Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: nilo on January 25, 2011, 10:01:30 am
-
Hi,
I played a bit with the color density slider in order to get more color laid down on the paper. I played with the 3880/K3VM and Canson Baryta Photographique.
Before I continue to waste a lot more paper and ink, I would really appreciate your insights and knowledgeable advice.
I noticed that, with this particular paper/printer combination, as you increase the color density, the paper very quickly starts to undulate. Humidifying the back of the paper just before the printing resolved that issue and I could print +50% without any waves. At such levels a heavy loss of details occurs.
What are you settings for various papers? Maybe you share some tricks?
kind regards
EDIT: SO, right now, I am particularly concerned about loss of detail. Do you just dial back the density or adjust sharpening? Or a mix of both?
-
I'd increase the color density in Photoshop then send the print to the printer/paper combo using the normal profile.
-
Randy, this is not what I want to achieve! If I understand you, you suggest to increase saturation/vibrancy and so on, in PS. This increase in PS has no influence on the amount of ink laid down on the paper. More ink on the paper does not have to increase saturation (especially if, like in my case, you profile for that), but should produce "thicker" and "richer" color, with deeper blacks for example.
regards
PS The aim is to get after profiling, the same colors, values an tones just "thicker" and shinier
-
I'm old school, so my approach my be different than most. I find putting the attention into time of capture more important than PP adjustments and printing settings. Starting with good color makes it easier to end with good color. When I'm evaluating a scene for angle, I look not only for compostion, but color elements. Sometimes I'll add something to the scene to punch the color. This is an old slide trick. We used to add something red to Kodachrome or something blue to Fujichrome. With digital, you can use either.
Also, certain lenses simply render richer colors. Certain cameras also produce richer colors, which is one of the reasons I switched to Pentax a few years ago.
A CPL also makes colors richer, but I don't use them often because there usually is a trade-off of detail. Underexposing slightly also makes colors richer. If you have a decent camera and shoot at the lowest ISO, noise is not a problem. Attached is a fall shot I underexposed considerably because the colors were all muted and rather dull. I also bumped the red and blue channels slightly in PP, but only slightly. The less you do in PP the better. When you bump a color channel, for example, it doesn't just effect that color, but the hues of all colors.
For this shot, I also used a GND to enhance the sky, which became an intregal part of the shot. I've had this shot enlarged to 20X30, and it actually looks better on the wall than the dummed down version on this screen.
-
This is not my question, though I agree with you Ron, and in regard to adding color directly to the scene when photographed, it's one of my constant routines ;-)
My question is about increasing color density, the amount of color laid down on the paper and that even WITHOUT increase of overall saturation. PS saturation and/or composition, exposure ... is NOT the topic!
I'm old school, so my approach my be different than most. I find putting the attention into time of capture more important than PP adjustments and printing settings. Starting with good color makes it easier to end with good color. When I'm evaluating a scene for angle, I look not only for compostion, but color elements. Sometimes I'll add something to the scene to punch the color. This is an old slide trick. We used to add something red to Kodachrome or something blue to Fujichrome. With digital, you can use either.
Also, certain lenses simply render richer colors. Certain cameras also produce richer colors, which is one of the reasons I switched to Pentax a few years ago.
A CPL also makes colors richer, but I don't use them often because there usually is a trade-off of detail. Underexposing slightly also makes colors richer. If you have a decent camera and shoot at the lowest ISO, noise is not a problem. Attached is a fall shot I underexposed considerably because the colors were all muted and rather dull. I also bumped the red and blue channels slightly in PP, but only slightly. The less you do in PP the better. When you bump a color channel, for example, it doesn't just effect that color, but the hues of all colors.
For this shot, I also used a GND to enhance the sky, which became an intregal part of the shot. I've had this shot enlarged to 20X30, and it actually looks better on the wall than the dummed down version on this screen.
-
Sorry, there seems to be a confusion about my undertaking:
We are talking about a fully profiled work flow, where I get the colors of the screen to match those of the print.
In this situation we want to increase color density to the maximum possible, before detail is lost and/or the paper undulates too much. At that point we'll profile again. (So, to get back to the previous posts, we could get under-saturated prints with increased color density).
Your knowledgeable experience and insight in this matter would be of great value to me (and, I am sure, a lot others)
kind regards
-
Nino,
I understand your question. You are talking about the option in the epson driver to increase density. There was an article on this site about doing this:
http://www.luminous-landscape.com/essays/in_search_of_the_ultimate_inkjet_print.shtml
The way I do it is two fold.
I bump it up in increments and look at the output and compare. When I see the image get soft I tone it down and find a happy medium.
I also found that I have to increase drying time at each pass when I bump it to high. Epson exhibition Fibre seems to puddle very quickly when you increase ink density...so I played with the settings and watched the printer print it and found the proper time. You can actually see it dry on this paper to get your dry time.
hope this helps
-
Thank you for that link natas!
What are the usual gains in density that you obtain?
regards
Nino,
I understand your question. You are talking about the option in the epson driver to increase density. There was an article on this site about doing this:
http://www.luminous-landscape.com/essays/in_search_of_the_ultimate_inkjet_print.shtml
The way I do it is two fold.
I bump it up in increments and look at the output and compare. When I see the image get soft I tone it down and find a happy medium.
I also found that I have to increase drying time at each pass when I bump it to high. Epson exhibition Fibre seems to puddle very quickly when you increase ink density...so I played with the settings and watched the printer print it and found the proper time. You can actually see it dry on this paper to get your dry time.
hope this helps
-
It really depends on the paper. Last week I got a new paper in from breathing color (Luster paper). I got my ink density up to 20 and stayed there.
I am not in front of my printer right now so my wording maybe off. One thing you also need to do is make sure your paper thickness is set right. Manufactures will give you the canned settings but many times they can be wrong. A perfect example here is for the new paper I got yesterday. The manufacture said to set paper thickness in the driver to 3 but when I did a print test from the printer that shows you the proper thickness I found a setting of 2 was better (via a loupe).
-
I too use an increased ink load (with third-party inks/papers) to get the optimum black level and colour gamut. I increase the ink limt until I start to see blocking of shadows (measured with a spectro), ink pooling, or excessive dot gain (bleeding). This has worked well for me and given me useful improvements.
But. my logic tells me that after profiling with the new ink limit the amount of ink used to produce any given in-gamut colour should stay the same. The profile will "demand" sufficient ink to produce any given colour - no more and no less. If you were to put down more ink for a given image colour then the printed colour would change. So, while you may benefit from increased Dmax and colour gamut there no change in "colour density" until you get to the gamut boundary. Or am I being too simplistic?
Sorry, there seems to be a confusion about my undertaking:
We are talking about a fully profiled work flow, where I get the colors of the screen to match those of the print.
In this situation we want to increase color density to the maximum possible, before detail is lost and/or the paper undulates too much. At that point we'll profile again. (So, to get back to the previous posts, we could get under-saturated prints with increased color density).
-
Or am I being too simplistic?
Sorry, I probably am. If you increase the gamut then colours will be rendered with slightly increased saturation, at least with Perceptual rendering intent.
-
This may prove useful in addressing the color density issue: http://www.on-sight.com/2008/04/04/how-to-determine-the-optimal-media-selection-for-any-paper/ Even though the focus is on media selection for third party papers, it should also apply to increased inking.
-
In addition to the above mentioned visual media testing, you can also make several profiles at different color density settings and compare the results in a gamut graphing application looking for Dmax and gamut volume.
Be aware that there is only one, exact setting that is optimal! Go below that (less density) and you'll get weak blacks and a smaller gamut. Go above that (more density) and you'll get blocked up shadows and a smaller gamut. Increased density does not always equate to increased saturation! When we graph these two things out we find that they increase together for a while until saturation plateaus and then starts to fall off while density continues to increase.
Whatever you do, use the same Color Density setting that the profile was made with. If you're wanting increased print saturation it's better to apply that to your images at the application level prior to printing.
-
This may prove useful in addressing the color density issue: http://www.on-sight.com/2008/04/04/how-to-determine-the-optimal-media-selection-for-any-paper/ Even though the focus is on media selection for third party papers, it should also apply to increased inking.
One more interesting link! Thank you Alan!
regards
-
Increased density does not always equate to increased saturation! When we graph these two things out we find that they increase together for a while until saturation plateaus and then starts to fall off while density continues to increase.
Whatever you do, use the same Color Density setting that the profile was made with. If you're wanting increased print saturation it's better to apply that to your images at the application level prior to printing.
I agree...
Also note that increasing or decreasing the ink density when making a profile target will tend to get the increase/decrease profiled out. There is an optimal settings and that will work the best for making profiles.
With Epson paper, it's really not useful to do anything with ink settings other than subtle tweaks...with 3rd party paper, it is useful to work the settings to optimize the ink density prior to making profiles because 3rd party papers don't necessarily fall into the Epson Media settings exactly. But simply increasing the ink density to pump more in onto the paper will not always be useful (and often causes more problems).
-
[...]
Be aware that there is only one, exact setting that is optimal! Go below that (less density) and you'll get weak blacks and a smaller gamut. Go above that (more density) and you'll get blocked up shadows and a smaller gamut. Increased density does not always equate to increased saturation! When we graph these two things out we find that they increase together for a while until saturation plateaus and then starts to fall off while density continues to increase.
[...]
I suppose you go through this procedure when testing new paper. Before doing such a series of tests one can't really asses the gamut and other capabilities of a given paper-printer combination.
(Did you try with Baryta Photographique?)
regards
-
[...]
Also note that increasing or decreasing the ink density when making a profile target will tend to get the increase/decrease profiled out. There is an optimal settings and that will work the best for making profiles.[...]
I suppose that all those findings are kept secret. Even though one would have to tweak and fine tune color density, paper thickness and drying time settings, it would be VERY nice if some could give hints for values, in order to get close to it (x880 Epsons with K3VM is my combo ;-). It's really not that I'm lazy, or would not like to do it, it's just that I outsource the profile making to the very best people I can find, 'cause they always do a better job than me.
regards
EDIT: In this vein Mark Dubovoy writes Just to give the reader a general feeling for my settings, I have found a number of semi-gloss papers that produce better prints using Ultrasmooth Fine Art for the Media setting plus adding +5 % or even +10% to the color density at the same time. For some papers, this will also require a small increase in the drying time per pass.
For my current two favorite papers, I use the Ultra Premium Semigloss media setting and I add +10% to the color density. Luckily, no extra drying time is needed.
In Search Of The Ultimate Inkjet Print http://www.luminous-landscape.com/essays/in_search_of_the_ultimate_inkjet_print.shtml
-
I suppose that all those findings are kept secret. Even though one would have to tweak and fine tune color density, paper thickness and drying time settings, it would be VERY nice if some could give hints for values, in order to get close to it (x880 Epsons with K3VM is my combo ;-).
I don't know that it's secret...I've only ever done it for 1 3rd party paper that I don't use any more and I don't remember. I know I was going back and forth between an Epson watercolor setting and Enhanced Matte. I don't remember which setting I ended up using nor exactly what I did but I'm pretty sure I actually reduced the ink density due to over inking.
EDIT: In this vein Mark Dubovoy writes In Search Of The Ultimate Inkjet Print http://www.luminous-landscape.com/essays/in_search_of_the_ultimate_inkjet_print.shtml
But I don't know if that was pre or post target making...I suspect it might have been post target. But a 10% boost is very mild compared to what you were talking about at +50 which to my mind would be over inked (as you indicated regarding the loss of shadow detail which won't be helped by sharpening but might be by lightening the shadows which kinda defeats the purpose of the ink increase).
-
In addition to the above mentioned visual media testing, you can also make several profiles at different color density settings and compare the results in a gamut graphing application looking for Dmax and gamut volume.
Can you suggest a relatively inexpensive application for the Mac that will do this?
Thanks.
Terry.
-
EDIT: SO, right now, I am particularly concerned about loss of detail. Do you just dial back the density or adjust sharpening? Or a mix of both?
Normally this is work that is done by the printer/paper manufacturer with the help of a kind of RIP and then translated to the media presets in the normal driver. With only the driver at hand your only choice is to try out different media presets and see what the paper can handle without warping, the coating can handle without bleeding. The parameters to measure are Dmax on the black and chroma on the color channels. There are some 21 step greyscale targets on my website that can be converted to color to check the detail and measure the Dmax and chroma. You should start with the media preset that has the highest ink load. It is a clumsy method, in fact starting with a RIP would give it the right base, the normal driver just isn't the way to do it. On the other hand the HP Z drivers allow much more control and HP documents tell what the media presets are based on, what their ink limit is and HP also gives a summary of third party papers/media preset combinations. So with an HP you are halfway to a RIP. Epson is just more secretive with information like that.
More density in color doesn't have to create a wider gamut as written in this thread already. An overload of magenta ink will measure a lower chroma than the right amount, depending on the transparency of the ink. This was very obvious with early pigment inks versus dye inks, the last will increase chroma with every ML added. I did print some hard edge art prints that needed a lot of blue/purple density. So I did that by running it twice through the printer, Epson 10000 loaded with MIS 7600, Photorag. It delivered a color impossible to make with any of the normal paper settings. Added register tabs etc on the printer to align the images. The ink layer was extremely delicate, a lot still didn't aligh properly but I got there though it was a financial disaster. With that method you couldn't print a decent photo though.
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +230 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
Ernst
can you please give us a link to the HP documents that you are talking about?
Interesting method - double printing - I will try it on my Z with slight change on procedure, I will try it like this:
load roll
print
cutter off
unload roll
load roll
print again
if Z load roll at the same position every time this may work - not that I need it but I am curious what is effect, also how precise Z loads paper.
Damir
-
Ernst
can you please give us a link to the HP documents that you are talking about?
Interesting method - double printing - I will try it on my Z with slight change on procedure, I will try it like this:
load roll
print
cutter off
unload roll
load roll
print again
if Z load roll at the same position every time this may work - not that I need it but I am curious what is effect, also how precise Z loads paper.
Damir
I don't think all are there, you have to go into the maze of HP's webpages to see more recent versions:
http://z3100users.wikispaces.com/HP+Tech+Newsletters
Little chance to get it right with a roll loaded. Put two register tabs at the front of the printer, blue line aligned, load paper sheets a bit tilted so the printer will ask you to align them to the blue line, align them to the tabs and to the blue line at the right. The last doesn't need a tab as the sensor on the head will measure the edge anyway. In that case the sheets will be in register. You can adjust one tab a bit to get the edge parallel to the image which is important if this method is used to print double sided. In that case the print page should be horizontally symmetric on the sheet too. There must be a thread somewhere with pictures of the tabs I made.
The paper load method for deckled edge sheets will ask you right away to align to the blue lines but has a wider tolerance in media position measuring.
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/
-
High DMax is not a Holy Grail. It is more important that the shadows look deep and rich. This is partly facilitated by very good profiles that keep colours as pure as possible , not murky. Lots of factors come in here such as necessity to multiple read dark patches, use of polarizing filters, and how finely defined are the spectral captures.,. Most commercial units don't come even close to what is possible. Multiple printing is a great idea. I think this was possible with Iris printers.On Epsons you can lay more ink by using the highest resolution setting. Most undulating paper hassles are solved with a few days drying. Of course if you have pooling there is not much point.
If anyone knows how to achieve exact registration on an Epson or Canon printer I would love to hear from them. Ernst's procedure on HP is interesting. Ernst, how accurate is this method? Do you get any slew? On an Epson I get better results with rolls. Sheets maintain horizontal register but drift vertically.
Cheers,
Brian
www.pharoseditions.com.au
-
I suppose that all those findings are kept secret. Even though one would have to tweak and fine tune color density, paper thickness and drying time settings, it would be VERY nice if some could give hints for values, in order to get close to it (x880 Epsons with K3VM is my combo ;-). It's really not that I'm lazy, or would not like to do it, it's just that I outsource the profile making to the very best people I can find, 'cause they always do a better job than me.
As other posters have said, the optimum depends entirely on the printer/ink/paper combination, so my results probably will not help you. But, FWIW, I am using InkjetFly ink in my 3800 and my optimized setting for Harman Gloss Baryta is +20%. With Ilford Smooth Pearl I can only go to +10% before I begin to see micro-pooling in the print, at least without increased drying time. As you can see, these increases are quite modest but they do give worthwhile increases in Dmax and overall shadow "quality".
I use a sort of hybrid method to find my optimum but it does involve spectro measurements so probably not useful for you. First I print an ink separation file with K only using QuadTone RIP. This tells me the maximum black (measured) that I can achieve with the particular ink/paper combination. I then use the method described at http://www.hermitage-ps.co.uk/#q7 to do test prints at different ink levels. For each one I compare the black level with my maximum achieved in QTR and also ensure that no other problems are visible at that level. I could of course use this method for all the colours, but I am more concerned with B&W, and my experience is that measuring the black ink works well enough for colours too.
I thnk the problem you may have without suitable measuring equipment is in determining maximum Dmax and chroma, although you could probably use a scanner to do this. However you do it, it's not a trivial exercise, and until you profile the combination you can never be certain that it gives you any improvement.
-
Can you suggest a relatively inexpensive application for the Mac that will do this?
Apple's ColorSync utility is unfortunately not very good at this. Although I'm fond of a few expensive applications (ColorThink and GamutWorks), the PerfX 3D Gamut Viewer is a decent free alternative. http://mac.sofotex.com/downloads/d136148.html
-
High DMax is not a Holy Grail. It is more important that the shadows look deep and rich. This is partly facilitated by very good profiles...
Yes, I like what you're saying! Different profiling packages handle the "approach to black" differently which plays an effect upon the perception of Dmax, shadow detail, and overall contrast. For years I've been impressed with how Monaco Profiler deals with this "approach to black" relative to ProfileMakerPro and other profiling applications. It's hard to talk about it here but when you've got side by side comparison prints made with these two applications you can really see what we're talking about. The internal curves inside ICC profiles can have a big effect on making shadows "look deep and rich" as you put it.
On Epsons you can lay more ink by using the highest resolution setting. Of course if you have pooling there is not much point.
With pigmented inks (but not solvent!) there is an exact point at which you've achieved the greatest saturation point (chroma) with each ink without causing pooling or loss of shadow detail. Any more and you'll get disappointing results. Finding this exact point is tricky, and there are various ways of doing so - several of which have been mentioned here. I'm actually a big fan of using ColorBurst's chroma graphing tool for finding this point, regardless of what printing process is being used (driver, RIP,etc). I've encouraged a few software developers to come up with a tool for doing this but, so far, there's not much interest in such a niche product.
-
I use a sort of hybrid method to find my optimum but it does involve spectro measurements so probably not useful for you. First I print an ink separation file with K only using QuadTone RIP. This tells me the maximum black (measured) that I can achieve with the particular ink/paper combination.
Great for B&W, not great for color.
I thnk the problem you may have without suitable measuring equipment is in determining maximum Dmax and chroma...
When analyzing ink limits for color channels DMax is irrelevant - it's all about chroma. The differences can be subtle but important. I wish there were some commonly available simple GUI tools for this. I use ColorBurst's chroma linearization tool to analyze this, even when I'm printing with a driver without ColorBurst.
-
High DMax is not a Holy Grail. It is more important that the shadows look deep and rich. This is partly facilitated by very good profiles that keep colours as pure as possible , not murky. Lots of factors come in here such as necessity to multiple read dark patches, use of polarizing filters, and how finely defined are the spectral captures.,. Most commercial units don't come even close to what is possible. Multiple printing is a great idea. I think this was possible with Iris printers.On Epsons you can lay more ink by using the highest resolution setting. Most undulating paper hassles are solved with a few days drying. Of course if you have pooling there is not much point.
If anyone knows how to achieve exact registration on an Epson or Canon printer I would love to hear from them. Ernst's procedure on HP is interesting. Ernst, how accurate is this method? Do you get any slew? On an Epson I get better results with rolls. Sheets maintain horizontal register but drift vertically.
Cheers,
Brian
www.pharoseditions.com.au
Like Onsight mentions a good profiler program will handle the shadows, if the Dmax black has no dotgain/bleeding the shadows should fit properly. With matte papers I like to have the highest Dmax possible on that paper, the HPs do that nicely. With gloss papers Dmax is less an issue as the black density is in general at a high level already and I have some doubts whether all spectrometers measure correctly beyond 2.4 D.
Accuracy is excellent for dual sided printing (0.5 mm at most), good enough for hard edge art (silkscreen simulation) double run, the last not enough for photo quality in my opinion. For dual sided printing I can start with rolls, let the printer pull some length first and reroll again, cut the leading edge of the roll too and then start printing the print pages that have some extra white at the top so they can be used again as sheets on the machine. Printer cuts accurate enough. The other run as sheets I place the cut leading edge against register tabs at front and make sure the print page is centered correctly left-right. There are some tricks for the last. If I would use rolls for both sides I can be sure that the print page shifts left or right up to a mm and on the length it wouldn't work either.
met vriendelijke groeten, Ernst Dinkla
New: Spectral plots of +230 inkjet papers:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
Also note that increasing or decreasing the ink density when making a profile target will tend to get the increase/decrease profiled out. There is an optimal settings and that will work the best for making profiles.
Jeff is exactly correct on this.....increasing the print density/ink volume and re-profling will only cause the profile conversion to compensate for this and reduce the ink levels accordingly. The only time where this is NOT true is if the increased print density results in a wider gamut print profile....and an actual image that will take advantage of this extra gamut.
With Epson paper, it's really not useful to do anything with ink settings other than subtle tweaks...with 3rd party paper, it is useful to work the settings to optimize the ink density prior to making profiles because 3rd party papers don't necessarily fall into the Epson Media settings exactly. But simply increasing the ink density to pump more in onto the paper will not always be useful (and often causes more problems).
What I do is run through several media settings that I think will generally be in the ballpark and then look at the resulting profiles in something like ColorThink Pro. Gamut volume, gamut boundaries and tonal distribution in the primaries/secondaries will usually tell you what is the correct media setting for your paper.
As far as increasing "density", I'll reiterate what Scott Martin pointed out.....increasing DENSITY does not necessarily increase COLOR. In more correct "colorimetric" terms, once you've increased the ink volume to the point where the chroma (saturation) has peaked and starts to fall off, increasing the ink volume further only results in more ink "darkness" (density, L*, whatever) and only serves to oversaturate the paper with ink (wrinkles like you experienced).
The only think I can suggest if you want to go this route is getting a tool such as ColorThink Pro that will allow you to look at profiles with different ink adjustments to help you determine where the optimum or peak chroma values occur.
If you REALLY want control over the ink volume, you should consider getting yourself a CMYK RIP that allows direct control over per-channel ink limiting and such. With good technique and some knowledge on your part, a CMYK RIP can give you finer control over ink settings and not restrict you to a vendor's driver settings. ColorBurst X-Photo is one such RIP that gives you that kind of control.
Terry
-
Thank you all for all those knowledgeable posts!
I will try to find this sweet spot where the increase in color density and other media settings will produce the largest gamut. I will try to build different profiles and look at them with a gamut viewer.
As I outsource the profile making, I´
ve only got a scanner and Monaco EZcolor to do that for he moment. I definitely plan on investing in an i1 Pro in the future. Is there any chance that I can "see" something with the combination that I have got right now? I would still let my profiles be made.
regards
-
Just be careful that when you're increasing Color Density (or other similar settings in the driver) that you're not making color gradations rougher. For example, transitions between deep yellows and deep greens (e.g., a large smooth transition between an out-of-focus yellow tulip and green background) can be problematic with too much ink.
-
If one focuses on find the chroma/saturation plateau, everything else will follow, including color gradations. Excellent color profiles, of course, play a huge role in these color gradations. Less than the best profiling software can produce disappointing color gradations at several color density settings.
Assuming one is using the best color profiling technology available (which IMO means Monaco Profiler and i1Profiler), one only needs to focus on finding the chroma/saturation sweet spot while avoiding pooling and "ink smudging". My experience suggests that all other subjective visual observations should be ignored, including color transitions and the apparent loss of shadow detail that the profile will deal with. Of course, colorimetricly analyzing and finding the chroma/saturation sweet spot is a tricky thing for most people, because it can't be done by eye - it requires spectrophotometers, software and willingness to conquer the learning curve.
People like Terry Wyse here know this well, because the process of determining per channel and total ink limits, linearization and profiling with a RIP is a real craft that takes years to master. In the process, it teaches you more about inkjet technology that one can ever get from just using printer drivers that have these ink limits, linearization curves and separation parameters baked in.