Luminous Landscape Forum
Raw & Post Processing, Printing => Colour Management => Topic started by: scott morrish on February 07, 2012, 06:12:32 pm
-
Hello everyone.
Does anyone know if Colorburst can get anything out of the x900 printers (in terms of image quality), that the supplied Epson software can not achieve?
Does anyone actually manage to get better results (colour / smoothness etc) from the 7900 / 9900 via the standard epson Rip than they can via colorburst?
Guess i am wondering if Rips (like Colorburst) are now only needed for contract proofing / productivity issues? ... assuming that they are still needed for these?
Regards,
Scott
-
Guess i am wondering if Rips (like Colorburst) are now only needed for contract proofing / productivity issues? ... assuming that they are still needed for these?
I have a friend, Douglas Dubler, an NYC fashion shooter who claims he got better with Colorburst with his 9880, but I have not heard if he is still using it for his 9900. Mark Dubovoy still likes ImagePrint from ColorByte. But I'm perfectly happy using the Epson driver for my 9900 with really good profiles (which is really the key).
If you need to do ANY CMYK proofing then Colorburst is the way to go. But if you are printing RGB files, then no, I don't think a 3rd party rip is needed–particularly if you are printing out of Lightroom 4 and using soft proofing.
-
Thanks Jeff.
Appreciated.
Accepting that profiling is the key here... is SpectralVision Pro (the profiling software bundled with Colourburst) and an Eye1 (non UV), as good as it gets for profiling Epson's UCHDR inks, or are there tangible advantages in using other profiling software... or indeed other spectrophotometers? I guess it is enough for CMYK, but what about for RGB?
Regards,
Scott
-
This is partly an issue of which platform you're on: Windows version handles all channels and linearizations but Mac version relies on Epson's LUT's and there is no linearization. I haven't upgraded from version 6 because of that and now use the Epson drivers.
If you need the rasterization features, you can take a look at ColorBurst Overdrive which uses the Epson drivers after processing the files itself.
Jim
-
Jim,
I am not sure that i fully understand what ColorBurst is doing in terms of the differences between Mac and PC - but i am under the impression that Windows does not use all of the inks in the way one expects either. If you look at orange in a linearisation... there is a lot of M&Y in it: if you look at green, there is a lot of cyan and yellow. This may / may not matter... but it does mean that you can not use ColorBurst to control each of the six ink limits completely independently (which is what i had hoped to be able to do). Instead, in order to reduce the ink limit for orange for example, you have to change the ink limits for M and Y. On top of that, there is limited functionality regarding blending of the different shades of the ink colours. All around, i preferred ColorBurst with the 9800... but that is long gone now.
Of course... if you are proofing CMYK... O&G don't matter anyway...
Scott
-
Hi Scott,
Check your Lin again carefully. It can only be valid if each channel prints it's own ink and no others.
If you have an Info IO (what colors going in and colors going out) that should tell you.
How do you know that O is made up of M and Y?
Len
-
To see what is happening on the 9900 via ColorBurst:
I print the 6 colour linearisation with O&G disabled. Then I can see just how much of the other colours are being used by CB.
Equally, if I turn off CMY... and print the 6 colour linearisation target again... I can see what the O&G inks are actually doing.
It was different in the past with the 9800 etc.
If i am wrong here i'd love to know... as this limits Colourburst to Proofing only (for me).
-
I think you may be somewhat confused.
Can you post a photo of your Lin chart with all 6 colors?
A cell phone pic should work.
-
Possibly... and it is also possible that i have simply not written clearly enough?
Let me look again: everything is switched off for today... so it will be tomorrow.
From memory, if you have a 6 colour (10 ink) environment, and reduce O to 0% (via the Ink limiting settings...) you will still see a whole load of orange and green on the 6 colour linearisation, only it is not made up of the orange ink... it is made up of M & Y.
This is (i think?) why if you want to change the density of Orange when setting up (or modifying) a print environment, you actually have to change the ink limits for M and Y.
Via ColorBurst the orange ink is not doing all of the work in the darker parts of images... it is only really working in the lighter areas.
If i am completely wrong, and simply confused about this, i will find out tomorrow.
Naturally i will let you know.
Scott
-
Linearization is intended to limit individual inks prior to profiling, therefor all color management must be turned off when printing the chart.
BTW, your chart looks like a step wedge for each individual color, right?
-
Hopefully the image attached will make sense of what i have written earlier in the thread?
It is printed via the Print Linearisation target option in CB... so all colour management is disabled for the creation of these prints... (as far as i understand it).
I have done this quickly... so if i am confused... i'd be very happy to have my understanding straightened out.
Regards,
Scott
-
Hi Scott,
It looks like you reversed the ink limits in what you wrote.
Check to make sure all ink controls and color management are turned off.
The reason I mention that is because all patches on your bottom sheet (the one you said O and G were limited to 50%) look too darn good.
It's as if you are printing the patches through a previously linearized calibration. If you read that set of patches I'll bet your curve is rather straight.
When you use CB in a straight forward "Plain Vanilla" set-up are you not happy with the results?
-
Hi Scott,
It looks like you reversed the ink limits in what you wrote.
Check to make sure all ink controls and color management are turned off.
The reason I mention that is because all patches on your bottom sheet (the one you said O and G were limited to 50%) look too darn good.
It's as if you are printing the patches through a previously linearized calibration. If you read that set of patches I'll bet your curve is rather straight.
When you use CB in a straight forward "Plain Vanilla" set-up are you not happy with the results?
The writing on the charts is correct.
And, as far as I know, colour management is disabled by ColorBurst when printing the linearisation target.
In the first linearisation chart, the Ink limits for Orange and Green are both set to Zero... yet there is still colour!
In the second linearisation, the Ink Limits for Orange and Green are set to 50, to illustrate my findings.
In another, third linearisation (attached), you can see what is left if C, M and Y are set to Zero: O & G don't look quite so good without support.
Scott
-
Are you new to CB?
-
Are you new to CB?
Feels like it...
-
Hi Scott,
Make a test print from Photoshop and Colorburst.
Which method made the better print? Hopefully they match:)
Which workflow do you like?
An observation, FWIW.
Typically you would print a linearization thru a new configuration that is not yet set up. First prints rarely look as linear as yours. It seems that you are changing an existing configuration to print these.
Is that the case?
An existing configuration has a name like " Premium Nice Glossy".
In CB can you see the Lin file (as a Graph?) Does the plot kind of hug the diagonal?
If you've been noodling around and changing anything in these configurations then I would reload the originals or (if it's easy as pie) uninstall and re-install ColorBust.
Doing this will get you back to the best possible ColorBust experience for this comparison.
It really is about what you like better:)
Cheers
PS: Don't you luv that typo?
-
I am not sure it matters wether i started with a pre-set Printing Environment or not.
The orange and green inks are either printing without contamination (from other inks), or they are not.
In this instance - yes... i am working from an existing Print Environment, and yes i am changing Ink limits (for my objectives) ... and re-linearising.
But does it make any difference to how Orange is printed? ... or green for that matter?
I don't think it does... but i am still happy to be corrected.
From what i can see at this point in time, and accepting that i am not an expert in these matters, I do not think the user has a choice over this matter: orange will be printed with Magenta and Yellow. Implemented perfectly, this might not matter. However if Ink curves are not blended carefully... there can be clunky steps in areas of tonal gradation through the colour in question. Because of the way this is implemented in CB... users can not correct this in a simple or logical way. With other Inks, on other printers... this was not a problem.
Scott
-
Hi Scott,
Holy crap on a cracker!
Did you really say that?
-
I'm not sure there are very many, if any, colors that will print with only O or G i n a normal printing situation. Those inks when used correctly seem to be added in addition to the others when running out of gamut. In fact, this is what you want- oranges made from Y and LM dots are smoother and less dotty than from O dots. Same applies for Y + LC vs G. It's desirable to have those inks introduced wisely and only as needed. The only way to see those inks completely independently is with linearization printouts from a 6 channel RIP, or with a true 6 channel file dropped into same RIP, with areas using only those channels. I can do this with StudioPrint, bot have no working knowledge with Colorburst.
Hope that makes sense.
Tyler
http://www.custom-digital.com/
-
I am not sure that i fully understand what ColorBurst is doing in terms of the differences between Mac and PC
Scott,
From the Mac OSX version of the SpectralVision Pro manual:
"ColorBurst utilizes the Epson halftone module (HM) for several printer models, including the Epson Stylus Pro 4900, 7900, 9900, and 11880. The halftone module makes linearization and ink limiting unnecessary, so the profiling process for HM printers in ColorBurst is much simpler."
So for the x900 series printers on the Mac side, you just make a new profile when you feel the need. You can't linearize or limit inks. On the PC side, those controls exist for the x900 printers
Jim
-
From the Mac OSX version of the SpectralVision Pro manual:
"ColorBurst utilizes the Epson halftone module (HM) for several printer models, including the Epson Stylus Pro 4900, 7900, 9900, and 11880. The halftone module makes linearization and ink limiting unnecessary, so the profiling process for HM printers in ColorBurst is much simpler."
So for the x900 series printers on the Mac side, you just make a new profile when you feel the need. You can't linearize or limit inks. On the PC side, those controls exist for the x900 printers
Jim
For what it is worth, I have CB running on a PC.
I adjusted Ink limits (for my objectives), re-profiled (via SpctalVision Pro), and am simply recounting what i have observed. Again... definitely worth highlighting, i know i am not an expert in these matters.
I think, going back to my original post, i am settling on Colorburst for CMYK proofing, and LR via Epson for RGB as Jeff suggested.
I would have liked to be able to use Colorburst for RGB too, for my own workflow reasons as much as anything, but unless ColorBurst can use Orange and Green more effectively (in my opinion, and based on the understanding i have attempted to explain in the rest of this thread)... i'll simply keep it for CMYK work.
I'm not sure there are very many, if any, colors that will print with only O or G i n a normal printing situation. Those inks when used correctly seem to be added in addition to the others when running out of gamut. In fact, this is what you want- oranges made from Y and LM dots are smoother and less dotty than from O dots. Same applies for Y + LC vs G. It's desirable to have those inks introduced wisely and only as needed. The only way to see those inks completely independently is with linearization printouts from a 6 channel RIP, or with a true 6 channel file dropped into same RIP, with areas using only those channels. I can do this with StudioPrint, bot have no working knowledge with Colorburst.
Hope that makes sense.
Tyler
You may be right. It just seems counter-intuitive to add M&Y to Orange, at the point of establishing the separate Ink Channels (to my mind). I'd rather keep each of the colours pure, and let the profile, (or my separations) determine how to achieve certain colours. The way it appears to work at the moment seems unduly complicated, can introduce problems that are not easy to identify or address, and as far as i can tell... gives the user less control and flexibility.
Perhaps i am trying to do things that Colorburst is not meant to do?
I know that on restricted gamut printers (9800 / 9880 etc) none of these problems occur. It is just a question of how the additional colours (orange and green) are used... and wether it is optimal.
Probably only worth thinking about this at all if one of the reasons for using a 9900 is for its increased gamut.
I realise that it is completely irrelevant if all that matters is CMYK.
-
Based on experience with the Wasatch SoftRip 6.2 driving the HP Z3100 versus the HP drivers for that printer I recommend you take a microscope or a strong loupe to see what actually is laid down on the paper. I guess you can use a CB OEM profile + media settings for an Epson paper and use the Epson driver with its own media settings + Epson profile.
In my experience only the HP driver uses the extra ink hues RGB where they function best: on the hue angles between the CMY inks and only at the highest saturations. On lower saturations you will see the grey inks appear and cM(m)Y mixes. No RGB inks at the neutral spine and actually no CM(m)Y composite greys there either. None of that you will see in the prints with the Wasatch Softrip, cM(m)YRGB is all over the place including in the neutral spine, Grey inks are hardly used, the black generation is very much a black generation and does not stretch to 255 RGB. This is with the Wasatch SoftRip 6.2 profiles delivered for some HP papers. My own attempts with some external N-Color profile creators + the ink control of the Wasatch did not bring a solution and on request Wastach made some alternatives that were not better either. That a RIP can linearise/calibrate all ink channels of a printer (the Wasatch can use the integrated calibration of the Z models) does not make a difference. If the N-color profile separation does not adequately separate the hues on hue angle + saturation and the black generation of the profile creator + RIP ink settings is not running from 0-100% you will see cM(m)Y ink dots on places where you would not like them. Which increased ink use an gave more "metamerism" in my case.
There have been comments that the Z3100 was not as "rich" in color as the Epsons at its introduction, I think it could have been related to that more severe separation with frugal ink use. The Z3200 is more like the Epson gamut, partly a result of the Chromatic Red but I see a change in gamma for the other hues too.
With CB relying on the Epson dithering/weaving (as I understand from this thread) I would expect that the RIP has to obey to hue separations that resemble the Epson driver ones to suite that way of laying down inks. So there may not be a similar difference I see in the Z output.
I do not use the RIP but Qimage Ultimate + the HP driver + either Color Center or APS for the profile creation with the integrated spectrometer. The occasional CMYK jobs are handled through a Photoshop P2P conversion. I am told I should actually do that with the HP Postscript driver of the Z3200-PS. A RIP like GMG for proofing should be the best solution around though. Ergosoft's RIPs have more control too.
met vriendelijke groeten, Ernst
Shareware too:
330+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
-
those are very interesting facts that ernst state here and i realized the same through a lot of testing the last months!
i worked with EFI, wasatch and onyx.. to try to be better than the built in driver.. as i always want to get the 110% percent quality for my art prints.
and i could gain some ... but lost others.. and so far i have to realize.. i dont see a way around!
the screening is the most important difference...!.. even with onyx who i think it the most advanced one for profiling n-channel inkjets.. (own n-channel ICC profiling swatch) the screening kills it!
on a matte fineArt paper you just see the dots.. no matter if you use stochastic, FDRP (onyx) enhanced.. veriable dot sizes.. man . compared to the contone driver i see the dots with my eyes!
not on canves.. so i dont have a problem there.. but the colorprofile for the the daguerre canvas i get with 4096 atkinson chart is so close to the one with onyx. is it worth it?.
i gain shadow details!.. def..
and what ernst recalls.. seperations: the rip introduces the RGB colors as soon, as they can replace other combinations.. the advantage with it tho is to reduce ink !.. dis are the dots..
the gray axis with onyx is perfectly balanced.. and i can use settings where there are no CMY introduced before like 40%.. so i dont see dots there.. not with wasatch i couldnt.
but.. is it worth the whole deal?.. i dont know.. its fun doing it.. ;-)
@ernst.. the PS driver works awesome for CMYK.. if your RGB profile is good.. which it will be of course.. and you have the pdf in fogra27.. just send it through the web server.. the colors will get soo close. i even get an ok on a fogra wedge. not as precise as EFI.. but without costs and hassle.. even the pantone emulation is there!