Is most color grading the grading of "TONE"?

[jasontb]

I should have said what I meant beacuase it seems nobody understood what I meant:

I don't know if diamonds should be color graded table up or down. I don't know enough to comment on that.

Now maybe I misunderstood, but I thought strmrdr was saying that maybe diamonds should be color graded by what color they 'look like' not by what color they 'are'. Such as: Evenly colored rough is split in half and two diamonds are cut. One is cut well and one is cut poorly. One gets a better color grade because it looks whiter face up. That seems wrong to me. Just tell me what color it is and I'll figure the rest out.

It reminds me of when AMD started marketing their processors with 'Pentium Equivalent' speeds. Don't call it an 1200 because it's as fast as a Pentium 1200-Mhz. Be honest and tell me it's a 800-Mhz processor. I am smart enought to understand that there's more to it than just the clock speed. Or how some (crummy) digital cameras tell you they have 3 million equivalent pixels. If it's a 1 megapixel camera with software that gives me an interpolated 3 million pixel image - I'd like to know that.

I guess the jist of my comment is that any color grading should be independant of other factors. And I though strm was advocating the posibility of inclucing other factors into color grading.

 

[valeria101]

Date: 3/31/2006 7:38:12 PM
Author: jasontb


Now maybe I misunderstood, but I thought strmrdr was saying that maybe diamonds should be color graded by what color they ''look like'' not by what color they ''are''. Such as: Evenly colored rough is split in half and two diamonds are cut. One is cut well and one is cut poorly. One gets a better color grade because it looks whiter face up. That seems wrong to me. Just tell me what color it is and I''ll figure the rest out.

That is precisely what happens with fancy colors. Say, yellow... and not rarely because from well formed crystals it is often the case that two diamonds are cut - a small and a large one. Not only cut, but also the volume influences observed color (this, regardless of how or if the piece is cut) - so the large piece is always substantially deeper color than the small.

I am afraid that what you say (grading color as a tone per unit of volume) is not near current practice. And if it would be done, would lead to grades totally removed from the appearance of the stones.

Do grades that are not meaningful visually (i.e. are not correlated with the appearance of the diamonds) make any sense ?

Maybe that could make sense for someone from some different perspective than jewelry usage (dunno, investment, or industrial applications).

Quick question for everyone:

Is there anywhere to be found a good report of how diamond grading evolved? I''d bet that the reasoning behind these standards makes a fantastic Renaissance story.

 

[valeria101]

Date: 3/29/2006 8:35:55 PM
Author:oldminer

Let''s consider ONLY grading the color of diamonds from D through M. There are several hues that are graded in the standard GIA scale which include Yellow, Brown Yellow, Brown, Greenish Yellow and all the slightly more mixed hues of these color types too.

Wondering if this problem is not particular to automated color grading: as if a sensitive device can discern hues and tones at extremely low saturation when they are hardly observable, but obviously does not have the ability to interpret them as humans instinctively do. As if the matter of color is exposed by technology where beforehand, intuition and practicality had done away with it without express analysis. If so, the problem opened on the thread could be placed in context as a choice to incorporate existing practice in the interpretation of automated grading or not (when technology offers alternative options, and procedural transcription of common practice is quite a bit of headache).


Would I like to know what color J-M diamonds are? Yes, sure. It would be fun and probably help some diamonds improve their appeal too. Even in the near colorless range it would be nice, although I can easily see myself staring hopelessly at a ''greenish yellow'' G.

 

[oldminer]

Just to clarify. The grading system we now use, the GIA system, grades the color of D to Q round diamonds from the side view. It grades large diamonds, 1.5ct and larger as if a small chip was nicked off and that little part graded alone in order to give the color of the material in spite of the appearance of the color as absorption of light darkens the larger view through the entire stone. This really gets noticeable as diamonds get 3ct and more. Why do this, one might ask? The reason is financial. The rough of high color stones is very costly and the economics of cutting make it better to cut larger diamonds than smaller ones, yet diamond dealers and cutters want the grade to be what the material is, not as it looks to be in a large size due only to absorption of light. The GIA saw the merit of this argument and that''s how the system has worked.

The problem is that no other substance is color graded in this way. Can anyone think of an example of color grading a transparent material, or selecting a color of a material based on a small sample of it? I can''t, and have thought over this quite along time.

Color grading face-up is done on fancy colors since the way a diamond is cut definitely can enhance the intensity and visibility of a color. It can also diminish the visibility of color. Would you really like colorless diamonds graded by face-up appearance instead of actual body color? Would you still agree to allow a subjective or algorithmic adjustment for the size of the diamond as it got larger?

These are rather complex issues and there is no agreement anywhere on how to move forward except we must admit the GIA system, however subjective or flawed, works well for marketing diamonds. To needlessly disrupt the market will bring economic consequences. I would not want to do such a thing without a lot of forethought. I feel certain that meaningful improvements to the present color grading scheme cold be brought forward, but without market understanding, it would be unwelcome, misunderstood, and criticized harshly. Its the kind of problem I enjoy thinking about as well as eventually doing something about.
Getting feedback on these subjects brings the process slowly forward. Thanks for all the participation.

 

[diamondlil]

Dave,
I just have to say thank you for this thread. Although I have nothing to add from a technical standpoint, this kind of conversation is EXACTLY why I joined PS in the first place so many years ago. There seems to be less and less educational discussion lately, and when a topic such as this pops up, I''m fixated on it.

Thank you.

 

[strmrdr]

to clarify my position:

Yes the material color is the color of the material but:
tone does not equal color.
a purple sapphire and a blue sapphire can have the same tone and not be the same color.

face up is more relevant than face down color.

 

[oldminer]

Strmrdr;

If you stick to sapphire or tanzanite of one hue and had a GIA color system that graded the depth of color with letters of the alphabet, then the slightly darker sapphire would be a lower letter of the alphabet. With diamonds, we look at colors that are somewhat similar, but not identical, like tanzanite and sapphire, yet we use a single, linear scale to report color depth. If this is not a judgement of "tone", what do you think it is? Obviously the hues, the colors, may be different, but to say which is "darker" is a tone judgment, correct?

 

[Serg]

Chroma ( and colorfulness ) is close to human color grade .
Roughly Chroma = Saturation * Lightness

 

[Serg]

Labs are doing color grade colorless diamonds from side view, because grader can not receive stable result in face up position for current light scheme. In side view position central part of round diamond is close to plate( plate can be easy graded(Relative grade) in any light condition) . I think there is possible to change light scheme and methodic for receive stable face up human color grade.

Labs are doing color grade fancy color diamonds from face up position because anybody can easily see (for fancy diamond ) that side view grade is not adequate. Color is depends from cut and light scheme of course.

 

[denverappraiser]

Date: 4/1/2006 9:41:06 AM
Author: oldminer
Strmrdr;

If you stick to sapphire or tanzanite of one hue and had a GIA color system that graded the depth of color with letters of the alphabet, then the slightly darker sapphire would be a lower letter of the alphabet. With diamonds, we look at colors that are somewhat similar, but not identical, like tanzanite and sapphire, yet we use a single, linear scale to report color depth. If this is not a judgement of ''tine'', what do you think it is? Obviously the hues, the colors, may be different, but to say which is ''darker'' is a tone judgment, correct?

People don''t view tone in the same way over the entire range of possibilities of combinations for hue and saturation. Eyes and brains are very complicated things and as you point out, it''s a fairly involved question. A single linear scale is not a complete description, but it may be sufficient. In sapphire or fancy colored diamonds, it would clearly be inadequate and the question at hand seems to be if it''s adequate for D-M diamonds. Does the tone alone describe enough about the color of a diamond to allow buyers to differentiate one stone from another in a meaningful way? This is the position that GIA has taken and it has merit dispite the observation that a yellow L and a pink L will appear quite different, and if measured with more scientific instruments may have very different tones. The objective in most grading (both diamonds and other things) is not to identify all possible differences, it''s to identify the attributes that are important to the transaction at hand.

Neil Beaty
GG(GIA) ISA NAJA
Professional Appraisals in Denver

 

[Serg]

Date: 4/1/2006 10:07:24 AM
Author: Serg
Chroma ( and colorfulness ) is close to human color grade .
Roughly Chroma = Saturation * Lightness

For aria Chroma=constanta the metric Saturation*( 1-transperency) is not bad.

D-Z has similar Chroma near zero. Vivid has maximum chroma

 

[valeria101]

Date: 4/1/2006 7:10:37 AM
Author: oldminer


Would you really like colorless diamonds graded by face-up appearance instead of actual body color? Would you still agree to allow an algorithmic adjustment for the size of the diamond as it got larger?


Sure. I wonder if this might be an issue to be accounted for by the color grades or the cut grades... or yet something else.

It seems to me that the dependence of apparent color on cut is already part of the folklore about premium cut diamonds.

From where I stand, the fact that diamond pricing and grading is so much more dependent on the properties of the rough than those of cut diamonds is quite funny: as if the history of the trade is more important for the grading and valuation than the object at stake

...Its the kind of problem I enjoy thinking about as well as eventually doing something about.

Surely pleasant to think of and talk about.

Would it be possible to make a credible ''what if'' scenario of what the commercial value of the gap between the current grading and a hypothetical face up color grading could be? It wouldn''t have to be more precise than a war game.

With such data at hand, it may become more clear what could be an acceptable presentation for the grading adjustment (if other than a challenge of the current overly rough-friendly grades).

It sounds like there is more than enough information available to put up such a bit of research.

If not done already.

 

[WinkHPD]

Date: 4/1/2006 9:09:59 AM
Author: strmrdr
to clarify my position:

Yes the material color is the color of the material but:
tone does not equal color.
a purple sapphire and a blue sapphire can have the same tone and not be the same color.

face up is more relevant than face down color.

Storm, when cutting affects the face up color so much and it is nearly impossible to grade the stones consistently from the table due to the "sparkle" factor it really is very difficult to do what you are suggesting with the human eye, partially because the master stones will also be affected by the cut that they have. You are wanting to open a HUGE can of worms when the system we already have is difficult enough.

Tom Tashey of PGS has suggested that stones should have both the table up apparent grade and the normally graded actual grade. I think he may have some merit in his suggestion, but I question how it would accurately be done outside of a lab.

I actually agree with you that from the consumer point of view face up is more relevant, but you will take away the vendor''s opportunity to serve his clientele by finding those stones that face up better than they cost if you get your way.

If an H that looks like an E is now graded face up as an E, then you can be sure it will be sold as an E with the concomitant pricing, and flush, there goes one of the competitive edges that so many of us cut nut pricescope vendors have...

Wink

 

[WinkHPD]

Date: 4/1/2006 1:40:37 PM
Author: valeria101

Would it be possible to make a credible ''what if'' scenario of what the commercial value of the gap between the current grading and a hypothetical face up color grading could be? It wouldn''t have to be more precise than a war game.

With such data at hand, it may become more clear what could be an acceptable presentation for the grading adjustment (if other than a challenge of the current overly rough-friendly grades).

It sounds like there is more than enough information available to put up such a bit of research.

If not done already.

I see you are already thinking along the lines that I am, that if we do this there will be a large increase in the value of the well cut stones that will not be justified by the cost of the rough, but if we get that increase in value, then DeBeers will raise the cost of that rough so that they can get the profit on that new profitability based on the potential to get a higher grade if well cut. Sheesh, they never leave much for the rest of us, and the only one hurt here will be the consumer as we shall surely have to pass on our higher costs to the consumer. Oh my, how I hope we don''t ever do that. The current system is both fair and fairly efficient. Some things do not need to be changed just for changes sake...

Wink

 

[valeria101]

Date: 4/1/2006 7:28:35 PM
Author: Wink

Date: 4/1/2006 1:40:37 PM
Author: valeria101

Would it be possible to make a credible 'what if' scenario of what the commercial value of the gap between the current grading and a hypothetical face up color grading could be?

I see you are already thinking along the lines that I am, that if we do this there will be a large increase in the value of the well cut stones that will not be justified by the cost of the rough, but if we get that increase in value, then DeBeers will raise the cost of that rough

He, he

I doubt a small bit of research could topple that big cart, even if most diamonds would be ideal cut. There sure is a premium on better cut for other reasons already ... Besides, this story sounds intricate enough to expect surprising conclusions anyway.

In fact, I wonder if there is no such work already done... if there is any research about cut and apparent color (Serghey mentioned something like this on an older thread).

Obviously day dreaming here!

 

[Garry H (Cut Nut)]

Date: 4/1/2006 12:43:07 PM
Author: Serg

Date: 4/1/2006 10:07:24 AM
Author: Serg
Chroma ( and colorfulness ) is close to human color grade .
Roughly Chroma = Saturation * Lightness

For aria Chroma=constanta the metric Saturation*( 1-transperency) is not bad.

D-Z has similar Chroma near zero. Vivid has maximum chroma

Sergey I for 1 do not fully understand your posts - so i think some others do not either.

Re cut effects - the length of ray paths (Sergey and yuri have made posts of work they are doing here recently) and the amount of darkenss in a stone affect the perception of body colour (see GIA article extract).

When graded (as we do) thru the side of the stone a 1ct H round well cut looks to have less color than a 1ct h radiant. The comments in the paragraph above account for some or most of those differences.

Perhaps a good consistent lighting and background color / shade / brightness could make it possible to grade from face up.


Perhaps also we could move to a more linear scale where D-F might be one grade since very few people can see this difference in a set stone front up anyway. H-I seems a wider grade than D-F

 

[strmrdr]

Wink, and everyone else.


When a vendor says this I diamond faces up like a G in today's system its wrong it faces up like a well cut I but there is too little relation between the lab color and real world color as it stands today so that saying came into use.
The problem is that it opens the window for less than honest people.
Lets say that a consumer reads on PS that a well cut I faces up like a G, they can watchyourwallet diamond company and the shoe salesman says yea this I faces up like a G it will rock your world dude!
They are going to think hmmmm the folks at PS said its possible kewl ill buy the egl-Israel 65% deep I,,, way kewl!!



questions,,

I don't really disagree with you that it could be a mess but the question is would it be a worse mess than today's system?

The current diamond grading system was designed by the trade for the trade correct?

Who benefits the most from a lot of the unpublished exceptions?

 

[Paul-Antwerp]

Going back to the original question, I cannot help but wonder how you consider the way in which HRD grades colour.

The normal colour-grades are used for yellowish hue. For a brownish hue, and probably for other colour hues too, the stone gets an equivalent colour-grade. So, a stone with a slight brownish hue, tone J, will get a grade ''J-eq'', in which the ''eq'' indicates that it is not a yellowish hue.

When first working with American labs, I was not aware that they did not use equivalent grades.

What is weird, is that the trade discounts equivalent HRD colour-grades. This means that there either is a loophole, in which intelligent traders can make a lot of money, or there is a big difference in grading.

Take possibility 1. an HRD J-eq would get a GIA J. In that case, one can buy a lot of equivalent colours with HRD-grades, at a discount because of the equivalent grade, have them re-graded by GIA, and sell them at the normal price.

Take possibility 2. an HRD J-eq would get a GIA K or L. In that case, pricing is correct, but it clearly shows a difference in grading-systems. Would that mean that an American lab degrades non-yellowish hue by one or two grades?

I do not know which of the two possibilities is correct. Maybe some of you know.

Live long,

 

[Garry H (Cut Nut)]

Date: 4/2/2006 4:16:45 PM
Author: Paul-Antwerp
Going back to the original question, I cannot help but wonder how you consider the way in which HRD grades colour.

The normal colour-grades are used for yellowish hue. For a brownish hue, and probably for other colour hues too, the stone gets an equivalent colour-grade. So, a stone with a slight brownish hue, tone J, will get a grade ''J-eq'', in which the ''eq'' indicates that it is not a yellowish hue.

When first working with American labs, I was not aware that they did not use equivalent grades.

What is weird, is that the trade discounts equivalent HRD colour-grades. This means that there either is a loophole, in which intelligent traders can make a lot of money, or there is a big difference in grading.

Take possibility 1. an HRD J-eq would get a GIA J. In that case, one can buy a lot of equivalent colours with HRD-grades, at a discount because of the equivalent grade, have them re-graded by GIA, and sell them at the normal price.

Take possibility 2. an HRD J-eq would get a GIA K or L. In that case, pricing is correct, but it clearly shows a difference in grading-systems. Would that mean that an American lab degrades non-yellowish hue by one or two grades?

I do not know which of the two possibilities is correct. Maybe some of you know.

Live long,

Hi Paul,

1. many brown stones come from Argyle.
All Arglye browns flouresce blue

Therefore (according to Marty Haske) fluoro brown stones will grade as a better higher color because gIA do not filter out UV light. (Unless HRD also filter UV light?)

 

[Sunshines]

Hi Garry and rest,

What about AGS grading? do they also not differentiate between Yellow and brown hue ie two stones with different hues and tones will be given the same colour irrespective and regardless of the tone each carries?

 

[Garry H (Cut Nut)]

Date: 4/2/2006 10:42:47 PM
Author: Sunshines
Hi Garry and rest,

What about AGS grading? do they also not differentiate between Yellow and brown hue ie two stones with different hues and tones will be given the same colour irrespective and regardless of the tone each carries?

In regards to clarity and color I think of AGS as being people well trained by GIA and following more the gIA philosophy than say the HRD methods.
Being a smaller lab in Vegas, with a good pool of stable staff to draw on, I think they have less challenges than say GIA''s NYC lab and much higher staff churn rates. i think GIA do a pretty good job - but they seem to have staff training issues in regards consistency. For example in that sense my experiance has been that IGI is a very consistant lab - but i see few stones graded by IGI in USA - most are from Mumbai where employing good reliable young eyes is very easy and the staff loyalty / turn over is far lower. However IGI have to manage grading consistency between about a dozen labs world wide (although I feel their color masters are perhaps 1/4 grade higher than GIA''s) - running so many labs - its a challenge that I believe they do rather well - visa vee the reputation for EGL LA run by Tom Tashi +5 years ago was better than the EGL NYC - and that perception still exists as i often see LA in the comments section on B2B listings of diamonds.

But in answer to your question - Sunshines - I dont know how AGS differentiates between different hues - but will ask

 

[Richard Sherwood]

Date: 4/2/2006 4:16:45 PM
Author: Paul-Antwerp
Going back to the original question, I cannot help but wonder how you consider the way in which HRD grades colour.

The normal colour-grades are used for yellowish hue. For a brownish hue, and probably for other colour hues too, the stone gets an equivalent colour-grade. So, a stone with a slight brownish hue, tone J, will get a grade ''J-eq'', in which the ''eq'' indicates that it is not a yellowish hue.

When first working with American labs, I was not aware that they did not use equivalent grades.

What is weird, is that the trade discounts equivalent HRD colour-grades. This means that there either is a loophole, in which intelligent traders can make a lot of money, or there is a big difference in grading.

Take possibility 1. an HRD J-eq would get a GIA J. In that case, one can buy a lot of equivalent colours with HRD-grades, at a discount because of the equivalent grade, have them re-graded by GIA, and sell them at the normal price.

Take possibility 2. an HRD J-eq would get a GIA K or L. In that case, pricing is correct, but it clearly shows a difference in grading-systems. Would that mean that an American lab degrades non-yellowish hue by one or two grades?

I do not know which of the two possibilities is correct. Maybe some of you know.

Live long,

Hi Paul. I think your "possibility 1" is more correct. I don''t see GIA grading equivalent colors one or two grades less.

Which brings me to my next question for you and the rest of the guys. How much of a discount do you feel is attached to "equivalent" colors, if any?

Let''s say the stone is a "J" or "K" color. What discount, if any, would be attached to a top light brown of the same color?

 

[oldminer]

Two diamonds of distinctly different colors, one yello and one brown, will be graded with the SAME letters of the alphabet as representative of their GIA assigned color grade. This is a grading of their tone, since their hue is not detailed in the GIA system.

My mentor, Dr. Aggarwal, has told me the concept I have about tone may well be true for a majority of the common color ranges of diamonds, but he has warned me there could be some exceptions lurking out there which would defy such simple grading. We don''t have examples to give, but one might expect exceptions. Rather than make a statement that might be proven wrong with just one opposing example, one might offer that tone might be how grading is done when comparing or judging diamonds of differing hues for assigining a linear color grade scale, such as the one GIA uses, but that exceptions to this process may exist. Being able to readily identify these exceptions will be very useful.

Of course, the idea of automated color grading is the goal. Efficiency of grading is another major goal.
The best way is to make a machine that can reliably color grade most diamonds and automatically reject any that cannot be graded by machine. In that way, the time for grading most diamonds will be minimized and human grading will only be required for the exceptional cases.

 

[Garry H (Cut Nut)]

Date: 4/3/2006 10:14:49 AM
Author: Richard Sherwood

Date: 4/2/2006 4:16:45 PM
Author: Paul-Antwerp
Going back to the original question, I cannot help but wonder how you consider the way in which HRD grades colour.

The normal colour-grades are used for yellowish hue. For a brownish hue, and probably for other colour hues too, the stone gets an equivalent colour-grade. So, a stone with a slight brownish hue, tone J, will get a grade ''J-eq'', in which the ''eq'' indicates that it is not a yellowish hue.

When first working with American labs, I was not aware that they did not use equivalent grades.

What is weird, is that the trade discounts equivalent HRD colour-grades. This means that there either is a loophole, in which intelligent traders can make a lot of money, or there is a big difference in grading.

Take possibility 1. an HRD J-eq would get a GIA J. In that case, one can buy a lot of equivalent colours with HRD-grades, at a discount because of the equivalent grade, have them re-graded by GIA, and sell them at the normal price.

Take possibility 2. an HRD J-eq would get a GIA K or L. In that case, pricing is correct, but it clearly shows a difference in grading-systems. Would that mean that an American lab degrades non-yellowish hue by one or two grades?

I do not know which of the two possibilities is correct. Maybe some of you know.

Live long,

Hi Paul. I think your ''possibility 1'' is more correct. I don''t see GIA grading equivalent colors one or two grades less.

Which brings me to my next question for you and the rest of the guys. How much of a discount do you feel is attached to ''equivalent'' colors, if any?

Let''s say the stone is a ''J'' or ''K'' color. What discount, if any, would be attached to a top light brown of the same color?

I have not heard back from AGS yet, but had a chat with IGI director Marc Brauner. This is the gist of it:

IGI follow the method that Paul outlined for HRD: they grade tone (being color blind could help) but for K and lower colors they add the modifier – but not in coded letters – but helpfully in in simple words. e.g. K faint pink, L faint gray, M faint brown, but Pink below ‘M’ intenstity probably corresponds to Fancy Pink. K and lower cape or yellow series stones (the vast majority) have no modifier added.

Then it is up to the market to set its own pricing based on that markets preferences (which can and do change over time and locality).

When there is a real borderline dilemma and 5 graders can not agree, they look they also look at the stone face up – in this case say a pear shape might be downgraded and a round with EX-EX and H&A often shows better in face-up; in such cases they aknowledge the relation between make and color that we all know exists.

Marc said occasionally they get a real dilemma – say a Fluoro Brown – it looks completely different in different lighting (I could picture him throwing his arms in the air – French is his first language

).

 

[Richard Sherwood]

Thanks Garry, great information.

As an appraiser, I'm curious about discounts. Does there seem to be a trade standard of discount for brown versus cape?

The way I see it, there is no discount to a small discount in the fainter colors, such as JK, and possibly L. Then the discount increases when you get to MNOP etc.

How do you find it? Anyone care to throw any numbers out, just for discussion?

 

[zhuzhu]

I hope the following thoughts won’t bother anyone, but given the advanced technology in our days, wouldn''t it be sensible and more objective to design machinery and computer program to standardize the grading of diamonds? I would think both inclusion and color can be read objectively across the board by applying a grading algorithm. After all the "cut quality table" has been designed using primarily mathematical equations, why not apply the same theory towards inclusion and color grading and have computer do the job?

Just my 2 cents...

 

[Richard Sherwood]

That seems to be the way things are headed, zhuzhu, and everybody''s working on it (or watching with avid interest).

There''s an awful lot of variables and red herrings though. This topic illustrates one.

 

[Garry H (Cut Nut)]

I am impressed that Marc Brauner has been very willing to share vital detailed information (which is what Pricescope is all about).

I drilled down on the Fluoro topic – he believes they use grading lamps with minimal or no Fluoro and has not followed GIA’s lead to what appears a deliberate strategy to include fluorescent wavelengths in the grading lamps (as per Marty Haskes revelations that a Strong or Very Strong Blue Fluoro GIA H is sometimes given F).
Marc explained:

“We do sometimes upgrade the profile color when fluorescence is Strong or Very Strong: example profile view is ‘G’, Fluor. is ‘Strong’: in face-up position stone looks E-F: Result is F.

This will usually not be done when the profile color is middle-to-low (within a single color grade), example ‘G’ –only when the stone is a high G. Ok, what I am saying here is that fluorescence of strong or very strong intensity make the stone look whiter (and sometimes bluer) than the same stone without such strong fluorescence: as my example indicated, when a diamond is graded ''G'' in profile position (which is the correct color grading position) and it has no or very little fluorescence, then the face-up view will show accordingly to what we would expect.
If you have strong fluorescence with a stone having the same profile ''G'', its face-up view will comparatively look like it is of a higher color: sometimes profile ''G'' looks as high as ''E'' in face-up when fluorescence is strong: when this occurs we tend to ''upgrade'' the color grade one grade up from assessed profile grade: this could mean - in my example - the report would show COLOR: F, Fluorescence: STRONG. (This kind of "Fluorescence-upgrade due to face-up view" cannot be done if the profile color was assessed as a "low G": this means we must first determine the profile color as Low, Middle or High (internally we express a color grade on scale of 100 - each color starts at 0 and ends at 100): if middle-to high, we can possibly consider the relationship between profile color and face-up color and grade according to previous procedure) When a diamond has strong fluor., we find that there is a gap somewhere between how the stone "speaks" in face-up and the value discrimination because of the ''Strong'' factor. We do not take value into acount but at the same time we feel the stone must receive the grade it deserves and in such situation the face-up view can play an influential role for the final color grade. In a similar way, we often see Excellent-Excellent diamonds (without fluorescence) that show much better in face-up than profile though here we cannot realy upgrade: for the real connoisseurs only, I could say that if a round brilliant is a true borderline between 2 colors, say ''H'' or ''I'': at the end of the day a decision will have to be taken, ''H'' or ''I'' ? and perhaps, in some cases, face-up appearance thanks to extremely fine make could be a determinant factor.

Borderline grades undergo strict procedures in our labs: usually these stones are put asside and are looked at later, sometomes the day after and by different people. The most vital part of this procedure is correct calibration: when someone is grading color, after a few stones the eyes calibrate and fine-tune to very high sensitivity: if at any moment the grader''s eyes experience doubt, the procedure is to move on and keep the doubt-stone asside. If after that no conclusive grade can be given to the said stone, the stone will have to be put asside for further inspection. The easy way would be to give the lowest grade at that moment but we train our graders not express a grade when there is a doubt: leave it open and for later review.
What I say to them during training is " to be a professional grader is to know when you don''t know - and then simply follow procedures". Perhaps these and alot more of similar procedures that we have in place ensure a certain level of consistency. Let me know if more is needed.”

On the phone Marc explained the difference between theirs and GIA’s master sets.

GIA has stones placed at the border – say a G master – it is just at the F/G border - a top ''G''.
IGI has its equivalent master slightly lower at what he describes as 80% toward F, but not quite at the border - a high ''G''.

GIA’s masters apparently can and are used by any and all graders – but the longest period a grader can do color is 1 ½ to 2 hours. IGI masters are only used by a small number of graders – each grader or small group of graders ONLY works with their own set. They become personally connected to and calibrated to that set. He believes that makes it easier to use, less stressful and makes for faster grading.

They have one chief Quality Control color grader with 25 years experience who sits in Antwerp with his own master set.

Summary - IGI default to Tone with colors other than yellow, but describe the color below K in simple words. In a small percentage of cases IGI will upgrade borderline stones with Blue Fluoro stone or an exceptionally well cut round stone over its profile graded tonal shade.

 

[Amandas_Jewels]

I just want to say THANK YOU to all of you, for this wonderful information! We never stop learning.

At least I don''t!

 

[adamasgem]

Date: 4/3/2006 11:40:53 PM
Author: Garry H (Cut Nut)

I am impressed that Marc Brauner has been very willing to share vital detailed information (which is what Pricescope is all about).

I drilled down on the Fluoro topic – he believes they use grading lamps with minimal or no Fluoro and has not followed GIA’s lead to what appears a deliberate strategy to include fluorescent wavelengths in the grading lamps (as per Marty Haskes revelations that a Strong or Very Strong Blue Fluoro GIA H is sometimes given F).

Hi Garry.. For those who haven''t read it , my commentary regarding fluorescence is at GIA Fluorescence

Part of the color grading problem is the knowing the position of master stones. My data indicates that the one sigma uncertainty in color position referenced to the nominal position of the GIA blessed master is about 16 to 17% of a color grade range, based on "best" estimate data (quantized at 10% of a color grade range by GIA)

Tone or depth of color is the primary sensory indocator for the nominal cape series stones. Masters should have facetd girdles otherwise they will accumulate metattlic particles from tweexzers and other crud, which can draw the color. Bruted girdle masetrs should be boiled in acid daily to weekly if they are in constant use. I have seen master stones three grades off because the "professionals" that had them hadn''t cleaned them since the day they were cut. I won''t name names to protect the guilty, but I have a lot of data on assorted master sets from tests run in Tuscon. GIA, I believe as well as AGS now require masters to have facted girdles so they don''t pick up crud.

Masters ar supposed to be cape series stones, and I have rejected stones that had distorted spectra, from things like natural radiation damage.

Inclusions can effect perceived color as well as distort spectra used iin color grading by machine. The human has to be in the loop...