Zum Hauptinhalt gehen

⚠️ Please note that this topic or post has been archived. The information contained here may no longer be accurate or up-to-date. ⚠️

Output to 16 bit TIFF - embedded camera profile or Adobe RGB

luc1
Hello,

What would be the advantages and disadvantages in Capture One 4.1 of outputting TIFF 16 bit files for local adjustment in CS3 with the following:
1) Camera embedded profile (Pentax K10D and K20D in my case) no conversion to workspace profile in CS3
versus
2) Adobe RGB
3) Prophoto RGB
The final output will be:
sRGB JPEg for web
TIFF to canon printer profile for 8*10 prints

Thank you for your input.

Cheers,
0

Kommentare

16 Kommentare

Datum Stimmen
  • thowi
    [quote="lbenac" wrote:
    What would be the advantages and disadvantages in Capture One 4.1 of outputting TIFF 16 bit files for local adjustment in CS3

    advantages:
    - you save the file in it's native color space (or better: phase ones color space). this color space does not contain less nor more colors as needed.
    - as a consequence you save the file in it's original TRC (gamma... which is gamma 1.8 in the case of phase ones camera profiles)
    - you can adjust the image in Photoshop with regard to another certain color space with the help of color warning
    - you can store the 16bit Tif in your archive. This is more or less a 1:1 copy of the raw file but in a data format that will be compatible to future computer systems, too (in the case of proprietary raw file formats this is rather uncertain)

    disadvantge:
    from my point of view just the bigger file size...

    C1 Output to AdobeRGB:
    very common color space for preproduction of prints.
    AdobeRGB does not store all the colors of the original file. But in most of the cases this might be minor as a capture never fills up the entire color space of the camera profile. So most of the subjects colors captured will be stored in AdobeRGB as well. But again - in some cases not all of the colors.
    Nevertheless you convert the greyscale from gamma 1.8 to gamma 2.2 which may cause some loss of information. In a 16bit workflow this should be minor. In an 8bit workflow the loss caused by quantization is not unimportant.

    C1 Output to Prophoto RGB:
    Why? Okay, if a client want to have a file in ProPhotoRGB send him a copy in ProPhotoRGB. For the rest you can take no advantage of ProPhotoRGB. As C1 always converts the camera profile colormetric to ProPhotoRGB you will store exactly the same colors but in a much bigger color space. And especially this color space ... is a challenge as it contains a wide range of theoretically colors.

    sRGB JPEg for web:
    depends on where you do the final adjustment. If you finalize the image in C1 you can do a second otuput of the image straight from C1. If you finalize the image in Photoshop than you will make a copy in sRGB 8bit over there...

    Finally: if you use color spaces bigger as sRGB (or better bigger as the color space of your display) you will not see all of the images colors! You always should take a look at the histogram, the color values and in Photoshop you should activate color warning.
    0
  • NN8893232
    thowi,

    Thanks for the info, it makes interesting reading.

    Although I have a question.

    You mention outputting a tif in the native camera colorspace to retain colour gamma. Surely if your destination is CS3 for PP, then CS3 will immediatley do a conversion to AdobeRGB (or other) on import.

    So in your view, is C1 better or worse than CS3 for such a conversion?
    0
  • thowi
    [quote="John_Nevill" wrote:
    Surely if your destination is CS3 for PP, then CS3 will immediatley do a conversion to AdobeRGB (or other) on import.
    Depends on the color settings.
    If you set "preserve embeded profiles" (sorry, not quite sure about the correct english terms) for the 3 color modes (rgb, cmyk and grey) and than deactivate color converson for "profile mismatches" in the two options "ask when opening" and "ask when paisting"... than Photoshop will keep the embedded profiles.
    http://home.arcor.de/thowidaten/fc/pscprefs.jpg
    Best Regards.

    edit:
    So in your view, is C1 better or worse than CS3 for such a conversion?
    Basically it's the same. In Photoshop I have the option to use color warning int the proof option... in C1 I have just the histogram. Too, in Photoshop I can use colormetric conversion even in tablebased profiles - in C1 (on Windows!) you can not set different rendering intends (but C1 V4pro will probably have that options). So C1 on windows works with the profile tags (camera profiles are set to perceptual): if the target profile is tablebased C1 uses perceptual RI, if the target profile is a matrixprofile ("display" profiles such as AdobeRGB, sRGB, ProPhotoRGB ... ) C1 uses colormetric as matrix based profile are not able to do perceptual conversion.
    0
  • luc1
    Thank you for your detailed answer. Embedding the camera color space seems to make a lot of sense. I was not doing that until 4.1 as this was broken in 4.01 (output was sRGB instead).
    I like the concept of keeping the color space as is until final output has been decided (sRGB, printer profile) and minimize as much as possible profile conversion.
    This is a little bit easier to do in CS3 with the proofing and warning tool.

    Regarding gamma within C1 (1.😎 and RGB matrix space (2.2 ?), how does it translate when you see the image on a monitor? Mine is calibrated to sRGB (close to 2.2) and within CS3. I am assuming that it make sense to keep CS3 and monitor gamma to 2.2 as it is closer to the final input? or am I wrong there and it would be better to calibrate everything to 1.8?

    Cheers,

    Luc.
    0
  • NN8893232
    thowi,

    I've always used AdobeRGB as the working space in CS3, but what you say makes sense. Afterall my dSLRs probably have a wider gamut than AdobeRGB, so I'm throwing color away.

    Hence, keeping the working space as "embedded camera profile", removes the need to do unnecessary conversions and therefore reduces potential colour loss.

    I'll use this method in future, great tip, many thanks!.
    0
  • thowi
    [quote="lbenac" wrote:
    Regarding gamma within C1 (1.😎 and RGB matrix space (2.2 ?), how does it translate when you see the image on a monitor? Mine is calibrated to sRGB (close to 2.2) and within CS3. I am assuming that it make sense to keep CS3 and monitor gamma to 2.2 as it is closer to the final input? or am I wrong there and it would be better to calibrate everything to 1.8?

    In a color managed workflow the different TRCs are converted. So you will not take notice of the differences in C1 or Photoshop (well... in technical gradations you may see some banding but not in real pictures.. not in a 16bit workflow).
    Basically the TRC of the display should the same as the TRC of the preferred working space. I would recommend to calibrate to gamma 1.8 ONLY if the display is used without exception for pre-press. On the other hand: if you use without exception color managed software (eg. Firefox 3 as webbrowser)... you can calibrate to gamma 1.8... no problem... in C1 or Photoshop this does not make a real difference. Me for example I calibrate to the linear TRC L* although I always keep the camera profile (and in the end I convert to a printer color space which has ~ gamma 1.8, too). But I set ECI-RGB V2 as working space in C1 and as softproof in Photoshop... so I always simulate my "working space" ECI-RGB V2 (which has a "linear gamma") and this is why L* as TRC for the display makes sense for me.

    [quote="John_Nevill" wrote:
    so I'm throwing color away.
    In the particular case that the captured colors of the subject have a wider color range: yes.

    Hence, keeping the working space as "embedded camera profile", removes the need to do unnecessary conversions and therefore reduces potential colour loss.
    Yes... the most serious losses in color conversions are caused by the conversion of the gradtion (TRC/gamma).
    (On the other hand... in pratice a single conversion is not a big deal.)
    0
  • NN8893232
    thowi,

    I did some testing tonight and there's no doubt that I'm getting an out of gamut warning when proofiing AdobeRGB in CS3 from an embedded profile C1 output.

    In fact on my NEC wide gamut 2690 LCD I can actually see a colour shift in the primary and secondary colours. The NEC is hardware calibrated using L*

    I've also proofed the ecRGB workspace and the out of gamut is negligible in comparison.

    Well, thanks again for the insight, I'm now sticking with the embedded camera profile when sending Canon 1 and 5 series raws to CS3.
    0
  • thowi
    [quote="John_Nevill" wrote:
    I did some testing tonight and there's no doubt that I'm getting an out of gamut warning when proofiing AdobeRGB in CS3 from an embedded profile C1 output.
    So in this case the capture obviously contains more colors as AdobeRGB can cover (certainly in dark tonalvalues...).
    I always do it this way: adjustments in C1 very moderate so that I preserve a safe space in the histogram. So it's white balance and maybe some slight gradation and maybe some color noice reduction - but no luminance noise reduction and definitely no sharpening.
    In Photoshop I do all the adjustments on layers (ECI-RGB set as proof color with color warning). So I do adjust with regard to ECI-RGB... but I save the ontouched 16bit in the camera profile on the botommost layer.
    By the way... I forgot to mention a disadvantage... this workflow does not save time and disk space ... the opposite 😊
    0
  • luc1
    I have not used Firefox 3 yet but I will download it tonight (Safari did not work for me as I am on Vista 64 bit).
    I will try to calibrate my monitor to Gamma 1.8 if I can (limited adjustments on Dell 2001FP).
    You mentioned that you setup C1 working space as ECI-RGB? I am not sure how to setup the working space in C1. I assumed that it was fixed by the choice of the output space i.e. Adobe RGB output space will make the histogram and the preview in C1 use Adobe RGB,....?
    How do you know what gamma a printer color space is using? I have a Canon iP6700 with a custom profile but I am not sure where to find that information.
    Thank you for all your help, you sure know a lot about color space.

    Cheers,

    [quote="thowi" wrote:

    Basically the TRC of the display should the same as the TRC of the preferred working space.. On the other hand: if you use without exception color managed software (eg. Firefox 3 as webbrowser)... you can calibrate to gamma 1.8... no problem... in C1 or Photoshop this does not make a real difference. Me for example I calibrate to the linear TRC L* although I always keep the camera profile (and in the end I convert to a printer color space which has ~ gamma 1.8, too). But I set ECI-RGB V2 as working space in C1 and as softproof in Photoshop... so I always simulate my "working space" ECI-RGB V2 (which has a "linear gamma") and this is why L* as TRC for the display makes sense for me.
    0
  • thowi
    [quote="lbenac" wrote:
    I have not used Firefox 3 yet but I will download it tonight
    After installation you have to type in the adress: about:config
    Than look for the entry "gfx.color_management.enabled". Double click on the entry (boolean value has to switch to "true").

    I will try to calibrate my monitor to Gamma 1.8 if I can (limited adjustments on Dell 2001FP).
    manual adjustment or with measurement device / probe?
    Anyhow... in this case - as you have to calibrate via graphics card - I would stay with Gamma 2.2... because it's probably very close to the native gamma of the display. But if you have a probe and a good calibration software you can give it a try.

    You mentioned that you setup C1 working space as ECI-RGB? I am not sure how to setup the working space in C1. I assumed that it was fixed by the choice of the output space i.e. Adobe RGB output space will make the histogram and the preview in C1 use Adobe RGB,....?
    Yes, the color space set as output is the "working space". Actually in C1 V4 there is no need to set any color space if you embed the camera profile in the end. In C1 V3 I set my working space but without changing anything I set "embed camera profile" in the output options.

    How do you know what gamma a printer color space is using? I have a Canon iP6700 with a custom profile but I am not sure where to find that information.
    Open a new file in Photoshop in the printers color space. Create a smooth gradtion / greyscale from black to white.
    Now assign (not "convert to"!) a color space with gamma 1.8... for example ProPhotoRGB. If the gradation is getting brighter significantly... the printer profile is maybe around gamma 2.2. If the greyscale stays just about similar... it's around gamma 1.8.
    In the case of printer profiles it does not make so much sense to talk about "gamma". It's the "builtup of black color" (sorry... really don't know the technical term in english) that is similar to a certain "gamma". For example offset printers (CMYK) have a dispersion of luminance and black that is close to gamma 1.8. Most of the professional RGB-Photoprinters, too. You should not worry about that too much...
    0
  • Pascal Despeaux
    Calibrate your monitor with the 1.8 gamma was the rule in the past, with the CRT monitor. All professional used it with the old color space ColorMatch RGB (5000K gamma 1.8 ). But it's now like a dinosaurus. 6500 K and 2.2 gamma is actualy the rule. With the modern LCD and a good video card if you use a 5000K and gamma 1.8, you destroyed posibility of it to visualize some colors. About a TIFF 16bits, for me the first advantage is if you want to change somethings in Photoshop. Color, contrast, saturation,.... all is more moderate in 16bits. And other point is for the future, I agree that. Now, you can have a problem with the 16bits, because you can not visualize all colors on you monitors.. And at the end you can make some mistakes in optimization. Recently, I had this problem to reproduce some work of art about a painter who likes a very saturate colors. I used a Phase One back and I ceated my own camera profil ( with ProfilMakers ) to optimize the result. At the beginning I developped all the picture on AdobeRGB 16bits but the result was that I can't evaluate the perfect optimization and when I check to print the fine art, that was all the time not perfect.. 🤬 I checked with the spectrophotometer the color of painting and the color in my picture on CS3 and that was nice but not for the print.. At the end I had a really good result to process in AdobeRGB 8bits and print.... And for the printer I have my own paper profil... The histogram on CS3 for the picture in 16bits was funny. 🙄 we could not see the top of the peaks, in made all exceeded the window. Ok, that's special because this painting is very saturate..
    0
  • thowi
    [quote="Pascal D." wrote:
    6500 K and 2.2 gamma is actualy the rule.
    The "rule" is that the whitepoint of the display should look equally to a D50 lightbox (the visible whitpoint is important, not the measured whitepoint). And this is mostly a whitepoint somewhere between 5200K and 5800K. So the "rule" is 5000K - 6000K.

    [quote="Pascal D." wrote:
    With the modern LCD and a good video card if you use a 5000K and gamma 1.8, you destroyed posibility of it to visualize some colors.
    The wider you move away from the native (generic) whitepoint and gamma of the display the more colors you "destroy". Even in hardware calibration. Via graphics card it depends... if you have a display with internal 10bit LUT you can adjust to a certain whitepoint without loosing (too much) colors. With a 8bit display you should stay with its native whitepoint.
    0
  • Pascal Despeaux
    Maybe, we don't speak of the same think about the "rule".... I don't say it's clever or not. I say, in my country 🙄 , 100% of professional use the 6500 K gamma 2.2, today. In the past that was 5000 K or 5500 K gamma 1.8. That's it... And for the native whitepoint of your monitor is more or less 9300 K with a lot of monitor.. But that true that a few years the idea that a good choice was, for the prepress, to use 5000K because it's near the natural color. In fact, when you use this you have a dull and yellow visualization. That's the reason that now a lot of expert and pro- profiler calibrate her clients in this TC and gamma.
    0
  • thowi
    [quote="Pascal D." wrote:
    in my country 🙄 , 100% of professional use the 6500 K gamma 2.2, today.
    That's interesting. Under which lighting conditions? If you use D65 ambient light that might be okay for the display (to eye up prints definitley not as especially photographic papers have a lot of optical bighteners... the higher the color temperature the bluer the paper seems to be).

    And for the native whitepoint of your monitor is more or less 9300 K with a lot of monitor..
    This is why photographers should invest a little more money than anyone 😉

    [quote="Pascal D." wrote:
    But that true that a few years the idea that a good choice was, for the prepress, to use 5000K because it's near the natural color. In fact, when you use this you have a dull and yellow visualization. That's the reason that now a lot of expert and pro- profiler calibrate her clients in this TC and gamma.
    D50 ambient light (and D50 lightbox to eye up prints) and 5000K on the display... sure, the display is yellow. This is why one should calibrate to a visaul equivalent whitepoint. But this definitely not 6500K. I use D50 lightbox and ambient light and the whitepoint is matching around 5600K.
    The TRC ("gamma") is another question...
    0
  • Pascal Despeaux
    [quote="thowl" wrote:
    I use D50 lightbox and ambient light and the whitepoint is matching around 5600K
    .

    That was the rule here in the past, yes. 😉 I was responsible fof a digital retouch service in big photographic laboratory in Paris. And all the station was calibrate at 5500K gamma 1.8. For drum scanner, Kodak LVT printers, and sur retouching station. And I now you can make a good job with this choice of calibration. For the lighting work condition, that is important for me, work on moderate light is the must. I don't say it's better for your eyes 🙄 but better to evaluate colors 😎 And, you are right with this lighting condition TCR are less important because the eyes have faculty to make the white point independently of the adjustment screen. Only important thinks is to optimize the visualization color space on your monitor because you never check really all the colors you have on your file. We need to lose some the least possible. Thus it is necessary to adapt these choices to the possibility of the screen. And, for the moment, if you ask to Eizo or at LaCie it indicates a TCR to 6500K and gamma 2.2. But, with arrived of the screens with a larger space colorimetric the things could change well. 😂
    0
  • thowi
    [quote="Pascal D." wrote:
    [quote="thowl" wrote:
    I use D50 lightbox and ambient light and the whitepoint is matching around 5600K
    .
    That was the rule here in the past, yes. 😉
    Whether there is a rule or not... display and prints are matching.

    work on moderate light is the must
    do you refer to to the luminance level herewith?

    Only important thinks is to optimize the visualization color space on your monitor because you never check really all the colors you have on your file. We need to lose some the least possible.
    In the case that the images color range is really wider as the displays gamut. In the case that the images color range (prepared for print) is inside the intersection of display gamut and print file gamut you certainly can see all of the printed colors.

    And, for the moment, if you ask to Eizo or at LaCie it indicates a TCR to 6500K and gamma 2.2. But, with arrived of the screens with a larger space colorimetric the things could change well. 😂
    Well... in Germany that's different. Eizo recommends a gamma of 2.2, yes. Because the displays are calibrated to gamma 2.2 from production - and even with hardware calibrated displays from a standpoint of best quality in gradation one should stay with the generic gamma of the display. For the whitepoint everyone refers to ISO 12646.
    But - in your particular case - if you measure submittals (eg. artwork like paintings or other products) it's often impossible to measure the subjects colors under D50 conditions. You have to evaluate the colors of the subjects under D65 conditions or maybe in real daylight. But if you return to the display and prepare the images for print ... you have to switch to the standard for print... which is D50.

    Anyhow... it's important that display and prints are matching... if you get best results with 6500K and TRC gamma 1.8 and me with 5600K, TRC L* at a luminance level adjusted to the lightbox (~115-120 candela in my case) ... everything is fine 😉
    0

Post ist für Kommentare geschlossen.