Spyder3Print SR device vs. Spyder3Print device

[drwillie]

Does anybody here know if there is any quality difference in the ability to read profiling targets and generate printer profiles between the Spyder3Print SR device and the Spyder3Print device (Datacolor Spectrocolorimeter Model #1005)? I know that the "SR" device can scan an entire row at a time while the earlier model can only read one color patch at a time. I own the Model #1005 and was wondering if the "SR" model can produce better profiles.

 

[Petruska]

There's a Yahoo DataColor group with CB Tobie and David Miller, two of the top two guys at DataColor, answering questions.

In the past they satated that you would probably see no profile image quality differences between using the new and old colorimeters.

I have the Spyder3Print and went many rounds with these two guys, their support group, to try to eliminate my very slight yellow color cast on prints. They basically said that my data was good and that was the best results I could obtain. Two kind souls on this forum made me custom profiles using their expensive X-rite I1 Pro auto scanning systems. Results, no yellow print cast on prints. I did my homework and found out that the older X-rite Pulse ColorElite system is basically the same quality as the current I1 Pro spectros. The Pulse system is not longer produced and supported by X-rite but their web site still has a support page with applications and drivers. The Pulse will work with Windows, XP and Vista, not Windows 7 directly. I installed Windows 7 Vitual XP mode and run the Pulse just fine on Windows 7 within the XP mode. The Pulse produces profiles that are spot on, I have yet to need to adjust a profile. The scanning of 729 patches takes less than 5 minutes, no scanning technique required, you can scan fast or slow and never a mis-scan.

You can usually find the Pulse on ebay for $200 or less, just make sure that you get the application disk with serial number, then just contact X-rite, tell them that you bought it used, and they will give you the needed registration info.

Now I have also compared the Pulse generated profiles against the X-rite ColorMunki (also profiles monitors) and the profiles were extremely close to the Pulse. I now highly recommend that users buy the ColorMunki so that they can calibrate both their monitors and printers fairly cheap and with ease, to obtain very good, if not great, results.

Bob P.

 

[drwillie]

Bob P. -

I appreciate your response to my question. I have the older model Spectro from Datacolor and I wanted to know if it would be advantageous to upgrade to the new "SR" model. The speed of patch reading is not that important to me because I rarely have a need to make profiles. I am concerned with the profile quality, though. Thanks again!

David

 

[Petruska]

I guess my bottom line is that I personally don't care for the profiles that DataColor products generate, they are not as accurate as X-rite product. With the Spyder products you can edit foever and not get it right and I don't believe that one should need to edit profiles if measured and generated properly. With X-rite products I have yet found a reason to edit a profile and that's the way I want it all to happen.

Bob P.

 

[sjwolfhope]

David,

I have the Datacolor Spyder SR and have been happy with the results I have gotten. I haven't gotten the yellow cast to pure white that Bob P. has gotten and many of my prints have major white areas - clouds and white covered bridges. The only major problem with Spyder profiles is a blue-purple shift that can be taken care of with a slight color shift of the targets. Datacolor agrees to the blue problem and has promised an update - but that has been long in coming.

I really find it hard to believe that X-rite profiles don't require an occasional edit. The reviews I read about the Colormunki before deciding on the Spyder SR talked about editing and how involved it is. Editing with the Spyder is much easier than with the X-rite - at least the Colormunki. As I understand it the Colormunki requires a new target print from the first profile, a drying period and rescan. With Spyder make the edit, generate a new profile and do a test print.

An update to the SR version would make generating profiles faster with the ability to strip scan. One plus to the Spyder SR software is the ability to make several passes and average the data results - minimizing errors due to a bad reading. I also put my scan data into a spreadsheet and can measure the Delta-E of each measurement for two passes. Large Delta-Es for a patch reading mean a bad reading. Sure these extra scans and such take time but the resulting profiles have been very good to my eye.
Just my $0.02 worth.

Steve W.

 

[Marcamera]

I use the Spyder3, and cannot get the same colour on Epson papers as generic (Epson 3880). very slight difference, but still.

Perhaps this is a silly question:

Could colour cast (yellow in this case) be the result of ambient light getting into a gap between the tip sensor and paper during measurment?

On spot measurements, the tip rests against the paper but there still may be a gap resulting from not a complete contact. When using strip method, the tip almost touches the paper, there may be ambient light ingress. Is it possible that this is the reason?

Or, as they recommend , profiling should be done on two underlying sheets of paper of the same kind to eliminate bleed through from the surface under the paper. Could this introduce colour cast?

rgds

 

[Sailor Blue]

Or, as they recommend , profiling should be done on two underlying sheets of paper of the same kind to eliminate bleed through from the surface under the paper. Could this introduce colour cast?

Definitely!

 

[Petruska]

I don't use the ColorMunki, I just ran a test to compare against my X-rite Pulse and it was very close. The X-rite Pulse is basically the I1 PRO , I have yet to edit a profile and I have generated at least 50+ profiles, if not more, for my Canon PRO9000, IP4300, IP4700 printers along with a slew of papers using HobbieColors, Image Specialists, and OCP inks. The prints match my NEC monitor almost perfectly (the color physics limitations), I couldn't be more happy! The Pulse scanning system is lightning fast, as fast as you want to scan without any scanning errors. The new I1 Pro doesn't have this great Pulse scanning guide and is slower. These Pulse systems are so great as I'm concerned that I bought two more in the last 2 months on Ebay as they don't show up too often and buyers are now aware what they a great job that they do at a much lower cost than the I1 Pros. Also the Spyders are Colorimeters, the X-rite products are true Spectros...

Could I have had a bad Spyder colorimeter giving me the yellow color cast, yep! I requested a few times from DataColor to allow me to send in target prints, have them scan them and email me the profile to prove it good or bad. They wouldn't take me up on that offer....thus the end of Datacolor products for me. And as you have stated the purple color issue and that was over a year ago, no update to the application. With the Pulse I don't need to shift target colors to obtain the correct profile colors.

Bob P.

 

[sjwolfhope]

I would also recommend doing the scans in a area with lower ambient lighting which would lessen the chance of any leak through.

Steve W.

 

[BenHere]

@David, Bob P is correct that there is for the most part is no perceptible difference between profiles created with older 1005 device and the new strip reading Spectro.
Just be sure you are running the latest version of Spyder3PrintSR.

@Steve W., we did just release an update to Spdyer3PrintSR. Use the software update feature to install the update and let us know if this helps.

@Marcamera, As long as you are scanning on a flat surface light leakage shouldn't be an issue. Using two sheets of the same paper underneath is to eliminate a color cast possibility. For you I would recommend creating a support ticket and we can help you out in detail related to your specific yellow color cast.
You can create a support ticket from here:

http://support.datacolor.com/index.php?_m=tickets&_a=submit&languageid=1&group=colorvision

You can reference my name in the title and then I will be the one to start helping you.

@Bob P., Sorry we weren't able to help you to your satisfaction...

--
Ben Vaccaro
Technical Support
Datacolor AG
5 Princess RD,
Lawrenceville, NJ 08648, USA

 

[tony22]

BenHere, as you are visiting from Datacolor I'd like to ask a question. I bought the S3P SR after reading many reviews. While the program does have a great profile editing capability, it does seem like it almost needs to be used to tweak a created profile. While I can live with it, it does kind of bother me. Can you detail what improvements have been incoporated into the latest software release and where we will see those improvements in terms of profile accuracy?

 

[Dominic.Chan]

@Steve W., we did just release an update to Spdyer3PrintSR. Use the software update feature to install the update and let us know if this helps.

I can't see any difference in the profile generated using 4.1.1 and 4.2.1, including the banding in the Cyan ramp.

 

[tony22]

I can't see any difference in the profile generated using 4.1.1 and 4.2.1, including the banding in the Cyan ramp.

I've never gotten any banding in the Cyan areas, but after I installed 4.2.1 just for kicks I did a 225 patch profile for Canon's own Photo Paper Plus Semi Gloss (I already had the patch pages printed from a number of days ago so they were fully set). The profile had to be tweaked to +10 Brightness and -6 (really +6 into) Cyan. Like I said, I love the ease with which I can tweak profiles, but why do I have to do it ? And on Canon's own paper no less.

Before anyone asks, yes my Sony CRT monitor is profiled, 6500K, gamma 2.2 - with a Spyder3!

 

[Dominic.Chan]

I've never gotten any banding in the Cyan areas,

I only noticed the cyan banding in a synthetic cyan ramp (on the Outback test image). I have not encountered that in "real" images.

The profile had to be tweaked to +10 Brightness and -6 (really +6 into) Cyan. What viewing light are you using, and at what intensity? Most of the so-called 6500K lights have a lower colour temperature than 6500K, when measured.

Also, have you compared print-to-print (with a Canon profile), in addition to comparing with the screen?

why do I have to do it/? And on Canon's own paper no less.

Most papers have optical brightening agents that will make the paper look white (but blue to the spectro/calorimeter). Spyder3Print has an option "Match paper grey" that you can trying turning on/off to see if it makes any difference.

 

[tony22]

Dominic, yes my home office does use the 6500K bulbs, but I have not measured them to see how accurate they are. Having said that, I'm not sure even if they were off by a little bit that it would explain the level of tweaking required in these cases. Almost every profile I've made so far requires some tweaking, and not in the same direction (varies from paper to paper, providing a bit more support to the probability that it's not the lights). The stock Canon profile, BTW, is way too Cyan although the luminosity is correct. It's ironic that the S3P SR profile needs Cyan in order to bring it back into line with a neutral target.

I did read the PDF for S3P SR and did use "Match Paper Grey" when creating the profile since PPPSG uses optical brighteners.

 

[dismalhiker]

The profile had to be tweaked to +10 Brightness and -6 (really +6 into) Cyan. Like I said, I love the ease with which I can tweak profiles, but why do I have to do it ? And on Canon's own paper no less.

I assume you mean that the profile had to be tweaked to +10 brightness in order to match what you see on your profiled monitor. Is this right?

If so, then it seems likely that the fault is not with Spyder hardware/software, but is much more likely that your monitor is too bright. If your monitor is too bright, then it makes sense that the profile would have to be tweaked to bring up the brightness of the print. Modern monitors are capable of producing a lot of light.

What is the luminance of your monitor? I had the same problem as you describe until I followed Eric Chan's advice and reduced the target luminance to 100 cd/m2 during monitor calibration. The Spyder3 software has the ability to measure the luminance and adjust it precisely.

Now my profiles produced with Spyder3Print are great without any tweaking. Having said this, my main photo-editing monitor is not bright set at 100 cd/m2. I wouldn't say "dim", but its noticeably not very bright. I have gotten used to it and I keep my 2nd monitor on the system a bit brighter.

Finally, to my knowledge, using these relative inexpensive monitor cal+print profile systems, the best one can do is "calibrate" the monitor in the sense of having optimal gamma, luminance and color temperature. However, this so-called "calibration" is independent of the printer. In my view, so-called monitor calibration gets the monitor settings "approximately right" within tolerances that most normal people will accept. However, the monitor cal is not linked to the printer output directly (to my knowledge). Spyder3print determines the discrepancies between a standard color space test sample vs. what comes out the printer and generates corrections. However, it seems to me that these color corrections (profiles) have nothing what-so-ever to do (directly) with what is seen on the monitor.

So... it isn't clear why one would expect the software-generated profiles to be perfect for all (so-called "calibrated) monitors. It seems to me that a normal, expected part of the profiling process is to either tweak the monitor to look like the print (e.g., reduce luminance) or tweak the profile (e.g., + brightness) so the print looks like what is on the monitor.

PS... I'm not an expert, just someone who had the same problem and found how to fix it. I found that the luminance measurement and adjustment makes a huge difference. Still learning...

John (dismalhiker)
--

http://www.flickr.com/dismalhiker
http://ouachita.dismalhiker.org/

 

[tony22]

If so, then it seems likely that the fault is not with Spyder hardware/software, but is much more likely that your monitor is too bright. If your monitor is too bright, then it makes sense that the profile would have to be tweaked to bring up the brightness of the print. Modern monitors are capable of producing a lot of light.

What is the luminance of your monitor? I had the same problem as you describe until I followed Eric Chan's advice and reduced the target luminance to 100 cd/m2 during monitor calibration. The Spyder3 software has the ability to measure the luminance and adjust it precisely.

John, did you miss where I said my monitor was calibrated? The luminance is 99.6 cd/m2. I use a Spyder3, and have been using Colorvision / Pantone / Datacolor devices to calibrate my monitor since the original Spyder. Been doing this a while.

 

[Dominic.Chan]

Dominic, yes my home office does use the 6500K bulbs, but I have not measured them to see how accurate they are. Having said that, I'm not sure even if they were off by a little bit that it would explain the level of tweaking required in these cases. Almost every profile I've made so far requires some tweaking, and not in the same direction

One problem with the S3P software is that it does not seem to make any attempt to detect misread patches. As Steve W pointed out in his post, his workflow is to read the patches twice and compare the results. I use a similar workflow (and also examine the shape of the 3D profile to confirm that there are no irregularities that would indicate misread patches).

The stock Canon profile, BTW, is way too Cyan although the luminosity is correct. It's ironic that the S3P SR profile needs Cyan in order to bring it back into line with a neutral target.

I believe you have a Canon Pro9000II. If everything is working properly, there should be no reason for the print to be "way too Cyan", especially if you're using a profile with the correct Canon paper. If the S3P profiles needs to correct a heavy Cyan cast, it's actually not entirely surprising that it may overcompensate in some areas.

I did read the PDF for S3P SR and did use "Match Paper Grey" when creating the profile since PPPSG uses optical brighteners.

Actually I have seen cases where turn off that option produces a more neutral print. You can quickly test to see if it makes any difference (no rescanning required). However, as pointed out above, the problem may lie elsewhere if you're getting a strong Cyan cast with Canon profile.

 

[dismalhiker]

John, did you miss where I said my monitor was calibrated? The luminance is 99.6 cd/m2. I use a Spyder3, and have been using Colorvision / Pantone / Datacolor devices to calibrate my monitor since the original Spyder. Been doing this a while.

Tony22, I didn't miss that your monitor was "calibrated". OK... but you didn't say what the luminance setting is. Many people "calibrate" monitors and set the luminance too high. The Spyder3 calibration procedure does not require any particular luminance setting. The statement that you calibrated it could mean many things.

Experience aside, your post sounded like you expect to "calibrate" the monitor, profile the printer and then never have to tweak a profile. You even put into italics "why do I have to do it?" Why would anybody expect that just because the monitor is "calibrated", the profile would be perfectly matched, in all characteristics, to the rendition portrayed by the the monitor? I think the kind of tweaking of profiles you describe is completely expected.

John

--

http://www.flickr.com/dismalhiker
http://ouachita.dismalhiker.org/

 

[tony22]

Experience aside, your post sounded like you expect to "calibrate" the monitor, profile the printer and then never have to tweak a profile. You even put into italics "why do I have to do it?" Why would anybody expect that just because the monitor is "calibrated", the profile would be perfectly matched, in all characteristics, to the rendition portrayed by the the monitor? I think the kind of tweaking of profiles you describe is completely expected.

John

If Colorvision is any example then I would have to agree. When I think back on my days using ProfilerPlus (with an IT-8 calibrated flatbed scanner as the "measuring" device), I still had to adjust the profiles. Recent comments, though, by Bob P would seem to suggest that the Colormunki requires less of this sort of thing. I won't try to compare that or the S3P to an i1Photo Pro, but in all the forums I participate in I can't say I've seen many posts at all talking about having to tweak i1 profiles. And if Bob is right about the -munki then maybe it's something about Datacolor's process that has a lower percentage of success with generated profiles.

The luminosity for both Moab Lasal and Ilford Classic Pearl, BTW, are right on the button. Go figure.