Spectrophotometer recommendation?

[paulr]

I’m on my 3rd Epson 3880 (don’t ask). The previous ones matched each other, and matched the screen perfectly when using generic profiles. This one’s a bit off, but works perfectly well otherwise.

I’m thinking this means it’s time to make some custom profiles. I’m looking at the i1 pro series, probably an older generation. My budget is maybe $400.

Are there good choices in this price range? Any considerations or warnings? What about software?

Thanks for any and all tips.

[sanking]

I’m on my 3rd Epson 3880 (don’t ask). The previous ones matched each other, and matched the screen perfectly when using generic profiles. This one’s a bit off, but works perfectly well otherwise.

I’m thinking this means it’s time to make some custom profiles. I’m looking at the i1 pro series, probably an older generation. My budget is maybe $400.

Are there good choices in this price range? Any considerations or warnings? What about software?

Thanks for any and all tips.

I use an iOne Pro and am very pleased with the operation. You can use the Measure Tool, part of Profile Maker 5.0, to create profiles. Profile Maker does not work with MAC systems later than Tiger so I use it with an older MAC laptop. You can find directions for making profiles in the QuadToneRip folder, along with gray scale targets. I believe you can use it with Lion with software called Color Port but am not familiar with the software.

You should be able to buy an iOne for about $400. It originally came in a small suitcase with a number of accessories. You want to make sure that you get everything that came with the unit because X-Rite does not sell these items any longer. Ruler, for example.

Sandy

[paulr]

Thanks Sandy.

Is ProfileMaker the only software that works with the i1 pro? Big drag if I need a dedicated computer to use the thing.

[Tyler Boley]

no I believe there are various price points available for i1Profiler, which will work on newer OSs. Check the xrite sight.

[Nathan Potter]

There are at least three other software packages that will work with the xrite i one Pro.

LaCie blue eye pro
ColorEyes Display Pro
BasICColor Display

The three I listed seem not to be favored generally, but as usual much of the unfavorability can be related to unfamiliarity. Don't know offhand what operating systems are functional for any of the above.

As of about a year and a half, the newer xrite i one Pro II is available and can be obtained as an upgrade from a registered i one Pro, but I've not taken advantage of that so don't know the cost.

Nate Potter, Austin TX.

[xavier deltell]

For me, the best profiles I've done with the old Monaco EZ Color (xrite)

[Peter De Smidt]

I have a Spyder Print (http://spyder.datacolor.com/portfolio-view/spyderprint/). It works well. I use it for color profiles as well as making custom QTR profiles for BW prints. Measuring large targets is tedious, but I don't change papers very often.

[paulr]

Thanks Peter. I haven't seen Spyder Print before. Any thoughts on what compromises you make with this $350 system? I'm comparing to $1500+ systems from x-rite, or $400+ used x-rite systems.

My choice based on what I now know would be between:

-A used i1Pro (previous generation) for around $400, plus some as-yet-determined profiling software (free? Expensive? I don't know the software choices)

-A new Spyder print system, hardware and software package, for $290.

(Color Munki should probably be on the list, but I recall reading some middling reviews a while ago. And it's spelled "munki."

[sanking]

Thanks Peter. I haven't seen Spyder Print before. Any thoughts on what compromises you make with this $350 system? I'm comparing to $1500+ systems from x-rite, or $400+ used x-rite systems.

My choice based on what I now know would be between:

-A used i1Pro (previous generation) for around $400, plus some as-yet-determined profiling software (free? Expensive? I don't know the software choices)

-A new Spyder print system, hardware and software package, for $290.

(Color Munki should probably be on the list, but I recall reading some middling reviews a while ago. And it's spelled "munki."


For the iOne Pro the software is basically free since you can download Profile Maker for nothing, and you don't need a license to use the Measuring Tool. However, as I mentioned, this sofware only works with MAC through 10.6.8, but the method is well-documented.

For newer versions of MAC OS with the iOne you might check out this tutorial, which is based on Color Port.

https://groups.yahoo.com/neo/groups/QuadtoneRIP/conversations/topics/10629

Sandy

[Peter De Smidt]

Hi Paul,

I don't have any experience with the i1pro.

Here's a decent review of the spyder print: http://www.northlight-images.co.uk/reviews/profiling/spyder3print.html

I've been happy with mine.

[Tyler Boley]

the only free software for building color profiles is argyll, it no doubt supports the i1 but has a learning curve unless you are used to command line work.
T

[paulr]

So, I settled on an i1pro kit that comes with a key code for the full software package (compatible with current Mac OS, btw).

It also comes with a monitor calibration attachment, which is not what I bought the thing for. But this raises the next question.

I have an NEC spectraview monitor with NEC’s own colorimeter. Spectraview is compatible with the i1pro as well. Any sense which is the better one to use?

Here’s what I’ve gathered so far:
-the i1, being a full-on spectrophotometer, and expensive, is a more sophisticate instrument, at least as far as color goes.

-But, the colorimeter that comes with the monitor is tuned specifically for the monitor’s gamut.

-And, one source I read suggests the i1 has worse low light performance than a typical colorimeter, making it less capable for calibrating gamma / linearity.

Any thoughts?

[Peter De Smidt]

How old is your Spectraview/monitor? Is the display an extended gamut type? Some of the older measuring devices don't do well with Adobe 98 capable displays. The current Spectraview II software was designed around the i1 Display Pro, although it can use other measuring devices. It shouldn't be too hard to try both and compare.

[paulr]

Mine's a few years old, and is a wide gamut model. The puck it comes with I believe is a rebranded and tweaked i1 display pro, as you said.

I realize I can compare profiles from both side by side, I don't know how I'd determine which is more accurate.

[Peter De Smidt]

Everything I've read has been very positive about Spectraview II and a i1 display pro. Mine is arriving on Monday, along with a 27" NEC.

[Greg Miller]

I realize I can compare profiles from both side by side, I don't know how I'd determine which is more accurate.

You should be able to see DeltaE from each profile and get a rough idea if either profile is meaningfully better than the other.

[paulr]

You should be able to see DeltaE from each profile and get a rough idea if either profile is meaningfully better than the other.

Can you explain? So I generate two profiles and one shows a higher Delta E. How do I know which is more accurate?

[Greg Miller]

Can you explain? So I generate two profiles and one shows a higher Delta E. How do I know which is more accurate?

A higher DeltaE means the colors are further from their target than lower DeltaE. I believe the iOne Pro gives you an average DeltaE and Max DeltaE for a profile (Max DeltaE meaning the worst DeltaE value within the entire profile). It's possible you could have one profile with a better average DeltaE but worse Max DeltaE, which would make it hard to say which is better. But if one profile has a significantly better average DelatE I would be inclined to say that profile is better. If the two average DeltaE's are similar than it's a pick'em scenario.

[paulr]

I think I'm failing to understand on a very basic level. Isn't the Delta E showing the measured difference between the Monitor's actual state and its ideal, which is then compensated for by the profile? And no matter what it is, isn't it being measured by the very devices we're comparing?

[Preston]

The SpectraView software will show you the average and max Delta E if you look at the 'Color Tracking' tab in the 'Info' window (Tools>Info). My P221W currently shows an Average Delta E of 0.73, and a max of 1.3. Values under 3.0 are considered to be good, but lower values are better.

One thing I've discovered is it's a good idea to run through the calibration process twice. The second run typically results in a much better (lower) Delta E, at least on my system.

I agree with Greg, that if the two Delta E values are quite similar, it is a 'pick one' scenario.

--P

[Tyler Boley]

Paul, your new software is basically i1publish with some of the more advanced and unneeded features disabled. For RGB printer profiles you will find it excellent. There are only a few gotcha's, one being the v4 vs v2 output profile build option.. stick with v2, v4 never went anywhere and some applications can't see those profiles. V4 may be the default, hence my warning.
The perceptual rendering is based on the old Monaco, so gorgeous, but with a new underlying technology in how the profiles are built.. very good.

[paulr]

Thanks Tyler. I remember about v2 vs. v4 ... I had some trouble with a paper manufacturer's v4 profiles.

The one other potential gotcha I see is in the ambient light profiling. I've always had good results with a not-quite-kosher method of using standard profiles, proofing under 3000K print viewing lights, and with my monitor calibrated to 6500K. This doesn't sound like it should work but it always has. I don't and won't have a prepress-style viewing booth, and 5000K screen calibration just doesn't work for me.

I didn't realize ambient light was built into the profile. What should I use, or at least start with?

[Tyler Boley]

D50 which should come up as the default. That will be as any profile we've made before given these options. I have yet to utilize this, nor hear of anybody but a tweaker making good use of it. If I was making prints for an unusual environment with something awful like sodium vapor lighting, I'd look into it more and give it a try, perhaps. The cool thing is you can save your chart builds, settings, and measurements, at every step along the way and come back and easily build different versions if need be.

[paulr]

Great, thanks Tyler. That's the answer I was hoping to hear. I may be a tweaker, but preferably not a profile tweaker.

[paulr]

So, I got the kit, and it’s really impressive. It came with a license for the all-features-unlocked version of profile-maker, so it can calibrate scanners, cameras, cmyk printers, projectors, etc.. I’m especially happy that there’s a colorchecker for camera calibration. And Profile Maker is Mavericks compatible.

I had to beat my head against the thing for a half hour to get it to read the patches. Turns out that when it says “try going more slowly” it really means “go much faster.” But after that hurdle it’s really quick to use, and the x-rite software is well designed. I spend a lot of time bitching about software. But this was really a breeze.

I made a profile last night just to make sure things are working. I’ll measure again today after the ink has had a day to dry.

I also tried using the spectrophotometer to calibrate/profile the NEC monitor. This was really interesting. I don’t understand what the Delta-E value is measuring … I had assumed it was showing the measured error that the profile corrected, so it didn’t make sense that you could use this to evaluate the precision of your profiling job. Can someone explain?

At any rate, delta-e with the NEC colorimeter was .74, and with the spectrophotometer was .34. Much more interesting is difference in reported color gamut. With the colorimeter, I always get a few more blues than Adobe RGB, but far fewer reds. With the spectrophotometer the monitor gamut almost completely engulfs Adobe RGB. If this is because of measuring limitations of the colorimeter, it’s strange … I’m using the NEC rebranded one that has a custom filter set for this monitor.

I’m a bit concerned about dark values. The spectraview software used to to report a number for the black point of the monitor. With the spectrophotometer it reports “NA”. And the delta E values for the darkest tones are very high.

I had read that spectrophotometers are less sensitive to very low light values. Considering this, would the colorimeter be a better choice for grayscale profiling, like with QTR?

[Greg Miller]

DeltaE desribes the difference between what a color should be, and what the profile is able to achieve (given the limitations of the hardware). No output device will be able to match colors exactly, so you will never achieve a DeltaE of 0.0.

[Peter De Smidt]

From what I can tell, the delta-e in the vary darkest tones isn't that important. My delta e average is 0.27, with a max of 0.40, not counting the darkest measurements. Counting everything , I get an average of 0.49, with a max of 1.68, the later of which is the black reading. Apparently anything under 2 is a very good reading. You can correct for the color but you give up a bit of contrast. [NEC PA272W, Spectraview II, i1 Display Pro.]

If I understand this correctly, delta-e quantifies the difference from ideal after the calibration and profile are taken into account. The curves display will show how the monitor was adjusted from pre-profile to profile.

I'm surprised that you get a result much different from Adobe 98 with the NEC puck. My measuring device is similar, and the results are very close to covering the full Adobe 98 gamut.

[paulr]

Yeah, it seems strange. Possibly something wrong with the puck. I bought it used with the monitor, and it's always shown a reduced red response.

What still confuses me about the delta E value is why the profile can’t completely correct for any variance the hardware/software is able to measure. At least within the realm of what it’s possible for that hardware/software to measure.

Another way to state my confusion: if a measuring device is evaluating itself, how can it ever detect an error? All 12” rulers say they’re exactly 12” long …

[Peter De Smidt]

The measuring device isn't evaluating itself. It's evaluating the monitor's response difference from an ideal response. There are just some colors, a very small amount, that the system can't produce, no matter what settings, calibrations, and profiles are used.

http://au.nec.com/en_AU/support/faq/each-time-i-calibrate-my-monitor-i-receive-a-different-delta-e-result.html

[Greg Miller]

What still confuses me about the delta E value is why the profile can’t completely correct for any variance the hardware/software is able to measure. At least within the realm of what it’s possible for that hardware/software to measure.

In addition to what Peter said, it's actually quite difficult to get all patches exactly right at the same time. It would be trivial to get each patch exactly right by itself. But it has to try to get ALL patches exactly right at the exact same time, using only 3 variables (R,G, & B). So as it adjusts for one patch, it might throw another patch, that was already correct, off.

[paulr]

So, I've started working with the first profile I built for the 3880. The colors seem dead-on but the print is dark. I'm using the same monitor calibration settings and viewing conditions that I always use. It wouldn't be hard to tweak the prints but it seems like something's a bit off. Does this sound like the result of a profiling mistake?

[Peter De Smidt]

What's your monitor intensity set at?

[paulr]

100 cd/m2, gama 2.2, 6500K.

This is what I've always used, and what gave predictable results with the generic profiles on my previous two Epson 3880s.

[Peter De Smidt]

100 cd/m2 should be ok.

[Tyler Boley]

I'm not sure what the problem may be here, and I do believe something is up because I have not experienced this with profile creation no matter the software. I can't think of anything in the build settings that effect overall luminosity lighter or darker. Your monitor settings are probably fine, though I actually set mine cd/m2 to 80, but all I do is printing. I don't think being at 100 should show a very discernible difference compared to a print than 80. Is it significantly different on paper than using the same profile in soft proof?
The remaining variable is always the problem of how to print a chart free of interference from today's helpful software engineers at adobe and apple. It's become absurd, I can't advise about that as I always print everything, including targets, through a RIP on a PC that sidesteps all of that.
When I have made profiles for friends, I always make them download the adobe print utility, and I'm told by others in the know that the new QTR print tool sidesteps all of that too, but it's not free.
Wish I had some suggestions...
Tyler

[sanking]

.......
The remaining variable is always the problem of how to print a chart free of interference from today's helpful software engineers at adobe and apple. It's become absurd, I can't advise about that as I always print everything, including targets, through a RIP on a PC that sidesteps all of that.
When I have made profiles for friends, I always make them download the adobe print utility, and I'm told by others in the know that the new QTR print tool sidesteps all of that too, but it's not free.
Wish I had some suggestions...
Tyler

There is a new version of the QTR Print Tool, now called simply the Print Tool. Not free, but not very expensive either, and sure makes printing a lot more trouble free if you are using adobe software and apple hardware. The Print Tool allows one to print with No Color Management, an attribute that is invaluable in my work.

Sandy

[Nathan Potter]

100 cd/m2, gama 2.2, 6500K.

This is what I've always used, and what gave predictable results with the generic profiles on my previous two Epson 3880s.

Paul, how are you measuring your monitor intensity?

Nate Potter, Austin TX.