User Tag List

Thanks useful information Thanks useful information:  8
Page 1 of 2 12 LastLast
Results 1 to 20 of 21

Thread: Soft proofing

  1. #1
    Ausphotography Regular Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)

    Soft proofing

    Following on from this thread (http://www.ausphotography.net.au/for...d-to-Adobe-RGB) I thought I'd post a few observations and questions and see if anyone has anything helpful to add, including good sources of simple info.
    Sorry if it's too long, if you don't want to get your colour/calibrating/soft proofing geek on, then stop reading now.

    This all stems from my research into whether I want to buy an Eizo monitor or not. It seems to me that if all I want to do is work in the sRGB colour space and display on web/screen, then a standard screen like a Dell Ultrasharp will do.
    But if I want to print images then such a screen will not adequately represent the contrast, saturation or brightness of the finished print so using one will involve a certain amount of "hit and miss" re getting a good print.
    Colour will be fine if the screen is calibrated, except that if it doesn't support the aRGB colour space I won't be able to see, and therefore edit, colours outside of sRGB. To be honest the colour space side of things doesn't worry me too much. Yes, sometimes you can get banding or lose subtle grading of tones if you use sRGB instead of aRGB, but generally a calibrated screen will represent the colours well for printing.
    What I'm more concerned about is the fact the monitor may not represent the contrast, saturation and brightness of the final print.

    The reason I ask about soft proofing is that I don't understand why Eizo man says that the only way to see contrast, saturation and brightness as they should be, is to use an Eizo monitor. Why can't I softproof in PS and have the software reduce the contrast, brightness and saturation to levels representative of a print (i.e. pretend to be an Eizo monitor). I started playing with this last night and when I softproof in PS and pick a printer/paper ICC profile, while the image does indeed change, it doesn't seem to change as dramatically as when an image is moved from a mac screen across to an Eizo. So maybe Eizo man is right?

    Has anyone got a simple how to on softproofing in LR or PS that will lead to the screen showing me, and allowing me to edit, what will be seen on print, or is getting an Eizo the only way? Eizo man tells me it is because, for a start, all other screens will be way too bright.

    Given 99 % of my work is on the web it's no good me getting an Eizo screen that shows me what prints look like but not what the image looks like on the web. It's a bit like getting a 4x4 because every couple of years I drive on sand.
    But when I do print I'd like to see what I'm going to get on screen. Do I have to accept that without an Eizo (yes I know there's NEC too) printing will be a little more difficult to get correct? Maybe using a test strip for example?

    As an aside, it would appear that El Capitan will now output in 10 bit (http://flip.it/Rqhsc) , which is something Mac OS has not done before. To be honest whilst the difference in colours between millions, billions and trillions of colours sounds impressive I'm not convinced that the ability for humans to print or distinguish the difference makes it that important.
    But it's interesting that while Eizo advertise their 10 bit/16 bit/whatever LUT capability, they don't mention that if you're a Mac user, you can't take advantage of it...until now.

    Please note that I write the above as what I think I know, but this is all open for anyone to tell my why they think I've got something wrong or misunderstood; or even agreeing would be helpful
    Last edited by Hamster; 02-11-2015 at 3:56pm.
    My Flickr Site
    Instagram _alex_ham_

    Gear - Canon 5D mkIII, 16-35 f2.8L, 24-70 f2.8L, 70-200 f4L IS, nifty 50, 75-300 f4-5.6. Sigma SD Quattro H, Sigma 35 mm Art, Sigma 85 mm Art, Canon G1X MkII, Panasonic Lumix DMC LX3, iPhone.


  2. #2
    Arch-Σigmoid Ausphotography Regular ameerat42's Avatar
    Join Date
    18 Sep 2009
    Location
    Nthn Sydney
    Posts
    16,827
    Mentioned
    21 Post(s)
    Tagged
    0 Thread(s)
    Just a quick Q, Hamster. I can't recall your PS experience and guess it is fairly good, but have you used the PS
    View commands of Proof setup, Proof Colors, Gamut Warnings? (Cited for humble CS2.)
    CC, Image editing OK.

  3. #3
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by ameerat42 View Post
    Just a quick Q, Hamster. I can't recall your PS experience and guess it is fairly good, but have you used the PS
    View commands of Proof setup, Proof Colors, Gamut Warnings? (Cited for humble CS2.)
    Yes am, this was what I was playing with last night. I had one image with saturated greens that was giving me gamut warnings and another with reds etc. The problem was I think I still need to get my head around what is going on exactly when I change profiles. For example I think I was getting gamut warnings because I had an image in the pro photo rgb colour space and this was wider than the colour space of the profiles I was using. But I need to take a look at the pro photo range to confirm.
    The thing that confused me though, was that the soft proof version with the smaller gamut didn't look different to the original, which I can't get my head round. I would expect it to be different because the gamut warning told me it couldn't cope with the gamut of the original.
    I was probably changing too many things at once and not being methodical. More playing required.

  4. #4
    Arch-Σigmoid Ausphotography Regular ameerat42's Avatar
    Join Date
    18 Sep 2009
    Location
    Nthn Sydney
    Posts
    16,827
    Mentioned
    21 Post(s)
    Tagged
    0 Thread(s)
    Hmm! Well, just a thought: you said you looked at the wider gamut before looking at the sRGB one and found them [the same].
    Perhaps it would not have been so unexpected if you had done it the other way: looked at the sRGB first.

  5. #5
    Moderately Underexposed I @ M's Avatar
    Join Date
    04 May 2007
    Location
    Marlo, Far East Gippsland
    Posts
    4,911
    Mentioned
    3 Post(s)
    Tagged
    0 Thread(s)
    Several points from me ----

    For web use I feel that any "good" screen will do an adequate job if calibrated correctly.
    Eizo are no longer the be all and all of monitors. NEC in particular and Dell in general have either caught up or surpassed the edge in performance that an Eizo once had.
    If you want to output your image for printing then a monitor that covers at least 98% of your chosen colour space will be an invaluable assistant.
    The above point only applies if the printer ( either on your desk or a commercial lab ) are able to print the same colour gamut.
    I can't advise on photoshop / lightroom for printing needs, their output seems more and more these days suited to solely web use but that is just my perspective.
    Ultimate print brightness relies on a careful match between what you / I see on our screens in a working environment and what the final print appears as in real life viewing conditions when you hang it on the wall. Typically, what you see on your monitor as far as brightness goes will probably follow when the image is viewed by people on the web. You will probably have to do test prints to assess brightness and adjust your final output for printing to suit the image that you have in your hand.

    My reasoning behind this is ---

    My desired output for all images is the printed form.
    I use a monitor / graphics card calibrator.
    I have established a brightness level that I apply to print images based on experimental prints. That level hasn't changed much over the years.
    If I see a colour difference between that which I see on the monitor and that which is returned to me from the printer I know that it is my fault. Such colour stuff ups have happened twice in the last 5 + years.
    The monitor is calibrated regularly and as of last week the calibration software was reinstalled on a new PC and the set up process run again. To my jaded eyes the colours on the new machine pre calibration were way off, post calibration they appear to be back to normal ( as per previously edited / printed images ) and even though I haven't printed anything this week I am confident that the final result will be fine.

    Checking the monitor info, it has done a tad under 15,000 hours, the calibration does not shift markedly and it doesn't look like quitting any time soon.
    It was not the top of the range monitor when it was bought, it came at a bargain price secondhand. It isn't even an IPS screen. It does work well and if it were to die I probably will go Dell or NEC hunting before I consider another Eizo ----
    Andrew
    Nikon, Fuji, Nikkor, Sigma, Tamron, Tokina and too many other bits and pieces to list.



  6. #6
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by ameerat42 View Post
    Hmm! Well, just a thought: you said you looked at the wider gamut before looking at the sRGB one and found them [the same].
    Perhaps it would not have been so unexpected if you had done it the other way: looked at the sRGB first.
    Yep, I agree. If I'd gone from a small colour space to a larger one I'd have been less surprised at see no difference. Although I understand that's not always the case. The colour science link talks about using the smallest gamut your image fits in. Otherwise stuff gets spread to thin to fit the wider space (my interpretation :-))

  7. #7
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by I @ M View Post
    Several points from me ----

    For web use I feel that any "good" screen will do an adequate job if calibrated correctly.
    Eizo are no longer the be all and all of monitors. NEC in particular and Dell in general have either caught up or surpassed the edge in performance that an Eizo once had.
    If you want to output your image for printing then a monitor that covers at least 98% of your chosen colour space will be an invaluable assistant.
    The above point only applies if the printer ( either on your desk or a commercial lab ) are able to print the same colour gamut.
    I can't advise on photoshop / lightroom for printing needs, their output seems more and more these days suited to solely web use but that is just my perspective.
    Ultimate print brightness relies on a careful match between what you / I see on our screens in a working environment and what the final print appears as in real life viewing conditions when you hang it on the wall. Typically, what you see on your monitor as far as brightness goes will probably follow when the image is viewed by people on the web. You will probably have to do test prints to assess brightness and adjust your final output for printing to suit the image that you have in your hand.

    My reasoning behind this is ---

    My desired output for all images is the printed form.
    I use a monitor / graphics card calibrator.
    I have established a brightness level that I apply to print images based on experimental prints. That level hasn't changed much over the years.
    If I see a colour difference between that which I see on the monitor and that which is returned to me from the printer I know that it is my fault. Such colour stuff ups have happened twice in the last 5 + years.
    The monitor is calibrated regularly and as of last week the calibration software was reinstalled on a new PC and the set up process run again. To my jaded eyes the colours on the new machine pre calibration were way off, post calibration they appear to be back to normal ( as per previously edited / printed images ) and even though I haven't printed anything this week I am confident that the final result will be fine.

    Checking the monitor info, it has done a tad under 15,000 hours, the calibration does not shift markedly and it doesn't look like quitting any time soon.
    It was not the top of the range monitor when it was bought, it came at a bargain price secondhand. It isn't even an IPS screen. It does work well and if it were to die I probably will go Dell or NEC hunting before I consider another Eizo ----
    Agreed, I&M. I think colour reproduction is not really an issue. And brightness can be addressed. As Ricktas said in that other thread, it's the difference between viewing using transmitted LED light on a screen, and reflected on a print. I'm not going into the whole 200 lux exhibition lighting vs sunlight vs .....though.
    But how do represent the lower contrast and saturation of a print on your screen? So that you can increase both of these for printed media?
    Re Eizo vs Dell vs Nec. I believe the Eizo and Dell panels are the same. Nec is no doubt equivalent but works out a similar cost after buying a compatible calibrator and software (for me), plus there's no local support.
    Dell - having tried on several occasions over the last week to get questions answered by their support, is living up to its reputation for having crap customer/tech support. I can't even get them to tell me if their calibration software is Mac compatible.
    Why no love for Eizo from you?

  8. #8
    Moderately Underexposed I @ M's Avatar
    Join Date
    04 May 2007
    Location
    Marlo, Far East Gippsland
    Posts
    4,911
    Mentioned
    3 Post(s)
    Tagged
    0 Thread(s)
    Re contrast and saturation. I don't have a problem with either in prints. I don't find that I have to compensate for what the screen displays. I feel that when I view the printed image mounted and hanging on the wall it matches what I saw on the screen. I don't add or subtract contrast or saturation to the finished on screen image before it is printed. Maybe I am just lucky.

    It is not that I have no love of Eizo products. I admire the quality and build of them and the reputation behind the image quality they display. I also very much like that Eizo still offer a square format monitor as the wide screen style does not suit the way I like to work with one monitor carrying a stand alone image and the other screen hosting the editing tools. ( pic example below )
    I just feel that to achieve the same quality that I experience now from an Eizo product, should it die and need to be replaced, I can probably source an equal product from Dell in particular and NEC in general at a lower price level than purchasing from the Eizo range.
    The monitor we have is a FlexScan L768 and the current monitors of similar ( although much newer and better ) specs run from $1500 upwards.
    There are some very good Dell and Eizo monitors that will equal or exceed the capability of our old Eizo at a far more attractive price point.

    I have never experienced bad service from Dell but we have only ever dealt with them for 2 purchases and both occasions the experience has been satisfactory. If they are getting a reputation for poor service then that is a worthy consideration if you are handing over wads of cash for costly product.

    If the Eizo range has a product that suits you, the service behind it is good and you are happy with the price then I will be the last person to try to dissuade you from the purchase.


  9. #9
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Thanks for your reply @I & M;. I wonder if the reason that you get correct contrast and saturation is because you use an Eizo which represents these properly. I guess that's what Eizo man would say.
    I was back in the camera shop today, and asked if it was possible to set the device profile as a standard sRGB monitor when using the Eizo. I figure if I can set it to represent a printer and paper combination why not a screen. In this way I could use that setting for every day use for web output and switch to a printer/paper profile for printing. Strangely they had never been asked this question and thought this was unusual as it seems an obvious one. Apparently the Eizo rep is coming tomorrow, and they're going to ask him.

  10. #10
    Moderately Underexposed I @ M's Avatar
    Join Date
    04 May 2007
    Location
    Marlo, Far East Gippsland
    Posts
    4,911
    Mentioned
    3 Post(s)
    Tagged
    0 Thread(s)
    I had not thought about separate profiles until now. So I started thinking.
    Take these thoughts as purely applying to the monitor that I am using.
    The monitor has a "mode" button which switches between presets and those presets are -- text, picture, movie, srgb and custom. There are very visible differences between the modes particularly in brightness, contrast and tones. Several of the presets such as srgb do not allow you to alter the white balance setting and although the colours look ok compared to the "picture" mode that I use as the base for editing which is also the mode used when calibrating, the contrast and detail look "flat" when viewing in srgb mode.
    I feel that once the monitor is set by the calibration software, altering the screen presets will simply give you a different set of tone curves to begin with.
    I don't understand all the science behind monitors and calibrating so hearing what the Eizo rep has to say about it all will be enlightening.

  11. #11
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Yes, I'm keen to hear it too. My assumption is that I've missed or misunderstood something. But I can't for the life of me work out why an Eizo can't be made to represent a "normal" LED s RGB monitor and a normal monitor can't be made to represent a printer/paper output just as well as an Eizo (given the correct profile to translate). This second one seems to be a definite no according to Eizo man, but then it would be, wouldn't it :-)

  12. #12
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    I had a chat to Steve Hamilton over at Eizo who called me back and was very helpful. Basically you CAN set up a profile for web work using an sRGB profile, colour temp of 6500 etc. along with a couple of commonly used printer/paper combos and flick between them. When I asked about why a Dell couldn't emulate a printer/paper combo as effectively the answer seemed to be more around the even/consistent reproduction of tones, saturation and brightness rather than anything. So I think basically the answer is, that if you buy a wide gamut screen it will do the job, probably to a level that most will find adequate. And even then you can fix issues with some test strips/prints.
    I'll probably get an Eizo soon, mainly because dealing with Dell over the last week has been painful, and I can be confident I'll have everything set to perfection, it's excellent local support.

  13. #13
    A royal pain in the bum! arthurking83's Avatar
    Join Date
    04 Jun 2006
    Location
    the worst house, in the best street
    Posts
    8,188
    Mentioned
    2 Post(s)
    Tagged
    0 Thread(s)
    My experience:
    I've only printed small 6x4 test images at home on the various printers I've had. They were more for the purpose of testing the print quality to assess it's durability to UV exposure over time.
    I still have those images(actually one was printed A4) .. and they have faded a fair amount being in a well lit kitchen facing the window, but with never any direct exposure to the sun.
    Due to the angle of the kitchens' window, light was always filtered in, or reflected. But those prints have still faded.

    Anyhow, I have had other prints made, some large 30" jobs and my calibration tester 12x10 randomly selected print.
    The 12x10 was printed a the local office works only for the purpose of giving me one hard set lowest common denominator comparator to judge my calibration against.
    It never comes out of the envelope and drawer it currently sits in.
    Even tho I do calibrate my image, and it can be calibrated well enough(for the human eye) .. I still compare the calibrated screen against this 12x10 print.
    It has a good range of tones from contrasty to less contrasty and colours to make an informed assessment.

    So with that image by my side, I then make any larger higher quality prints, so far only 3 at 30" .. and all given away(or asked for).
    Those 30" prints were done by a reputable lab where the staff preferred the use of 16 bit tiff files saved in aRGB.
    My standard workflow is to use sRGB all the way from raw file to saved file. But for those specific 30" prints, I did convert them to aRGB in the raw format before conversion to the tiff file.
    My software of choice tho is CaptureNX2, and I know little to nothing of Ps to compare with that.
    Only in one situation did I convert to aRGB(remember raw file!), and I had to re edit the image. That image was predominantly green, and on switching from sRGB to aRGB, the green in the aRGB version of the raw file looked strange. I remember it either looked more blue .. or more vivid or something(we're talk'n 2008 here .. and I can barely remember my name let alone what I did 8 years ago! )
    But this moment I remember just because I had to somehow change the unexpected effect.
    Note my screen is a typical 8 bit sRGB(barely!) type and graphics card was only 8 bit capable too. So, I technically cant' see the aRGB colourspace either.

    I suppose on a technical note, when you convert from one colourspace to another you really shouldn't see any shift in the colours/tones unless you have push processed the image a fair whack!
    This is where the benefits of the wider colour spaces comes into it.

    Imagine it this way. If you have a very well exposed aRGB image, and it didn't need much processing .. and the colour range was within the sRGB gamut, then changing from the aRGB gamut to the sRGB one should not have any visual impact!
    If the aRGB file was only just out of the sRGB gamut, then you may see the difference .. but again the point of conversion is that it maps one colour/tone to a new one in a translation process.
    That is, it supposed to mask the tonal change .. so you really shouldn't see any.

    I tried to show some images in that other thread, but not only got distracted, but now I can't replicate the changes as I saw back in 2008 on that (mostly)green 30" print.
    I did see some strange effects tho, and this was due more so to the software used to display the images.

    on one image, which was predominantly blue and red(in fact only blue and red), there was a change in the brightness level of the blue in going from sRGB to aRGB which was easily seen in all software I have.
    But a green image I had as the test file, showed weird results:
    In ViewNX2, I can see the change in the green-ness of the green, but only in the thumbnail of the image in the film strip. The actual file rendered in the main edit window hasn't changed!
    (that's weird!)
    In CaptureNX2, there is no change in the way the file is rendered.
    In FastStone Viewer, I see no change either.
    But in XnView I can see the change.
    The method I used to notice the change in the green-iness of the sRGB vs the aRGB was that I created two jpg files set to those colourspaces.
    Only in XnView can I see the actual change in green tone when flicking from one file to the other.

    But the idea of the colour spaces and the ability to convert from one to the other is that we should really see no change unless the image has some wild and impossible colour range.
    Which brings us back to the point of heavy processing of the image. if you push an aRGB image to the edges of the colour tone range, (mainly in the green!) then convert to sRGB, then obviously that conversion to sRGb will have some anomalous effect place on it.
    Think of it in terms of numbers.
    If the machine understands the numbers from 1-100(ie, not 0 or 101), how do you describe 0 or 101 to that machine .. it will have a hard time trying to understand it.

    So if you don't see a change in converting(and make sure you do a conversion and not assigning a different colourspace!) that probably only means that the colour tone range in one matches or is within the bounds of the other colourspace ..

    Again, going back to the man from the reputable printing service. His advice was that while an aRGB is great, as long as you don't overly process your images, you shouldn't have issues with colour tones or gradients. Colours and tones may not be exact matches, but most human eyes would hard pressed to detect the differences.
    But it's the heavy processing effects that you won't really see.
    Again if you think about it logically it makes sense:

    aRGB file .. you push it hard, say you have captured a pale yellowy green tone and want to push it to a more vivid green saturated that renders much darker.
    Pushing that will be ok as long as you do so on an aRGB capable screen .. you should see the effect of any posterisation or other weird colour effect as you process.
    Do that on an sRGB screen and obviously this is where you can't see the effect of the anomalies that may present themselves
    But, if you work in the sRGB space and push as hard as you like in the processing phase, you should see the effects easily .. your monitor is matched to the working colourspace.
    So why the need for aRGB? .. easy! That's what the man at the shop told me to do! (it's really that simple for me).
    I guess their workflow is done in batches, or is predefined ... so they just want consistency or something.

    Why do I see these strange colourspace conversion effects in the image I recently tested(ie. the weird unchanged preview image but re-rendered thumbnail in ViewNX2 situation .. etc) I have no idea.
    I'll see if I can post some screen shots that work to show the effect to others.

    Lastly, on the topic of 10bit Mac OSX. This is good, but only if the hardware is 10bit capable.
    That is, if your Mac has this OSX that now does 10bit colour, but any of the hardware can't(ie. graphics/screen) then it's no good to you.
    You need 10 bit hardware for that to be an improvement.
    So screen and graphics card need to be.
    All this news is about, is an announcement that they now have the drivers for 10bit colour compatibility.

    There's other gotchas about so called 10bit colour stuff that folks should be aware of .. but I've already blathered on for too long. If you're thinking of acquiring any new hardware to take advantage of all this, just be sure to do some in depth research on it all.
    Nikon D800E, D300, D70s
    {Nikon} -> 50/1.2 : 500/8(CPU'd) : 105/2.8VR Micro : 180/2.8ais : 105mm f/1.8ais : 24mm/2ais
    {Sigma}; ->10-20/4-5.6 : 50/1.4 : 12-24/4.5-5.6II : 150-600mm|S
    {Tamron}; -> 17-50/2.8 : 28-75/2.8 : 70-200/2.8 : 300/2.8 SP MF : 24-70/2.8VC


  14. #14
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Some good stuff there ak83, thanks. I think that all makes sense to me, except the mix of visible and non visible differences between programs and colour spaces. I can't explain that either.
    I have an image with lots of saturated greens that I get out of gamut warnings for in PS (blues too), maybe this is a similar phenomenon, but I can't understand why I get these warnings when I select an aRGB destination, if the raw file is aRGB. it may be because I've set PS to the prophotoRGB space and it has stretched the RAW up to this and is then trying to fit it back into aRGB. I just haven't sussed out the process.
    Re the 10bit, I agree. The GT750m card in my MBP has always been able to output 10 bit but the OS hasn't supported it until now. I just find it amusing that Eizo, for example, have been selling the greater than 8 bit capability of their monitors when all their Mac using customers (which must include a fair amount of pros) have been unable to use it. So, as you said, everything needs to line up.

  15. #15
    A royal pain in the bum! arthurking83's Avatar
    Join Date
    04 Jun 2006
    Location
    the worst house, in the best street
    Posts
    8,188
    Mentioned
    2 Post(s)
    Tagged
    0 Thread(s)
    VNX2_aRGB_153-210-106.jpgVNX2_sRGB_121-212-97.jpg

    I tried to turn those two images into a looped gif, but couldn't get it small enough to upload to AP, and had trouble linking to the gif too.
    Sorry for the extra low quality .. but what I was referring to can clearly be seen in the ViewNX2 window.
    The three leafy green images are all the same primary raw file and then converted to jpg in both aRGB and sRGB. sRGB is the last file in the film strip.
    You can see that the thumbnail is more green/blue looking. If you could cycle through the two images over each other, in the preview window nothing changes(but the thumbnail still shows a different rendering! )
    If I set the mouse cursor to a location in the image(in the preview pane), it shows an XY coordinate detail with RGB brightness levels at that point.
    When I screen capped those two images, the mouse icon was set to XY coords 481,434(you can't really see it well, but it is given in the info bar above the image, just to the right of the yellow [i] info indicator.
    For the aRGB image at that coord RGB point levels were 153,210,106(respectively), and for the sRGB image they are 121,212,97.
    So while the preview image clearly doesn't display differently, according to the pixel data at that point(and all other points!) they are different.
    i.e. .. the image should display more green-blue than it does. Technically, it should display less red(ie. down from 153 to 121).

    What I'm thinking is: because my screen can't display aRGB, you'd expect not to see a difference in rendering when trying to view the aRGB version of the file!
    Question is, why can I see a difference in the thumbnail only!

    And then comes the display shown in XnView:

    XNV_aRGB.jpgXNV_sRGB.jpg

    While it's probably a bit more difficult to see the differences, they can be seen. I saved the images to as small a file size to create animated gifs, but the gifs were still too large to upload, and the gifs show the differences clearly even at those low resolutions.
    But now XnView shows both thumbnail and preview as you'd expect it to look, considering the colourspace differences.

    CaptureNX2 and ACR/Bridge/Ps(that I had for a while) both didn't show any differences in the way the two different images were rendered.
    Neither does FSViewer.

    So I dunno what's going on. I doubt that the green tones are 'out of gamut' in the aRGB file, barely any processing was done to it.
    It was chosen because it was the most recent and easily found green looking image I had to play with. As I said in the other colourspace thread, where sRGB and aRGB differ is more in the green/cyan colour range, so this is where you should see the most differences.
    And speaking of that, in the list of images, there are three red/blue(glowing rock) images.
    Those three are the predominantly red/blue images I tested with that I mentioned. The blue in those images changed with just the change from sRGB to aRGB but one was done whilst in the jpg state while the other was done in the raw file state.
    So that is, the image was already set to my default sRGB mode, and I converted both the jpg and raw image to aRGB mode to see if there is any difference.
    There was a slight drop in the brightness of the blue channel in the raw file conversion.
    The two images(of that red rock series) on the LHS are an sRGB jpg made from a raw file, and an aRGB jpg made from a jpg file(the one on the RHS is the aRGB made from a raw file).

    When cycled between the sRGB(left) and aRGB(middle) images, there are no differences at all in the way it renders .. even tho one is sRGB and the other aRGB, and that's what you'd expect to see(even on an aRGB capable system!) It would map the different colours to the correct values in the aRGB file, but in reality those aRGB colours can't exist as they've never existed in the first place(sRGB is smaller than aRGB) hence trying to increase it's perceived quality was a waste of time.

    For more info and a good downloadable file for assessing true 10 bit image capability there is a 10 bit graduated file available for download in psd format.
    As I no longer have use of the trial version of Ps/ACR etc I can only view it via FSViewer and XnView at the moment.
    The file and webpage is HERE at Imagescience
    the file is called 10bit test ramp(somewhere at the lower end of the page, in a zip format and when you unzip it it's a psd file.
    Because my PC runs only in 8bit mode only .. I see the graduated grey image as a series of step wedges and not a smooth gradient as it would be in a true 10 bit environment.
    If your entire computer system, from OS to VGA card to screen, is true 10bit capable you should(in theory) see it as a smooth graduated greyscale image and no posterisation or step wedges.
    It's a good test file to assess your system's ability for 10 bit mode. (note that the file is 1mb or so as a zip and about 6Mb when extracted).

    On the topic of Dell, I've had(not myself personally, but via assisting others) both good and bad.
    Helping my ex mother in law, I put in a claim for a faulty RAM module. With no questions asked, they sent a new one next day with a return shipping bag for the old one!
    A while later when trying to get a PC for my sister, nothing but trouble in getting the order there. after over two weeks of not getting the PC delivered to her, she cancelled the order.

    I've been looking to get a new high quality monitor for a few years now(but keep getting distracted along the way ).
    My image stuff is not critical enough to warrant rushing into it tho, so I can wait. I'd probably go with a Dell, but with a keen eye on their service.
    I've seen I@M's Eizo screen and it looks very nice, especially when I've got home and compared it to mine .. and I'd love one, but they are serious money for something I'd never really make full use of(ie. value for money). My other options are a high qulaity Samsung or LG screen(afterall, it's usually either Samsung or LG that actually make the panels for them all! )

    hope that helps(but note that colourspaces still make my head spin!)

  16. #16
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by arthurking83 View Post

    (but note that colourspaces still make my head spin!)
    You and me both But your post is helpful.

    I'm wondering if your statement of " What I'm thinking is: because my screen can't display aRGB, you'd expect not to see a difference in rendering when trying to view the aRGB version of the file!"
    is the problem I'm having. i.e. some changes, or lack of changes are due to the fact I'm trying to see the effect of different colour spaces on a screen that is limited to sRGB.
    Which maybe where the gamut warning steps in. If that is to show which colours in your colour space are not going to be reproduced effectively in the colour space of your output device then great, I can understand that. What I can't understand is why I get warning of some small areas being out of gamut when I have an image in aRGB and I proof for a ProphotoRGB space. aRGb fits entirely in to the ProphotoRGB space, so there cannot be any colours in the image that require a gamut warning. Another confusing thing is that when I then proof for an aRGB space (i.e. I am going from aRGB to aRGB) I get no gamut warning (so far so good, this I expect) but I do get a change in the appearance of the image. The colours look less saturated in the proof, and yet it's the same image in the same colour space. Note - I'm not proofing for an aRGB paper/printer combo via an ICC profile, I'm just simulating an aRGB device.
    All the above is using relative calorimetric not perceptual as the rendering intent.

    Re screens, I'm pretty similar to you on this. Other things pushing me away from Dell is that I can't seem to easily find out about the software development kit that is used for calibration, plus I would need to spend $300 on an i1pro calibrator from x-rite (as the colormunki display I have isn't supported). I will be pushing myself to print more and enter competitions next year, so I want to be sure I have the best solution, but I do think the Dell is probably better value hardware wise. The extra expense for the Eizo would cover the lower hassle factor I reckon (for me at least)

  17. #17
    A royal pain in the bum! arthurking83's Avatar
    Join Date
    04 Jun 2006
    Location
    the worst house, in the best street
    Posts
    8,188
    Mentioned
    2 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by Hamster View Post
    .... What I can't understand is why I get warning of some small areas being out of gamut when I have an image in aRGB and I proof for a ProphotoRGB space. aRGb fits entirely in to the ProphotoRGB space, so there cannot be any colours in the image that require a gamut warning. ....
    Until I get a full 10bit system and workflow .. can't even begin to help solve something like that.
    Have you downloaded the 10bit test ramp file?
    This should help you determine just how 10bit(ie. aRGB) your computer is.

    Re screens, I'm pretty similar to you on this. Other things pushing me away from Dell is that I can't seem to easily find out about the software development kit that is used for calibration, plus I would need to spend $300 on an i1pro calibrator from x-rite (as the colormunki display I have isn't supported)... [/QUOTE]

    Had a quick peek at imagescience the other day and they have a reasonably priced NEC screen for about $2.5K. It's not what I'd prefer(a 4K 10 bit 30" screen!) .. but anyhow ...

    Re the Dell screens: any calibrator will work on those screens without any issue. Where the iPro is handy to have is that it can 'hardware' calibrate the screen, so it's independent of any software to load the LUT into the VGA card.
    This is supposed to be ideal in a few ways, but not really a necessity!
    I think the major advantages are that you don't have cluttering software to set the screen calibration during bootup into the OS .. both time and resource usage here.
    You can effectively use that screen on any uncalibrated other computer and know that the screen is still well calibrated .. etc.

    I wouldn't discount the Dell solely on that basis anyhow .. I'd probably plan it so that at some point in the future, getting the iPro calibrator will help .. but you wouldn't be limited with what you have now.

    Speaking of which, I have a Spyder 3, and while I know many folks(here and elsewhere) are more than happy to run the Spyder software, on my PC it was flaky, dodgy, unstable, unreliable, miserable, and I think in the end inaccurate. In the end I trialled and then purchased BasICColor for calirating my screen. Does a good job, is almost instantaneous and seamless in loading the calibrated rendering. On Win10 tho it does now take a sec or two to finally load up, where on Win7 it was loaded before the desktop started to respond.
    But I actually prefer the way it now works on Win10, as I know the LUT is being loaded up.
    The only reason I use a desktop image is to view some of my (known) images as a sort of reference point in terms of calibration points.(Otherwise I hate them(desktops with distracting images).
    On Win7, you had to assume that the calibration software was doing its thing. You never got to see the screen rendering change which basically confirms calibration.
    But now that the LUT is being loaded just after the desktop settles(so an image is on screen) .. I now see it go from a looking image of horrendous proportion, to what it's supposed to look like

    Out of gamut warnings .. can't really offer any advice or help .. as in I have no experience with it.

    While you say that your computer is 10bit compliant, I've been reading up on the topic and there is a fair amount of conflicting info about whether there is or isn't.
    It's not as clear cut as there is, or there is not!
    Even tho some nvidia cards are 10bit capable, they aren't all of the time, and in some instances a true 10bit capable VGA card is needed(eg. an nvidia Quadro card or similar).
    But I know even less than nothing about the Apple world of hardware software and where to search for reliable info about the environments they create.

    eg. (in a Windows environment) You can have a 10 bit capable nvidia GT750 card But apparently many applications(Adobe's included) don't actually use 10bit mode with any VGA card(even tho it's 10 bit capable, and set in this way, because of the way the software's API is written. Nvidia's website has a bit of info about it.
    And also, you may have a 10 bit capable card, and 10 bit capable software, but if you run Win7 with the normal Aero theme(the default), 10 bit mode is disabled.

    A lot of the info found is a bit old, not much between '14 and '15, so maybe al the modern(new) graphics cards/chips now run 10 bit without restriction?
    Maybe Win 8 to Win10(for us Win .. gers!) runs natively without restriction in 10 bit mode too?
    You'd kind'a hope that these software-hardware conflicts are slowly becoming a thing of the past.

  18. #18
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    Quote Originally Posted by arthurking83 View Post
    Until I get a full 10bit system and workflow .. can't even begin to help solve something like that.
    Have you downloaded the 10bit test ramp file?
    This should help you determine just how 10bit(ie. aRGB) your computer is.

    Re screens, I'm pretty similar to you on this. Other things pushing me away from Dell is that I can't seem to easily find out about the software development kit that is used for calibration, plus I would need to spend $300 on an i1pro calibrator from x-rite (as the colormunki display I have isn't supported)...

    Had a quick peek at imagescience the other day and they have a reasonably priced NEC screen for about $2.5K. It's not what I'd prefer(a 4K 10 bit 30" screen!) .. but anyhow ...

    Re the Dell screens: any calibrator will work on those screens without any issue. Where the iPro is handy to have is that it can 'hardware' calibrate the screen, so it's independent of any software to load the LUT into the VGA card.
    This is supposed to be ideal in a few ways, but not really a necessity!
    I think the major advantages are that you don't have cluttering software to set the screen calibration during bootup into the OS .. both time and resource usage here.
    You can effectively use that screen on any uncalibrated other computer and know that the screen is still well calibrated .. etc.

    I wouldn't discount the Dell solely on that basis anyhow .. I'd probably plan it so that at some point in the future, getting the iPro calibrator will help .. but you wouldn't be limited with what you have now.

    Speaking of which, I have a Spyder 3, and while I know many folks(here and elsewhere) are more than happy to run the Spyder software, on my PC it was flaky, dodgy, unstable, unreliable, miserable, and I think in the end inaccurate. In the end I trialled and then purchased BasICColor for calirating my screen. Does a good job, is almost instantaneous and seamless in loading the calibrated rendering. On Win10 tho it does now take a sec or two to finally load up, where on Win7 it was loaded before the desktop started to respond.
    But I actually prefer the way it now works on Win10, as I know the LUT is being loaded up.
    The only reason I use a desktop image is to view some of my (known) images as a sort of reference point in terms of calibration points.(Otherwise I hate them(desktops with distracting images).
    On Win7, you had to assume that the calibration software was doing its thing. You never got to see the screen rendering change which basically confirms calibration.
    But now that the LUT is being loaded just after the desktop settles(so an image is on screen) .. I now see it go from a looking image of horrendous proportion, to what it's supposed to look like

    Out of gamut warnings .. can't really offer any advice or help .. as in I have no experience with it.

    While you say that your computer is 10bit compliant, I've been reading up on the topic and there is a fair amount of conflicting info about whether there is or isn't.
    It's not as clear cut as there is, or there is not!
    Even tho some nvidia cards are 10bit capable, they aren't all of the time, and in some instances a true 10bit capable VGA card is needed(eg. an nvidia Quadro card or similar).
    But I know even less than nothing about the Apple world of hardware software and where to search for reliable info about the environments they create.

    eg. (in a Windows environment) You can have a 10 bit capable nvidia GT750 card But apparently many applications(Adobe's included) don't actually use 10bit mode with any VGA card(even tho it's 10 bit capable, and set in this way, because of the way the software's API is written. Nvidia's website has a bit of info about it.
    And also, you may have a 10 bit capable card, and 10 bit capable software, but if you run Win7 with the normal Aero theme(the default), 10 bit mode is disabled.

    A lot of the info found is a bit old, not much between '14 and '15, so maybe al the modern(new) graphics cards/chips now run 10 bit without restriction?
    Maybe Win 8 to Win10(for us Win .. gers!) runs natively without restriction in 10 bit mode too?
    You'd kind'a hope that these software-hardware conflicts are slowly becoming a thing of the past.
    no, I'm still to download that file, but I'm guessing that since I only have an sRGB screen I may be limited in what I see.

    Thats a good point re the calibration. Agreed I can software calibrate, but I was thinking I'd hardware calibrate, hence the ref to the i1 pro. I've been looking at the cs270 at about $2k as the Nec seemed to work out slightly more by the time I'd bought Spectraview (can't remember if I needed a different calibrator too). I'll have another go to see how Mac compatible and easy the hardware calibration of an ultra sharp is.

    I haven't delved too deeply in to 10 bit capabilities, just noted the link in a previous post, which contradicts what Apple Store staff said. Not too worried about it really given all the Mac using pros that have been managing fine without it for years. By the time that software conflict disappears, no doubt there'll be a new generation of conflicts
    Last edited by Hamster; 11-11-2015 at 1:00am.

  19. #19
    Member
    Join Date
    01 Dec 2010
    Location
    salisbury
    Posts
    566
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    SOFT PROOFING PHOTOS & PRINTS by Cambridge in colour,has a good article covering this subject in depth.

    Jack.
    Pentax K5iis, k7 plus lenses from 18mm-600mm.

  20. #20
    Ausphotography Regular
    Threadstarter
    Hamster's Avatar
    Join Date
    20 Feb 2012
    Location
    Perth
    Posts
    908
    Mentioned
    7 Post(s)
    Tagged
    0 Thread(s)
    That's it, I give up


Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •