AGS introduces cut grading on DQR reports

[strmrdr]

Date: 10/20/2008 10:05:34 PM
Author: Garry H (Cut Nut)

Date: 10/20/2008 9:34:39 PM

Author: Allison D.


D-Ideal and R-Ideal,

Hey man, that's cool, this one - she's De' Ideal


No Sir, you are wrong, this one is the Real ideal - you can tell because it is writen in plain English and not some code.

I'm starting to like the terms:
Real AGS Ideal
Fake AGS Ideal or kiss of death avoid at all cost or kodaaac

AGS koi or kind of ideal
AGS MBI or might be ideal.

 

[jasontb]

If you can''t beat ''em, join ''em. And use ridiculously confusing terminology while you''re at it.

 

[adamasgem]

Date: 10/21/2008 12:52:01 AM
Author: Serg




Allison, Marty

Firstly I need add what
If same real polished diamond will receive AGS GOLD Ideal grade and just AGS Platinum 2 grade is not Misleading for market. It is not problem( not perfect symmetry is simple explanation. All Marty post explains it OK, we are in sync there


Misleading if diamond receives TOP grade in TOP AGS grading system and just third grade in NEW ASG grading system.
Some Consumer could start think what NEW AGS grading system is more strict than AGS Platinum( 3d light performance ASG Brand) . But new ASG GOLD grading system is more softer. Now I see what you are saying as a possible mis- conception, I don''t consider it misleading, but ust requiring education.

 

[adamasgem]

Date: 10/21/2008 12:52:01 AM
Author: Serg





Allison, Marty

My opinion what ASG GOLD Ideal should include ASG Platinum 0.
but some points Top Platinum grade is outside Top Gold grade range

That is logical, as without all information being used, AGS seemingly feels that it couldn't give it a top grade.


AND some points Top Gold grade is outside Top Platinum grade range.As a result of a looser or different criteria being applied....



"It is creating chaos in understanding, it is reduce consumer confidence to ASG cut grade ."

As Storm pointed out, and should be understood and taught, that the consumer can have more confidence in the results of the Platinum report than the Gold, such that:

it is a AGS Platinum 0 with a 95% certainty
or a AGS Gold Ideal with a 85% certainty AND
it might be a GIA EX with 80% certainty

The concept that needs to be understood is that you have different levels of trust and higher error rates in the verbally stated results with the differing methodologies, and is no different or quite similar to the AGS/GIA issue, but difficult to explain as well as assigning a specific level of trust or certainty.

You have different levels of trust (or implied error rate) in different report methodologies, just as you have different levels of trust (or implied error rate) in what different labs say for the same supposed parameter, such as color or clarity grade.

It is similar in concept that the same person might do a better job on a specific project if he were allowed more time or given better tools to do the job.

 

[Garry H (Cut Nut)]

Date: 10/21/2008 5:32:08 AM
Author: adamasgem

It is similar in concept that the same person might do a better job on a specific project if he were allowed more time or given better tools to do the job.

e.g. GIA could have done better if they had an additional 10 years and more money for tools?

What would Homer Simpson say?

 

[adamasgem]

Date: 10/21/2008 5:44:25 AM
Author: Garry H (Cut Nut)

Date: 10/21/2008 5:32:08 AM
Author: adamasgem

It is similar in concept that the same person might do a better job on a specific project if he were allowed more time or given better tools to do the job.

e.g. GIA could have done better if they had an additional 10 years and more money for tools?

What would Homer Simpson say?

Sorry Garry, I wasn''t considering the outside influences $ that might effect the impetus to do a better job regardless at how much was thrown at it

 

[strmrdr]

Date: 10/21/2008 5:55:37 AM
Author: adamasgem

Date: 10/21/2008 5:44:25 AM

Author: Garry H (Cut Nut)

Date: 10/21/2008 5:32:08 AM

Author: adamasgem

It is similar in concept that the same person might do a better job on a specific project if he were allowed more time or given better tools to do the job.

e.g. GIA could have done better if they had an additional 10 years and more money for tools?


What would Homer Simpson say?

Sorry Garry, I wasn''t considering the outside influences $ that might effect the impetus to do a better job regardless at how much was thrown at it

And now AGS is doing the same thing GIA did. Outside influence $$$$$$$$$, trumps science.

 

[strmrdr]

Date: 10/21/2008 3:31:01 AM
Author: jasontb
If you can''t beat ''em, join ''em. And use ridiculously confusing terminology while you''re at it.

That pretty much sums it up.

 

[adamasgem]

Date: 10/21/2008 6:15:46 AM
Author: strmrdr

And now AGS is doing the same thing GIA did. Outside influence $$$$$$$$$, trumps science.

Storm, I didn''t mean to imply that in the case of AGS; their new system is looser, with better science behind it, and Yes, it appears to be influenced by the ability to grow, without the tax advantages that the 800# Gorilla has, and has usurped and corrupted.
This is a replacement for the older AGS parameter based grading (now obsolete and not used by the lab), different, with more science behind it, and something that now can directly comparable to the GIA grading for the most part, but still results in much tighter grading. Maybe less confidence than the PGS system, but I think driven by the ability of the manufacturers to grasp and have a higher confidence in obtaining the results they desire.

There may be some inconsistencies there between the two systems, but there different criteria applied, and there is the critical assumption in the charts that the stone is perfectly symmetric,just as GIA has done.

Both systems will evolve over time, and this appears to be an OPEN (non proprietary), simpler, essentially no cost methodology, over the cost of the measurement technology. It will be somewhat marginally less work for the lab, and probably allows for the use of faster (more errored measurements) without effecting the end results as much as the more sensitive PGS methodology.

More later, off to jury duty.... Last time I did it 25 years ago it was for a month, this time it is one day or one trial, hopefully simple and short.

 

[Regular Guy]

Date: 10/20/2008 9:24:34 PM
Author: Allison D.

Date: 10/18/2008 7:01:27 PM
Author: Regular Guy

One way or the other, regardless of exactly why AGS is bringing the new Ideal & AGS system forward, it remains to be seen why a shopper will end up being motivated to select this new AGS Ideal, vs an available AGS0.

I don't see this step as being driven by a desire to motivate customers. Rather, I see it as a desire to motivate the cutters who are submitting the stones.

The things that represent the most value to cutters submitting their stones is 'AGS Ideal' or 'GIA EX'.....it's the top label that matters. Because GIA's 'top label' includes approximately 5x the number of stones that AGS' top label does, there's more incentive to go to GIA.

If now all 5 stones earn an AGS Ideal or a GIA Ex, that levels the playing field. AGS can still preserve their 'ultimate ideal' by still carving out a class of ideals that are supreme ideals.

Allison...never mind that your point was largely mine, third from the top, page 1 (well...part of it anyway...personally...I'd like to see Garry's chart blown up a bit to document this better)...regardless of what is driving the new system...THIS shopper could be interested in what circumstance might make me consider gold instead of platinum, and when.

And...Ms. WF, do your employers have any insight as to how this evolution of systems will reflect on future practices for identifying ACAs vs ES diamonds?

BTW...kudos to Marty for playing Obama in these discussions (not his usual temperament)...and helping to add light where there was more heat.

 

[Allison D.]

Date: 10/21/2008 9:54:29 AM
Author: Regular Guy

Date: 10/20/2008 9:24:34 PM
Author: Allison D.

Date: 10/18/2008 7:01:27 PM
Author: Regular Guy

One way or the other, regardless of exactly why AGS is bringing the new Ideal & AGS system forward, it remains to be seen why a shopper will end up being motivated to select this new AGS Ideal, vs an available AGS0.

I don't see this step as being driven by a desire to motivate customers. Rather, I see it as a desire to motivate the cutters who are submitting the stones.

The things that represent the most value to cutters submitting their stones is 'AGS Ideal' or 'GIA EX'.....it's the top label that matters. Because GIA's 'top label' includes approximately 5x the number of stones that AGS' top label does, there's more incentive to go to GIA.

If now all 5 stones earn an AGS Ideal or a GIA Ex, that levels the playing field. AGS can still preserve their 'ultimate ideal' by still carving out a class of ideals that are supreme ideals.

Allison...never mind that your point was largely mine, third from the top, page 1 (well...part of it anyway...personally...I'd like to see Garry's chart blown up a bit to document this better)...regardless of what is driving the new system...THIS shopper could be interested in what circumstance might make me consider gold instead of platinum, and when.

And...Ms. WF, do your employers have any insight as to how this evolution of systems will reflect on future practices for identifying ACAs vs ES diamonds?

BTW...kudos to Marty for playing Obama in these discussions (not his usual temperament)...and helping to add light where there was more heat.

Ira, I didn't see it as being precisely the same point, but glad we are in agreement.

To your question, I don't expect this evolution of systems to alter our practices at all. One of Brian's requirements for inclusion in the brand is earning the most stringent AGS cut grade, and that won't change. Our A Cut Above stones easily meet the AGS0 cut grade requirements under the 3D light performance methodology.

As many here know, ES stones typically come with either the DQR report or a GIA report, and that will continue to be so moving forward.

 

[michaelgem]

Most of the discussions here have the underlying assumption that the PGS software is the ground truth of diamond light performance. But AGS has said that their grading system would continue to evolve as new insight or new metrics, such as scintillation measures, are added.

Those believing that the “hearts and arrows”, optically symmetric, Tolkowsky Ideal (40.75, 34.5) is the “pinnacle of light performance” may note that the PGS software will give the same Ideal 0 grade to a diamond without H&A symmetry, cut to (41.5, 32). Because of continuing research, Ideal, as defined by the output of the current PGS software, is not quite the end of the story.

AGS grades as Ideal 0, those proportion sets that give similar enough light performance and spread, under the metrics in the PGS software, to the Tolkowsky Ideal (40.75, 34.5). The underlying premise is rounds cut to these theoretical angles are the standard of Ideal optical performance and beauty. Proportion sets with optical performance close enough to the Tolkowsky Ideal (as cut with “minor facet” proportions by such manufacturers as Lazar Kaplan, Paul, Brian and Hearts on Fire) are given the Ideal 0 cut grade. This grade includes craftsmanship constraints which go beyond and do not effect optical performance, such as the requirement for Ideal polish.

Also, there are points in both GIA''s and AGS''s grading systems where observable differences in light performance are not given a different grade, but are said to be matters of personal taste. Deciding where differences of light performance raise or lower the grade and cease being “matters of personal taste” leads to many of the arguments between laboratories as to where the boundaries lie. Especially the most important boundary between the best (AGS Ideal 0 or GIA Ex) and everything else.

All this is mainly to say that although there is a desire for a cast in stone standard that tells you with certainty that the Ideal 0 diamond you bought is better than any other, such a standard is, like diamond beauty, in the eye of the beholder. That “eye” is too often influenced by marketing and the dictates of those we decide to look to for answers.

By introducing a gold standard, that differs from AGS Ideal 0 PGS grading, with less restrictive requirements on aspects that do not materially effect light performance (such as Ideal polish), AGS is retaining in this new standard the important aspects of Ideal from the perspective of light performance and beauty.

For those like HOF, Paul and Brian, AGS still retains the toughest platinum standard with the highest requirements for craftsmanship. If seen and explained in this light, it could be a win-win for AGS and everyone. (Or not, depending on your point of view)

Michael D Cowing

 

[adamasgem]

Date: 10/18/2008 12:55:07 AM
Author: Garry H (Cut Nut)

Date: 10/18/2008 12:24:12 AM
Author: adamasgem

But I do not understand when you say that: neither you nor I (nor Sergey) can do the arithmetic to compute the resultant PBS (parameter based system) metric. why not? if you have the software then you can run the proportions and make a table - or you can take the tables that AGS do and send out to manufacturers. Did I miss something? In the past AGS said the manufacturers charts were conservative and had a safe edge to allow for some sym deviations etc - so why is the parametric system different - did Sergey make a mistake - if so can someone please give him the info - I am sure he will accept it like a gentleman.

What I meant was that besides taking many runs to create the charts, we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario.

 

[Serg]

Date: 10/21/2008 1:57:19 PM
Author: adamasgem

Date: 10/18/2008 12:55:07 AM
Author: Garry H (Cut Nut)

Date: 10/18/2008 12:24:12 AM
Author: adamasgem

But I do not understand when you say that: neither you nor I (nor Sergey) can do the arithmetic to compute the resultant PBS (parameter based system) metric. why not? if you have the software then you can run the proportions and make a table - or you can take the tables that AGS do and send out to manufacturers. Did I miss something? In the past AGS said the manufacturers charts were conservative and had a safe edge to allow for some sym deviations etc - so why is the parametric system different - did Sergey make a mistake - if so can someone please give him the info - I am sure he will accept it like a gentleman.

What I meant was that besides taking many runs to create the charts, we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario.

Marty,

re:we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario


Do you mean what Light metric for AGS Gold is better(at least different) than for ASG Platinum?
Sorry, Your last post is not clear for me. Could you please rewrite it on more simple English?

 

[michaelgem]

AGS may have done itself a disservice by describing their new “Gold Standard” grading system as only “proportion based” while calling their “Platinum Standard” AGS 0-10 system “performance based”. The reality is both of their systems are performance based, and use the same evaluation metrics to arrive at a different 6-level word grade instead of the 11-level numerical 0-10 grade. Calling one system proportion based and the other performance based is semantics originally used to imply the superiority of their “performance based” grading system to that of GIA''s, which, for several reasons, including necessary averaging and rounding, AGS and others referred to as “proportion based”.

The underlying truth is that both GIA''s and AGSLAB''s grading systems are both performance based. That is, they are based upon the three-dimensional (3-D) light performance resulting from the 3-D interaction of light with a diamond''s proportions. This is in contrast to the old 2-D ray trace analysis of proportions done in the past by diamond cut investigators, the most celebrated of whom was Marcel Tolkowsky. His theoretical angles were developed from ray tracing using only a 2-D cross-sectional slice of the 3-D round brilliant cut diamond.

Performance based means that the AGS system evaluates the 3-D light performance of a diamond to arrive at a grade, by raytracing a virtual model of its proportions obtained from a Sarin scan. This is compared to the GIA system, which takes the same proportions from the scan, judiciously averages them, and arrives at an estimated grade with their “performance based” system embodied in the “GIA Facetware”. Facetware contains “Cut Grade Reference Charts” and “Cut Grade Estimation Tables” which are published in GIA''s Diamond Grading Lab Manual for use in cut grade estimation by gemologists, appraisers and cutters. These charts and tables are a simple method of communicating their grading system.

This method of estimating a cut grade using look-up tables, which, through usage, has incorrectly come to be called proportion based, is winning the day for GIA because a diamond cutter can determine from them how to best optimize the cutting of a rough diamond crystal to get the best grade and yield.

Seeing the GIA''s success, and recognizing this need on the part of the cutters and others desiring the simplest way to understand their grading system, AGSLAB has developed and published their “Gold Standard” grading system with their own charts and tables. Lab director, Peter Yantzer feels this new system “is much easier to use, easier to understand, easier to teach and easier to explain". AGSLAB has developed with this Gold Standard, a mainstream way of communicating their grading system using the word rather than numeric terminology: Ideal, Excellent, Very Good, Good, Fair and Poor. Some craftsmanship issues not effecting light performance, such as polish grade have been relaxed. However, underlying this new Gold Standard is the same AGSLAB''s performance based grading system. It is expressed and communicated with this simpler word scale, and with charts and tables similar to those successfully employed by GIA.

It is true that when you quantize the diamond''s parameters to use the charts and tables rounding error is introduced. The effect of rounding error is minimized by finer quantization of the parameters that most impact light performance like the pavilion main angle. This makes for a very usefull and workable grading system for the cutter, the gemologist and appraiser.

Just remember that underlying both GIA''s grading system and AGSLAB''s two grading systems are each laboratories'' grading metrics, which are both 3-D light-performance based. In addition, GIA, because of the urging and efforts of Al Gilbertson, did and continues to do comparison testing to refine and validate their performance metrics. What could be more performance based than human evaluation comparing the light-performance of pairs of diamonds side by side in the same lighting?

Michael D. Cowing

 

[adamasgem]

Date: 10/22/2008 3:41:38 AM
Author: Serg

Date: 10/21/2008 1:57:19 PM
Author: adamasgem

Date: 10/18/2008 12:55:07 AM
Author: Garry H (Cut Nut)

Date: 10/18/2008 12:24:12 AM
Author: adamasgem

But I do not understand when you say that: neither you nor I (nor Sergey) can do the arithmetic to compute the resultant PBS (parameter based system) metric. why not? if you have the software then you can run the proportions and make a table - or you can take the tables that AGS do and send out to manufacturers. Did I miss something? In the past AGS said the manufacturers charts were conservative and had a safe edge to allow for some sym deviations etc - so why is the parametric system different - did Sergey make a mistake - if so can someone please give him the info - I am sure he will accept it like a gentleman.

What I meant was that besides taking many runs to create the charts, we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario.

Marty,

re:we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario


Do you mean what Light metric for AGS Gold is better(at least different) than for ASG Platinum?

Sorry, Your last post is not clear for me. Could you please rewrite it on more simple English?

The light metric is DIFFERENT because it

1) The GOLD report assumes perfect symmetry based on a limited set of Table/Crown/Pavilion combinations, and they ran something like 36 ray traces to establish the Grade value for each box, as opposed to the two or three ray traces done on the ACTUAL stone BUT, CONTRARY to the FRAUD that GIA perpetrates, real averaged numbers from the Sarin Out file are printed on the report, so that the consumer knows the ACTUAL AVERAGES of the stones angles instead of the implied cutsy FRAUD where GIA shows a profile and says the profile shown matches the averaged ROUNDED numbers and DIRECTLY OR INDIRECTLY IMPLIES that these are the ACTUAL PROPORTIONS of the stone, WHICH THEY ARE NOT.

The rounding is used to see which box (category) the stone falls into.

2) The Platinum report uses the ACTUAL Sarin scan, no averaging EXCEPT when they give the average proportions on the profile plot, NOT averaged and ROUNDED as GIA does.



3) Since it would take on average about 17 to 18 times the current processing time to process a PGS report by doing face up and 35 ray trace in the same manner (using the actual proportions), I don''t think many manufacturers would have the processing power to do it, currently.

It certainly would theroetically give a finer perspective if you used the actual proportions, as I presented back in a PriceScope discussion in the 2005 time frame where I showed, using the GIA methodology, the 70% localized dips in weighted light return, that both you and Garry participated in whereI loked at the GIA methodology averaged over all azimuths for each tilt, and vice vera.

As to which methodology is better, I''d have more statistical confidence in the Platinum, as I believe AGS has.

IF, AND ONLY IF the STONE WAS PERFECTLY SYMMETRIC, AND PERFECTLY MATCHED THE NUMBERS IN THE POSITION OF THE AGS BOX, then I might go with the GOLD report, as it characterized the stone over a wider range of observation conditions.


A look up table is way-way simpler and MUCH MUCH faster than a ray trace...


And no, I don''t think that, in effect, using rounded averages is statistically is in any way superior to using actual proportions.

 

[adamasgem]

Date: 10/22/2008 7:13:52 PM
Author: michaelgem

If I responded to Cowings GIA mantra, then Irv might take back the nice things he said about me

 

[Regular Guy]

Date: 10/22/2008 9:11:31 PM
Author: adamasgem

Date: 10/22/2008 7:13:52 PM
Author: michaelgem

If I responded to Cowings GIA mantra, then Irv might take back the nice things he said about me

It suddenly occurred to me...I bet he means ME!

BTW...reading here of late, I feel very much like Tevya on Fiddler on the Roof. In that (movie version or any? version) there was animated discussion, with many people disagreeing with each other. (And I actually also find Michael''s recent discussion clear headed too...also I think entirely missing any discussion of a sweet spot). Tevya said "I agree" with both of the people in disagreement, and when a third said it''s not reasonable to agree with both people in disagreement...he said I agree with that too!

 

[strmrdr]

Date: 10/22/2008 7:13:52 PM
Author: michaelgem

Long quote deleted.. see above...

Michael D. Cowing

The fact that thousands of combos will get gold ideal and not 3d ideal even discounting pol/sym blows that argument out of the water.
Either the 3d ideal is right or the gold ideal is right, they both can not be right because they do not agree.

 

[adamasgem]

Date: 10/22/2008 11:32:26 PM
Author: strmrdr

Date: 10/22/2008 7:13:52 PM
Author: michaelgem

Long quote deleted.. see above...

Michael D. Cowing

The fact that thousands of combos will get gold ideal and not 3d ideal even discounting pol/sym blows that argument out of the water.
Either the 3d ideal is right or the gold ideal is right, they both can not be right because they do not agree.

Storm.. Don''t you UNDERSTAND that they can''t allways agree, GOLD is a looser and a DIFFERENT LOOSER standard, by definition, just by virture of the lower polish standard (which I happen to agree with as too harsh and iin general, unimportant to the public, especially as soon as the stone is worn

) or assumption of symmetry that may not exist, especially on or boundaries between grades (ridges, hills, cliffs)

When PGS adds to the Platinum standard scintilation and fire aspects, it too, will probably change in specific cases.

The earth was flat one time.. or so we we told in history.

 

[michaelgem]

Date: 10/22/2008 11:32:26 PM
Author: strmrdr

Date: 10/22/2008 7:13:52 PM
Author: michaelgem

Long quote deleted.. see above...

Michael D. Cowing

The fact that thousands of combos will get gold ideal and not 3d ideal even discounting pol/sym blows that argument out of the water.
Either the 3d ideal is right or the gold ideal is right, they both can not be right because they do not agree.

Not sure which arguement is blown out of the water. The main point I belabored was that both Gold and Platinum AGS systems and GIA''s are performanced based and calling any of them proportion based is an incorrect allusion to the square boundaries of yesteryear when AGS graded each parameter independently by how much it deviated from the "American or Tolkowsky Ideal". I noted on page 3:

I think the DQR-Ideal is being positioned as tighter than the GIA Excellent but more forgiving than the AGS Ideal 0. And the DQR-Ex seems to have a wider, more forgiving range than the GIA EX. Can anyone who has checked this confirm or deny?
By using the GIA word grades with the topper of Ideal, AGS is equating the Gold Standard to the GIA grades, not necessarily to their numeric 0-10 Platinum System. So a Gold Ideal may not be intended to have the same boundaries as an Ideal 0. It just has to be tighter than the GIA EX to be a higher grade. And where do you put the boundary for Gold EX? Somewhere around the AGS 2-3 boundary? I suspect the boundaries were chosen, not to necessarily correspond to the Platinum numeric boundaries, but instead to have meaning compared to the GIA''s boundaries for the same word designations of Ex, VG, G, F, P with the additional tighter than Ex boundary of Ideal.

No?

Michael

 

[Garry H (Cut Nut)]

Date: 10/23/2008 12:39:15 AM
Author: michaelgem


and GIA''s are performanced based
Michael

 

[Garry H (Cut Nut)]

Date: 10/22/2008 7:13:52 PM
Author: michaelgem

The underlying truth is that both GIA''s and AGSLAB''s grading systems are both performance based. That is, they are based upon the three-dimensional (3-D) light performance resulting from the 3-D interaction of light with a diamond''s proportions. This is in contrast to the old 2-D ray trace analysis of proportions done in the past by diamond cut investigators, the most celebrated of whom was Marcel Tolkowsky. His theoretical angles were developed from ray tracing using only a 2-D cross-sectional slice of the 3-D round brilliant cut diamond.

Michael D. Cowing

Michael you need to get me some of that stuff you are using please.

In which way is GIA''s system any way related to their ray tracing results for performance?

For example - here is the GIA''s light return ray trace study results from 1998 (where they were completely oblivious to the inverse relationship between crown and pavilion angle - I added the lines and they argued I was nuts when I pointed it out) which clearly indicate their steep deep proportions did not make the grade.

 

[Serg]

Date: 10/22/2008 9:08:09 PM
Author: adamasgem

Date: 10/22/2008 3:41:38 AM
Author: Serg

Date: 10/21/2008 1:57:19 PM
Author: adamasgem

Date: 10/18/2008 12:55:07 AM
Author: Garry H (Cut Nut)

Date: 10/18/2008 12:24:12 AM
Author: adamasgem

But I do not understand when you say that: neither you nor I (nor Sergey) can do the arithmetic to compute the resultant PBS (parameter based system) metric. why not? if you have the software then you can run the proportions and make a table - or you can take the tables that AGS do and send out to manufacturers. Did I miss something? In the past AGS said the manufacturers charts were conservative and had a safe edge to allow for some sym deviations etc - so why is the parametric system different - did Sergey make a mistake - if so can someone please give him the info - I am sure he will accept it like a gentleman.

What I meant was that besides taking many runs to create the charts, we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario.

Marty,

re:we don''t necessarily have the breakpoints used to determine the grade, as to the weighted average of the front view and multiple tilts, based on an averaged parameter set, as this was, and I don''t think is, part of the PGS grading software, which is based on the actual stone, face up view and I believe a limited tilt scenario


Do you mean what Light metric for AGS Gold is better(at least different) than for ASG Platinum?


Sorry, Your last post is not clear for me. Could you please rewrite it on more simple English?

The light metric is DIFFERENT because it

1) The GOLD report assumes perfect symmetry based on a limited set of Table/Crown/Pavilion combinations, and they ran something like 36 ray traces to establish the Grade value for each box, as opposed to the two or three ray traces done on the ACTUAL stone BUT, CONTRARY to the FRAUD that GIA perpetrates, real averaged numbers from the Sarin Out file are printed on the report, so that the consumer knows the ACTUAL AVERAGES of the stones angles instead of the implied cutsy FRAUD where GIA shows a profile and says the profile shown matches the averaged ROUNDED numbers and DIRECTLY OR INDIRECTLY IMPLIES that these are the ACTUAL PROPORTIONS of the stone, WHICH THEY ARE NOT.

The rounding is used to see which box (category) the stone falls into.

2) The Platinum report uses the ACTUAL Sarin scan, no averaging EXCEPT when they give the average proportions on the profile plot, NOT averaged and ROUNDED as GIA does.



3) Since it would take on average about 17 to 18 times the current processing time to process a PGS report by doing face up and 35 ray trace in the same manner (using the actual proportions), I don''t think many manufacturers would have the processing power to do it, currently.

It certainly would theroetically give a finer perspective if you used the actual proportions, as I presented back in a PriceScope discussion in the 2005 time frame where I showed, using the GIA methodology, the 70% localized dips in weighted light return, that both you and Garry participated in whereI loked at the GIA methodology averaged over all azimuths for each tilt, and vice vera.

As to which methodology is better, I''d have more statistical confidence in the Platinum, as I believe AGS has.

IF, AND ONLY IF the STONE WAS PERFECTLY SYMMETRIC, AND PERFECTLY MATCHED THE NUMBERS IN THE POSITION OF THE AGS BOX, then I might go with the GOLD report, as it characterized the stone over a wider range of observation conditions.


A look up table is way-way simpler and MUCH MUCH faster than a ray trace...


And no, I don''t think that, in effect, using rounded averages is statistically is in any way superior to using actual proportions.

Marty,
You mixed Light metric and models. Models could be different , For PGS you can use scanned models or ideal models.
Type of model is not issue . Symmetry is not issue too.
Issue is: Did AGS used different( from PGS light metrics) light metrics TO CREAT AGS Gold cut grading?
Why was it necessary? ( if AGS used new metrics to Create AGS Gold cutters guidelines )
Could we discuss only light metrics ? ( what is reason to show different between ideal model and scanned model again and again? Do you really think what I do not know it well?)

 

[adamasgem]

Date: 10/23/2008 7:38:29 AM
Author: Serg




Marty,
You mixed Light metric and models. Models could be different , For PGS you can use scanned models or ideal models.
Type of model is not issue . Symmetry is not issue too.
Sergey, you have a conceptual or a political problem. PGS is entirely dependent on the actual scanned stone.
An Ideal model is irrelevent other than illustrative and is not applicable practically as NO stone I have ever seen has had perfect symmetry. And I would predict that there is NO stone in AGS's or GIA's data base that has PERFECT symmetry.




Issue is: Did AGS used different( from PGS light metrics) light metrics TO CREAT AGS Gold cut grading?
Why was it necessary? ( if AGS used new metrics to Create AGS Gold cutters guidelines )

As John previously said anyone can view the Gold system here
http://www.agslab.com/trade_research_and_development_proportion_based_cut_grade.php

I stated before that the AGS used a weighted average of face up and a tilt range (1-35 degrees?) of PGS results to set the Parameter based grade, all based on a PERFECTLY symmetric stone.

I believe i also stated before that I showed years ago, based on GIA metrics, that there seems to be areas of tilts with sharp dropoffs in light performance (over a small range of tilt angles), with performance about 30% of peak.

I have always advocated some weighted average of tilts to be used in a performance metric and AGS used face up and 15 degree tilt (based on actual symmetry).

It is currently time wise, and too computer intensive to look at the stone over all tilts and all azimuth perspectives, which should happen.

Think of running raytraces at 1 degree increments of tilt, 90 tilts, rotating the stone about the table/culet axis 1 degree and doing the same thing 360 times and then taking a weighted average based on some probability of looking at the stone from that perspective. If you have a perfectly symmetric RBC, you can cut that down by a factor of 8.

Why, because, how does the stone look to the wearer in the normal viewing position, and HOW is it mounted in azimuth, BECAUSE when the stone is on your hand, the FACE UP position is NOT THE ERGONOMICALLY MOST CONVENIENT and COMFORTABLE viewing position; the wearer has to twist their wrist or arm and strain to see their stone in the face up position, so they are USUALLY viewing the stone in a position OTHER THAN FACE UP. We have gone over this before.

I'll find the link to the previous Pricescope thread from 3 years ago where I posted the proof of my thesis why these things should be considered, and the very clear possible change in metrics. You seeming understood then, and Garry even posted a DiamondCalc example that was visual proof of the thesis, so you must have forgotten, or....



Could we discuss only light metrics ? ( what is reason to show different between ideal model and scanned model again and again? Do you really think what I do not know it well?)

See above

 

[Serg]

Date: 10/23/2008 10:47:02 AM
Author: adamasgem

Date: 10/23/2008 7:38:29 AM
Author: Serg






Marty,
You mixed Light metric and models. Models could be different , For PGS you can use scanned models or ideal models.
Type of model is not issue . Symmetry is not issue too.
Sergey, you have a conceptual or a political problem. PGS is entirely dependent on the actual scanned stone.
An Ideal model is irrelevent other than illustrative and is not applicable practically as NO stone I have ever seen has had perfect symmetry. And I would predict that there is NO stone in AGS''s or GIA''s data base that has PERFECT symmetry.





Issue is: Did AGS used different( from PGS light metrics) light metrics TO CREAT AGS Gold cut grading?
Why was it necessary? ( if AGS used new metrics to Create AGS Gold cutters guidelines )

As John previously said anyone can view the Gold system here
http://www.agslab.com/trade_research_and_development_proportion_based_cut_grade.php

I stated before that the AGS used a weighted average of face up and a tilt range (1-35 degrees?) of PGS results to set the Parameter based grade, all based on a PERFECTLY symmetric stone.

I believe i also stated before that I showed years ago, based on GIA metrics, that there seems to be areas of tilts with sharp dropoffs in light performance (over a small range of tilt angles), with performance about 30% of peak.

I have always advocated some weighted average of tilts to be used in a performance metric and AGS used face up and 15 degree tilt (based on actual symmetry).

It is currently time wise, and too computer intensive to look at the stone over all tilts and all azimuth perspectives, which should happen.

Think of running raytraces at 1 degree increments of tilt, 90 tilts, rotating the stone about the table/culet axis 1 degree and doing the same thing 360 times and then taking a weighted average based on some probability of looking at the stone from that perspective. If you have a perfectly symmetric RBC, you can cut that down by a factor of 8.

Why, because, how does the stone look to the wearer in the normal viewing position, and HOW is it mounted in azimuth, BECAUSE when the stone is on your hand, the FACE UP position is NOT THE ERGONOMICALLY MOST CONVENIENT and COMFORTABLE viewing position; the wearer has to twist their wrist or arm and strain to see their stone in the face up position, so they are USUALLY viewing the stone in a position OTHER THAN FACE UP. We have gone over this before.

I''ll find the link to the previous Pricescope thread from 3 years ago where I posted the proof of my thesis why these things should be considered, and the very clear possible change in metrics. You seeming understood then, and Garry even posted a DiamondCalc example that was visual proof of the thesis, so you must have forgotten, or....




Could we discuss only light metrics ? ( what is reason to show different between ideal model and scanned model again and again? Do you really think what I do not know it well?)

See above

reGS is entirely dependent on the actual scanned stone.

Marty,
1)It is very well known
2) it is not issue.
If AGS changed weighed for Tilt during develop ASG Gold.

They should do some changing light metric( weighed Tilt in same way) for AGS Platinum( yes it could be much more computer time costly do it for each scanned model). But Otherwise light metrics for ASG Gold and AGS Platinum are different.

And most probably light metric for ASG Platinum is worse ( less correct) than for ASG Gold
But according AGS Platinum Brand , ASG platinum light metric should Best( most correct at least among all ASG cut grading systems)
I think i have not "conceptual or a political problem". But because you defiantly do not like listen about this ASG conceptual problem, you could have some political reason( problems) do it.

 

[adamasgem]

Date: 10/23/2008 11:37:46 AM
Author: Serg



reGS is entirely dependent on the actual scanned stone.

Marty,
1)It is very well known
2) it is not issue.

If AGS changed weighed for Tilt during develop ASG Gold.

They should do some changing light metric( weighed Tilt in same way) for AGS Platinum( yes it could be much more computer time costly do it for each scanned model). But Otherwise light metrics for ASG Gold and AGS Platinum are different.

And most probably light metric for ASG Platinum is worse ( less correct) than for ASG Gold
But according AGS Platinum Brand , ASG platinum light metric should Best( most correct at least among all ASG cut grading systems)

I think i have not ''conceptual or a political problem''. But because you defiantly do not like listen about this ASG conceptual problem, you could have some political reason( problems) do it.

Serg.. I think if you read the attached pdf, which are copies of data published 3 years ago and are somewhere in the PS archives, you may realize that you do have a conceptual problem, or else you haven''t read or understood what I have said. What more can I say, I think the data speaks for itself in supporting what I have said, and you certainly haven''t presented anything contrary. The designations refered to those describing the GIA from their article on fire. They are on PS somewhere also.

Those limited runs, represented a hell of a lot of number crunching using a symmetrical model, and Garrys contribution showing how a stone went from light to dark with a small change in tilt supported the theoretical work, so stop trying to start an argument or confuse the issue. Facts are facts.

 

[michaelgem]

Date: 10/23/2008 6:19:04 AM
Author: Garry H (Cut Nut)

Date: 10/22/2008 7:13:52 PM
Author: michaelgem



The underlying truth is that both GIA's and AGSLAB's grading systems are both performance based. That is, they are based upon the three-dimensional (3-D) light performance resulting from the 3-D interaction of light with a diamond's proportions. This is in contrast to the old 2-D ray trace analysis of proportions done in the past by diamond cut investigators, the most celebrated of whom was Marcel Tolkowsky. His theoretical angles were developed from ray tracing using only a 2-D cross-sectional slice of the 3-D round brilliant cut diamond.

Michael D. Cowing

Michael you need to get me some of that stuff you are using please.

In which way is GIA's system any way related to their ray tracing results for performance?
For example - here is the GIA's light return ray trace study results from 1998 (where they were completely oblivious to the inverse relationship between crown and pavilion angle - I added the lines and they argued I was nuts when I pointed it out) which clearly indicate their steep deep proportions did not make the grade.

You are right, Garry. In 1998 GIA was oblivious. The ray tracing results you displayed of their WLR brilliance metric in hemisphere lighting is no longer part of their current performance based system. It has been replaced with a metric accounting for head obstruction that agrees with the statistical analysis of their comparison testing.

I like to think that folks like you, Sergey and I and certainly Al Gilbertson had an influence on their thinking. My over ten year interaction with folks like Al and Peter at GIA and AGS gives me a somewhat unique perspective that I will try to relate with the following chronology. You will note that I believe Al Gilbertson is the unsung hero at GIA in the development of their current system, which is much changed from 1998.

1. GIA published in 1998 their WLR metric for brilliance, which used hemisphere illumination, which I pointed out in the Journal in 2000, gave incorrect results for brilliance, because it did not account for observer obstruction.

2. Al Gilbertson was hired by GIA around that time frame, and at his urging, GIA funded him to do the comparison testing to validate the results of their performance metrics for brilliance and fire.

3. Having known Al from his early days with Diamond Profiles, I had discussed this issue of observer obstruction with him. I knew that he too had long been aware of these sorts of issues concerning diamond and gemstone light performance. I had no doubt, and had told Al so, that the analysis of the comparison testing would not agree with their WLR, because it did not account for the reality of observer obstruction. I was never privy to the results. In any event, the testing resulted in additions to and modification of their metrics.

4. This in turn resulted, along with a lot more testing and analysis, in the creation of the current grading system as incorporated in software (Facetware) and embodied in their charts and tables published in their lab manual.

So, in the final analysis, their system, which is represented by their charts and tables, resulted from their light performance comparison results and their revised light performance metrics, along with many of the constraints similar to AGS, like spread, polish, symmetry and girdle thickness. This is why I say that it is performance based, just as is AGS's new Gold Standard, and their Platinum Standard.

If anyone can get over my use of the tennis and golf term of “sweet spot” to indicate the range of Ideal rather than the single point at Tolkowsky's or Morse's angle combinations, they can see in the Accordance article that, except for the steep/deep region, GIA and AGS are largely in sync with respect to the range and target center and inverse slope (Ideal axis of -4.5:1) of the top performance proportions of the EX or Ideal cut round brilliant.

Michael

 

[strmrdr]

There is nothing scientific about having 2 cut grade system with the same name top grade from the same lab that one is much much wider than the other.
Playing around with combos AGS Gold Ideal goes out to ~AGS3 on the Plat standard without considering cheating averages and painting/digging.
With those considered it dips well into an estimated ags5 or worse range.

Both can not be "Ideal" in any scientific manner.
They are apples and oranges!

The biggest difference is that even if the range was the same a 2d system using averages is much much more prone to cheating!
Therefore its confidence level is less than half of a 3d system.

 

[adamasgem]

Date: 10/24/2008 3:24:02 PM
Author: michaelgem

This is why I say that it is performance based, just as is AGS''s new Gold Standard, and their Platinum Standard.

OK, Michael, I''ll agree with you partially, FARCEWARE(TM) is performance based, the very lowest level of a partial performance metric, combined with a bad taste test, not to mention Certifigate type probable payoffs influencing the ranges; a system that is everything to everyone, world wide, except, of course, the consumer, a system meant mainly for the trade''s benefit, and not very selective at all.

The hand writing was on the wall when Boyajian wrote his executive preface after their initial, well intentioned, but problematic, "Brilliance Study".

Sure you will find a good stone there, but the object is to weed out the swindled dogs, which it doesn''t do.

It is like calling you a pro if you can break 100 in golf...

Give some people credit for common sense.