Leica M8 and Pro photo color space

[woodyspedden]

Advertisement (gone after registration)

FWIW, when I was first working with the DMR and the M8, I actually wrote to Joseph Holmes about what the best colourspace he recommended based on some of the profiling I did... and on the bit depth of the Leica files.

 

And based on what Joseph wrote back to me, Jack is right on the money: DCAM 4 or even DCAM 3 is what he recommended and what I normally use for capture spaces in C1.

 

The utility of DCAM 3 is that it's a pretty darned good print output space as well for most printer profiles.

 

See, what some people don't realize about a tremendously large colourspace like ProPhoto is that while it's good for archiving, and maybe even necessary for some editing of incredible saturated colours, it's not, unfortunately, usually the best output profile for printing, since it arranges, for instance, primary colours in very weird places in the profile.

 

So you can actually get a print from a wide colour space that the colour management engine, print driver or RIP output can't handle (not to mention inks!), and important colours end up being clipped on output (the colour can end up like mud). IOW, you need to know exactly what you're doing with really wide spaces.

 

Now fortunately Photoshop is excellent about transforming (converting) from one colourspace to another; other applications though--including print driver applications and profiles--and YMMV is an understatement.

 

So as a matter of course, unless you have your own print profiles and routines, you should generally convert from ProPhoto into some more normal colour space for printing. There is simply not point in trying to make the printer print stuff it can't physically print.

 

This is why, if you know where you're printing and to what, a narrower space is often best.

 

So DCAM 3 is a nice tradeoff for minimal conversions. However, if I was prepping for a traditional CMYK pres, there are colours in DCAM 3 that would still be out of gamut.

 

BTW--Joseph also sells a DCAM 5 profile, which, IIRC, is actually wider than ProPhoto Again, for archiving purposes, it's probably the best out there.

 

But your output profile differs from the capture profile, and that really does depend on the content.

 

Workflow wise, for me= camera profile-->capture profile-->print profile (where print can be screen or actual paper profile)

 

EDIT: what Jack says about modern printers is correct: their spaces are, particularly in some colours, often larger than aRGB. Heck, modern Epson printers aren't really even CMYK, they're CCMMyKKK (and growing--don't the new ones have even more cyans and magenta? That's all to print blues, you know )

 

Anyway, this is why a custom output / paper profile can be so important. But you still don't want a wider profile than necessary when printing.

 

Last November I was fortunate to attend the Digital Printing Summit where Joe Holmes was the guest instructor and spent loads of time on colorspaces. If I could give a one liner bottom line it was that you should use a colorspace that is large enough to hold all the necessary colors but not much larger. So he would never recommend his DCam5 for other than archiving purposes. For those of us who were experimenting with modern sensors e.g. Kodak sensors used in DMR and M8, we found Dcam4 to be nearly ideal. We were printing with the Canon (9900 I believe) and the first generation new HP vivera ink based 13x19 printers and differences could be seen when sending different color spaces. Not much science on our parts (though lots of previous science from Joe) but our eyes told us the story. So I have now standardized on DCam4 until better sensors and printers demand something new.

 

Woody Spedden

[Eoin]

Jack, well you did edit your post after I posted and perhaps that's where I miss read. Sorry!

 

Sincerly thank you for putting flesh on your comments and now I understand where your coming from.

[Advertisement]

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

[Jack_Flesher]

If I could give a one liner bottom line it was that you should use a colorspace that is large enough to hold all the necessary colors but not much larger. ~~~ For those of us who were experimenting with modern sensors e.g. Kodak sensors used in DMR and M8, we found Dcam4 to be nearly ideal.~~~ So I have now standardized on DCam4 until better sensors and printers demand something new.

 

 

Well said Woody, and precisely the right reasons to be using it!

 

One more point of clarification -- Joe's DC-4 space is very (very) close to Prophoto in total size and gamut. ProP has a tiny bit more magenta and DC-4 has a tiny bit more green, and both should be used in 16-bit mode. The main benefit to using DC-4 is his proprietary tone curves better match his variant sets for DC-4 than his Prophoto sets do. I personally use Prophoto because it is more standard from an archiving perspective. Except for how his saturation variants get applied, the real-world differences in the two spaces are nominal.

 

Cheers,

[Eoin]

Re Aperture, the points are 1) it remains unclear how Aperture is (or even if it really is) converting to Prophoto, and 2) the fact Aperture currently does NOT allow one to convert to a custom space like DC-3 or 4. That's all we are saying. In that respect, Aperture is not a full-featured converter while C1 is.

 

I was curious about this and went and downloaded his free ekta space profile. I installed the profile in both colorsync and for Adobe and exported from aperture a 16bit tiff in the ekta space. I opened it up in CS/2 which reports the profile as ekta space.

 

The same applies if I switch the export profile to ProPhoto and open in CS2 reports it as Prophoto. On screen you can see subtle differences between the two tiffs.

 

Just goes to show with the help of others I can learn something new. I have not invested as of yet in the DC-3/4 profiles but it looks like lowly old aperture can do this.

 

BTW I agree 100% with you that Apple are not very forthcoming on any information as to what is going on under the hood of Aperture.

[jquimby]

colour management is not specific to the M8 !!

but for those who (like me) are old fashioned enough to read books, try:

'Colour Management' by Bruce Fraser, Chris Murphy and Fred Bunting

(I have no connection to the above...)

Guy

 

Just wanted to make sure that this post wasn't lost or overlooked. Bruce Fraser who probably the single greatest source for information on color past away last year. I highly recommend anything he has written no matter how good you are at this process.

 

The only thing more amazing than Bruce's knowledge regarding color and how to manage it was his willingness to share it. For starters here are some links:

 

(disclaimer: i have no affiliation or monetary interest in any of these)

 

In memoriam, Jack Nack, Adobe

John Nack on Adobe: In Memoriam: Bruce Fraser

 

a couple of books and a bunch of free articles on all things color

creativepro.com - Bruce Fraser - Creativepro.com Authors

 

all of his books from his publisher peachpit

Peachpit: Bruce Fraser

[jquimby]

I was curious about this and went and downloaded his free ekta space profile. I installed the profile in both colorsync and for Adobe and exported from aperture a 16bit tiff in the ekta space. I opened it up in CS/2 which reports the profile as ekta space.

 

The same applies if I switch the export profile to ProPhoto and open in CS2 reports it as Prophoto. On screen you can see subtle differences between the two tiffs.

 

Eoin, what you are seeing has been duplicated before. Actually a link to part of the discussion appeared on page 1 of this thread. Here is the summary of the testing, hope this helps.

 

From the Apple color sync mailing list:

 

Source: Re: Working color space of Aperture/Lightroom

 

"One conclusion therefore could be that Aperture uses internally Adobe RGB (or something based on Adobe RGB having the same gamut). If a file with colors outside Adobe RGB is imported it clips it to Adobe RGB since the Lab values it display match those of the gamut of Adobe RGB. However, upon exporting into ProPhoto, Aperture again remembers that the original file had a larger gamut and produces again a file that can fill out the gamut of ProPhoto."

 

-joe

[Eoin]

Advertisement (gone after registration)

Joe, thanks. I think you'll find I originally posted that link on page two, however it makes interesting reading but is pure speculation in the absence of cold hard facts.

[woodyspedden]

I was curious about this and went and downloaded his free ekta space profile. I installed the profile in both colorsync and for Adobe and exported from aperture a 16bit tiff in the ekta space. I opened it up in CS/2 which reports the profile as ekta space.

 

The same applies if I switch the export profile to ProPhoto and open in CS2 reports it as Prophoto. On screen you can see subtle differences between the two tiffs.

 

Just goes to show with the help of others I can learn something new. I have not invested as of yet in the DC-3/4 profiles but it looks like lowly old aperture can do this.

 

BTW I agree 100% with you that Apple are not very forthcoming on any information as to what is going on under the hood of Aperture.

 

Eoin

 

Joe's Ektaspace gamut was optimized for his 4x5 ektachrome trannies. The overall gamut is also significantly smaller than prophoto or DCam4. Either of these will serve you much better than EktaSpace.

 

Woody Spedden

[jquimby]

Joe, thanks. I think you'll find I originally posted that link on page two, however it makes interesting reading but is pure speculation in the absence of cold hard facts.

 

See i knew somebody mentioned on the previous page, should of checked who!! Sorry about that!

 

( so in support of what you said i'm quoting what you said back to you. there is some fine logic there:p )

 

Also i posted this over at apple discussion forum hoping that Aperture product manager Joe Shorr might chime in and shed some light on this. He does pop in from time to time and answer questions

 

you can play along at home here:

Apple - Support - Discussions - Aperture working color space?? ...

[Eoin]

Thanks Woddy,

The only reason I DL that was because it's free, I didn't fancy dropping €72 to see if custom profiles worked. But I may consider the DCam 4 seeing as many seem to be raving about it. If all else fails I could always use it in C1 Pro..

[papimuzo]

Once more, I would like to THANK all of you generous experts giving advices particularly in this thread. What I learned about colour management is incredible (or better I was so dumb on that at least ). I discover what C1 is able to do .

Thanks again Jack, Guy, Jamie,...etc... This forum is really a gold mine for us, average users;) .

[ho_co]

I append my thanks to Marcel's, but I'm below his level of "average user." Still have many questions, but to help me along:

 

Eric--In your http://www.l-camera-forum.com/leica-forum/leica-m8-forum/31019-leica-m8-pro-photo-color-space-3.html#post326251, was that all a quotation from the Aperture manual? It's differently formatted from your other Aperture instructions quotes, but there aren't any quotation marks, so I'm not sure whether that's all Apple talking, or whether at some point you state your opinion.

 

Just dense; sorry. Seems to be clear to others, but I'm having a hard time trying to follow.

 

Thanks!

 

--HC

[wlaidlaw]

Thanks Woddy,

The only reason I DL that was because it's free, I didn't fancy dropping €72 to see if custom profiles worked. But I may consider the DCam 4 seeing as many seem to be raving about it. If all else fails I could always use it in C1 Pro..

 

Eoin,

 

As a free alternative you can try the various versions of the ECI colour spaces, which are supposed to be an improvement/somewhat larger than Adobe RGB 1998. They are downloadable free from various sources but took a bit of finding. I am using, I think, the latest one eciV4. If you look at the diagram, it looks rather like Prophoto. If you can't find them but want them, PM me and I can send you v2 and v2/v4

 

Wilson

[ericperlberg]

I append my thanks to Marcel's, but I'm below his level of "average user." Still have many questions, but to help me along:

 

Eric--In your http://www.l-camera-forum.com/leica-forum/leica-m8-forum/31019-leica-m8-pro-photo-color-space-3.html#post326251, was that all a quotation from the Aperture manual? It's differently formatted from your other Aperture instructions quotes, but there aren't any quotation marks, so I'm not sure whether that's all Apple talking, or whether at some point you state your opinion.

 

Just dense; sorry. Seems to be clear to others, but I'm having a hard time trying to follow.

 

Thanks!

 

--HC

 

Howard,

 

In the first post I separated the quote from Apple's manual for Aperture that I downloaded from the net and then separated my opinion from it with a ____________________.

 

The second and third posts from the Aperture Manual were simply copied from the manual I had downloaded, no opinions.

[sandymc]

I don't think this is quite true. I believe that Aperture and Lightroom (like ACR) apply an internal profile based on the camera (or the DNG matrix in the RAW file) on import.

 

The DNG matrix takes you to the CIE color space (which covers everything the eye can see).

 

As I understand it, for LR the process is firstly use the DNG color matrix to go to CIE color space. Then for displaying to the screen, convert to the intermediate Melissa color space to compute histograms, RGB value displays, etc, and then convert from Melissa to whatever the color space of your screen is to actually display the image. Because Melissa has a broarder gamut than any screen, that isn't a problem. However, when printing or exporting, conversion is straight from CIE to the colorspace of the output.

 

What this means is that regardless of what your display or printer is, the histograms, etc in LR will always be the same, because they always refer to Mellisa space. Otherwise, they would change everytime the output device(s) changed.

 

Again, as I understand it, in Aperture's "Soft proofing" mode, that intermediate space is the output space........

 

Sandy

[ericperlberg]

Andrew Rodney (Color Management for Photographers) explains Lightroom's internal working arrangement: Color Management in Lightroom. (It's a pdf)

[Eoin]

....Again, as I understand it, in Aperture's "Soft proofing" mode, that intermediate space is the output space........Sandy

 

I'm not 100% sure on this, the soft proof (on screen proofing) is a user selected option that will try to display the image in the colour space you select for on screen proofing. However this profile is not automatically set or updated as the output profile. One has to manually set the output space desired in the export presets.Or use one previously saved for the intended output profile.

 

The use of on screen proofing as far as I understand it is to evaluate the effect of say printing / paper profiles and the effect they will or may have on the image output if chosen compared to the image displayed on screen in the normal Aperture working space. One can see posterization in shadows for example with some Ilford profiles that are not present in normal working space screen display. But on output one can set the profile say for printing to be in whatever profile they wish with presets. That's why it's important within aperture to make sure that prior to printing you check your image with a soft proof profile that is the same as your intended preset output profile otherwise what you see on screen may not match exactly what you get from the printer.

[ho_co]

In the first post I separated the quote from Apple's manual for Aperture that I downloaded from the net and then separated my opinion from it with a ____________________.

 

The second and third posts from the Aperture Manual were simply copied from the manual I had downloaded, no opinions.

Gotcha! Thanks for the clarification! All clear now!

 

--HC

[Jamie Roberts]

The DNG matrix takes you to the CIE color space (which covers everything the eye can see).

 

{snipped}

 

Exactly. It's how they get from the DNG matrix to the LAB:CIE edit space that is the 'issue' (it's not an issue at all if you like the rendering). Certainly the differences we've seen over the last months in ACR and Lightroom are happening right there.

[Advertisement]

Hallo,

Look here: Leica M