Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: StevenSzabo on April 12, 2010, 11:31:40 pm
-
So I decided to bite the bullet and purchase my own profiling solution, reviewed some options and decided on Spyder3 Studio SR, ordered it from B&H and anxiously awaited the arrival of my friendly neighborhood UPS man. It arrived today and I basically tore in to the package and started reading the easy start guides. Started by profiling the monitor, then progressed and started profiling with Epson luster (I figure it gives me some leeway if I mess it up being a cheap paper and all).
I printed out the 729 target profile, and carefully worked my way through the scanning until I had a end result that I THOUGHT looked consistant. The reader is definitely reading more muted than the "pure" results, but even from the online manual I read on their own site it says that is basically to be expected.
Saved my profile, and opened up my desired image in CS4. Printed it off 11 X 17 and was immediately aware that something was horribly wrong. the image is of a wheel with a brake caliper underneath that is VERY much blue. Only issue with that being that in the print, the brake caliper was very much purple. I'm not talking a few shades off from the monitor either, I'm talking PURPLE.
I'm printing using the following:
-Macbook Pro
-OSX Snow Leopard
-Epson 9900
-Epson Luster
-CS4
For experimentation's sake I tried printing out of lightroom too, same results. In CS4 I used saturation rendering intent. In lightroom I used perceptual, since saturation was not available.
has anyone had similar issues when profiling using spyder3 or any other method? I'm quite frustrated here, I'm doing nothing but waste paper and get mad here. I mean, honestly, I only got this because I feel that the canned profiles I've been using are to high contrast and I lose a lot of the shadow tones to black, which is of course no good.
-
You have to make sure color management is turned off in the printer driver when you print the targets and your image. The rendering intent should remain the same for both target and image.
It looks as if you are double profiling somewhere in your workflow.
-
Hi,
There are some issues with profiling on Leopard/Snow Leopard.
Check this: http://www.luminous-landscape.com/tutorials/solving.shtml (http://www.luminous-landscape.com/tutorials/solving.shtml)
I'm in a fight with Snow Leopard, Color Munki, LR and CS3 right now. To me it seem that my prints are to dark. I used Babel Color to generate a synthetic Color Checker printed it and fed the measured values back to
Babelcolor. The screen dump shows my print using my own color profile on the left 'original' in the centre and "Epson Manages Color" on the right.
The bottom row are comparisons. It seem that color is quite OK but the L values differ a lot. Could it be a "gamma related issue"?
It seems that I get similar results in Photoshop CS3 and Lightroom 2.6. I also feel I have lived with the problem a long time on different OS-es and using both ColorVision Print Fix Pro and the Color Munki.
Sorry, not any of this solves your problem, but at least can add some fuel to the discussion.
[attachment=21467:Screen_s...27.11_AM.png]
Best regards
Erik
So I decided to bite the bullet and purchase my own profiling solution, reviewed some options and decided on Spyder3 Studio SR, ordered it from B&H and anxiously awaited the arrival of my friendly neighborhood UPS man. It arrived today and I basically tore in to the package and started reading the easy start guides. Started by profiling the monitor, then progressed and started profiling with Epson luster (I figure it gives me some leeway if I mess it up being a cheap paper and all).
I printed out the 729 target profile, and carefully worked my way through the scanning until I had a end result that I THOUGHT looked consistant. The reader is definitely reading more muted than the "pure" results, but even from the online manual I read on their own site it says that is basically to be expected.
Saved my profile, and opened up my desired image in CS4. Printed it off 11 X 17 and was immediately aware that something was horribly wrong. the image is of a wheel with a brake caliper underneath that is VERY much blue. Only issue with that being that in the print, the brake caliper was very much purple. I'm not talking a few shades off from the monitor either, I'm talking PURPLE.
I'm printing using the following:
-Macbook Pro
-OSX Snow Leopard
-Epson 9900
-Epson Luster
-CS4
For experimentation's sake I tried printing out of lightroom too, same results. In CS4 I used saturation rendering intent. In lightroom I used perceptual, since saturation was not available.
has anyone had similar issues when profiling using spyder3 or any other method? I'm quite frustrated here, I'm doing nothing but waste paper and get mad here. I mean, honestly, I only got this because I feel that the canned profiles I've been using are to high contrast and I lose a lot of the shadow tones to black, which is of course no good.
-
is that not a CS3 specific issue? I'm printing directly out of the Spyder3 utility, which I would assume is dealing with the color management on its own.
But it did appear that the color management MAY have been on in the driver in the original print...I will take a look and see what I end up with this time around.
-
Hi,
I'm much confused about Color Sync. Found this:
[attachment=21468:Screen_s...48.49_AM.png]
This seems to be assigned to color sync, although I never used that profile (AFAIK). Does the Bruce Frazer book (Real world Color Management) still apply to todays ColorSync?
Best regards
Erik
Hi,
There are some issues with profiling on Leopard/Snow Leopard.
Check this: http://www.luminous-landscape.com/tutorials/solving.shtml (http://www.luminous-landscape.com/tutorials/solving.shtml)
I'm in a fight with Snow Leopard, Color Munki, LR and CS3 right now. To me it seem that my prints are to dark. I used Babel Color to generate a synthetic Color Checker printed it and fed the measured values back to
Babelcolor. The screen dump shows my print using my own color profile on the left 'original' in the centre and "Epson Manages Color" on the right.
The bottom row are comparisons. It seem that color is quite OK but the L values differ a lot. Could it be a "gamma related issue"?
It seems that I get similar results in Photoshop CS3 and Lightroom 2.6. I also feel I have lived with the problem a long time on different OS-es and using both ColorVision Print Fix Pro and the Color Munki.
Sorry, not any of this solves your problem, but at least can add some fuel to the discussion.
[attachment=21467:Screen_s...27.11_AM.png]
Best regards
Erik
-
I mean, honestly, I only got this because I feel that the canned profiles I've been using are to high contrast and I lose a lot of the shadow tones to black, which is of course no good.
Hi Steven,
I think Jerry is probably on to your problem, the targets need to be printed without a profile. Also, make sure you've updated Snow Leopard to the latest version (10.6.3). Previous versions had issues with printing an untagged document in Photoshop.
Aside from that I have a feeling that you probably won't get a better profile by using a Spyder 3 system. Building profiles is not an instant gratification sort of deal. My experience has been that the profiles that the manufacturers build are usually quite decent. If you're after something better, sometimes there is room for improvement but getting those results requires lots of experimentation and specific tools. I have an i1 Xtreme (which uses the i1 Pro with Eye One Match) and I very quickly discovered that the Eye One Match software leaves a bit to be desired. It's a very basic software that is designed to be easy to use and as such it provides very little in the way of user input. It just wants you to select a target, print it, scan it, and be done. There are no real options to setup. That brought me to using Bill Atkinson's profiles and using the measure tool from the Profile Maker demo (it's a bit of a hack workaround but it gives you more options). I tried his 5202 patch profile figuring it would yield a more accurate result with the paper I was using but found out it created more problems (I suspect the i1 Pro might not be the most consistent spectro on the planet). I then moved down to about 900 patches and got BETTER results so... sometimes less is more. Then I got into the hack to change the gamut mapping with Eye One Match by editing the config file. This is getting me closer to the results I'm after but..... I'm still not satisfied so now I'm looking at Profile Maker or Monaco Profiler software. Also, sometimes one profile might be more appropriate than another depending upon the subject matter. I would highly recommend using Bill Atkinson's profile test images. You can download them from here (they're free).
http://homepage.mac.com/billatkinson/FileSharing2.html (http://homepage.mac.com/billatkinson/FileSharing2.html)
If you really want to build your own profiles I recommend.... patience and persistence. LOTS of it. You're going to go through a lot of sheets of paper experimenting. Use a decent test image for profile evaluation (I think the Atkinson file is a great resource) and also use a Granger Chart. Tools like ColorThink can also be useful as well for comparing profiles.
If you just want to get on with it and start printing I highly recommend the profiles that Booksmart Studio makes. I've bought a few of them in the past and they're very nice. Definitely better than the manufacturer's profiles in a lot of situations (although sometimes a certain profile might favor a certain image better than another) and much better than anything you're going to get out of the Datacolor system as the tools they're using are more sophisticated but also (and this plays a bigger role) they have a lot more experience in building profiles and that counts for A LOT.
http://www.booksmartstudio.com/ (http://www.booksmartstudio.com/)
I mean, honestly, I only got this because I feel that the canned profiles I've been using are to high contrast and I lose a lot of the shadow tones to black, which is of course no good.
I don't know where you're at with monitor profiling, viewing, etc. but what are you editing with and soft proofing on and what are you using to profile that screen? You mentioned you have a MacBook Pro but you didn't say anything about your monitor. The MacBook Pro is not really an appropriate screen to be doing critical editing with (I'm typing on a 17" MBP with the matte screen right now). For a laptop it's decent but it's just not a good tool for editing or soft proofing when you're making critical adjustments (and if you're a picky person and particular about the results you get, which it sounds like you may be, the problem gets compounded). I use an NEC 2690 for soft proofing. If you're seeing a mismatch between shadow detail and contrast I'd be looking hard at the MacBook (if that is what you're using) as the source of your problems. Also what light source are you using to proof your prints with? Most artificial light has spikes and gaps in the spectrum. If you're using an artificial light source there's going to be certain things that you're not seeing in your print. If you don't already have some SoLux bulbs I'd check 'em out. http://www.solux.net (http://www.solux.net) The monitor you use and the light source you use are SO SO SO critical in this process. If they're not absolutely accurate (or as accurate as possible) then you can spend all the time in the world trying to build printer profiles but if the adjustments you're making to your files are flawed because your monitor isn't telling you the whole story then you'll be going in circles.
Good luck!
Cheers, Joe
-
I'm in a fight with Snow Leopard, Color Munki, LR and CS3 right now. To me it seem that my prints are to dark.
You probably need to adjust the display unless the prints are really, really too dark no matter where you view them (and that has not been my experience with the Munki making profiles, which are usually excellent). It isn't a CS3 print issue (since you print the targets out of the host app).
-
I understand that there's probably better profiles available out there. I'm looking for something I can show that basically says that I'm pretty damn close. I do not use the macbook display for critical adjustments, and the contrast related issues that I'm experiencing are not a minor issue that can be related to monitor brightness. It's a HUGE difference from what I'm seeing on my displays to what I'm seeing in print, and that's unacceptable to me. I'm talking a total loss of detail in shadows, just clippinng to black, which I know for a fact the 9900 should not be doing.
And as said, this isn't a minor color shift with my profile, I'm going from a royal blue, to a royal purple, not even close. If I tried to deliver it to a customer they would laugh at me. Are you saying that the money I spent on the Spyder3 was a waste, because from what I've seen it should get me a decent profile. I'm not expecting it to be perfect...but something I can work from seems to be a reasonable expectation.
-
as an update I reran the targets, rescanned them, and ended up with a profile that I feel is significantly better than the canned luster profile using the 256 target scan. the blue channel is now DEAD on, and the shadow definition is much better. excellent, as said, I'm sure there's better options out there, but I'm much happier.
-
I'm happy with the Spyder3 Print system, It works well for me also. As mentioned earlier, with any print profiling system there is a learning curve. Turning off color management in the printer driver is a critical step in printing the targets.
With Datacolor Spyder3 Print you can further adjust paper profiles previously created to compensate for different lighting situations, depending on where you display the finished print.
-
And as said, this isn't a minor color shift with my profile, I'm going from a royal blue, to a royal purple, not even close.
Good point there... definitely wouldn't be a monitor issue in terms of the color. Issues with the exposure and shadow detail very easily could cause problems with the other issues you're having so I wouldn't rule that out completely. To get on the same page I'd be curious to know how bad the shadows are... if you could run a print of the lab test page from the Bill Atkinson link I'm familiar with those images as are a few others here. It might be useful for us to have a standard frame of reference.
If I tried to deliver it to a customer they would laugh at me. Are you saying that the money I spent on the Spyder3 was a waste, because from what I've seen it should get me a decent profile. I'm not expecting it to be perfect...but something I can work from seems to be a reasonable expectation.
I'd hesitate to say it was a waste but at that price point I'd be skeptical that it could deliver the same quality as one of Booksmart's profiles (last time I bought one of their profiles was 2 years ago, unless anything has changed I suspect they're still quite good). It doesn't cost much to try... they're about $4 (again that was 2 years ago) for their semi-custom profiles. For a bit more you can print the targets and send them in and they'll build one specific to your printer BUT... the 9900 is a really fantastic machine and it should be pretty consistent from unit to unit so I'd be surprised if you'd benefit from that. Used to be that the tolerances from printer to printer were rather sloppy but inkjets are so much more consistent now. Also... until you figure out that blue to purple problem you're not looking at a good profile.
I had a few more thoughts... you mentioned you were printing the profiles through the DataColor app. I'd maybe try printing through Photoshop. The profile as mentioned should be untagged and in the driver where it asks for the ICC profile you should set it to none. Everything else should be set up exactly how you'd set up your driver settings as if you were going to make a print. Same quality, same print method (bi-directional vs. uni-directional if you have the option), same media type, etc. Also, once printed give the targets time to dry. That still wouldn't explain the radical color shift you saw but the properties of the print will change over time. Usually 24 hours is the recommended amount of time. Another thought... make sure there's no strong source of light in the room when you're scanning the targets. It's unlikely this caused you any problem but if another source of light mixed in with the illuminant from your puck or if perhaps your hand accidentally lifted just a smidge while scanning over a certain patch allowing external light to leak in then that could potentially have an impact on your profile.
Cheers, Joe
-
as an update I reran the targets, rescanned them, and ended up with a profile that I feel is significantly better than the canned luster profile using the 256 target scan. the blue channel is now DEAD on, and the shadow definition is much better. excellent, as said, I'm sure there's better options out there, but I'm much happier.
Didn't see this post before I responded. Excellent! Congrats. As for better options I think that this price point the next step would be significantly more expensive. I've never done a comparison between the Spyder 3 Print and the i1 Xtreme system. The Datacolor software seems to be quite a bit more functional than the Eye One Match software which is REALLY basic (fortunately there are a few undocumented tweaks and tricks you can use that allow you to use different targets and different gamut mappings). I think the next step in terms of "better" would be a full spectrophotometer and a more powerful software package. At this point though you're talking about $2850ish bucks for the software and a spectrophotometer.
Cheers, Joe
-
Hi,
Thanks for good advice, just a few comments:
My prints are not horribly dark, it's just that they are significantly darker than than a real CC. This picture illustrates pretty much what I see:
[attachment=21493:Screen_s...27.11_AM.png]
Top row: Left: printed with CM-profile ------------------mid: "synthetic CC" right: Printed "with printer manages color"
Bottom row: CM-profile compared with synthetic checker. --------------------------------Right "printer manages color" compared with ColorChecker
All the above images were produced from CGATS files or spectrum data measured by the Color Munki.
Most of the deviation seems to be in the L-channel.
I also tried to convert the synthetic CC to the "Color Munki" profile but saw very little change from the original. My expectation would be that if I print a synthetic Color Checker the print would be very close to a color checker card.
Best regards
Erik
You probably need to adjust the display unless the prints are really, really too dark no matter where you view them (and that has not been my experience with the Munki making profiles, which are usually excellent). It isn't a CS3 print issue (since you print the targets out of the host app).
-
I think Jerry is probably on to your problem, the targets need to be printed without a profile. Also, make sure you've updated Snow Leopard to the latest version (10.6.3). Previous versions had issues with printing an untagged document in Photoshop.
Is this issue finally fixed in 10.6.3? Has the fix been documented by Apple? I have not seen this reported anywhere else.
-
Is this issue finally fixed in 10.6.3? Has the fix been documented by Apple? I have not seen this reported anywhere else.
Apple only lists printing reliability in the release notes. However several have tested this and for Epson printers as well as i1 match it appears to be resolved. As far as other drivers I can't say. See this thread.
http://luminous-landscape.com/forum/index....=41940&st=0 (http://luminous-landscape.com/forum/index.php?showtopic=41940&st=0)
-
Probably I'm now trying to kick in an open door:
when measuring the printed target: put some unprinted, white paper underneath to avoid that the spectro picks up the color of the desktop.
-
I've never done a comparison between the Spyder 3 Print and the i1 Xtreme system. The Datacolor software seems to be quite a bit more functional than the Eye One Match software which is REALLY basic (fortunately there are a few undocumented tweaks and tricks you can use that allow you to use different targets and different gamut mappings).
David Miller from Datacolor say: "if I measure a ColorChecker (or any other target print, with an arbitrary set of color patches) with an EyeOne and one of our devices, and compare the results, the average delta-e is less than one."
I've compared a Spyder3Print SR vs an i1Pro (and an i1Pro UVcut) for X-Rite ColorChecker and got the following result:
avr. dE00=3.72, max dE00=5.63
-
David Miller from Datacolor say: "if I measure a ColorChecker (or any other target print, with an arbitrary set of color patches) with an EyeOne and one of our devices, and compare the results, the average delta-e is less than one."
I've compared a Spyder3Print SR vs an i1Pro (and an i1Pro UVcut) for X-Rite ColorChecker and got the following result:
avr. dE00=3.72, max dE00=5.63
With something like ColorThink, you can line up the deltaE report and see if the high delta's are only in the Lstar which I kind of suspect based on the above screen capture. It would be interesting to sort by highest to lowest detlaE and then see where those high delta's lay. The values can be telling. For example, a 1 deltaE in aStar going the opposite direction of a 1 deltaE in the bStar can look a lot worse than you'd expect from just a report because they are going in an opposite direction. IOW, a 2 deltaE in one area can often appear to be "worse" than a 1 deltaE in differing directions in a and bstar because you compound the visual errors.
-
With something like ColorThink, you can line up the deltaE report and see if the high delta's are only in the Lstar which I kind of suspect based on the above screen capture. It would be interesting to sort by highest to lowest detlaE and then see where those high delta's lay. The values can be telling. For example, a 1 deltaE in aStar going the opposite direction of a 1 deltaE in the bStar can look a lot worse than you'd expect from just a report because they are going in an opposite direction. IOW, a 2 deltaE in one area can often appear to be "worse" than a 1 deltaE in differing directions in a and bstar because you compound the visual errors.
No, I didn't use the Lstar mode in Spyder3Print software.
The result sorted from highest to lowest detlaE is shown below (from top to bottom, from left to right). Patches with red border have color difference dE00>3; with yellow border - de00>1.5
The stats are shown on another figure.
P.S. I measured the ColorChecker three times with each instrument and averaged the results.
-
BTW half a year ago I compared a ColorMunki spectro with my i1Pro. For ColorChecker target the color difference was: avr. dE00=0.57, max dE00=0.96
-
No, I didn't use the Lstar mode in Spyder3Print software.
That's not what I'm suggesting. I'm suggesting that the deltaE error is mostly Lstar (light to dark) instead of color shifts. We need to know where the deltaE errors are happening, I'm thinking, based on what I'm seeing in the screen captures that its a luminance error.
-
Hi,
I have had limited experience with the Spyder Studio, but I find it to be a struggle to use compared to using the iOne iMatch with iOnePro. Here in Australia the iOne profiling solutions is more the 3 times more expensive. but I can see myself with more headaches using the Spyder Studio then the iOne solution. The spyder was slow and cumbersome, the results were no where near the canned profiles and miles away from the iOne. Which gave me great results.
From my limited experiences with it, I would not choose the spyder solution. I was very tempted by its price, but I had to return it. I am wondering if the ColorMunki is any better?
Henrik
PS: I am not saying I know everything about the Spyder and therefore dismiss it, but the experience I had was not encouraging to continue to use it.
-
We need to know where the deltaE errors are happening, I'm thinking, based on what I'm seeing in the screen captures that its a luminance error.
OK, look at the tables below.
-
I am wondering if the ColorMunki is any better?
The quality of the print profiles are easily on par and some could suggest superior (with the new core technology and iteration options) than EyeOne Match.
-
OK, look at the tables below.
It be useful if the software would sort by deltaE (how hard can that be), but yes, the big errors I see just glancing here appear to be the Lstar values which is odd and interesting at the same time.
-
It be useful if the software would sort by deltaE (how hard can that be), but yes, the big errors I see just glancing here appear to be the Lstar values which is odd and interesting at the same time.
Wait, I looked too quickly. What I really hopped to see, sorted by deltaE is the reference compared to the measured data if possible. I see you have the measured data from both products although, if you say the X-Rite package is closer, then it still points to a luminance error or issue with the Spyder. Or is the EyeOne values the measured data from the target itself (which is kind of like our reference)?
How many patches did you measure to build the Spyder profile? Could the instrument be iffy?
-
Wait, I looked too quickly. What I really hopped to see, sorted by deltaE is the reference compared to the measured data if possible. I see you have the measured data from both products although, if you say the X-Rite package is closer, then it still points to a luminance error or issue with the Spyder. Or is the EyeOne values the measured data from the target itself (which is kind of like our reference)?
How many patches did you measure to build the Spyder profile? Could the instrument be iffy?
I've got original X-Rite ColorChecker (X-Rite don't supply a ColorChecker with individually measured reference data) and measured it with i1Pro (using MeasureTool from ProfileMaker) and with Spyder3Print SR. And then compared results.
-
OK, look at the tables below.
Hi ProbeP,
Looking over the table, I see a tendency of your Spyder3 to be reading the L* channel too high. Note, for example, that on the A4 white patch of the Colorchecker you were reading 99.8 with the Spyder3. That patch is not that bright and should be coming in around 96-97 like the i1 reads it. This suggests you may be having a little trouble calibrating the Spyder unit on it's white tile. It tends to be a little snug, at least on the two units I've used, so you want to make sure that when you stick the unit's snout into the trough of the calibration tile, that you are applying enough downward pressure to get it correctly into contact. If you don't, the unit will "see" the tile as darker, and then treat that lower reading as the reference value. The net effect is that your subsequent measurements will read high, which would explain your results.
I have two Spyder3printSR units (one I bought, and the other kindly sent to me by Datacolor because I teach some courses here at AaI&A). They match my Spectrolinos pretty close, only a little further off on the green colors than I'd like to see, but not as much as you are experiencing, and definitely not enough to cause a serious error with profiling. Measurements of the neutrals are very close indeed, which is why I suspect you may just have a minor problem seating your unit into it's tile when making the initial calibration.
One other comment for new SpyderPRintSR users. I was definitely "user-challenged" initially when using the scan mode on the SpyderPrintSR unit. I had to train myself to do it right, and the way I ended up teaching myself how to get the correct hand-eye coordination was to build an Excel Spreadsheet, take two sets of measurements, and then compare delta E's in my spreadsheet. At first, I had some real fliers in the two data sets, but as I got into the groove, I found instrument repeatability was very good, with an average delta E between the two sets of measurements less than 0.5. Not bad for an inexpensive colorimeter (i.e., spectrocolorimeter, or whatever one calls it).
cheers,
Mark
http://www.aardenburg-imaging.com (http://www.aardenburg-imaging.com)
-
MHMG
Yes, you are right about high L*.
But I calibrated the Spyder3 each time before each measurement. BTW I measured the ColorChecker 3 times and then averaged results.
OK, maybe my measurements were incorrect. I'll re-measure the ColorChecker. In the dark room.
BTW profiles created with the Spyder3Print SR are good enought. And the difference between the i1Pro and the Spyder3Print SR for non-original ColorChecker (printed copy of the ColorChecker) is much better than for original CC. Maybe my Spyder3Print SR doesn't "like" original CC?
-
MHMG
Yes, you are right about high L*.
But I calibrated the Spyder3 each time before each measurement. BTW I measured the ColorChecker 3 times and then averaged results.
OK, maybe my measurements were incorrect. I'll re-measure the ColorChecker once more. In dark room.
BTW profiles created with Spyder3Print SR are good enought. And the difference between i1Pro and Spyder3Print SR for non-original ColorChecker (printed copy of the ColorChecker) is much better than for original CC. Maybe Spyder3Print SR don't "like" original CC?
Yes, but I'm just saying if you don't press down firmly on the unit when calibrating, it may not seat correctly, and therefore not calibrate correctly. Just letting it rest on its own weight is definitely not good enough to get proper contact with the calibration tile. Press firmly! I suspect this is the key source of your measurement error, at least as far as the hand readings you are making on the CC.
As for profiling, I initially generated a very poor profile with the Spyder3Print SR unit using the scan mode and reading an "extended greys" target. The profile had rather bizarre behavior in the shadows resulting in a sort of posterized appearance. When I began to investigate my measurements as the possible source of the problem, I realized that I was indeed making very poor measurements. Bad measurements build lousy profiles. It's not the fault of the software profiling engine nor the instrument accuracy itself. It's operator error caused during the "get acquanted" phase of learning how to use the product correctly. Datacolor should perhaps give more "heads up" advice on how to use the instrument, but apparently the vast majority of its customers aren't a user-challenged idiot like me Anyway, it wasn't the instrument itself. It was my initial scanning technique that introduced the large errors. In fact, in subsequent conversations with the folks at Datacolor, it became clear there had been an encoding error in the instruction video that leads the new user into thinking the appropriate scan speed is significantly different than what it should be. I hope Datacolor has since fixed this error in its tutorial regarding correct use of the device. When you scan way too fast or way to slow, the software will alert you, but when you scan slightly too fast or slightly too slow, you pick up leading or trailing edges of each patch in a way that the software doesn't flag, and then you have big errors in your measurements used to build the profile. So, get that sorted, and then it works great and builds very nice profiles. But that said, if you haven't "trained" yourself correctly then garbage in garbage out.
BTW, I understand the Colormunki can also do a great job building profiles as Digital dog has noted, but it will have definite issues with its iterative measurement technique when used to profile dye-based printers. Short term drift of the dye/paper combo can lead to large errors if you don't slow the whole profiling process down to allow for reasonable "drying time" of the dyes. Not a big problem with pigments (they settle to final color in a matter of minute/hours), but dye-based systems can take many days to settle into their final color. Thus, the Datacolor product, with it's single target measurment process, has an advantage over the Colormunki when profiling dye-based systems (you only have to wait for one sufficient dry-down period, not two). I"m just saying... the low end packages can do a very nice job if you work with them patiently, but the higher end packages like the i1 pro are also money well spent if you can go the distance on the higher price.
cheers,
Mark
-
When I began to investigate my measurements as the possible source of the problem, I realized that I was indeed making very poor measurements. Bad measurements build lousy profiles. It's not the fault of the software profiling engine nor the instrument accuracy itself. It's operator error caused during the "get acquanted" phase of learning how to use the product correctly.
I wouldn't let Datacolor off the hook all together here Mark. They did design the instrument and process so if a number of users are experiencing what you and the OP report, that's bad. It should at the very least report a questionable measurement or, like the EyeOne and Munki, report errors to the user. Bad enough it requires a one patch at a time process which sounds like agony (from someone that did this back in early 1990's using a ColorTron <g>) but to end up with a lousy profile is kind of inexcusable.
-
Yes, but I'm just saying if you don't press down firmly on the unit when calibrating, it may not seat correctly, and therefore not calibrate correctly. Just letting it rest on its own weight is definitely not good enough to get proper contact with the calibration tile. Press firmly! I suspect this is the key source of your measurement error, at least as far as the hand readings you are making on the CC.
Yes, you are absolutely right. I've re-measured X-Rite ColorChecker with Spyder3Print SR folowing your advice. Moreover I measured the ColorChecker 7 times, then averaged results. Yes, it was better: the difference between MY Spyder3Print and MY i1Pro for my ColorChecker was avr. dE00=1.08 (max dE00=2.89).
-
I wouldn't let Datacolor off the hook all together here Mark. They did design the instrument and process so if a number of users are experiencing what you and the OP report, that's bad. It should at the very least report a questionable measurement or, like the EyeOne and Munki, report errors to the user. Bad enough it requires a one patch at a time process which sounds like agony (from someone that did this back in early 1990's using a ColorTron <g>) but to end up with a lousy profile is kind of inexcusable.
Well, there is truth in what you say. The Spyder3PrintSR does have a scan mode that, once you get the hang of it, is reasonably convenient. But, its "scan" mode is certainly more primitive than the Xrite method for the i1 and Colormunki. As I understand it, the Xrite instruments take multiple readings as the unit moves over each patch and then Xrite uses the software and reference values to isolate the "good data" that is read somewhere within the area of each patch. Spyder3 printSR uses a very basic IR detection scheme that looks for the black/white stripes at the leading edge of each patch which tells it that the next patch is coming. Then it further relies on a simple timing mechanism (something like 0.5 seconds after the IR stripe detection) to trigger the color reading of the patch.
The obvious weakness in the Spyder3PrintSR scan mode method is that if you are moving the instrument at a steady pace forward but nevertheless going too slow or too fast, the single color reading doesn't occur in the middle of each patch. The software does detect when you clearly miss, but it doesn't easily detect a situation where, for example, the aperture has not made it fully onto the patch and a little of the stripe is still being integrated into the reading. Hence an error, but not so big that the software can detect. Ditto, if you move to fast and the reading takes place with the aperture now starting to hit the trailing edge stripe. Thus, for users who haven't taken the time to figure out this end-user scanning pace like I did, the Datacolor product does invite some user-skill problems. With the bigger "EZ targets" there is more forgiveness, with the smaller "Expert" targets, I consider this scan mode technique to be challenging at best. Better to read the smaller-sized Expert patches with the single-patch reading mode rather than the new "scan" mode.
Also, if you follow my logic on how the Spyder3PrintSR detects patches, you will conclude as I did that a way to improve one's odds of landing in the middle of each patch at the right time is not to go at a steady scan speed forward, but rather to move quickly over each stripe then pause deliberately in the middle of each patch and wait for the "click" sound taken as it measures, then rinse and repeat as they say. This quick-slow scan technique improved my measuring success greatly, and I was even successful with this measuring technique on the smaller Expert patches. Anyway, I suggested to the Datacolor team that they build into their software a feature like Xrite's Measure tool that let's you compare two sets of readings. Then offer the ability right in the software to go back and remeasure the patches where the two readings don't agree). I was told that this feature would be too complicated for the typical user, but ironically, the current Spyderprint software does support advanced features like averaging of two data sets and exporting data to text files. Go figure!
So, to be fair, the Datacolor product does build decent profiles, and has many nice features in the software that photographers will like, but it also demands some skill and patience from the user that some people may find frustrating. Simply put, make sure you are are getting the instrument fully in contact with the calibration tile by putting a little additional downward pressure on the head, and also take the time to train yourself on the correct pace when using the instrument's new "scan mode" feature. Once I had these two issues figured out, then I was able to get very good and repeatable results. That said, I heartily recommend the "compare two sets of measurements" routine for optimum results, but this perfectionist's approach does mean having to take two complete sets of readings, and also having some knowledge of Excel since you can't do the delta E comparison in the software at this time.
-
Yes, you are absolutely right. I've re-measured X-Rite ColorChecker with Spyder3Print SR folowing your advice. Moreover I measured the ColorChecker 7 times, then averaged results. Yes, it was better: the difference between MY Spyder3Print and MY i1Pro for my ColorChecker was avr. dE00=1.08 (max dE00=2.89).
Good to hear. Your inter-instrument agreement may not be quite up to Laboratory grade instrumentation, but for the price, one should not expect it to be. Nonetheless, it appears that your Spyder3PrintSR unit is providing more than adequate color accuracy to build very good custom profiles. They will be much better than the hit-and-miss world of generic profiles. Step outside the OEM printer-paper-ink combinations with tuned media settings and carefully matched profiles, and I believe it's well worth owning an instrument like the Spyder3PrintSR or the Colormunki. In addition to building custom profiles, they are great little tools for periodically checking that your printers, inks, and papers aren't drifting away from their normal output. That said, the i1 is probably the most versatile color measuring tool on the market today for serious photographers and printmakers who can afford the higher price tag.
cheers,
Mark
http://www.aardenburg-imaging.com (http://www.aardenburg-imaging.com)