leica chose.

Kodak made the sensors for the M9 and the 645D so it makes a little sense that there's a similar closeness with the 645Z to the 645D. The Sony MF sensor are based on the same "family" of sensors as the A7 series. Might just be the curve that the Sony uses that is more appealing to me than the CMOSIS chips.
 

Egor

Member
I will crosspost some crude screen grabs and give you a dropbox you can download some raws from:
I have a couple of screen grabs for now. Sorry folks, long day again

Raw Files Dropbox

We found that the IQ250 tends to shoot, right out of the gate at any rate, more accurate color and better detail in highlights and shadows, not as good as the Credo in micro contrast and mid tones...all probably easy to correct;and it looks like a light on the left misfired, so lighting not exactly the same; but I found it interesting...
Crop sensor difference quite noticeable
Check out those highlights and fur detail in the IQ250! Also, that chestnut color on the one boot in the middle, the IQ nails it while the Credos have issues with warm colors for some reason. Perhaps that skin tone thing Credo is always talking about is not good for accurate color when warms and yellow/reds are concerned...I really want to use the "standard film" profile from the IQ250 on the Credo 80.....why can't I?
 

dougpeterson

Workshop Member
I really want to use the "standard film" profile from the IQ250 on the Credo 80.....why can't I?
The processing pipeline each go through are quite different. One cannot simply carry over profile+curve from one to the other. You'll notice even if you set both to linear scientific, and No Color Correction ICC profile that they will still look different.

One was meant to honor the legacy and heritage of the Leaf Look and the other was meant to provide continuity of the Phase One look. Different purposes = different methods = different results.

You can, of course, try to homebrew a profile for either/both backs which will bring them closer together. But such target-generated profiles are usually quite frail - they work well to reproduce the target but in real world use spot colors not present in the target can often become strange, and any change in lighting or post-processing style can exacerbate issues with such in-situ target-generated profiles.

By the way, I think this is a great example of a file where sharpening-threshold can/should be reduced; the texture of the fur is greatly aided by such a change.

The difference you're noticing in micro contrast in some areas seems to be because the IQ250 is focused a hair further back in the scene (not for instance the tip of the rear black boot toe) and any remainder is likely due to where each back's respective curves place additional contrast. Adding a slight curve to the midtowns on the phase image, or using the clarity/structure tool at a low setting on the IQ250 file (e.g. 10 structure/5 clarity) does a great job of increasing perceived midtone contrast without misrepresenting the product.

Regarding focus the use of the IQ250 wireless live view (enabled today via Capture Pilot update that Apple was taking their sweet time to approve for the App Store) should make a nice tool for focus assist. Imagine a iPad mounted to your tripod, for in-situ immediate, low latency, high res live view. You can do that with either back via USB3 of course, but usually (not always) the relevant monitor is a neck-turn away and not easy to move with the tripod as you reposition or change sets, whereas the iPad (or iPad mini or iPhone 6/6plus etc) can come along for the ride.

Naturally both backs would be providing a bit more bitty detail if shot at something lower than f/22, but I fully understand the practical requirements of such a shoot and getting
 
Last edited:

Egor

Member
OK, then I guess I will have to build a custom profile or adapt an existing Leaf profile to try and match?
I just thought that Leaf and Phase being same company and all that, same sensor and all that, most same electronics and all that....they might wanna give the option to emulate each other?
I am aware that some people like vanilla flavored coffee and some like hazelnut flavored coffee; we like coffee flavored coffee here at our studio :)
We have a hard time with the Leafs and yellow/reds mostly in art-repro and product like the chestnut boot there. No biggie, we fix :)
 

dougpeterson

Workshop Member
I made an edit to my post about the same time as yours about custom profiles - might want to reread. IMO they are a black hole for your time vs having the results you want in the form of a couture hand-adjusted profile. You can read about the profiling for the IQ250 in my article about IQ250 - CMOS Fully Realized.

There will definitely not be any "emulation" mode provided in the future - the raison d'etre for each back's look is different and requires different approaches. We've done some custom-hand made profiles for Phase backs to look more like Leaf backs for portrait/people shooters, but this is much simpler as the starting point of the Phase [profile+curve] is more neutral to begin with. Even so I can only say, with expertise in such work and with many hours of time invested, that I can only get an IQ to look similar to a Credo - I definitely cannot "match" it. Making a Leaf Credo look like a Phase IQ is even more tricky. It's easier to add spice to a stew than to try to take it out. It would also defeat the purpose for the Leaf Credo which is to provide the famous Leaf Color, especially its handling of skintones.
 

Egor

Member
Thanks, Doug!
Yes, I have found that it is easier to make Lab adjustments in C1 to standardize warm product colors that we shoot. Everything else is pretty neutral.
I guess we came into this MFD world with you and Lance with Leaf Aptus and have tended to stick with Leaf ever since. If I had to do it over again and the significant price difference wasn't a factor (it is) I would go with PhaseOne backs over Leaf because I like to start with neutral and if I want it warm and glowing I know how to do that :)

Yes, real world demands compromises like using F22 even though I am fully aware of diffraction. The first thing we learned about MFD was that it had VERY limited DoF and to deliver what my clients wanted, we needed to close down the lens or start using tech-cameras and lenses. We tried both and the production value of tech-cams vs DSLR (DF+) was too great. In other words, much easier/faster/more productive to use DF+ and close down to get DoF than fumble all day with tech cams and then there is the costs....

The crop sensor format surprised me a little (I don't know why, I knew about it, just didn't consider until we had it in here) I am considering checking out a used 55-110 zoom to see if it will work for us here to emulate the range we need with the crop sensor. 75-150 is probably in our future but the range is slightly too on the telephoto side to be useful for our needs.

Staying on topic with this thread, does anyone notice any major differences between the CCD (Credo80) and CMOS (Credo50 and IQ250) ?
 

vjbelle

Well-known member
Staying on topic with this thread, does anyone notice any major differences between the CCD (Credo80) and CMOS (Credo50 and IQ250) ?
I have come to different conclusions than you have. I have always found my IQ180 to be warmish just as I did my P45. I don't attribute this to the sensor but more how C1 renders the file. My Credo 50 is much less warm and, to my eyes, more accurate shooting in daylight. Without any WB the differences are fairly dramatic but even after WB to each the IQ180 renders towards the warmish side for certain colors.

Victor
 

carstenw

Active member
Hmm, I don't know if this can ever be resolved, but if so, then through images which speak clearly enough to convince all (or most).

In the IQ250 vs IQ260 image, I like the tone mapping of the IQ260 better, but the sparkles in the noise are quite objectionable, and I am guessing that the colour is quite a lot less accurate, inaccurate white balance aside.

In the 645D vs 645Z images, the red tones in the cloth at 3 O'Clock have quite a lot more separation on the 645D, whereas the blue cloth at 10 O'Clock has more separation with the Z.

Interesting, but not conclusive.
 

Egor

Member
Hmm, I don't know if this can ever be resolved, but if so, then through images which speak clearly enough to convince all (or most).

In the IQ250 vs IQ260 image, I like the tone mapping of the IQ260 better
Carstenw, do you mean my credo 80 vs iq 250? Or is there another test I didn't see?
 

ErikKaffehr

Well-known member
Hi,

Some points to keep in mind on the Pentax 645D and the 645Z.

- That comparison is done using test samples from ImagingResource, taken several years apart.

- I generated a DNG profile for each using another sample with a good ColorChecker.

- Exposure in Lightroom was adjusted to match on second brightest field of the small colour checker in the test image.

Doug Peterson has published raw files from the "library shots". That was a very good test including the IQ 250/260/280, but unfortunately was not really useful for judging colour.

Lance at Capture Integration posted a ISO comparison between Leaf Credo and IQ-250 and posted raws from that.

As Doug Peterson indicated, Leaf is intended to have a different colour reproduction from Phase One and he clearly said there will not be a Leaf emulation in C1.

It is possible that Capture One and Leaf use different CFA (Colour Filter Arrays), Doug Peterson has mentioned in his article on the development of the IQ-250 that Sony offered two different CFA designs.

What Doug also told in that story is that Capture One's profiles are in part built by the Image Professor using thousands of real world pictures.

I cannot say anything about that, but I made a test on my P45+ for colour accuracy and I compared the studio flash and portrait profiles. The differences were incredibly small, but flipping the images differences were perceptible, some very fine tuning there.

Best regards
Erik


In the 645D vs 645Z images, the red tones in the cloth at 3 O'Clock have quite a lot more separation on the 645D, whereas the blue cloth at 10 O'Clock has more separation with the Z.

Interesting, but not conclusive.
 

ErikKaffehr

Well-known member
Hi,

The ImagingeResource pages also contain one of the popular colour reference images often called musicians. I guess they come from Kodak.

I processed those files with three color profiles:

  • Adobe Standard
  • DNG Profile editor (which I made from another part of the same image)
  • Embedded - which I guess is the DCP profile coming from Pentax

The images are here: http://echophoto.dnsalias.net/ekr/Articles/P645D_vs_P645Z/Musicians/

And a screen dump is below, I would suggest checking them link above, though, as colours are probably more correct.


Best regards
Erik
 

VICTOR BT

Member
with film, colour or black and white, one can use emulsions that on paper are very similar, like fuji acros 100 and kodak t-max 100, for example, or fuji 160 and kodak portra. conducting the print optically (darkroom), even when the photograph is done in controled light, there is still a significant signature, vibe, character to each of these films. their tactility is distinguishable, the tactility of the object, like metal, skin, or fabric have slightly different feel even when the luminosity (and contrast) is very similar.

in digital however, the stream from capture to conversion (up to the appearance of the file in the software) are the determinative elements. then, the software (Lightroom, C1, Phocus etc) allow a great measure of further manipulation to achieve the desired "look".
there is no anti-aliasing filter, like in common dslrs, which is a significant element altering the potential "look" of the photograph even under extensive manipulations in the software.
i suppose at an extent, the polarity of CCD and CMOS is traced back to that difference and what derived from that.

but now, with the same lens, with no anti-aliasing filter, with great extent to which the file can be manipulated after the conversion, should there be an issue ? like it is with film, no matter how one manipulates it within the possibilities given in optical printing.

and most importantly, considering that the main and intended advantage of medium format is for print, how the CCD/CMOS difference comes out eventually ?

what i mean is not the technical aspects (like live-view, or issues with tech-cameras, iso, battery etc) but the look and the feel, which might be subjective, nevertheless they might be there.

for example, how hasselbald 40 (33x44), 50c (33x44) and the bigger sensor 50 differ ? same about P1 and leafs etc? does it come out significantly different after the conversion ? is it possible to bring them to desired "look" more fluently with one or the other ?
i see the hasselbald 40 and 60, and they look slightly different (mainly for the size and resolution), but aesthetically and inherently, not as different in "feel" as tmax100 or acros100 (both equally refined films without the more authentic characteristics of traditional films). same about colour films like fuji/kodak 160.
 

carstenw

Active member
Carstenw, do you mean my credo 80 vs iq 250? Or is there another test I didn't see?
No, I think it was Doug at CI (at the time?) who did the test, where this image came up. Sorry, I just realised that I followed a link and didn't make a note of that:

Login
 

satybhat

Member
No, I think it was Doug at CI (at the time?) who did the test, where this image came up. Sorry, I just realised that I followed a link and didn't make a note of that:

Login
Sensor morgulis (all sensors must die),
Sensor dohiris (all sensors must serve) !!

Is it time to exhume this issue?

I gave myself 90 seconds to make the IQ3 100 shot look like the IQ280 shot.
This is what I could manage.
Any opinions on the CCD/CMOS issue in the 100MP era?

Cheers :grin:

 
Top