Color profile for Leica M (M10, 262) with ColorChecker

[MikeMyers]

Advertisement (gone after registration)

Thanks; the watermark seems to have fixed itself.  I also found I can start editing an image in DxO PhotoLab4, then send it to Topaz Sharpen AI, and when I finish it brings me back to editing in PL4.  Perfect.

One of the pages as I paid the money said something about it being valid for a year.  I can try to find it again, and copy it here so you can see it, but based on what you just wrote, it's irrelevant.  Thanks for the heads-up about buying it (but don't suggest too many things to me, or I'll go broke!).

[MikeMyers]

Found it:

I don't understand this - they are telling me I only have it for one year????

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jaapv]

no - you own the license outright and they throw in one year of free upgrades.

[zeitz]

11 hours ago, jaapv said:

you own the license outright

Yah.  I own the license, but the price to upgrade to V3.2 is $49.00 for this little piece of really nice software.  That is equivalent to 5 months of Adobe upgrades for a big complete set of really nice software.  If you own a larger suite of Topaz software and if you want access to Topaz' improvements, which are significant, be ready to pay a steep price.

[jaapv]

Maybe, but AI technology is not cheap. Besides, their helpdesk and customer interaction is about tenfold better than  Adobe. The price is well worth it to me as it has become an essential part of my workflow. 

[MikeMyers]

From what I think I've learned, much of it from watching the videos posted above, and considering my own "limitations" (real or imagined), my best option seems to be to continue with PhotoLab4, and when more sharpening is needed, if PL4 isn't adequate, to use Topaz Sharpen AI.  I've found that I can open an image in PL4, do much of my editing there, and at some point send the almost-finished image off to the Topaz app, do the sharpening, and it will be returned to me to finish things up in PL4, including adding my watermark.

I've only had one opportunity to use it so far.  I was walking to the store, and had my Fuji X100f with me, with an almost dead battery.  I came across some iguanas, and I started taking photos of one of them.  It climbed up onto this wall in front of Biscayne Bay, and it allowed me to get reasonably close, before it started scampering off to my right.  I panned my camera with the iguana, and got one photo at a "random time" that my finger selected for taking the image.  I got home, and the image was rather ugly, not only with the feet in motion being very blurry (as I expected) but the body of the iguana also being unacceptably blurred.  Having just learned about Topaz Sharpness AI, I purchased it, and edited the image selecting "motion blur", but while the body looked better, the "spikes" across the top looked awful.  I tried it again, this time using the option for a soft image, and got the photo I've copied below.  I take photos like this all the time with my Leica, hoping that the image will be sharp enough to be useful, but with enough blur (preferably in the background) to show the motion.

The concrete below is blurred because of the motion.  I need to learn how to 'mask' the image to do this even better.  But I ended up with an image I like.  It even shows how the iguana "walks", with two opposite feet working in unison.  I'm sure there are other solutions, maybe even the one I was thinking of before @jaapv sold me on the Topaz program.  In the meantime, this is a nice tool to have available, and it works perfectly when I use PL4 to edit an image, then send it off to another program to do something special, after which the revised image is sent back to PL4.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited August 7, 2021 by MikeMyers

[zeitz]

Advertisement (gone after registration)

I'm just saying, if you think this path is going to be cheap, it is not.  And I am not criticizing anyone's decisions or preferences.

[MikeMyers]

My opinion is that there are many paths, some expensive, some cheap, and some reasonable based on the person following them.

The end result is the photographs each of us create, and how satisfied we are with the results.

If I spend a gazillion dollars on the M11-RX with a gazillion pixels, I doubt my photos would be any better than with my M10.

On the other hand, the ability to sharpen images when needed, makes an obvious difference when needed.

My own path is mainly PhotoLab4, and there are already far too many options for me to learn.  They're explained nicely in the videos and webinars, just as the video up above explained how to use Topaz Sharpen AI.  I'm scared though about the term "cheap". In college, I considered $10 as cheap, and $100 as expensive.  Now I accept $100 as cheap, and $1,000 as expensive.  I used to dream of photo gear I couldn't afford, and I still do.  Meanwhile, I've lost track of the number of photos that I've taken that didn't look sharp enough, and with this new software, some (many?) of them were (are) fixable.  I don't think any of this goes against my personal definition of "photojournalism", but I still wonder about that.

I don't think I will ever upgrade from my M10, but ten years or so ago that's what I thought about my M8.2 - and when the M9 came out, I had that option, but didn't take it.  

That I purchased "Sharpen AI" doesn't mean or imply that I'm about to purchase any more of the Topaz software.  I don't see it being part of any path.  I do see following @jaapv as a path, but I'm mostly happy with my own paths already.

[Adam Bonn]

On 8/4/2021 at 4:43 AM, MikeMyers said:

It makes sense to me that when Leica ships a new camera, all the calibration and adjustments are already done and "correct".

Leica embedded profiles all work the same, in that they use the same mechanics to provide information to the RAW editing software.

However, the information these profiles contain is different.

Leica profiles contain two 'Color Matrices'

One for StA (tungsten) and one for D65 (cloudy noon light)

However, the numerical values contained in these 3 x 3 matrices (so that's nine values) are different for each model of camera

But in every Leica camera profile there are these two transformation matrices that convert XYZ values to reference camera native space values

This is a very, VERY small part of demosaicing a RAW file into an image.

How your RAW editing software reads and reacts to these matrices will vary depending on the brand of software, in fact AFAIK (...) most RAW apps will ignore them completely and use their own data tables to create an image for editing (even LR defaults to 'adobe standard' and you have to fish 'Mx Embedded' out of the drop down if you want to use it)

When you use a colorchecker/etc to create your own .DCP file, you're creating a new set of Color Matrices based on the native space value of your specific camera

So the camera is calibrated by Leica.... but the calibration is based on calculations and tolerance expectations of the sensor used, not the specific sensor in your exact camera.

[MikeMyers]

13 minutes ago, Adam Bonn said:

So the camera is calibrated by Leica.... but the calibration is based on calculations and tolerance expectations of the sensor used, not the specific sensor in your exact camera.

Isn't that the way it should be?  If I put a yellow filter on my lens, and shoot in 'dng', anyone opening the file would see a yellow image.  If my lens is very old, and perhaps discolored, ditto.

PhotoLab4 corrects for design defects in a lens automatically - but only if you allow it to do so.

PhotoLab4 corrects for design defects in a camera automatically - but again, only if you allow it to do so.

 

Are you suggesting that different camera manufacturers do this calibration differently, and that if we take the identical 'raw' image in a Leica, a Canon, a Nikon, a Fuji, and so on, the images will be significantly different?  I guess that makes sense, but I would assume that all Leica cameras being made at the same facility at any given time would be pretty much identical.  If I'm shooting my M10 tomorrow, and it stops working, and I borrow your (different) model of an M10, I wouldn't expect to notice any differences.  Am I wrong?

[Adam Bonn]

46 minutes ago, MikeMyers said:

Isn't that the way it should be?

Well IIRC (...) the Hasselblads are calibrated on an individual camera basis, but yes - in manufacture producing items within a tolerance spec is indeed the way it is

 

46 minutes ago, MikeMyers said:

PhotoLab4 corrects for design defects in a lens automatically - but only if you allow it to do so.

PhotoLab4 corrects for design defects in a camera automatically - but again, only if you allow it to do so.

And RAW editing apps will correct for colour errors automatically if you click 'auto' on the WB tool (sometimes it even looks ok too 😅)

 

46 minutes ago, MikeMyers said:

Are you suggesting that different camera manufacturers do this calibration differently, and that if we take the identical 'raw' image in a Leica, a Canon, a Nikon, a Fuji, and so on, the images will be significantly different?

You're putting too much gravitas on what the camera OEM does. Away from SOOC JPEGs it matters little - the RAW software does it's own thing and does not care about much (bar EXIF data) that the OEM burns into the RAW image

As a very loose helicopter view of what's going on, you take a picture. It burns something relating to the WB value used into the RAW data. This value is then used in conjunction with colour information (eg in Leica the 'Color Matrices or Capture One's ICC profile') to form the basis of the look of the picture (before you edit it to your tastes)

Leica use a WB tag called "As Shot Neutral" (a real example is 0.388679 1 0.899923)

I just looked at a historical Fuji RAF I had lying around

WB GRB Levels Standard          : 302 374 853 17 302 658 506 21
WB GRB Levels Auto              : 302 583 553
WB GRB Levels                   : 302 583 553

So yes both Fuji and Leica (and canikon etc) embed WB data, but no not in the same way. It will do a similar job though

"Significantly Different" well I'm afraid that's a bit of a non-sequitur really... Of course the skies will be blue, the grasses will be green, the red cars red... but sensor sizes, lens used, differences in how the camera processes the signal, the differences in the CFA and filter stack, the differences in the native tone operators driven by the dynamic range of each camera will lead to differences. Quite tangible ones really. (ever seen an M9 vs M240 colours thread?)

 

46 minutes ago, MikeMyers said:

 I guess that makes sense, but I would assume that all Leica cameras being made at the same facility at any given time would be pretty much identical.  If I'm shooting my M10 tomorrow, and it stops working, and I borrow your (different) model of an M10, I wouldn't expect to notice any differences.  Am I wrong?

I think this is a fair bet (that the same models will behave the same) But it would have to be the same model, after all an SL renders differently to an M10, which is different to an M10R, which is different to a Q2 and so on..

Of course if the manufacturer changes anything on that model during the production run... (perhaps for example M9 cover glass)

Edited August 7, 2021 by Adam Bonn

[Jeff S]

Roger Cicala (from Lensrentals) knows more about sample variation than the average bear, and it applies to both lenses and cameras. 
 

https://m.dpreview.com/articles/7333489584/variation-facts-and-fallacies

Even metering and shutter timing can be different between two bodies of the same camera model, which is why in film days I did film speed tests for each of my cameras, sometimes resulting in different ASA ratings for Tri-X for my different M bodies.  I still do testing and profiling of my current cameras, using my own lenses, within my specific shooting and print workflow. It reduces variables that I can control, with tools that make it relatively easy, and helps me better learn how my gear performs under different circumstances.

Jeff

 

[MikeMyers]

On 8/5/2021 at 1:25 PM, jaapv said:

I do indeed check sharpening @ 100%, but I do so in Sharpen AI Preview; much better than any editor. And it works perfectly well on a standard resolution screen. I obviously judge the full image at normal size, and to me the high-res screen looks awful with a properly sharpened image - as do prints that look OK on 4K.

Well, you've got me rather confused, in a good way.  I took a photo just after sunrise this morning, with the golden yellowish light that was slowly going away, and I expected the image to be sharp.  I thought I had the camera perfectly still, resting on my railing, but the "Motion Blur" tool made a huge improvement.  I have a feeling that Topaz Sharpen AI is going to become a part of my processing from now on.  I'm not sure if the screen-captured image below will show this properly.

I thought I knew what I was doing, but I'm always learning new things in this forum.  Thank you!!!!  

Time for me to watch the two videos posted up above again.  

One question - if "Motion Blur" gets the best results, is it safe to assume that the problem being corrected is that the camera moved during the exposure?  Both "Out of Focus" and "Too Soft" helped, but not as much as "Motion Blur".  I guess I need to take the same photo again, with the M10 on a tripod, and with me using the self timer or a cable release - and even then, I'm out on my balcony, and I don't know how stable the balcony is.  Maybe I'll put a glass of water out there, and see if there are any ripples.....

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Guest Nowhereman]

On 7/2/2018 at 8:43 AM, tom24 said:

Is it worth it to produce profiles with ColorCecher (X-Rite) or is it better to use the Adobe Standard for General Use - for best colors and general purpous for Leica M?...

I haven't wanted to make my own profiles but have found that there's now a set of new camera profiles that are at least as accurate as custom profiles that you can get made or make yourself. These are certainly more accurate than the Adobe Standard and the M10 embedded profiles in Lightroom. They are being sold by Cobalt-Image as a DNG Base Pack for each camera — and there are also emulation packs for various sets of films and for for various sets of cameras, that run on top of the DNG Packs. In  the thread that I started on this you can see the results here and, if interested, you can go back to the beginning of the thread if you want more information.

The developers of the Cobalt-Image packages state that the accuracy of their profiles is so high because they are based on using 600 samples  (against Xryte SG's 140 samples). I'm very happy with these profiles but don't know whether they are indeed better than what one can do oneself.
_______________________________________
Frog Leaping photobook and Instagram

[Jeff S]

Hopefully Tom has developed a successful workflow in the intervening three years.

Jeff

[Guest Nowhereman]

^ Of course it's obvious that the OP was from 2018 — and that is why I didn't address Tom in my post above: the "you" addresses the current readers of the thread — with the issue I raised being whether profiles like the ones made by Cobalt-Image (with 600 samples, etc.) can be more accurate than profiles that I could make myself for my specific camera. 
_______________________________________
Frog Leaping photobook and Instagram

[Jeff S]

You did address Tom, by quoting him.  We’ve seen your dedicated thread, which is easily linked, of course.

Jeff

[Guest Nowhereman]

^ No, it was not my intention to address Tom: it was only that the OP stated the issue. Anyway, what's the big deal?
_______________________________________
Frog Leaping photobook and Instagram

[Jeff S]

No big deal.  Nor is it a big deal to use a ColorChecker… still simple, quick and relatively inexpensive.  

Jeff

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[jaapv]

21 hours ago, Nowhereman said:

I haven't wanted to make my own profiles but have found that there's now a set of new camera profiles that are at least as accurate as custom profiles that you can get made or make yourself. These are certainly more accurate than the Adobe Standard and the M10 embedded profiles in Lightroom. They are being sold by Cobalt-Image as a DNG Base Pack for each camera — and there are also emulation packs for various sets of films and for for various sets of cameras, that run on top of the DNG Packs. In  the thread that I started on this you can see the results here and, if interested, you can go back to the beginning of the thread if you want more information.

The developers of the Cobalt-Image packages state that the accuracy of their profiles is so high because they are based on using 600 samples  (against Xryte SG's 140 samples). I'm very happy with these profiles but don't know whether they are indeed better than what one can do oneself.
_______________________________________
Frog Leaping photobook and Instagram

The problem with "pre-cooked"profiles is that they can be extremely accurate, but only in strictly defined light conditions. To be completely precise, one still needs to shoot a profile for a specific shoot. That way one can also match the colour output if one mixes cameras. Especially M cameras, with their high iR sensitivity need an ön the spot"profile in some instances.