Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: keith_cooper on May 07, 2008, 02:42:24 pm
-
Hi
I've just finished a review of what's in the latest version (V3.5) of the Spyder3 Print software
Spyder3 Print review (http://www.northlight-images.co.uk/reviews/profiling/spyder3print.html)
There are quite a few tweaks and improvements, especially for those who've complained about its measurement speed ;-) I've always been a little curious about the impatience of many who've complained in the past, almost as if there were a prize for speed. I can easily beat it for speed, but that's using a considerably faster i1 iSis (oh, they are a bit more expensive too ;-) The measuring device is not what I'd want to use to make profiles -every- day, but it isn't aimed at that market.
The profiles I looked at were all very good, and with the extra greyscale target, rather good for B/W as well.
The software is currently (I believe - check the web site) being provided as one of the currently fashionable 'Public beta' releases, although it didn't have -any- issues on the Mac I tested it on.
Since I've finally got working ColorMunki software and Epson have just sent me a 4880 to look at I suspect I'll be doing a bit more profile building than usual this coming weekend ;-)
-
We will be very interested in a direct comparison between the CM and the Spyder in terms of the profile quality.
Thanks for the work!
quote=keith_cooper,May 8 2008, 03:42 AM]
I've just finished a review of what's in the latest version (V3.5) of the Spyder3 Print software
Spyder3 Print review (http://www.northlight-images.co.uk/reviews/profiling/spyder3print.html)
[a href=\"index.php?act=findpost&pid=194198\"][{POST_SNAPBACK}][/a]
[/quote]
-
We will be very interested in a direct comparison between the CM and the Spyder in terms of the profile quality.
That's actually harder to do -well-, than might first seem.
I'll hopefully get an article covering the ColorMunki in the same depth as the Spyder one, but it won't have any direct comparisons.
I've always avoided doing comparative reviews since (IMHO) to be of real use they need to be done in a very rigourous and methodical way, and I don't think I've ever been called that ;-) But seriously, many so called comparisons tell you more about the biasses and likes/dislikes of the reviewer than real differences in products. I'd rather hope to present enough information so that a reader can look at both and see which factors matter to them, for their own buying decision.
Suffice to say, I will be doing lots of profiling and hopefully what I write will reflect that.
Of course anyone near my home in Leicester, England would be welcome to pop round and have a look. Then I can show actual prints and explain all the caveats and details that might get lost in any discussions, coloured by peoples' perceptions of the companies/personalities involved (even on a relatively polite board like this one ;-)
-
I tested the spyderprint a few weeks ago against the profiles from my 3800 which Epson delivers with the printer.
Making the profile was easy but the results were on the 3800 slightly to warm.
I compared the print with my monitor (calibrated argb monitor).
When I print with the supplied profiles the output was almost perfect.
With the new profile the output was slightly to warm (almost bronzed).
The problem I did find with the software was the print out of the test pattern, it would be nice to have a full size A4 output instead of the small centered version, it does work but full size would be great, although you can also print in A3 of course if you need bigger :-)
We tried it also on a cheap Canon 4300 and the results were stunning, much better than the supplied profiles.
Because in the studio we use Epsons we are now only using the spyder 3 for customers.
-
I tested the spyderprint a few weeks ago against the profiles from my 3800 which Epson delivers with the printer
Have a look at the new version of the software - one of the things it suggests is printing the patches larger to make reading easier.
You mention bronzing, if this isn't a problem with the Epson profiles then it suggests that a different media setting might be beneficial? I've been trying out some profiling this afternoon on the 4880 and found some quite subtle variations in output when trying different media settings in the media check section of the software. I've been testing a glossy Innova paper (IFA09) and found that TFA produces good results, followed by PQIJ setting. Interestingly, the suggested VFA setting blocks up shadows slightly - causing a loss of quality if you were to profile without a media settings check...
-
I compared the print with my monitor (calibrated argb monitor).
When I print with the supplied profiles the output was almost perfect.
With the new profile the output was slightly to warm (almost bronzed).
Could be the effect of optical brighteners which the CV product doesn't appear to be able to deal with. Its not a true Spectrophotometer (I'm not sure exactly what a SpectroColorimeter is supposed to be, does it measure spectral data or RGB data like a Colorimeter)?
-
Whatever it is (probably best to ask your question on the Colorvision message board) it works very well for the money. I will note it is more effective on photography style papers than limited gamut matte and canvas medias where purple shift is quite an issue.
Since using the Colorvision I have switched to the i1Pro and send off my profiles using colorport to a colorgeek with Monaco Profiler. Printing on limited gamut media like canvas and matte I am stunned at how much better the colors are with this higher end combination using perceptual rendering.
Thank you Scott Martin for a fairly inexpensive and fast workflow that is giving me the best of both worlds.
-Jonathan
Could be the effect of optical brighteners which the CV product doesn't appear to be able to deal with. Its not a true Spectrophotometer (I'm not sure exactly what a SpectroColorimeter is supposed to be, does it measure spectral data or RGB data like a Colorimeter)?
[a href=\"index.php?act=findpost&pid=194629\"][{POST_SNAPBACK}][/a]
-
Whatever it is (probably best to ask your question on the Colorvision message board) it works very well for the money.
I've asked them, never got an answer.
-
We use the professional equiptment from gretag (now owned by the same company as spyder I believe) in the past we got great responses now it's almost zero, it's bad.
It's not a huge difference by the way just a bit too warm, but I'm incredible sensitive for this.
-
It's not a huge difference by the way just a bit too warm, but I'm incredible sensitive for this.
Ah, then that can be one of the reasons to spend rather more on a profiling solution ;-)
The more you do this sort of stuff, the more you notice.
I have to remind myself sometimes that if the customers can't see any difference and I can, then I should carefully consider what I'm doing as part of the business, and what just to satisfy myself ;-) The two are -not- always the same.
When I'm doing talks of colour management for camera clubs and the like, I often point out that people who consider themselves perfectionists (I don't :-) should always be slightly wary, getting into printer profiling, lest they have shares in ink or paper manufacturers :-)
-
I've asked them, never got an answer.
Google is your friend, since it seems to be a relatively well known term :-)
AFAIK the Spyder2 Print one measures six portions of the spectrum with six narrow band light sources. It recombines the numbers to get L*a*b* values.
The spectro in the CM is a UV cut one, unlike the iSis which offers cut and uncut via different light sources.
...Of course the whole UV cut/no cut discussion is a different matter all together, I believe both approaches have their fans/detractors :-)
-
The first link is a hardware comparison chart including the Datacolor Spectrocolorimeter, Spectrolino, Eye-One Pro, DTP41 SeriesII Autoscan and Pulse:
http://spyder.datacolor.com/product-compare1005.php (http://spyder.datacolor.com/product-compare1005.php)
The second link is Datacolor Spyder3 Studio vs. Xrite ColorMunki:
http://spyder.datacolor.com/pdfs/CI-303%20Calibrating-S.pdf (http://spyder.datacolor.com/pdfs/CI-303%20Calibrating-S.pdf)
-
The first link is a hardware comparison chart including the Datacolor Spectrocolorimeter, Spectrolino, Eye-One Pro, DTP41 SeriesII Autoscan and Pulse:
http://spyder.datacolor.com/product-compare1005.php (http://spyder.datacolor.com/product-compare1005.php)
Well its a comparison (of questionable use). It doesn't tell us anything about the data gathering of this "technology" although the best info I've seen to date is from Keith (AFAIK the Spyder2 Print one measures six portions of the spectrum with six narrow band light sources. It recombines the numbers to get L*a*b* values.).
-
The spectro in the CM is a UV cut one, unlike the iSis which offers cut and uncut via different light sources.
...Of course the whole UV cut/no cut discussion is a different matter all together, I believe both approaches have their fans/detractors :-)
[a href=\"index.php?act=findpost&pid=194692\"][{POST_SNAPBACK}][/a]
The iSis is so far, the best Spectrophotometer in terms of qualities I've used, and after talking with Bill Atkinsion about the UV debate, I've decided to always use the UV cut in it. I didn't have any UV issues without a filter in the past on other non UV cut Spectrophotometer's but I'm using ProfileMaker Pro's approach to compensating in software (again, something useful when you gather true spectral data). I suspect CM is using the same approach.
-
Well its a comparison (of questionable use). It doesn't tell us anything about the data gathering of this "technology" although the best info I've seen to date is from Keith (AFAIK the Spyder2 Print one measures six portions of the spectrum with six narrow band light sources. It recombines the numbers to get L*a*b* values.).
Have a look at the photo and the quicktime movie not far from the top of the review I wrote when the PrintFIX PRO first came out. You can see the sequence of coloured lights as it reads a patch. It is this sequence that's been speeded up to increase the patch measuring speed.
www.northlight-images.co.uk/article_pages/printfix_pro.html (http://www.northlight-images.co.uk/article_pages/printfix_pro.html)
-
Have a look at the photo and the quicktime movie not far from the top of the review I wrote when the PrintFIX PRO first came out. You can see the sequence of coloured lights as it reads a patch. It is this sequence that's been speeded up to increase the patch measuring speed.
www.northlight-images.co.uk/article_pages/printfix_pro.html (http://www.northlight-images.co.uk/article_pages/printfix_pro.html)
[a href=\"index.php?act=findpost&pid=194887\"][{POST_SNAPBACK}][/a]
What interests me is that you're essentially using only 6 "measurements" within the spectrum with some possible overlap I suppose. The bandwidth of a Spectrophotometer varies too of course but its usually far finer (from 1nm to maybe 20). With a 1nm bandwidth, assuming you're measuring 380-700nm, that's 320 steps being measured. According the spec's for the EyeOne Pro that X-Rite provides (and isn't visible for the CV product that I can find), its optical resolution is 10nm over a range of 380-720nm. Further, in the case of actual Colorimeters which are measuring 3 primary colors, essentially an XYZ, there are filter matrixes which ideally are based upon what they have been designed to measure. That's why a Colorimeter that's mated to an actual display based upon these matrixes is the ideal instrument to use for display calibration. We're seeing how older Colorimeters are having difficulties with wide gamut displays and why, at least in the case of the NEC LED wide gamut, NEC OEM'ed an EyeOne Display with custom matrices.
It appears, and I could be totally wrong here, that the instrument in question is adding three additional sensors and probably has additional filter matrixes. I wonder if this accounts for the need to update assumptions about what its measuring in terms of papers and inks and why some report problems building printer profiles for some kinds of these inks and papers?
-
This appears to be the precursor to the Datacolor Spectrocolorimeter. The discontinued ColorMouse and ColormouseToo.
http://www.colorsavvy.com/PDFs/ColorMouseToo.pdf (http://www.colorsavvy.com/PDFs/ColorMouseToo.pdf)
http://www.colorsavvy.com/PDFs/Color_mouse_paper.pdf (http://www.colorsavvy.com/PDFs/Color_mouse_paper.pdf)
-
This appears to be the precursor to the Datacolor Spectrocolorimeter. The discontinued ColorMouse and ColormouseToo.
http://www.colorsavvy.com/PDFs/ColorMouseToo.pdf (http://www.colorsavvy.com/PDFs/ColorMouseToo.pdf)
http://www.colorsavvy.com/PDFs/Color_mouse_paper.pdf (http://www.colorsavvy.com/PDFs/Color_mouse_paper.pdf)
[a href=\"index.php?act=findpost&pid=194909\"][{POST_SNAPBACK}][/a]
Thanks for that, the second paper has a quite detailed description of the techniques in multi-band colorimetry along with the maths involved, and I believe addresses several of Andrew's concerns.
Add ten LEDs and you've got a 10 band spectrophotometer, or is it a ten band colorimeter :-)
With the advent of the CM I've noticed a general rise in the level of marketing fluff surrounding what are essentially different styles of measurements devices. Whilst I don't expect too much (commercial) detail from manufacturers, it behoves us who write about such stuff (IMHO) to try and sift out some of the more egregious marketing propaganda from the useful stuff ;-)
-
Andrew and Keith,
I have read a million words in sixteen threads. Could you please just give me the bottom line on this question?
I need a printer profiler solution for my Epson R2400 soon to be upgraded to a larger Epson. I want to spend around $500. I prioritize accuracy way over ease of use as I am an engineer and not bothered by technical complexity.
(Option 1) The Xrite Eye-One PhotoLT Color Calibration System is $730 witha $200 rebate = $530.
(Option 2) The ColorVision Spyder3Print Printer Calibration System from Datacolor is $500.
(Option 3) The Xrite ColorMunki Photo Color Management Solution is $450.
The prices of these solutions are close enough to be essentially a wash. Please advise which of these, in your opinion, would provide me with the most accurate profiles, assuming good technique and attention to detail in the profile creation process.
Anyone else with experience using multiple ones of these models, please feel free to provide opinions.
Many thanks,
Bruce
-
Door #1 provides you an upgrade path to an i0 (which is quite nice) and higher end software products should you wish to go that route. Its more proven solution.
-
I need a printer profiler solution for my Epson R2400 soon to be upgraded to a larger Epson. I want to spend around $500.
Given from what you say, if that is your budget, the the first one will just cost you more to do anything useful... as to the i0, oh yes it's nice to use, but hardly in the stated budget is it :-)
The second is the most flexible and complex in terms of what the software does - it offers lot's of features, and if you have the desire you can print lots of big targets and average readings.
The third may be a good instrument, but I think anyone who wanted to tinker or experiment with profile building would exhaust the capabilities of the software pretty quickly. It has been directed at a specific market, and I suspect you are not in it. I've been using it this afternoon and some of its 'features' certainly irritated the hell out of me. BUT, as I always point out, I'm not in its target market ;-) so YMMV
One other alternative for the true experimenter (and you will learn a lot of colour management in the process) would be to get the basic i1 LT and use it with a package like the Argyll CMS
www.argyllcms.com (http://www.argyllcms.com/)
Argyll is an open source, ICC compatible color management system. It supports accurate ICC profile creation for scanners, CMYK printers, film recorders and calibration and profiling of displays. Spectral sample data is supported, allowing a selection of illuminants observer types, and paper fluorescent whitener additive compensation. Profiles can also incorporate source specific gamut mappings for perceptual and saturation intents. Gamut mapping and profile linking uses the CIECAM02 appearance model, a unique gamut mapping algorithm, and a wide selection of rendering intents. It also includes code for the fastest portable 8 bit raster color conversion engine available anywhere, as well as support for fast, fully accurate 16 bit conversion. Device color gamuts can also be viewed and compared using a VRML viewer. Comprehensive documentation is provided for each utility, and a general guide to using the tools for typical color management tasks is also available. A mailing list provides support for more advanced usage.
-
Thank you for your input, Andrew and Keith; very helpful.
Keith, two questions relative to your comments:
"Given from what you say, if that is your budget, the the first one will just cost you more to do anything useful..."
(1) I don't quite follow what you mean by this. Do you mean that profiles created with the Xrite Eye-One PhotoLT Color Calibration System will be essentially useless? Or do you mean that the profiles will not be as good as those obtained with the ColorVision Spyder3 Printer Calibration System because the PhotoLT has a dumbed-down version of software?
(2) I like your idea of a migration path involving the Argyll CMS and the Eye-One spectro. But can the Argyll software do profiles for an RGB printer like the Epson R2400, 3800, 4880, etc.? My confusion here lies in that the Argyll write-up cites its capability with CMYK printers, not RGB printers. Could you please clarify that for me?
Thanks!
Bruce
-
The LT only offers small patch count profiling, which works for some well behaved papers on some well behaved printers, but it doesn't do much else in the printer profiling stakes.
See i1 LT review (http://www.northlight-images.co.uk/reviews/profiling/i1-lt.html) and other linked articles on the site to see just what you can do with the different i1 versions.
Ah yes, Argyll - you will need quite a lot of tinkering and colour management experience to go with this, not to mention more software to drive the printer (look to Gutenprint amongst others) This really is the find some land, grow your own, make your own cigarette papers and roll your own approach :-) Not practical for most, but I mentioned it to show there are alternatives in which you can swap a lot of work for spending money ;-) Once you have the i1 there are quite a few software packages that support it, but that sort of goes out of your budget zone...
-
The LT only offers small patch count profiling, which works for some well behaved papers on some well behaved printers, but it doesn't do much else in the printer profiling stakes.
See i1 LT review (http://www.northlight-images.co.uk/reviews/profiling/i1-lt.html) and other linked articles on the site to see just what you can do with the different i1 versions.
Ah yes, Argyll - you will need quite a lot of tinkering and colour management experience to go with this, not to mention more software to drive the printer (look to Gutenprint amongst others) This really is the find some land, grow your own, make your own cigarette papers and roll your own approach :-) Not practical for most, but I mentioned it to show there are alternatives in which you can swap a lot of work for spending money ;-) Once you have the i1 there are quite a few software packages that support it, but that sort of goes out of your budget zone...
[a href=\"index.php?act=findpost&pid=195320\"][{POST_SNAPBACK}][/a]
Keith,
Thank you for the link to your articles and reviews; very helpful. (I am beginning to get the idea.) It sounds like Argyll might keep me tinkering to the exclusion of shooting and printing; so that would not be good.
If I bought i1 LT mainly for the good spectro, is there some commercial profiling software at a cost point that, when combined with the cost of the i1 spectro, is less than the i1 Photo bundle and gives better results than the i1 Match software?
I ask this question for two reasons:
(1) To keep from spending as much as the i1 Photo bundle of ~ $1500 (if possible); and
(2) In your original review "Eye One Printer Profiling," you say:
"So, how good are the profiles I made? - good."
"Are they good enough for my own fine art prints? - not really."
I realize that the original review was from 2005 and that you later (March 2006?)updated it with a report of new features in i1 Match Vers. 3.6. (From the X-rite website it appears that the current version is 3.6.2; so no radical revisions since your 2006 report?) So, has some combination of the updated Match profiling software and/or newer Epson-family printers and drivers changed your assessment of the i1 spectro/Match combination, or would you continue to consider the combination insufficient for your own fine art prints?
I ask that because if I cannot buy a profiling solution that is adequate for the highest-quality fine art prints even if I spend $1500 then I probably need to rethink
the idea of doing my own profiling. I do not mind scanning the patches by hand using the straight edge (versus the automated I1Sis page scanner), but I cannot imagine paying $1500 for a profiling solution and going to all the associated trouble to obtain anything less than than profiles adequate to support the highest-quality fine art prints.
Sorry for all the primitive questions, but could you please help me to tweak my understanding of this?
Thanks again,
Bruce
-
I've left all the original reviews up since it can be quite confusing with all the different options available (thanks for noticing when it was written BTW ;-)
The 918 patch target in the i1 photo can actually give some very good profiles, particularly on newer printers. I think some of my concerns over profile quality at the time were from not properly testing media settings and getting optimal results -before- printing. That was fair, since Eye One Match really doesn't emphasise this at all - oh, and I've learnt an awful lot more about colour management in the three years since I wrote that review ;-)
So, thanks for asking... I'll go back through some of my older reviews (for current products) and add a few 'sideboxes' if I feel a point could do with clarification
The media settings issue is, I suspect, going to lead to less than optimal results for quite a few users of the ColorMunki. I tested a light artificial 'inkjet paper' yesterday with a 4880, and since it is a paper that I bought a pack of several years ago for testing, thought I'd print off A4 media test images for just about every media setting that the 4880 has. The variations are considerable and I wasn't able to predict the results at all. It happens to be a paper that takes some time to dry, so the two stage profiling process is not overly conducive to testing multiple profiles in a hurry. You can print multiple first sheets, but each second sheet can only be produced from a scan of the first sheet. I've seen various attempts at a workaround mentioned (printing to PDFs and the like) but you are fighting what the software wants to do, to an extent that just asks for errors and problems. It does what it does, and that's it.
I don't have any third party software I regularly use, however you could get an idea of the sort of stuff available if you look at the BasICColor site (http://www.basiccolor.de/english/index_E.htm) They are bringing out some new products soon, and I hope to have a look at some of them, since I've been asked about 3rd party profiling solutions several times now. BTW Don't hold your breath on this though, since I do have lots of 'real work' too and writing reviews of complex products is not a two hour job ;-)
-
I've left all the original reviews up since it can be quite confusing with all the different options available (thanks for noticing when it was written BTW ;-)
The 918 patch target in the i1 photo can actually give some very good profiles, particularly on newer printers. I think some of my concerns over profile quality at the time were from not properly testing media settings and getting optimal results -before- printing. That was fair, since Eye One Match really doesn't emphasise this at all - oh, and I've learnt an awful lot more about colour management in the three years since I wrote that review ;-)
So, thanks for asking... I'll go back through some of my older reviews (for current products) and add a few 'sideboxes' if I feel a point could do with clarification
The media settings issue is, I suspect, going to lead to less than optimal results for quite a few users of the ColorMunki. I tested a light artificial 'inkjet paper' yesterday with a 4880, and since it is a paper that I bought a pack of several years ago for testing, thought I'd print off A4 media test images for just about every media setting that the 4880 has. The variations are considerable and I wasn't able to predict the results at all. It happens to be a paper that takes some time to dry, so the two stage profiling process is not overly conducive to testing multiple profiles in a hurry. You can print multiple first sheets, but each second sheet can only be produced from a scan of the first sheet. I've seen various attempts at a workaround mentioned (printing to PDFs and the like) but you are fighting what the software wants to do, to an extent that just asks for errors and problems. It does what it does, and that's it.
I don't have any third party software I regularly use, however you could get an idea of the sort of stuff available if you look at the BasICColor site (http://www.basiccolor.de/english/index_E.htm) They are bringing out some new products soon, and I hope to have a look at some of them, since I've been asked about 3rd party profiling solutions several times now. BTW Don't hold your breath on this though, since I do have lots of 'real work' too and writing reviews of complex products is not a two hour job ;-)
[a href=\"index.php?act=findpost&pid=195431\"][{POST_SNAPBACK}][/a]
Thank you for the clarification, Keith. I will check back on your site periodically.
Bruce