Letter to the Editor of the Australian Gemmologist

[Garry H (Cut Nut)]

Date: 4/2/2007 12:47:58 PM
Author: Serg

Big part table in RD with table +60(65) try collect light sources in direction +45 degree . Usually light is on central part ceiling and in central part wall. Light sources in corner between ceiling and wall are to rare.

real light distribution is much more complex then hemisphere with square cosines or Blue-Red-Green ASET.

This site can help you understand what sergey means.

It is a very powerful tool and one that lead Sergey to the ETAS concept.
But unlike the AGS approach to fire, it still does not do what ETAS can - and that is adjust for human attributes

e.g. a flash of fire can be so bright that it flares and appears bigger to us than it actually would be if it were assessed digitally

edited to add the link: http://www.cutstudy.com/cut/english/comp/scint1.htm (it is nearly 10 year old site and Microsoft browsers might need some setting adjustments)

 

[RockDoc]

Garry

I don''t understand the image that the link takes you to.

The stone imaged has a 40 crown angle ? With a 60% table?

The pattern shown looks well organized, but is labeled scintillation.

How does it adjust for human variance perception? How do we replicate this in real life instead of a computer pattern? Is it possible.

For some stones I''ve seen similar patterns in the Gemprint machine. But not all stones have a succint pattern like that.

I''d like to be able to comprehend and understand this.

Rockdoc

 

[Garry H (Cut Nut)]

Date: 4/2/2007 10:15:22 PM
Author: RockDoc
Garry

I don''t understand the image that the link takes you to.

The stone imaged has a 40 crown angle ? With a 60% table?

The pattern shown looks well organized, but is labeled scintillation.

How does it adjust for human variance perception? How do we replicate this in real life instead of a computer pattern? Is it possible.

For some stones I''ve seen similar patterns in the Gemprint machine. But not all stones have a succint pattern like that.

I''d like to be able to comprehend and understand this.

Rockdoc

Roc it is interactive - it is like a game - you can change the proportions and see the difference.

The stone is under an imaginary sheet of tracing paper with a beam or ray shining on the stone, and the sparkles hit the tracing paper.
You can move the stone, change the steps in the proportions and all sorts of stuff

 

[michaelgem]

Sergey comments:

A big part of the table area (the table reflection) in the RD with a table over 60(65) tries to collect light sources in the +45 degree direction. Typical or usual lighting is on the central ceiling area and in the central part of the wall. Light sources are rare in the corner between ceiling and wall where the diamond when held face-up is trying to collect its light. Sergey

Thank you, Sergey.

This is exactly what I was commenting upon when I said to Strm:

“To answer your question, I do think this is a matter of a real downside. Isn't the diminishment in optical performance apparent when you look at this image in diamcalc's disco lighting?”

Look at the enclosed DiamCalc image comparison of the 65% table diamond and one with a 56% table. In the 65T diamond the table reflection area has become too prominent, and is negatively affecting the diamond’s optical performance. The typical Ideal table reflection area seen in the 56T is smaller and nearer one-third the table diameter.

In the 65T the inner ring pupil of the ‘eye of the diamond’ has dilated from the one-third Ideal size to 50+ percent. This table reflection area has the poorer optical performance properties accurately explained by Serg. It is diminishing the overall beauty of this diamond in comparison to the same diamond with a 56T.

You can see the enlarged darkness in the table reflection at 12 o’clock as it reflects from too low an angle where there is no light. The rest of the inner ring table reflection is uniform and lacks the desirable contrast aspect of brilliance. (That is why I said: “If you are going this route, I suggest longer pavilion halves" to help compensate for this diminishment in performance.)

This simple analysis of the diamond's image is what I refer to in my writing as direct assessement of aspects of diamond optical performance. Direct assessment reveals aspects of a diamond's optical performance that current metrics are not yet taking into account. (Of course, it is easy to misread these images, and it is important to use typical illumination and check other angles of observation.)

This diminishment in light performance and the perception that diamonds with table sizes noticeably greater than 60% have a thin crown with a flat, spread appearance that is not perceived by cutters, jewelers or consumers as Ideal are the two reasons why the AGS would be wise to exclude diamond’s with tables noticeably greater than 60% from the Ideal 0 grade.

Michael Cowing

 

[Serg]

You can find here PGS AGS result for 3% girdle bezel, 55% Star (45% upper girdle) and 80% lower girdle facet length (81.5% lower facet depth). Table 51%-66%

http://octonus.com/oct/mss/gia-agspgs.phtml

Here you can compare PGS result and AGS guideline for AGS0

http://octonus.com/oct/mss/gia-ags.phtml

Here you can compare FacetWare GIA and PGS AGS results

http://octonus.com/oct/mss/gia&agspgs.phtml

 

[adamasgem]

Date: 4/13/2007 12:01:59 AM
Author: Serg

You can find here PGS AGS result for 3% girdle bezel, 55% Star (45% upper girdle) and 80% lower girdle facet length (81.5% lower facet depth). Table 51%-66%

http://octonus.com/oct/mss/gia-agspgs.phtml

Here you can compare PGS result and AGS guideline for AGS0

http://octonus.com/oct/mss/gia-ags.phtml



Here you can compare FacetWare GIA and PGS AGS results

http://octonus.com/oct/mss/gia&agspgs.phtml

Sergey.. Nice summary and a lot of work BUT... Do the same, considering
1) GIA quantizes to nearest 5% for stars and LGF and nearest 0.5 degreee fro crown angle, etc, etc, etc

Look at the possible PGS results and resultannt VARIANCES when the stone is assymetrical WITHIN what GIA calls equalnumbers.. a 80% LGF GIA AVERAGE means the average LGF is between 77.5 and 82.5 and a 55% star has an average between 52.5 and 57.5, etc..

PGS IS NOT A PARAMETRICALLY BASED SYSTEM AND IS NOT BASED ON AVERAGES ..
I know you guys don''t like the GUIDELINE CHARTS, but I would like to see you look at what the gradients are for normal assymetries outside the initial guidelines. REMEMBER, even though you are inside the GUIDELINES with your averages, assymetries may not get you the percieved grade.

It is not as simple as some might like to portray it...

 

[michaelgem]

Here is the argument for constraints on table size and table reflection size:

Because the AGS metrics do not take some important aspects of the Ideal cut into consideration, their current grading system is not adequately accounting for their importance. A thin, spready crown with too large a table reflection is just one of these aspects. There are others.

Early on, spread was one of these aspects. At the Moscow conference I mentioned to Peter that one could solve the problem of non-Ideal diamonds with thick girdles or high crowns, that can have high optical performance. You do this by imposing tight constraints on spread. A diamond is not Ideal in spite of ideal performance if it looks noticeably smaller than expected for its weight. You expect, for example, a 1ct to be close to 6.49mm – 6.5mm. The PGS software now factors that into the overall cut grade metric as deductions for “weight ratio”.

The issue is that non-ideal cut diamonds with thin crowns and spready tables get high scores in their metric. The metric does not fully account for the diminishment in optical performance caused by the spready table and the large table reflection. This is solved as I have done by imposing constraints on the table size and the size of the table reflection.

This is no different or any less justified than the AGS imposed constraints on the overall diameter for a given weight.

The current AGS Diamond Quality Documents state they are “Performance-Based.” It is clear and a good thing that they are more than that. They are also based upon craftsmanship aspects such as proper spread, ideal polish, ideal symmetry, ideal facet alignment and meet points. There are also constraints on culet size, girdle thickness and durability.

It is a logical addition to include constraints on table size and table reflection size in addition to spread or weight ratio constraints. This eliminates thin crown, spready diamonds with their previously demonstrated performance deficiencies from the rank of Ideal 0. You justify this both from a craftsmanship perspective and a performance perspective.

These additional needed constraints are not in conflict with but rather augment the overall AGS cut grading system.

Michael Cowing

 

[Paul-Antwerp]

Date: 4/13/2007 10:31:44 PM
Author: michaelgem

Here is the argument for constraints on table size and table reflection size:

Because the AGS metrics do not take some important aspects of the Ideal cut into consideration, their current grading system is not adequately accounting for their importance. A thin, spready crown with too large a table reflection is just one of these aspects. There are others.

Michael,

One of the methods used in demagogy is that of half-truths. Another one is the use of ''apples and oranges''. In this case, which ''ideal cut'' are you talking about? Are you sure that ''Ideal'' should be reserved to that specific look of ''ideal cut'' only? Or could it be that you need to broaden your perspective and accept that there might be other ''ideal cuts'', which you have never ever seen or which never actually have been cut?

Date: 4/13/2007 10:31:44 PM
Author: michaelgem

Early on, spread was one of these aspects. At the Moscow conference I mentioned to Peter that one could solve the problem of non-Ideal diamonds with thick girdles or high crowns, that can have high optical performance. You do this by imposing tight constraints on spread. A diamond is not Ideal in spite of ideal performance if it looks noticeably smaller than expected for its weight. You expect, for example, a 1ct to be close to 6.49mm – 6.5mm. The PGS software now factors that into the overall cut grade metric as deductions for “weight ratio”.

Another favourite method of the demagogue is false authority. In this case, i was also present at the Moscow conference, and the factor of a spread-reduction was already factored in into the new AGS-system, that they were presenting there for the first time. You may have well mentioned this, but it is like telling me that I am European.

Date: 4/13/2007 10:31:44 PM
Author: michaelgem

The issue is that non-ideal cut diamonds with thin crowns and spready tables get high scores in their metric. The metric does not fully account for the diminishment in optical performance caused by the spready table and the large table reflection. This is solved as I have done by imposing constraints on the table size and the size of the table reflection.

A demagogue also often uses the technique of generalisation. In this case, you are saying that thin crowns and spready tables score high in the AGS-metric, while it entirely depends on the whole stone. You need that generalisation however, since there are basically no AGS-0-stones on the market with these spready tables. So, you need the generalisation to show that a spready table (which can be found on the market) actually does not perform that well. In this way, you end up with a demonization (another demagogic technique) of all spready tables.

Date: 4/13/2007 10:31:44 PM
Author: michaelgem

This is no different or any less justified than the AGS imposed constraints on the overall diameter for a given weight.

The current AGS Diamond Quality Documents state they are “Performance-Based.” It is clear and a good thing that they are more than that. They are also based upon craftsmanship aspects such as proper spread, ideal polish, ideal symmetry, ideal facet alignment and meet points. There are also constraints on culet size, girdle thickness and durability.

It is a logical addition to include constraints on table size and table reflection size in addition to spread or weight ratio constraints. This eliminates thin crown, spready diamonds with their previously demonstrated performance deficiencies from the rank of Ideal 0. You justify this both from a craftsmanship perspective and a performance perspective.

These additional needed constraints are not in conflict with but rather augment the overall AGS cut grading system.

Michael Cowing

So, my dear friend, your whole argument contains no beef. If you could only open up your mind to the possibility that what you have learned and observed over years of experience might not be complete. There could well be some uncharted terrain where nobody ever ventured into, because of unknown reasons, but traditions and the resulting economic marketing-consequences might well be at the root of this.

It is high time that you open up your mind, and start thinking outside of the box.

Live long,

 

[michaelgem]

Paul,

An individual on Pricescope commented to me the other day that he was amazed that you call me a demagogue. He said that my writing doesn’t rise to the level of what he sees as real demagoguery by some here on Pricescope.

There is room for all our ideas and opinions, and both your ideas and mine have validity in the context in which they are said.

Your statements “there might be other ''ideal cuts'', which you have never ever seen or which never actually have been cut?” and “There could well be some uncharted terrain where nobody ever ventured into, because of unknown reasons, but traditions and the resulting economic marketing-consequences might well be at the root of this.” certainly have validity.

It occurs to me that a metric such as the AGS ASET and PGS software is not likely to venture into that uncharted terrain or find other different ideal cuts, because it was designed to define the range of proportions that yield performance similar to the Tolkowsky Ideal. It gives high scores to diamond cuts with similar distributions of red blue and green.

To answer your question “which ''ideal cut'' are you talking about?” I thought I’ve made it very clear that I am talking about the proportions of the standard round brilliant that are perceived to have the best brilliance (both brightness and contrast), fire and sparkle in typical illumination circumstances. This is the Ideal cut I am talking about.

The AGS long ago decided to remain true to this original concept of Ideal. They have reserved the designation “Ideal 0” for this historically most important diamond cut that evolved from Morse’s and Tolkowsky’s contributions to diamond cutting.

I am sorry that you see me as an antagonist rather than a contributor to understanding of this concept of Ideal in the standard round brilliant.

Rather than an analysis of the way my writing comes across, I would like to hear from you or others as to why I am wrong when I demonstrate the diminishment in optical performance caused by too large a table resulting in too large a table reflection? Do you think this is or is not a reason to remove such diamonds from the Ideal 0 category? You are, after all, one of the expert cutters of this Ideal. So your opinion in this regard is as important as anyone’s. Are you cutting Ideals with table sizes noticeably greater than 60%? Do you find them as beautiful and high performing?

I expend a lot of time and energy writing here. Hopefully, many will appreciate the truth and knowledge about the Ideal cut that I am demonstrating and imparting. I have found that this knowledge is in accordance with cutters of the Ideal.

So let me repeat the “meat” of my last post:

The issue is that non-ideal cut diamonds with thin crowns and spready tables get high scores in the current AGS Ideal 0-10 grading system. Their metric does not fully account for the diminishment in optical performance caused by spready tables with large table reflections. This is solved as I have done by imposing constraints on the table size and the size of the table reflection.

This is no different or any less justified than the AGS imposed constraints on the overall diameter for a given weight, and the constraints on the allowed girdle thickness and culet size.

The current AGS Diamond Quality Documents state they are “Performance-Based.” It is clear and a good thing that they are more than that. They are also based upon craftsmanship aspects such as proper spread, ideal polish, ideal symmetry, ideal facet alignment and meet points. There are constraints on culet size, girdle thickness and durability.

It is a logical addition to include constraints on table size and table reflection size in addition to spread or weight ratio constraints. This eliminates thin crown, spready diamonds with their previously demonstrated performance deficiencies from the rank of Ideal 0. You justify this both from a craftsmanship perspective and a performance perspective.

These additional needed constraints are not in conflict with but rather augment the overall AGS Ideal 0 cut grading system.

Michael Cowing

 

[Garry H (Cut Nut)]

Michael have you ever wondered what a graing system might achieve if it did not start out with an intention to prove that one stone was best, and then allow a tolerance for what was acceptable as ''good enough'' ?

Why not design a system that describes the worst as 0.000, and lets anything that is better be say 1, and better than that is 2 etc?

We have seen that the AGS PGS software tolerances for ''preformance'' seem to be set at 0.0 plus or minus some tolerance or other for each of the factors they calculate or count colored pixels for does not get them a strict a grade as we imagine they might prefer (ideal may not be ideal).

Can we question such an idea as ''ideal'' as Sergey and Yuri have shown us at Moscow?

What are the effects that these terms and usages have and could have on the overall market?

 

[strmrdr]

Date: 4/14/2007 7:41:31 AM
Author: Paul-Antwerp

Date: 4/13/2007 10:31:44 PM
Author: michaelgem

Here is the argument for constraints on table size and table reflection size:

Because the AGS metrics do not take some important aspects of the Ideal cut into consideration, their current grading system is not adequately accounting for their importance. A thin, spready crown with too large a table reflection is just one of these aspects. There are others.

Michael,

One of the methods used in demagogy is that of half-truths. Another one is the use of ''apples and oranges''. In this case, which ''ideal cut'' are you talking about? Are you sure that ''Ideal'' should be reserved to that specific look of ''ideal cut'' only? Or could it be that you need to broaden your perspective and accept that there might be other ''ideal cuts'', which you have never ever seen or which never actually have been cut?

Date: 4/13/2007 10:31:44 PM
Author: michaelgem

Early on, spread was one of these aspects. At the Moscow conference I mentioned to Peter that one could solve the problem of non-Ideal diamonds with thick girdles or high crowns, that can have high optical performance. You do this by imposing tight constraints on spread. A diamond is not Ideal in spite of ideal performance if it looks noticeably smaller than expected for its weight. You expect, for example, a 1ct to be close to 6.49mm – 6.5mm. The PGS software now factors that into the overall cut grade metric as deductions for “weight ratio”.

Another favourite method of the demagogue is false authority. In this case, i was also present at the Moscow conference, and the factor of a spread-reduction was already factored in into the new AGS-system, that they were presenting there for the first time. You may have well mentioned this, but it is like telling me that I am European.

Date: 4/13/2007 10:31:44 PM
Author: michaelgem


The issue is that non-ideal cut diamonds with thin crowns and spready tables get high scores in their metric. The metric does not fully account for the diminishment in optical performance caused by the spready table and the large table reflection. This is solved as I have done by imposing constraints on the table size and the size of the table reflection.

A demagogue also often uses the technique of generalisation. In this case, you are saying that thin crowns and spready tables score high in the AGS-metric, while it entirely depends on the whole stone. You need that generalisation however, since there are basically no AGS-0-stones on the market with these spready tables. So, you need the generalisation to show that a spready table (which can be found on the market) actually does not perform that well. In this way, you end up with a demonization (another demagogic technique) of all spready tables.

Date: 4/13/2007 10:31:44 PM
Author: michaelgem


This is no different or any less justified than the AGS imposed constraints on the overall diameter for a given weight.

The current AGS Diamond Quality Documents state they are “Performance-Based.” It is clear and a good thing that they are more than that. They are also based upon craftsmanship aspects such as proper spread, ideal polish, ideal symmetry, ideal facet alignment and meet points. There are also constraints on culet size, girdle thickness and durability.

It is a logical addition to include constraints on table size and table reflection size in addition to spread or weight ratio constraints. This eliminates thin crown, spready diamonds with their previously demonstrated performance deficiencies from the rank of Ideal 0. You justify this both from a craftsmanship perspective and a performance perspective.

These additional needed constraints are not in conflict with but rather augment the overall AGS cut grading system.

Michael Cowing

So, my dear friend, your whole argument contains no beef. If you could only open up your mind to the possibility that what you have learned and observed over years of experience might not be complete. There could well be some uncharted terrain where nobody ever ventured into, because of unknown reasons, but traditions and the resulting economic marketing-consequences might well be at the root of this.

It is high time that you open up your mind, and start thinking outside of the box.

Live long,

Hey Paul,
I gota say im am extremely disappointed in what you wrote here.
You have the knowledge to answer the points and did give your opinion, but frankly the disrespectful name calling was uncalled for.
Now there are others id expect it from, but had more respect for you to think you were one of them.

 

[strmrdr]

Date: 4/14/2007 4:31:31 PM
Author: Garry H (Cut Nut)
Michael have you ever wondered what a graing system might achieve if it did not start out with an intention to prove that one stone was best, and then allow a tolerance for what was acceptable as ''good enough'' ?

Why not design a system that describes the worst as 0.000, and lets anything that is better be say 1, and better than that is 2 etc?

We have seen that the AGS PGS software tolerances for ''preformance'' seem to be set at 0.0 plus or minus some tolerance or other for each of the factors they calculate or count colored pixels for does not get them a strict a grade as we imagine they might prefer (ideal may not be ideal).

Can we question such an idea as ''ideal'' as Sergey and Yuri have shown us at Moscow?

What are the effects that these terms and usages have and could have on the overall market?

No matter what grading system someone comes up with it will be nothing more than their opinion on what makes for a well cut diamond.
For that very reason they all suck in one way or another because not everyone agrees on what the perfect diamond looks like or even what an acceptable diamond looks like.

riddle me this.. which grading system or tool will tell me that a diamond will look to me better than all other diamonds when they are on my finger?
IS,ASET-30,ASET-40 - a lot of diamonds will give different performance "readings" under each device. Which is right?
Which image will most match the performance curve of the diamond on my finger?
How much obstruction is the diamond on my finger under? which way are the lights coming from? What type of lights? whats the environment? am I moving my finger? if so how much?
No way to tell is there.....

 

[Garry H (Cut Nut)]

Date: 4/15/2007 12:04:25 AM
Author: strmrdr


riddle me this.. which grading system or tool will tell me that a diamond will look to me better than all other diamonds when they are on my finger?
IS,ASET-30,ASET-40 - a lot of diamonds will give different performance ''readings'' under each device. Which is right?
Which image will most match the performance curve of the diamond on my finger?
How much obstruction is the diamond on my finger under? which way are the lights coming from? What type of lights? whats the environment? am I moving my finger? if so how much?
No way to tell is there.....

That is a bit too easy Storm - HCA and Ideal-scope for round diamonds will show up diamonds that have leakage which is always worse when the diamond is dirty.
All the other systems penalize too much for obstruction (which is rarely an issue ''on the finger'' ) and favour deeper smaller looking diamonds.

 

[strmrdr]

Date: 4/15/2007 3:32:23 AM
Author: Garry H (Cut Nut)

That is a bit too easy Storm - HCA and Ideal-scope for round diamonds will show up diamonds that have leakage which is always worse when the diamond is dirty.
All the other systems penalize too much for obstruction (which is rarely an issue ''on the finger'' ) and favour deeper smaller looking diamonds.

Of so you feel 22% is better than 30%? Why(addressed above a little)?
personally I think 40 is bogus so that leaves 22 and 30.
There will be a center level of stones that look great under all 3 but it does move the edges of the system around.
Which is really what this argument is about, the edges.
How do you define those edges?
Even with the IS over the years here defining the edges of the acceptable image has been a huge issue.

Which system in your opinion defines the edge of acceptable better with rounds? and why?
AGS-0
GIA-EX
HCA(properly applied)+IS/ASET
HCA(properly applied)+IS/ASET+hearts
HCA(properly applied)+IS/ASET+hearts+full helium/sarin report.

For me there is a pretty clear answer which of the above is better for rounds.

 

[Garry H (Cut Nut)]

what is 22?

 

[Paul-Antwerp]

Dear Michael,

I am sorry about the way I approached you yesterday. It just shows the level of frustration that you give me by not reading and listening, and not opening up to other possibilities. I am not saying that you are wrong, you are just clinging on to certain beliefs and fixed concepts, that you cannot see or grasp the possibility of there being other options.

At the end of page 6 of this thread, Garry Holloway introduces a proportion-combo with a 65% table, which does NOT reach the AGS-0-grade. Ever since you are constantly repeating that AGS should have an extra rule in their system so that such a table-size does not get AGS-0. I truly wonder why you spend all this energy in writing these posts about a non-issue from the start. It is extremely frustrating.

Immediately, you put all ''larger-tables'' in the same basket, by describing them as low crown-height, shallow crown-angle. In the original example, it is a 33°-angle, which is just slightly shallower than the traditional ideal, but it is far from the real shallow angles of 30° and lower, which you most often find on currently marketed ''larger-tables''. So, you are just disregarding another important fact, in order to lump it together with proportion-sets that you know and have studied. Again, the question is why do you disregard another fact while spending so much energy into denouncing what does not even exist anyway?

Every time when you are under pressure about this, you invoke the ''holy'' Tolkowsky and Morse. The thing is: these people were not wrong, but they could well be incomplete. In the same way, Henry Ford was an amazing innovator in factory organisation, but I would advise you not to organise a seminar on his techniques nowadays. Or do you remember the great economic mind Keynes? Anyone interested in any keynesian economic regulations nowadays?

So, yes, Tolkowsky and Morse were great minds of their times, and it is amazing at how well their predictions were, knowing how limited their resources were in those days.

You are asking me now directly, as a kind of proof of your theory, why I personally am not cutting these ideals with higher table-sizes? What a single company chooses as a strategy is exactly a strategy, and is not a proof of a certain gemmological truth. I have often before said that within AGS-0, there are still several levels of cut-quality according to me. We have chosen our goal-proportion-set and (very important) our strictness in cutting in an attempt to achieve the best possible light performance. The fact that this ends up with an AGS-0-grade is a side-effect, not a goal.

And even if achieving the 0-grade would be our only goal, and we could cut some stones with a 62%-table, we would not do it. Why? Because there are so many traditions and set beliefs in the market, and it would be a gigantic task to educate wholesalers, retailers and consumers that this table-size can also warrant the label Ideal. I personally know how difficult it is to bring across the simple truth that in princess-cuts, depth is not inversely related to spread. So why would any cutter go against traditional set belief and cut such ideals: this is basically the main reason why you do not see these in the market.

So why am I so adamantly fighting your theory, if it does not concern our product in any way? The reason is because with your holding onto tradition, and rejecting new knowledge, you are actively blocking scientific progress. Especially since you are an authority in gemmology, with very high regards all over the U.S., I find it very important to stop you when you are wrong.

What is more, I totally do not understand how you keep on pressing about a non-issue, starting with incorrect information, and just adding incomplete info and theory in order to support your theory.

Again, in your last post, you are suddenly bringing up the case that the new AGS-system were designed starting from Tolkowsky and Morse. What do you base this statement on? I have seen countless presentations of the system by AGS, including the very first one in Moscow, and I have never heard this proposition. It is yet another example of incorrect or incomplete info in your posts.

Therefore, I plea to you, let it rest. Do not try to bring in extra info, which will again be incomplete or incorrect. As I showed in the beginning of this post, it is a non-issue to start with. Stop wasting our energy on this. Put it into studying new theories, do also study the princess-cut, because by stepping away from the round brilliant and applying what you think you know to another shape, you will learn a lot, while it also is a humbling experience.

Live long,

 

[Garry H (Cut Nut)]

Date: 4/15/2007 7:40:09 AM
Author: Paul-Antwerp
Dear Michael,

I am sorry about the way I approached you yesterday. It just shows the level of frustration that you give me by not reading and listening, and not opening up to other possibilities. I am not saying that you are wrong, you are just clinging on to certain beliefs and fixed concepts, that you cannot see or grasp the possibility of there being other options.

At the end of page 6 of this thread, Garry Holloway introduces a proportion-combo with a 65% table, which does NOT reach the AGS-0-grade. Ever since you are constantly repeating that AGS should have an extra rule in their system so that such a table-size does not get AGS-0. I truly wonder why you spend all this energy in writing these posts about a non-issue from the start. It is extremely frustrating.

Immediately, you put all ''larger-tables'' in the same basket, by describing them as low crown-height, shallow crown-angle. In the original example, it is a 33°-angle, which is just slightly shallower than the traditional ideal, but it is far from the real shallow angles of 30° and lower, which you most often find on currently marketed ''larger-tables''. So, you are just disregarding another important fact, in order to lump it together with proportion-sets that you know and have studied. Again, the question is why do you disregard another fact while spending so much energy into denouncing what does not even exist anyway?

Every time when you are under pressure about this, you invoke the ''holy'' Tolkowsky and Morse. The thing is: these people were not wrong, but they could well be incomplete. In the same way, Henry Ford was an amazing innovator in factory organisation, but I would advise you not to organise a seminar on his techniques nowadays. Or do you remember the great economic mind Keynes? Anyone interested in any keynesian economic regulations nowadays?

So, yes, Tolkowsky and Morse were great minds of their times, and it is amazing at how well their predictions were, knowing how limited their resources were in those days.

You are asking me now directly, as a kind of proof of your theory, why I personally am not cutting these ideals with higher table-sizes? What a single company chooses as a strategy is exactly a strategy, and is not a proof of a certain gemmological truth. I have often before said that within AGS-0, there are still several levels of cut-quality according to me. We have chosen our goal-proportion-set and (very important) our strictness in cutting in an attempt to achieve the best possible light performance. The fact that this ends up with an AGS-0-grade is a side-effect, not a goal.

And even if achieving the 0-grade would be our only goal, and we could cut some stones with a 62%-table, we would not do it. Why? Because there are so many traditions and set beliefs in the market, and it would be a gigantic task to educate wholesalers, retailers and consumers that this table-size can also warrant the label Ideal. I personally know how difficult it is to bring across the simple truth that in princess-cuts, depth is not inversely related to spread. So why would any cutter go against traditional set belief and cut such ideals: this is basically the main reason why you do not see these in the market.

So why am I so adamantly fighting your theory, if it does not concern our product in any way? The reason is because with your holding onto tradition, and rejecting new knowledge, you are actively blocking scientific progress. Especially since you are an authority in gemmology, with very high regards all over the U.S., I find it very important to stop you when you are wrong.

What is more, I totally do not understand how you keep on pressing about a non-issue, starting with incorrect information, and just adding incomplete info and theory in order to support your theory.

Again, in your last post, you are suddenly bringing up the case that the new AGS-system were designed starting from Tolkowsky and Morse. What do you base this statement on? I have seen countless presentations of the system by AGS, including the very first one in Moscow, and I have never heard this proposition. It is yet another example of incorrect or incomplete info in your posts.

Therefore, I plea to you, let it rest. Do not try to bring in extra info, which will again be incomplete or incorrect. As I showed in the beginning of this post, it is a non-issue to start with. Stop wasting our energy on this. Put it into studying new theories, do also study the princess-cut, because by stepping away from the round brilliant and applying what you think you know to another shape, you will learn a lot, while it also is a humbling experience.

Live long,

Well and politely said Paul

Michael as an example of what could be raised as an issue for a more positive discussion:
What size would be the minimum diameter for a Tolkowsky round below which a single cut style would look and perform better?
Or
Should table sizes on Tolkowsky crown and pavilion angles change with size - e.g. there are many people who believe 60% or bigger is better below say 0.10ct
Or
What should GIA do if there is any agreement with the Cut Groups letter to the editor?
Or
What are the size breaks where princess cuts should have 2, 3 or 4 or even 5 chevrons, and should that change with princess cut table sizes?
or 1,000 other things that need good answers

 

[strmrdr]

Date: 4/15/2007 6:00:40 AM
Author: strmrdr

Date: 4/15/2007 3:32:23 AM
Author: Garry H (Cut Nut)

That is a bit too easy Storm - HCA and Ideal-scope for round diamonds will show up diamonds that have leakage which is always worse when the diamond is dirty.
All the other systems penalize too much for obstruction (which is rarely an issue 'on the finger' ) and favour deeper smaller looking diamonds.

Of so you feel 22% is better than 30%? Why(addressed above a little)?
personally I think 40 is bogus so that leaves 22 and 30.
There will be a center level of stones that look great under all 3 but it does move the edges of the system around.
Which is really what this argument is about, the edges.
How do you define those edges?
Even with the IS over the years here defining the edges of the acceptable image has been a huge issue.

Which system in your opinion defines the edge of acceptable better with rounds? and why?
AGS-0
GIA-EX
HCA(properly applied)+IS/ASET
HCA(properly applied)+IS/ASET+hearts
HCA(properly applied)+IS/ASET+hearts+full helium/sarin report.

For me there is a pretty clear answer which of the above is better for rounds.

that should be 22(or 24 maybe I forget exactly which the IS is...) and 30 degrees not % .. thats what I get for posting in the middle of the night.

 

[michaelgem]

Paul and Garry,

I can’t begin to adequately attempt to address all the many points and good questions you have raised. The length of your posts and the clear effort you have made is evidence of the passion you both have for this field of diamond fashioning and grading. That Paul makes such an earnest effort to get me to “open up to other possibilities” is evidence of that passion. And he has succeeded. It just doesn''t look that way, as my efforts, like his, are directed at getting our own thinking accross.

The points and questions posed by Garry are subjects for many discussion threads. I have my answers to them as does Garry, but it is a daunting task to begin addressing them here.

I don’t have the time or energy to do more than address the concept of Ideal in the standard round brilliant.

I feel that my research and investigations have reached the goal of understanding and defining the concept of Ideal in this single but historically most important way to fashion a diamond.

Reporting and demonstrating that research and understanding is the goal of my writing. This may frustrate those who see my research and investigations not as that but rather as my close-minded opinions. I can’t help that.

When I was a consultant for Peter and AGSL through EightStar on fancy cut grading research, I said many times that we first need to refine the understanding of Ideal in the standard round brilliant, before we can adequately begin defining ideal in fancy shapes and other rounds.

This is what I believe that I have accomplished. It is a lot more than just my opinion. It is my science and understanding from my diamond cut studies and investigations. This understanding of Ideal in the standard round brilliant has application to fancies, but is only the beginning of the understanding of what is the best way to fashion most fancy cuts.

My writing is attempting to communicate a higher level of understanding as to why the Ideal cut has emerged as the dominant way to fashion the diamond crystal. I demonstrate to my clients on a daily basis why some modern variations of the Ideal will not take hold, because of clear demonstrable deficiencies in optical performance. One example is round brilliants with pavilion halves noticeably greater than 80%.

I have made my best attempt in the "cut accordance" article soon to appear in the British "Journal of Gemmology", to demonstrate and communicate this understanding, and to show that it agrees not only with the research of AGS and GIA but also the efforts and progress of the best cutters of the Ideal round brilliant.

These are not just my opinions when I can back them by demonstrating deficiencies in optical performance. I believe I have accomplished that in these posts with respect to standard round brilliants with thin crowns, large tables and the resulting large table reflections.

Michael Cowing

PS Paul misread me, as he says “you put all ''larger-tables'' in the same basket, by describing them as low crown-height, shallow crown-angle.” That, I did not say or imply.

I am discussing large table thin crown diamonds, not shallow crown-angle diamonds, which are another topic altogether. My example demonstration, comparing the 56 table to the 65 table, varies the table size keeping the same crown angle of 33 degrees. That results in a thinner, flatter looking crown as the table size increases. The crown angle stays the same.

Sergey’s cut group charts show the pgs software gives an Ideal 0 to this 41.4P33C65T cut. The same Ideal 0 is given to the 64T and even the 66T. There are four combinations of pavilion and crown angles that get Ideal 0 for 64T.

None of these are Ideal from my perspective, GIA''s, diamond cutters, jewelers, appraisers, or consumers.

 

[Garry H (Cut Nut)]

Dera Michael,

You just added 15 I''s and 17 my and me''s to the conversation, but your post again adds no real value.

Add some value please. It might be hard work - but what you wrote was hard work with no out come.

One of your comments

Date: 4/15/2007 1:23:00 PM
Author: michaelgem

My writing is attempting to communicate a higher level of understanding as to why the Ideal cut has emerged as the dominant way to fashion the diamond crystal. I demonstrate to my clients on a daily basis why some modern variations of the Ideal will not take hold, because of clear demonstrable deficiencies in optical performance. One example is round brilliants with pavilion halves noticeably greater than 80%.

Take a diamond that gets AGS 0 using the current PGS, with a table 53%, crown 37 and pavilion 40 - do you really believe such a diamond should have lower pavilion halves below 80% (and please always mention length(GIA) or depth (MSU).

Are you happy with the idea that diamond cut should cease to be a developing field? Or should we just make every piece of rough in to a perfect (''ideal'' or what ever that means) round?

 

[Yuri]

Date: 4/13/2007 12:01:59 AM
Author: Serg

You can find here PGS AGS result for 3% girdle bezel, 55% Star (45% upper girdle) and 80% lower girdle facet length (81.5% lower facet depth). Table 51%-66%

http://octonus.com/oct/mss/gia-agspgs.phtml

Here you can compare PGS result and AGS guideline for AGS0

http://octonus.com/oct/mss/gia-ags.phtml





Here you can compare FacetWare GIA and PGS AGS results

http://octonus.com/oct/mss/gia&agspgs.phtml

All these table comparisons clearly indicate that GIA and AGS still do not agree what is a good cut and what is a bad cut.

If cut of a particular stone is graded by highest score by one lab and downgraded by the other lab then the industry and buying public should recieve an explanation from both. It is common sence. How can we relate on these scores if we do not know the reasons? If a diamond cutter looks on these tables and find cut proportions with different grades what should he do?

 

[adamasgem]

Date: 4/16/2007 7:47:30 AM
Author: Yuri

All these table comparisons clearly indicate that GIA and AGS still do not agree what is a good cut and what is a bad cut.

If cut of a particular stone is graded by highest score by one lab and downgraded by the other lab then the industry and buying public should recieve an explanation from both. It is common sence. How can we relate on these scores if we do not know the reasons? If a diamond cutter looks on these tables and find cut proportions with different grades what should he do?

Yuri.. A simple answer would be to try for a region included in both sets, even though the attempted parameterization of the AGS system may get you into trouble unless one takes a real hard look at the cutting statistics (lack of assymetries) necessary to keep yourself within the "indicated" AGS performance goal(s), and your ability to maintain the symmetry consistency to do so.

 

[Yuri]

Date: 4/16/2007 1:09:15 PM
Author: adamasgem

Date: 4/16/2007 7:47:30 AM
Author: Yuri

All these table comparisons clearly indicate that GIA and AGS still do not agree what is a good cut and what is a bad cut.

If cut of a particular stone is graded by highest score by one lab and downgraded by the other lab then the industry and buying public should recieve an explanation from both. It is common sence. How can we relate on these scores if we do not know the reasons? If a diamond cutter looks on these tables and find cut proportions with different grades what should he do?

Yuri.. A simple answer would be to try for a region included in both sets, even though the attempted parameterization of the AGS system may get you into trouble unless one takes a real hard look at the cutting statistics (lack of assymetries) necessary to keep yourself within the ''indicated'' AGS performance goal(s), and your ability to maintain the symmetry consistency to do so.

But a region of intersection of GIA EX and AGS 0 is too small. Cutters also have their own considerations. A few of them choose 52 of 53 table size. Most combinations can be obtained by reducing the pavilion angle together with the bigger crown angle. This is not a big degree of freedom because most diamonds are fixed in the rough shape by diameter. Initially GIA and AGS cut grading systems were intended to give cutters more freedom, and I remember as GIA representatives explained this goal. But do you see how this freedom can be achieved from these charts? http://octonus.com/oct/mss/gia-agspgs.phtml

 

[adamasgem]

Date: 4/17/2007 7:44:27 AM
Author: Yuri

Date: 4/16/2007 1:09:15 PM
Author: adamasgem

Date: 4/16/2007 7:47:30 AM
Author: Yuri

All these table comparisons clearly indicate that GIA and AGS still do not agree what is a good cut and what is a bad cut.

If cut of a particular stone is graded by highest score by one lab and downgraded by the other lab then the industry and buying public should recieve an explanation from both. It is common sence. How can we relate on these scores if we do not know the reasons? If a diamond cutter looks on these tables and find cut proportions with different grades what should he do?

Yuri.. A simple answer would be to try for a region included in both sets, even though the attempted parameterization of the AGS system may get you into trouble unless one takes a real hard look at the cutting statistics (lack of assymetries) necessary to keep yourself within the ''indicated'' AGS performance goal(s), and your ability to maintain the symmetry consistency to do so.

But a region of intersection of GIA EX and AGS 0 is too small. Cutters also have their own considerations. A few of them choose 52 of 53 table size. Most combinations can be obtained by reducing the pavilion angle together with the bigger crown angle. This is not a big degree of freedom because most diamonds are fixed in the rough shape by diameter. Initially GIA and AGS cut grading systems were intended to give cutters more freedom, and I remember as GIA representatives explained this goal. But do you see how this freedom can be achieved from these charts? http://octonus.com/oct/mss/gia-agspgs.phtml

Yuri.. I am unaware of any such stated goal by AGS as highlioghted above.

My impression has always been that the AGS system was/ and is/ a performance based system with scientifically based criteria based on the optics, as both GIA and AGS realized that the old approach of not considering the stone as a whole was probably wrong.

Whether one agrees or disagrees with their (AGS) methodology and cutoffs is another question.

As to the GIA, I strongly believe that there were some scientifically flawed assumptions from the gitgo, and I have been very public about it.

The two systems are quite different, with GIA''s criteria apparently much looser than AGS''s, irrespective of their (GIA''s) doubletalk regarding the similarities and overlaps.

AGS does a fundamental raytrace on each stone, trying t o look at the primary facet interactions, where GIA apparently decided that that doesn''t matter, as they apparently wanted a parametrically based loosey-goosey system from the start, and there was an obvious disagreement between what their optics study and their "taste test" results so they melded the two to keep the trade very very happy, and eliminated the necessity for the GIA to do fundamentaly ray tracing on each stone.

The approaches were different and the results are going to be different, so it is difficult, to use an analogy, to fource a square peg in a round hole unless that square peg is smaller than the round hole...

The MSU results differ from both GIA and AGS, in the same manner, because of different starting assumptions and methodology, yet there is some overlap, which I would expect.

GIA''s system appears wholey parametric, and we know there are regions of steep performance related gradients, and AGS''s cutting guidlines were intended to warn cutters away from those areas of potential low performance. The AGS charts ARE NOT GRADING CHARTS, whereas, in general, GIA''s ARE.

Simple assumptions, like the differences in polish/symmetry requirements for the top grade, give you some indication of the looseness in the GIA system, relative to the AGS, one of which (polish), from an optical performance standpoint, I have always agreed with, and think that AGS may be a little too harsh there, yet it ultimately AGS''s decision to maintain their more stringent requirements.

As to specific regions of disagreement, their may be discussions, but as I said before, one has to do their homework first.

I surely don''t know how to factor in the GIA''s taste test data, because no one has it, so it ultimately it is up to the consumer to decide whetehr htey like the stone or not, irrespective of seller''s hype or representations that if the "world''s greatest authority" says it is so, it must be right.

Well, it is obvious that I and others disagree in part with the trade so called "G-D". I could give a damn about the marketing driven aspects of the GIA system, and up to this point, and probably will in the future, have chosen to totally ignore FARCEWARE(TM) results.

GIA needs to clean house and straighten out their own technical mess, and hopefully they will, with thier new leadership

 

[adamasgem]

Date: 4/17/2007 7:44:27 AM
Author: Yuri

But a region of intersection of GIA EX and AGS 0 is too small. Cutters also have their own considerations. A few of them choose 52 of 53 table size. Most combinations can be obtained by reducing the pavilion angle together with the bigger crown angle. This is not a big degree of freedom because most diamonds are fixed in the rough shape by diameter. Initially GIA and AGS cut grading systems were intended to give cutters more freedom, and I remember as GIA representatives explained this goal. But do you see how this freedom can be achieved from these charts? http://octonus.com/oct/mss/gia-agspgs.phtml

Yuri.. The charts that were generated by OCTONUS are like trying to compare apples and oranges.

The only thing you can say from the charts is, that in the GIA charts you DON''T NEED perfect symmetry for the grade and there is a range of conditions where they are apparently valid (i.e. Star +/- 2.5 %, LGF +/- 2.5%, Table +/- 0.5%, etc) while the AGS PGS charts are ONLY VALID for perfect symmetry and at the specific conditions indicated.

In the AGS case you have to look at the performance gradients and the probability of attaining the charts'' indicated possible grade, given asymmetries AND all possible combinations within the allowable half assed rounded GIA range.

 

[Yuri]

Date: 4/17/2007 12:46:44 PM
Author: adamasgem

Yuri.. The charts that were generated by OCTONUS are like trying to compare apples and oranges.

The only thing you can say from the charts is, that in the GIA charts you DON''T NEED perfect symmetry for the grade and there is a range of conditions where they are apparently valid (i.e. Star +/- 2.5 %, LGF +/- 2.5%, Table +/- 0.5%, etc) while the AGS PGS charts are ONLY VALID for perfect symmetry and at the specific conditions indicated.

In the AGS case you have to look at the performance gradients and the probability of attaining the charts'' indicated possible grade, given asymmetries AND all possible combinations within the allowable half assed rounded GIA range.

Marty, I am aware about your critique of GIA cut project which was starting maybe from Igloo lighting hemisphere remarks in 1999. In our Australian Gemmologist letter we made some statements that are very similar with your points. But here we are talking about other things.

Now you try to distinguish between GIA and AGS cut grading systems, and at the same time Sergey is making his point that they are clearly became more similar. The reason of this similarity is also evident because both labs loose consumer confidence if they constantly give different opinions for same stones. I do not agree with you what you write here about difference in these labs symmetry approach. Maybe you are right here or maybe wrong. WE CAN NOT CHECK IT because nither GIA not AGS made public their approach to symmetry grading.
Do you agree with me? Tell me for example what is the limit of the pavilion main facet angles deviation between Excellent and Very good GIA symmetry? Or maybe you know this limit between AGS 0 and AGS 1?

 

[adamasgem]

Date: 4/18/2007 1:28:38 PM
Author: Yuri

Marty, I am aware about your critique of GIA cut project which was starting maybe from Igloo lighting hemisphere remarks in 1999. In our Australian Gemmologist letter we made some statements that are very similar with your points. But here we are talking about other things.

Now you try to distinguish between GIA and AGS cut grading systems, and at the same time Sergey is making his point that they are clearly became more similar. I respectfully diagree there.

The reason of this similarity is also evident because both labs loose consumer confidence if they constantly give different opinions for same stones. I do not agree with you what you write here about difference in these labs symmetry approach. Maybe you are right here or maybe wrong. WE CAN NOT CHECK IT because nither GIA not AGS made public their approach to symmetry grading. Well, I believe they BOTH offer a qualitative symmetry assesment, but ONLY the AGS approach factors the light return aspects of assymetry as to the effects on light performance. A stone can be asymmetric and have great light performance because of the proper balance to generate enough optical symmetry, which is DIFFERENT than the physical symmetry, a judgement call, to this time.

Do you agree with me? Not exactly. Tell me for example what is the limit of the pavilion main facet angles deviation between Excellent and Very good GIA symmetry? I wish I and others knew!!! Or maybe you know this limit between AGS 0 and AGS 1? Same thing as my comments regarding GIA.

Right now, as far as I know it has been a subjective judgement call in both cases.

Bill Bray is the only one who has tried to quantify what is good and bad numerically for physical symmetry (workmanship) in BrayScore, parts of which I think all should look at.
I''ve discussed this with Bill long ago, that he should break up his total score into two parts, separating out his workmanship aspects from a cutters perspective (min max deviations or standard deviations, whichever he uses) and what we would call mean value judgements for what is, lets say, a proper mean crown angle because it is a multidemensional balancing act to fashion a stone and one set of angles depend on another.

Earlier in this thread I showed some statistics regarding GIA polish and symmetry assesments (judgements) based on databases, and implied that if they (GIA and AGS graders) were both taught the same such that graders were similar in their calls of EX or VG, that immediately that would narrow the range of possible AGS 0''s, because AGS REQUIRES what MAY BE only EX symmetry to get a 0 grade. I wish it were more quantifiable and less judgemental. I think AGS may be looking at something like this, as it does effect the ultimate grade.

 

[adamasgem]

Yuri.. One of the things I''d like to see is a full error covariance matrix from the scanner measurments showing the mathematical error correlations between the defined planes... Would be fun to play with to get a better understanding of measuremnet erros but so far, I have not seen one example from any manufacaturer, Helium, Sarin, OGI, etc.. Not even an estimate in the measurement error correlations... I think it is doable, but a bear,to solve the facet problem using applied optimal estimation techniques using Kalman Filtering (Actually Dick Battin came up with the idea at the same time but Kalman got the credit) and probably require close to 300 states to properly model a round brilliant.