CHROMiX

variation in measurements using different spectrophotometer

Hi everyone,

I’m new to this forum. I’ve been working in prepress for 13 years now and in the past few years I’ve been exploring the world of certified proofing.

It’s not always easy working with people who have always relied on their visual perception instead of measurements. Personally I believe in a combination of both. One thing bothers me though… the reliability of the measurementdevices.

Using an HP Z2100, Heidelberg ColorTool and an EyeOne I’ve managed to achieve mean deltaE values less than 1 (compared to the ISO 12647 standard), or so I thought. Visually the proofs looked okay, but not spectacular and way too red in the gray balance. When I measured the Fogra MediaWegde with an EyeOne I’d get good results, but when I measured the same MediaWedge with a Techkon SpectroDens I’d get a mean deltaE value of 2 or even more. And to my surprise the EyeOne gave contradicting values on individual patches compared to the Techkon.

I decided to make a new profile using the Techkon and now the proofs look a lot better. When I measure them with the Techkon I get great results (dE and dH).

My point is that the difference between these two measurement devices seems to be not only pretty big, but also very inconsequent. Both devices comply to the applicable ISO standards, but apparently this standard allows for differences that could make certified proofing a living hell.

The same goes for the ISO 12647 standard itself. Anyone is able to produce a ISO certified proof that is well within the margins, but looks like crap. A proof measuring a mean deltaE of 5 is still acceptable according to ISO.

Does anyone else here have experience in this field with different spectrophotometers?

edit: I see I’m not the only one: [url]http://one.imports.literateforums.com/t/spectrophotometer-calibration-and-tolerances/77/1]

What you describe is unfortunately not unusual. There are several recent studies, including the data obtained in the recent SWOP proofing study that indicate the state of inter-instrument agreement with spectrophotometers may be in conflict with the common perception of acceptable tolerances for proof verification. This is the case not only for spectros from different manufacturers, but also of the same manufacturer, and even the same make and model. The i1 pro devices can measure differently depending on rev level for example.

As far as your specific problem, its important to ensure the devices measure consistently, probably more so than they actually agree with each other. Ensure that both devices are using the same setting for illuminant and observer (D50/2 degree), are of the same filtration (I.e. Uvcut or none…your z2100s embedded spectro is uvcut if you happen to be using it) and are using the same backing method. Its not a bad idea to ensure both devices are within the manufacturers certification periods…at least initially.

The substrate itself can play a large roll in how instruments agree with each other. Often instruments that agree very closely on standard tiles have larger deviations on real world Substrates. The illuminants in the devices can excite the media differently and lead to deviation or surface reflection of the substrate can play a role.

These issues certainly cause difficultly and confusion when attempting to communicate proof verification data, and certain elements of the standards bodies have noted this and are looking into them.

FYI, ISO 12647 does not specify a mean delta E of 5, but rather the primary solids have this tolerance around them, and also a delta H of 2.5, which further limits the allowable hue deviation on the primaries. These values are useful for setting up a printer or press calibration, but not sufficient or intended for proofing tolerances in and of themselves. The swop proofing system certification tolerances are often adapted for day to day proof verification, and can often be used for such as long as the issues with instrument agreement are understood and the device data communicated to relevant parties.

mike

What you describe is unfortunately not unusual. There are several recent studies, including the data obtained in the recent SWOP proofing study that indicate the state of inter-instrument agreement with spectrophotometers may be in conflict with the common perception of acceptable tolerances for proof verification. This is the case not only for spectros from different manufacturers, but also of the same manufacturer, and even the same make and model. The i1 pro devices can measure differently depending on rev level for example.

As far as your specific problem, its important to ensure the devices measure consistently, probably more so than they actually agree with each other. Ensure that both devices are using the same setting for illuminant and observer (D50/2 degree), are of the same filtration (I.e. Uvcut or none…your z2100s embedded spectro is uvcut if you happen to be using it) and are using the same backing method. Its not a bad idea to ensure both devices are within the manufacturers certification periods…at least initially.

The substrate itself can play a large roll in how instruments agree with each other. Often instruments that agree very closely on standard tiles have larger deviations on real world Substrates. The illuminants in the devices can excite the media differently and lead to deviation or surface reflection of the substrate can play a role.

These issues certainly cause difficultly and confusion when attempting to communicate proof verification data, and certain elements of the standards bodies have noted this and are looking into them.

FYI, ISO 12647 does not specify a mean delta E of 5, but rather the primary solids have this tolerance around them, and also a delta H of 2.5, which further limits the allowable hue deviation on the primaries. These values are useful for setting up a printer or press calibration, but not sufficient or intended for proofing tolerances in and of themselves. The swop proofing system certification tolerances are often adapted for day to day proof verification, and can often be used for such as long as the issues with instrument agreement are understood and the device data communicated to relevant parties.

mike

-----Original Message-----
From: Raak
To: PrintProfilingForum@colorforums.com PrintProfilingForum@colorforums.com
Sent: Sat Jan 10 18:25:59 2009
Subject: variation in measurements using different spectrophotometer

Hi everyone,

I’m new to this forum. I’ve been working in prepress for 13 years now and in the past few years I’ve been exploring the world of certified proofing.

It’s not always easy working with people who have always relied on their visual perception instead of measurements. Personally I believe in a combination of both. One thing bothers me though… the reliability of the measurementdevices.

Using an HP Z2100, Heidelberg ColorTool and an EyeOne I’ve managed to achieve mean deltaE values less than 1 (compared to the ISO 12647 standard), or so I thought. Visually the proofs looked okay, but not spectacular and way too red in the gray balance. When I measured the Fogra MediaWegde with an EyeOne I’d get good results, but when I measured the same MediaWedge with a Techkon SpectroDens I’d get a mean deltaE value of 2 or even more. And to my surprise the EyeOne gave contradicting values on individual patches compared to the Techkon.

I decided to make a new profile using the Techkon and now the proofs look a lot better. When I measure them with the Techkon I get great results (dE and dH).

My point is that the difference between these two measurement devices seems to be not only pretty big, but also very inconsequent. Both devices comply to the applicable ISO standards, but apparently this standard allows for differences that could make certified proofing a living hell.

The same goes for the ISO 12647 standard itself. Anyone is able to produce a ISO certified proof that is well within the margins, but looks like crap. A proof measuring a mean deltaE of 5 is still acceptable according to ISO.

Does anyone else here have experience in this field with different spectrophotometers?

Post generated from email list