Just Suppose? A color grading POLL

[oldminer]

I have one of Martin Haske's excellent SAS 2000 devices. I would urge anyone who needs to use a spectroscope in their gemology to consider buying one. It runs circles around any other spectroscope offered to gemologists.....I isn't a cheap tool, but good tools cost bucks.

I do not rely on the SAS 2000 for diamond color grading. I find it is far better than the Austron or Gran colorimeters that I have owned, but still not better than my eye. That's the truth and there is no way to sweeten it. An investment in diamond masters is highly important to an appraiser / gemologist.

The device I am referring to is from Imagem, Inc. www.imageminc.com The accuracy of the color grading is 1/6 of the upper grades and probably much more precise, (more steps) in the medium to low grades that no one cares a lot about. We will report GIA color grades and possibly a numerical grade which the machine generates. Imagem and myself have not totally committed to how much change in the color grading status quo is most acceptable or welcome. My distinct feeling from this long thread is that ACCURACY is welcome, but that breaking down color grades into finer parts may prove unproductive, confusing, annoying, and disconcerting. Now, I can't say that I totally dislike controversy or being a bit of a leader, not just a follower. I was way ahead with the AGA Cut Class grades, but while useful, they have hardly become the accepted norm of the entire industry. I need to be professional and pragmatic, especially with the Imagem product. This is world class grading technology and I don't wish to marginalize its importance to the world trade by making it problematic. I want it to be palatable, and highly acceptable along with ACCURATE.

I appreciate all the feedback so far. I sort of knew accuracy was what people wanted, but there were some superb suggestions made within the thread which have brought clarity to me on how this product should best be used. Pricescope consumers and dealers will lead the way with this new technology as soon as it is in operation....

 

[strmrdr]

Dave I want to thank you for getting our opinion on it.
It would be far better is more people/orgs in the business did the same.
This thread is an exellent example of the right way to do it!

ps. your link leads to a this domain is for sale site.

 

[perry]

Dave:

A few more thoughts:

Almost by definintion if the new machine will grade to 1/6 increments you should be calling no more than 1/3 increments (measuring devices should be at least twice as accurate as the interval you are measuring).

With 1/3''rds you could say something was a G-, G, or G+ which would not be too confusing. I do however agree that there should be some kind of linear output provided with the hopes that in the future we transition away from DEF... (and who knows what happened to ABC).

However, any such increase in rating will in fact be reflected in pricing, with the D+ diamonds streaching way up into the stratosphere.

As you link did not work I was not able to look at the specs for the machine. However, I believe it to be very important that you are measureing color in stones without any UV to excite whatever fluoresence is there (that way the color reflects the color on a rainy day or in many indore settings). Marty Haske has provided good indications of that is how diamonds used to be graded.

Perry

 

[valeria101]

Date: 11/7/2004 5:47:13 PM
Author: perry

(and who knows what happened to ABC).

This may be an interesting precedent, actually.

"A, B, C" went off in dust when GIA introduces it''s grades. The three letters would have been a three-tier quality system, much like the one for pearls today - equally deregulated. By avoiding the grade names of the system they wanted to replace, GIA was hoping to reduce resistence to it''s new system. With different names, the two grading methods could opperate side by side during a manageable friction period.

In theory, the naming game moves the controversy from the practical level: instead of having diamond owners thunder about "who''s A is this" you would have them discuss "whose system is better" and how the grades correspond across the scale. Arguably, the second type of inquiry would be less conflictual and more constructive. Just theory, of course. But the same is applied elsewhere and the effect is recognized.

Making new grades of >+- sounds intriguing. This practice inherently recognizes that GIA grades are precise by themselves. Only the finess of the scale is increased three fold, as you say. At least, at first sight one (myself, for now) would imply that a stone that gets a H+ grade on the machine would otherwise be H in GIA terms only a "better H" by looks. Am I right ?

But there seems to be a match. Labs guarantee their grades to +- half a grade only. So it may sound legitimate that a new tool is able to close this gap labs already recognize (sort of). Up to the new grading procedure to claim an interval of confidence of it''s own.

 

[Diermint]

Date: 11/7/2004 9:22:53 PM
Author: valeria101

With different names, the two grading methods could opperate side by side during a manageable friction period.

Hi All,

Another option is to use the colorimeter to accurately determine the colour grade of the diamond and then to post the numerical result along with it, giving the consumers more information, eg: D(1.2) or D(1. (from a previous example in this thread by oldminer) . The colour scales can be numerically defined based on the existing scheme, hence the "D, E, F..." scheme needn''t change. In the future if we can measure to greater precision then we could simply put in a more precise number without changing the grading scale, eg: D(1.22) or D(1.7. The numerical scale could be adjusted to be, say, 0 for a colourless diamond and 100 at the other extreme.

As for pricing, how would this be different to carat weight? Carat weight is not quantised to 0-0.2 (A), 0.2-0.4 (B), 0.4-0.6 (C), 0.6-0.8 (D)... If this were the case people would still look for a large one in that range rather than a small one. This is what happens with colour now - some H''s and I''s look yellower than others and some consumers feel the need to be discerning. The range of diamonds found to be G in colour can still average the same price but the less coloured ones command a slight increase since they are closer to the bottom F''s and the more coloured ones a slight discount since they are closer to a H.

In an ideal world it would be nice to be able to look at colour specifications on a certificate and be as certain of it as we are with weight, and if it is too exact, who cares? Clarity will probably always be subjective but with colour we might be able to remove some of the subjectivity, which IMO is a good thing.

 

[valeria101]

Date: 11/8/2004 12:52:18 AM
Author: Diermint

Another option is to use the colorimeter to accurately determine the colour grade of the diamond and then to post the numerical result along with it[...]

In an ideal world it would be nice to be able to look at colour specifications on a certificate and be as certain of it as we are with weight, and if it is too exact, who cares?

Agreed... numbers are easier to handle than pre-made categories.

Even with the most precise color grades, there is till lots of room for interpretation ( say aparent color face-up influenced by cut, setting, fluorescence... what not). No need for the basic grades to remain debatable themselves.

You can imagine what the buzz around numeric color gardes could be. Even for weight there is some regulation for what accuracy level is admisible and then, "1 carat" diamonds seem to be between 1.00 and 1.05 but not 0.955 and 1.05 and soooo on. It is quite incredible what numbers become when used to describe diamonds

At least the issues remain few and rather predictable: the most often heard "is this good enough" or "does this difference matter". Much easier to talk numbers than diamonds, IMO.

 

[Diermint]

Date: 11/8/2004 1:50

3 AM
Author: valeria101

Even with the most precise color grades, there is till lots of room for interpretation ( say aparent color face-up influenced by cut, setting, fluorescence... what not). No need for the basic grades to remain debatable themselves.

Thanks Ana,

and if I may carry the analogy between colour and carat a bit further - some one carat diamonds appear to be larger than others and this is something we routinely accept. We look at additional data such as diameter, depth% etc. rather than relying on just the weight to determine how large it will appear.

In order to adequately characterise colour we will probably require additional info rather than just a colorimeter reading, such as from an absorption spectra (is this sensible experts?) which should show scientifically whether it is likely to be yellowish, brownish, greyish etc. rather than simply the intensity of colour.

 

[oldminer]

The link provided above should have been www.imageminc.com and it has been corrected. Thanks.

I believe the right way to handle color grading with this new technology is to give GIA grades. Why confuse anyone more? However we do want true accuracy, to the extent that can be provided. Thiough presently subjective, there is nothing wrong with GIA grading with D-Z. It is universally understood, so why rock the boat needlessly?

Surely, I would not want to make a less desireable stone category within each color grade, such as low end F or low end G. These are great stones and there is no reason to hurt them. Possibly one should indicate a diamond is nearly at the next color higher such as G* (nearly F). I see less downside to that approach.

The exact color numbers could be on a higher cost, premium sort of report, where engineer types could seek out the finest available details, those that other people don''t really want to suffer over.