shape
carat
color
clarity

Should GIA grade diamonds based on beauty?

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Here's an exchange from a different thread between my good friend Garry and I
Garry H (Cut Nut)|1481679632|4107028 said:
Rockdiamond|1481671942|4107011 said:
HI Garry-
Just a slightly different take:
For a shopper looking for a "normal" diamond, such as a 1 1/4ct H/SI- a category where there's a lot of options: is it a good idea for a consumer to purchase a diamond that will cost them hundreds in assessment fees right out of the box? Sending a stone to an appraiser would entail such cost.

About twining wisps- my experiences are different- I've not seen a lot of VS2's with twining wisps that affected brilliance. That's not to imply you have not- I'm sure you have.
But it does make my statement about looking in person and how an honest trained professional can earn their salary by examining a diamond together with a consumer, and showing them the imperfections and how they affect the diamond.
In such a case it might be entirely possible a professional would not use IS or ASET.
That will likely change more over time- but there's still plenty of knowledgeable , honest professionals that do not use IS/ASET

Hi David, for the sake of learning - here is an example of a stone that based on the inclusions visible under 10X the stone would probably be a low VS2 - but GIA graded it SI1 - and in my opinion that is how they grade stones which are slightly milky hazy as this one is. The photo was taken using ViBox, which you have and I was one of the developers - so we can see the life is down a bit.
We are not arguing however - I am just warning consumers that when a clarity grade looks too good to be true, there is almost always a hidden nasty.
I think GIA should be sued by all the people who bought these yucky diamonds for loss of sparkle.
GIA should come here and defend the fact they give this stone their top cut grade - which implies it sparkles, but use jargon that conceals the fact the stone has reduced life and sparkle. It really is criminal. It is done to enhance sales at the cost of consumer confidence.

It is a good subject and rather than threadjack...
My take is that GIA issues "clinical" grades.
As a someone who has spent a lifetime working as a diamond grader, this has always been something I felt good about. After all, you need a competent human to evaluate a diamond.
AS you point out Garry- prices too good to be true are generally based on issues with a stone's appearance, regardless of the GIA grade. In this way, the market adjusts for the inadequacy of a given stone- and the superiority of another.

The reason I do not see this as a case of bad intentions on the part of GIA, is that it is a lab that offers no value judgment. To this point GIA issues a fairly clerical, non judgmental report that identifies a diamond.
Clarity is a great example- some SI2's are eye clean, others have easily visible imperfections.
GIA does not grade the clarity based on if you can see it- rather if it exists.

I get Garry's point- but the alternative is for GIA to "dictate" what looks best.
Fancy cut grading is a great example of sometime that's virtually impossible to quantify inclusive of taste.

I also believe that if it can, GIA will go in the direction of trying to eliminate my job as someone who assesses diamonds for a living.

What do y'all think- should GIA include more "practical information" on the report- such as "eye clean", "foggy" or other descriptive terms?
How about value?
 

whitewave

Super_Ideal_Rock
Premium
Joined
Feb 29, 2012
Messages
12,331
Why they are sitting there looking at it, I don't see why adding eyeclean would be a big deal. It's helpful
 

whitewave

Super_Ideal_Rock
Premium
Joined
Feb 29, 2012
Messages
12,331

Garry H (Cut Nut)

Super_Ideal_Rock
Trade
Joined
Aug 15, 2000
Messages
18,423
Rockdiamond|1481682674|4107031 said:
Hi David, for the sake of learning - here is an example of a stone that based on the inclusions visible under 10X the stone would probably be a low VS2 - but GIA graded it SI1 - and in my opinion that is how they grade stones which are slightly milky hazy as this one is. The photo was taken using ViBox, which you have and I was one of the developers - so we can see the life is down a bit.
We are not arguing however - I am just warning consumers that when a clarity grade looks too good to be true, there is almost always a hidden nasty.
I think GIA should be sued by all the people who bought these yucky diamonds for loss of sparkle.
GIA should come here and defend the fact they give this stone their top cut grade - which implies it sparkles, but use jargon that conceals the fact the stone has reduced life and sparkle. It really is criminal. It is done to enhance sales at the cost of consumer confidence.

hazy_0.jpg
 

Garry H (Cut Nut)

Super_Ideal_Rock
Trade
Joined
Aug 15, 2000
Messages
18,423
Rockdiamond|1481682674|4107031 said:
My take is that GIA issues "clinical" grades.
As a someone who has spent a lifetime working as a diamond grader, this has always been something I felt good about. After all, you need a competent human to evaluate a diamond.
Not true David. GIA has been setadily replace humans with machines and devices.
AS you point out Garry- prices too good to be true are generally based on issues with a stone's appearance, regardless of the GIA grade. In this way, the market adjusts for the inadequacy of a given stone- and the superiority of another.

The reason I do not see this as a case of bad intentions on the part of GIA, is that it is a lab that offers no value judgment. To this point GIA issues a fairly clerical, non judgmental report that identifies a diamond.
Again, I strongly protest. If a diamond looks like a VVS under 10X and GIA grades it as VS2 because it is hazy or foggy - then they can and should say so. To do as they do is of itself a "value judgement" but not one that is stated. The market may 'adjust the price" but that is no excuse for not telling people what you know.
Clarity is a great example- some SI2's are eye clean, others have easily visible imperfections.
GIA does not grade the clarity based on if you can see it- rather if it exists.

I get Garry's point- but the alternative is for GIA to "dictate" what looks best.
Fancy cut grading is a great example of sometime that's virtually impossible to quantify inclusive of taste. Another good example - and I know GIA is working on this because their senior R&D guy told me - but it is certainly possible to give information like that that AGS generate - except they are really not going to escape my wrath either - because they set different standards for different cuts - where as open and transparent info would be to pick say the best practical 1ct round cut and give out comparitive info that they actually have as a percentage of those numbers - e.g. dispersion is 97%, light return is 75%. Even AGS's current system for rounds fails this simple test - because their software which you can buy works on pass or fail and if a parameter fails they tell you that it got xx for light return - but they never give the number for exceeding the pass limit - this means there is no incentive to OVER perform. That is really dumb. Imagine if some cutter works out how to get 110% fire, but gets dinged because they only get 95% light return! What a lost marketing opportunity.

I also believe that if it can, GIA will go in the direction of trying to eliminate my job as someone who assesses diamonds for a living.

What do y'all think- should GIA include more "practical information" on the report- such as "eye clean", "foggy" or other descriptive terms?
How about value? Careful, what about me mate Martin!
 

lissyflo

Brilliant_Rock
Premium
Joined
May 23, 2016
Messages
1,719
As a complete diamond 'lay-person', I agree with both of you!

I would expect a GIA report to include details on all areas that have had an impact on the clarity (or any other) grade given. So, if I'm reading your posts correctly, I would be confused with the example of a VVS louped diamond having a VS grade based on milkiness if nothing is said in the report about this aspect. I would assume the grade was based only on things stated in the report. And surely reports are there to help, not confuse, the end purchaser?!

I also think though, to give my view on your question about whether reports should assess eye cleanliness etc, that that would be one messy can of worms to open! What's the definition of eye clean?! For a professional body to give a statement on eye clean, the definition of what it is would need to be so tightly worded as to be meaningless anyway (eg. face on view, from x inches, with 20/20 eyesight). I have a VS2 diamond that's totally eye clean from the top but has an inclusion that is visible with the naked eye from the side if you view at a certain angle. It doesn't bother me in the slightest (in fact I like knowing it's mine!) but may bother some. Those who would be bothered by a side visible inclusion would still need to either see a diamond in person or have it viewed by a vendor they trust (or pay more for very high clarity grades!), so you're probably back to square one anyway, but with the diamond costing more as GIA have done more grading work on it. I guess it depends on whether there's a fixed definition of eye clean that covers enough diamond purchasers preferences' as to be meaningful to the majority.

But to give my view on your question in the header - should GIA grade based on beauty - no! Beauty is in the eye of the beholder; one person's 'perfect' ideal cut diamond may not be perfect for someone looking for a unique diamond with an individual character. Grading authorities should provide information on measurable characteristics that can contribute to beauty (like inclusions, angles, light return if they go down that road, and they should state explicitly anything that affects a grade) plus things that are subjective to within a small degree (like colour assessment); anything that helps the vast numbers of people buying sight unseen nowadays to focus their search to their preferences. But you can't grade overall 'beauty' on a one-size-fits-all system.
 

Karl_K

Super_Ideal_Rock
Trade
Joined
Aug 4, 2008
Messages
14,634
There are issues with the idea that the market compensates by price.
Where more diamonds being sold based on the grading report and the seller never seeing the diamond these
days the potential is there that the price will not be adjusted.
There is also the FACT of the savings not always being passed on at retail in the more traditional market.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
whitewave|1481684271|4107039 said:
Why they are sitting there looking at it, I don't see why adding eyeclean would be a big deal. It's helpful
OK< say GIA decides to use eye clean as a parameter.
1) How close to the eye?
2) How much lighting?
3) what is the vision of the person deciding?
4) from the bottom as well as the top?
5) we'd need some level of degree- a teeny black point versus a HUGE white booger

My point is, "eye clean" is not that simple a question
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Garry H (Cut Nut)|1481696296|4107089 said:
Rockdiamond|1481682674|4107031 said:
My take is that GIA issues "clinical" grades.
As a someone who has spent a lifetime working as a diamond grader, this has always been something I felt good about. After all, you need a competent human to evaluate a diamond.
Not true David. GIA has been setadily replace humans with machines and devices. Let's stipulate that parts of GIA grading are done by machine. All that does is further necessitate a human to validate the results generated by machine...if it's my $10k ( or whatever) I want to make sure I agree with the machine. If I wasn't a diamond grader, I'd still want a human I trusted to validate the findings on such an expensive purchase.
AS you point out Garry- prices too good to be true are generally based on issues with a stone's appearance, regardless of the GIA grade. In this way, the market adjusts for the inadequacy of a given stone- and the superiority of another.

The reason I do not see this as a case of bad intentions on the part of GIA, is that it is a lab that offers no value judgment. To this point GIA issues a fairly clerical, non judgmental report that identifies a diamond.
Again, I strongly protest. If a diamond looks like a VVS under 10X and GIA grades it as VS2 because it is hazy or foggy - then they can and should say so. To do as they do is of itself a "value judgement" but not one that is stated. The market may 'adjust the price" but that is no excuse for not telling people what you know.
That's a very good point Garry. Sometimes you look at the diamond and it takes quite a while to figure out how GIA came to the clarity grade. I chalk this up to the endless variation possible in nature, placement and size of imperfections- and how that relates to the stone overall. Yes there are easy judgment calls on clarity ( 99% of graders see it as an SI1 ie)- but some are exceptionally difficult to judge. Meaning different graders will come to different conclusions.
Clarity is a great example- some SI2's are eye clean, others have easily visible imperfections.
GIA does not grade the clarity based on if you can see it- rather if it exists.

I get Garry's point- but the alternative is for GIA to "dictate" what looks best.
Fancy cut grading is a great example of sometime that's virtually impossible to quantify inclusive of taste. Another good example - and I know GIA is working on this because their senior R&D guy told me - but it is certainly possible to give information like that that AGS generate - except they are really not going to escape my wrath either - because they set different standards for different cuts - where as open and transparent info would be to pick say the best practical 1ct round cut and give out comparitive info that they actually have as a percentage of those numbers - e.g. dispersion is 97%, light return is 75%. Even AGS's current system for rounds fails this simple test - because their software which you can buy works on pass or fail and if a parameter fails they tell you that it got xx for light return - but they never give the number for exceeding the pass limit - this means there is no incentive to OVER perform. That is really dumb. Imagine if some cutter works out how to get 110% fire, but gets dinged because they only get 95% light return! What a lost marketing opportunity.

I also believe that if it can, GIA will go in the direction of trying to eliminate my job as someone who assesses diamonds for a living.

What do y'all think- should GIA include more "practical information" on the report- such as "eye clean", "foggy" or other descriptive terms?
How about value? Careful, what about me mate Martin!

I can also confirm that GIA is indeed working on Fancy Shape Cut grading. But it's been literally years in the making. Who knws when they'll roll it out.
If and when they do this discussion will get a lot more intense, yes?
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Karl_K|1481731229|4107163 said:
There are issues with the idea that the market compensates by price.
Where more diamonds being sold based on the grading report and the seller never seeing the diamond these
days the potential is there that the price will not be adjusted.
There is also the FACT of the savings not always being passed on at retail in the more traditional market.
HI Karl- thanks for joining in!
In my experience, the cutter DOES look at the stone as they are pricing it.
Say they are being greedy and overprice a bad SI1.
Say it gets sold off a website with no picture.
Is the buyer going to actually look at it?
Do you know the return rate of sites with no pictures?
I can tell you the return rate is significant. Even for those sites with pictures, there's a lot of returns.
So, the market has very real ways of equalizing this situation.
Having said that, for sure there are deficiencies that may not be so easy for a consumer to see that get passed down the line.
But the market forces really do play an active role in lowering prices for dogs.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
lissyflo|1481712745|4107111 said:
As a complete diamond 'lay-person', I agree with both of you!

I would expect a GIA report to include details on all areas that have had an impact on the clarity (or any other) grade given. So, if I'm reading your posts correctly, I would be confused with the example of a VVS louped diamond having a VS grade based on milkiness if nothing is said in the report about this aspect. I would assume the grade was based only on things stated in the report. And surely reports are there to help, not confuse, the end purchaser?!

I also think though, to give my view on your question about whether reports should assess eye cleanliness etc, that that would be one messy can of worms to open! What's the definition of eye clean?! For a professional body to give a statement on eye clean, the definition of what it is would need to be so tightly worded as to be meaningless anyway (eg. face on view, from x inches, with 20/20 eyesight). I have a VS2 diamond that's totally eye clean from the top but has an inclusion that is visible with the naked eye from the side if you view at a certain angle. It doesn't bother me in the slightest (in fact I like knowing it's mine!) but may bother some. Those who would be bothered by a side visible inclusion would still need to either see a diamond in person or have it viewed by a vendor they trust (or pay more for very high clarity grades!), so you're probably back to square one anyway, but with the diamond costing more as GIA have done more grading work on it. I guess it depends on whether there's a fixed definition of eye clean that covers enough diamond purchasers preferences' as to be meaningful to the majority.

But to give my view on your question in the header - should GIA grade based on beauty - no! Beauty is in the eye of the beholder; one person's 'perfect' ideal cut diamond may not be perfect for someone looking for a unique diamond with an individual character. Grading authorities should provide information on measurable characteristics that can contribute to beauty (like inclusions, angles, light return if they go down that road, and they should state explicitly anything that affects a grade) plus things that are subjective to within a small degree (like colour assessment); anything that helps the vast numbers of people buying sight unseen nowadays to focus their search to their preferences. But you can't grade overall 'beauty' on a one-size-fits-all system.
Great post Lissyflo. Clearly we share the ideas about "eye clean"
About beauty: If you consider AGSL current fancy shape grading, it is claimed to be scientific....but bottom line, AGSL is grading for beauty, as they see it.
 

Karl_K

Super_Ideal_Rock
Trade
Joined
Aug 4, 2008
Messages
14,634
Here is why a cut grade will never be able to define beauty in fancies.
Below are examples of in my opinion awesome asschers.
One set is real diamonds the other are virtual designs.
If you base the parameters for beauty based on one of them it will eliminate the others.

edit: see next post for second set. Sorry about the 2 posts was having posting problems.

_38689.jpg
 

Karl_K

Super_Ideal_Rock
Trade
Joined
Aug 4, 2008
Messages
14,634
Second set of images.

Images removed because of watermarks.
My apologies.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Karl_K|1481754887|4107303 said:
Here is why a cut grade will never be able to define beauty in fancies.
Below are examples of in my opinion awesome asschers.
One set is real diamonds the other are virtual designs.
If you base the parameters for beauty based on one of them it will eliminate the others.

edit: see next post for second set. Sorry about the 2 posts was having posting problems.

I think we're on the same page Karl.
IMO, we can extend this AGSL grading of fancy shapes, for example.
 

Garry H (Cut Nut)

Super_Ideal_Rock
Trade
Joined
Aug 15, 2000
Messages
18,423
Karl_K|1481754887|4107303 said:
Here is why a cut grade will never be able to define beauty in fancies.
Below are examples of in my opinion awesome asschers.
One set is real diamonds the other are virtual designs.
If you base the parameters for beauty based on one of them it will eliminate the others.

edit: see next post for second set. Sorry about the 2 posts was having posting problems.
Karl I disagree - but acknowledge the process would not be simple. From the closing paragraph of the very long and complex article that the Cut Group published 3 years ago is copied below:

When we find two diamonds which have close metrics: size, brightness, flashes etc, and if experts evaluate the Performance of these two diamonds as the same, then it validates the system. But if the two diamonds with similar metrics and the Performance of this pair of diamonds were scored differently by experts, then it means that the system does not take into account some critical aspect of human perception. If two diamonds have close flash statistics, but different expert evaluation, this will mean incorrect counting of some important factor of flash statistics, or expert error. Having such statistics for at least 5-10% of diamonds sold annually will make it possible, within several years, to have data (a map) for totally automatic Performance evaluation for 99% of diamonds. This is because for gems with similar flash statistics prior evaluations can be used, while for 1% of gems with new and unique statistics such as totally new cut phenomena (or new 'DNA’), expert evaluation would be used. Expert evaluations should be constantly verified by Crowdsourcing methods based on consumer evaluations. We are planning to discuss these subjects in the third article. In the next, second article, we will discuss how we use metrics based on 3D cut models and special illumination models for rapid new High Performance cut development by means of automatically rejecting cuts with low optical metric values, and expert evaluation of cuts with high metric values based on stereo movies created with DiamCalc software (see DiamCalc webpage link in the Reference section).

Link to the full article http://www.gem.org.au/ckfinder/userfiles/files/GAA_Journal_V25_No3_web2(1).pdf
 

Karl_K

Super_Ideal_Rock
Trade
Joined
Aug 4, 2008
Messages
14,634
Garry, I disagree with your disagreement.
The major difference is patterns.
Some people will prefer one pattern over another.
Good luck trying to grade patterns in a fancy diamond grading system.
 

Garry H (Cut Nut)

Super_Ideal_Rock
Trade
Joined
Aug 15, 2000
Messages
18,423
Karl_K|1481758603|4107324 said:
Garry, I disagree with your disagreement.
The major difference is patterns.
Some people will prefer one pattern over another.
Good luck trying to grade patterns in a fancy diamond grading system.
You are trying to use a tough example to disprove a well thought out approach Karl.
In the case you are using it would take more iterations and perhaps a decade to refine a system to match human preferences - but it should not be impossible.
And remember that in the first instance it is easy for a trained artificial intelligence system to eliminate poor performance stones. Ones with say too many dark zones close to each other, or really poor light return.

You have seen the article before - as the only 100% english native speaking author, I can tell you the closeness and thousands of interactions in writing and editing this article has given me a deeper understanding of the meanings included.
Mark my words, this article will be read and re read for the next decade or more by the people and organisations that will ultimately achieve the goal of human cut quality and beauty assessment - and that will lead to the information systems that can create more beautiful diamonds.
And they will not be cookie cutter solutions - there will be amazing diversification.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Garry- if it takes a decade to refine, the results might not be relevant as tastes change.
What is the "best" angle and size for the corner on an Emerald Cut?
You'd need a ton of experts, consumers and others to validate a system that still could not account for differences in taste.....

btw- I have nothing but the utmost respect for the work. I believe that being a "devil's advocate" might make me seem like a pain- but hopefully I raise relevant points to the work for the long term.
IN the case of here and now, PS is a group where the discussion of what's best is extremely relevant.
Consumers come here every day wanting to get "the best".
 

Garry H (Cut Nut)

Super_Ideal_Rock
Trade
Joined
Aug 15, 2000
Messages
18,423
Rockdiamond|1481760079|4107331 said:
Garry- if it takes a decade to refine, the results might not be relevant as tastes change.
What is the "best" angle and size for the corner on an Emerald Cut?
You'd need a ton of experts, consumers and others to validate a system that still could not account for differences in taste.....

btw- I have nothing but the utmost respect for the work. I believe that being a "devil's advocate" might make me seem like a pain- but hopefully I raise relevant points to the work for the long term.
IN the case of here and now, PS is a group where the discussion of what's best is extremely relevant.
Consumers come here every day wanting to get "the best".

From Robert Parker on wine preferences and taste:

Scores, however, do not reveal the important facts about a wine. The written commentary that accompanies the ratings is a better source of information regarding the wine's style and personality, its relative quality vis-à-vis its peers, and its value and aging potential than any score could ever indicate.
— Robert M. Parker, Jr., The Wine Advocate Rating System[23] (emphasis in original)
No scoring system is perfect, but a system that provides for flexibility in scores, if applied by the same taster without prejudice, can quantify different levels of wine quality and provide the reader with one professional's judgment. However, there can never be any substitute for your own palate nor any better education than tasting the wine yourself.
— Robert M. Parker, Jr., The Wine Advocate Rating System[23] (emphasis in original)

In essence we wine buffs can read these reviews and the ranking number (between 50 and 100) and have a very good idea about the effect of the corner size and angle. A novice drinker will struggle, but then more diamonds are purchased by people with an addiction than by novices.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Although there are parallels, here's where I think the wine analogy falls short:
1) how many wine experts are there in the world?
How many diamond experts?
I think the winos outnumber the diamond lovers 50:1
This makes the sample size an issue.

2) Range of opinions will vary more with diamonds than wine. I agree that in shapes like Pear, Princess, Marquise, Oval there's going to be broad consensus on a fairly narrow range of what the shape, is supposed to look like. Maybe even Asscher cut. What about Radiant Cut, Heart Shape, Rectangular Emerald Cut. Cushion cut! There's many different methods for a creative cutter to achieve very good results- using different light paths and contrast patterns. So that two very well cut stones can perform differently
And even in Pears Marquise etc- as soon as we start varying from the classic ideal shape and proportions, the parameters become less linear- or non linear. Difficult to grade on a 1-10 scale.
I just think the consensus would be more broad on wine as opposed to diamonds.
Of course these are my opinions not facts.
I just don't see how a numerical, linear fancy shape grading system can allow for the myriad of options cutters can exercise with odd shaped rough.
I would envision a more subjective approach.
 

Garry H (Cut Nut)

Super_Ideal_Rock
Trade
Joined
Aug 15, 2000
Messages
18,423
Rockdiamond|1481770576|4107376 said:
Although there are parallels, here's where I think the wine analogy falls short:
1) how many wine experts are there in the world?
How many diamond experts?
I think the winos outnumber the diamond lovers 50:1
This makes the sample size an issue.

2) Range of opinions will vary more with diamonds than wine. I agree that in shapes like Pear, Princess, Marquise, Oval there's going to be broad consensus on a fairly narrow range of what the shape, is supposed to look like. Maybe even Asscher cut. What about Radiant Cut, Heart Shape, Rectangular Emerald Cut. Cushion cut! There's many different methods for a creative cutter to achieve very good results- using different light paths and contrast patterns. So that two very well cut stones can perform differently
And even in Pears Marquise etc- as soon as we start varying from the classic ideal shape and proportions, the parameters become less linear- or non linear. Difficult to grade on a 1-10 scale.
I just think the consensus would be more broad on wine as opposed to diamonds.
Of course these are my opinions not facts.
I just don't see how a numerical, linear fancy shape grading system can allow for the myriad of options cutters can exercise with odd shaped rough.
I would envision a more subjective approach.
1. There are way more people wearing diamonds than wino's in the world.
2. check out this infographic http://winefolly.com/tutorial/wine-descriptions-chart-infographic/ David.
I suppose we have brilliance (brightness plus contrast), fire, and pattern / scintillation contrast plus spread.
 

Karl_K

Super_Ideal_Rock
Trade
Joined
Aug 4, 2008
Messages
14,634
Garry H (Cut Nut)|1481759614|4107328 said:
Karl_K|1481758603|4107324 said:
Garry, I disagree with your disagreement.
The major difference is patterns.
Some people will prefer one pattern over another.
Good luck trying to grade patterns in a fancy diamond grading system.
You are trying to use a tough example to disprove a well thought out approach Karl.
Garry, it is not my intent to disprove.
I always throw the toughest challenges at any system to explore the limits of the system.
Even a limited system can be useful as long as the limits are explored and understood.
For example: reflector technologies are limited but can be useful when properly applied and considered.
 

Serg

Ideal_Rock
Trade
Joined
Mar 21, 2002
Messages
2,620
I would like clarify . First target ( step) For AI system is not final grading. First targets are searching, pre sorting systems.
You would ask them to find Similar diamonds as your reference , similar by optical symmetry , similar by optical performance ,..
Such systems does not to have 100% accuracy and repeatability but in same time they have bigger adding value for buyers ( end consumers and professional buyers ) than just grading systems.
 

oldminer

Ideal_Rock
Trade
Joined
Sep 3, 2000
Messages
6,691
Serg; You have made the shortest and most reasoned response in this thread. It is a great subject and surely controversial. I agree that systems will streamline and simplify some human grading efforts, but judging what individuals like the best or consider orders of beauty will continue to recognize and validate our personal and human decision making. Adding value to the product while at the same time making uniform descriptions of certain aspects of appearance and shape will help consumers and the trade to buy, sell and enjoy the wide range of diamonds that can be cut.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Serg said:
I would like clarify . First target ( step) For AI system is not final grading. First targets are searching, pre sorting systems.
You would ask them to find Similar diamonds as your reference , similar by optical symmetry , similar by optical performance ,..
Such systems does not to have 100% accuracy and repeatability but in same time they have bigger adding value for buyers ( end consumers and professional buyers ) than just grading systems.

Totally agree with my friend David Atlas on this.
Serg- that's an amazing idea which would deal with a lot of my considerations.
For example- say someone wants a rectangular Radiant-it makes no sense to compare light performance with a square stone.
If there are categories that encompass the outlying tastes and styles that could make an inclusive system that works for a broad range of tastes.
 

Rockdiamond

Ideal_Rock
Trade
Joined
Jan 7, 2009
Messages
9,711
Garry H (Cut Nut)|1481772500|4107381 said:
Rockdiamond|1481770576|4107376 said:
Although there are parallels, here's where I think the wine analogy falls short:
1) how many wine experts are there in the world?
How many diamond experts?
I think the winos outnumber the diamond lovers 50:1
This makes the sample size an issue.

2) Range of opinions will vary more with diamonds than wine. I agree that in shapes like Pear, Princess, Marquise, Oval there's going to be broad consensus on a fairly narrow range of what the shape, is supposed to look like. Maybe even Asscher cut. What about Radiant Cut, Heart Shape, Rectangular Emerald Cut. Cushion cut! There's many different methods for a creative cutter to achieve very good results- using different light paths and contrast patterns. So that two very well cut stones can perform differently
And even in Pears Marquise etc- as soon as we start varying from the classic ideal shape and proportions, the parameters become less linear- or non linear. Difficult to grade on a 1-10 scale.
I just think the consensus would be more broad on wine as opposed to diamonds.
Of course these are my opinions not facts.
I just don't see how a numerical, linear fancy shape grading system can allow for the myriad of options cutters can exercise with odd shaped rough.
I would envision a more subjective approach.
1. There are way more people wearing diamonds than wino's in the world.
2. check out this infographic http://winefolly.com/tutorial/wine-descriptions-chart-infographic/ David.
I suppose we have brilliance (brightness plus contrast), fire, and pattern / scintillation contrast plus spread.

I somehow missed your response Garry,
1) Interesting debate which, of course, can't be settled....Take France- in fact pretty much all of western Europe.
People drinking wine is common, people wearing diamonds over 1ct super rare.
But the point I raised is that there's more wine experts than diamond experts.

2) excellent chart, which as a total non wine drinker, proves my point in a way. Terms like Cliff Edge and Hollow have no meaning to someone who considers most wine "Austere":)

If we are using Brilliance, contrast etc as measured by a machine, we'll have a debate about the usefulness of such measurements.
The idea of categories, and different characteristics within those categories is very similar to the wine chart- and could be a great avenue
 
Be a part of the community Get 3 HCA Results
Top