Princess Apples to Apples on cut b/t GIA and AGS?

[ChunkyCushionLover]

Date: 5/19/2010 7:58:17 PM
Author: Karl_K
this one is much better and the resulting ASET image is very close face up.
But the scan errors will show up more as it is tilted.
Viewing the face up ASET images side by side you would not be able to locate the scan error unless you blew it up full screen.

Karl,

It is rare for us to request a .gem file normally we ask for just one static image (although I would ask for a .gem file more often now just to see what happens upon tilt).
Scan error will most often make the diamond look worse, it isn''t going to introduce a systematic positive bias, not much of an evil there.
I am comfortable with the vendor deciding if the simulated images look much worse than the photographed images and if a new scan should be taken.

It is in the interests of the vendor selling branded diamonds at a premium over unbranded to have the best images and scans available so that they can justify the premium price and educate consumers on the differences.

Am I worried about if the vendor selling unbranded diamonds will go to the same lengths to insure their images are up to the same standard? Not really it can only hurt the sales of these unbranded diamonds if they don''t.

 

[Karl_K]

Date: 5/20/2010 2:59:29 AM
Author: ChunkyCushionLover

Scan error will most often make the diamond look worse, it isn't going to introduce a systematic positive bias, not much of an evil there.

That is not so, scan error has an impossible to predict effect on scan based images/grades.
At times it will be better or worse there is no way to predict it.
On a super-ideal rounds scan error will often make them look worse but on fancies all bets are off.
Scanner tech is just not there.
Helium is slightly better but far from perfect. (sarin has closed the gap somewhat with the updates)

 

[Todd Gray]

Date: 5/20/2010 2:36:57 AM
Author: ChunkyCushionLover

Date: 5/19/2010 3:14:45 PM

Author: Todd Gray

Oh where to begin… I suppose that it is only appropriate to begin by thanking you for driving the nails into the coffin CCL, because your posts are a classic example of the very abuse and misinterpretation which I am suggesting needs to be avoided here on Price Scope!

Todd what I find most disagreeable with this thread is that you are rejecting all comparison tools and leading consumers to only the ''proven'' branded lines. While I don''t generally have a problem with your statements which often promote directly or indirectly Infinity Diamonds (that you happen to sell), I do however feel this thread crosses a line especially since there are so many factual inaccuracies and assumptions (see below) made in your statements.

Unless I am misinterpreting the intent of your statements and the examples which you use, you are making the mistake of using simulated images which are generated from scans taken using two different machines… for diamonds of two different facet structures… and trying to determine minutia. Apples and Oranges are both types of round fruit, but they are entirely different in all other ways, the only similarity between a princess cut and a solasfera cut is that both are square cut diamonds.

The examples chosen were both of well cut and characterized Princess cuts, I am well aware the one has two chevrons (Infinity) while the other has three (Solasfera).

But this is secondary and the examples chosen were only to to show why the faceup is a good overall indicator of the important tilted positions as well.

Do you not see how little difference slight tilting makes to where these two diamonds gather light from? The greatest difference is small areas show contrast(obstruction) which is favoured in some positions over others.

See the .gem video files if you don''t accept my captures.



All the AGSL is attempting to do is separate diamonds into broad cut grades, but what you’re doing is using one 2D snapshot of completely different kinds of stones to extrapolate aspects that the AGSL will tell you cannot yet be extrapolated from that category of data-set; I know this for a FACT because I sat next to AGS Lab Director, Peter Yantzer for three days at the Dallas Symposium which I attended last month and discussed this matter with him in great detail after having had the pleasure of watching his presentation on this very subject.

I will invite Peter to this thread I hope you will do the same. Regardless, the AGS research on the correlation between ASET and real world diamond performance published here and elsewhere stands on its own and came before the PGS.



I suppose that all of that could be considered by some to be irrelevant, yet I believe that the entire premise of your comparison is incorrect to begin with. I have great respect for Rhino, both as a friend and a competitor, and I understand that Rhino used the AGS PGS for his solasfera using the princess setting in an attempt to provide additional insight for his clients, however since the AGSL does not yet have a metric for the solasfera, the bottom line is that the results are inaccurate because they are based on a facet structure of a different configuration. Period. Once again, the only similarity between a princess cut diamond and the solasfera square cut diamond is that they are both square.

You have confused so many different things here.



1) Rhino did not produce the Solasfera nor is he their only vendor. Solasfera is a brand name and a supplier in its own capacity.

2) The Solasfera has attained AGS 0 on grading reports already see the DQD for this diamond

3) I have no idea what you mean by the AGSL does not have a metric for the Solasfera?

Attempting a head to head judgment for an 81-facet stone versus a 45-facet stone using these tools is not a practical comparison, in my opinion since the results will clearly be different based upon the distinct differences of diamond design. I’m not even sure that attempting to make such a comparison would be considered to be ethical by consumers if such as feat was attempted by a member of the trade (?) but I suppose that if the AGSL intended for their software to be used to evaluate the solasfera, that they would have built a custom metric for it to be measured by, however the fact of the matter is that they have not. And I know for a fact that Peter Yantzer and his team possess both the capability and the talent to create the metrics for doing so, they create custom facet structures for the plotting diagrams on lab reports for specialty cut diamonds all the time. It is a fascinating process. However I can’t figure out why the AGSL would be inclined to create such a metric for the solasfera since they seem to be primarily graded by the GIA; which I suppose means that the AGS PGS will never provide a sufficient metric for measuring the solasfera.

The number of facets is a personal preference much like LGF length in rounds and is a subjective beauty aspect not a light performance metric to be graded.

There are 8 more facets on the pavillion of an 3 chevron versus a 2 chevron princess but since the Infinity has 4 more crown facets the difference is only 4 facets! How did you come up with those numbers?

Please do not confuse how they generate new templates for their inclusion plot diagrams ( Sarin > Gemcad > Fit To Sarin Data >Plot) and the PGS software whose metrics do not change based on a new pavillion facet structure.


The AGS does grade princess cuts with different pavillion facet structures using the same light performance criteria. I would shutter to think what kind of credibility they would have if they bent their performance grading to suit new cuts or suppliers.



As implied by the graphic of the automobile silhouette which I provided as an example above, little can be determined about the potential for brightness between two diamonds for which the only similarity is that they are both square. The implications of angular spectrum are different for different configurations of facet structure and cut precision; period. For those who might not be aware, the AGSL uses separate metrics for configurations which I consider to be closer in similarity than the princess and solasfera, such as for round and oval cut diamonds and that is with a full ray-tracing suite (!) not just a few computer-generated snapshots taken from different scans on different machines… And you have 49 crown facets on the one stone! That is simply a scanner nightmare. Yeah I know that the helium is a great scanner and all, but sorry… I have too much experience to trust 3D scan data for stones with that many subdivided break facets… Been there and done that, it’s not happening.

Even if for a moment, I could convince myself to pretend that the comparison you have suggested is valid; which I can’t; you chose a single data set with a black background. Why would you do that? The white ASET background is much more useful for evaluating leakage versus contrast; especially for fancy shape diamonds. A professional diamond buyer would know this, a consumer probably wouldn’t and here again is my issue with “pro-sumers” running all over the forum telling less knowledgeable consumers “get an ASET; get an ASET” because purchasing decisions are being made on limited results and inaccurate assumptions. I realize that this is likely not to be a popular opinion amongst the pro-sumers on the forum, I thought about that before raising the issue to this height; however I believe that the continued assumption that a static view of a diamond as seen through an ASET scope is an extreme disservice to the very consumers who come to this forum seeking true insight.

The images above was meant to show one thing, tilt the diamond 15 degrees in either direction or anything less versus faceup the images are quite similar. You are right your opinion is not popular and you haven''t supported it by showing us false positives or examples of misrepresentation by other vendors. If anything poorly scanned sarin results and simulated images make stones look worse not better and it is in the vendor''s best interests to redo the scans if they feel this has occurred.

Since we’re all here to learn and this includes me… Take a look at the PGS simulation of the solasfera, with white background, and you’ll notice a lack of red at the edges, whereas the ASET of the Infinity Diamond on HPD, which is an actual photo by the way and not a computer generated scan;



Integrate the areas of the red regions of the two diamonds it is not even close the Solasfera has much more red.

Once again I don''t want to get into a comparison of the two but there are some videos on Vimeo and youtube of Solasfera versus other AGS 0 princess cuts if you are interested.



shows red at the long outer edges, even in the more-challenging white-backed metric... This alone contradicts your basic conclusions… However since as I stated above, attempting to compare results for a square solasfera and a princess cut diamond is far from an apples to apples comparison, I am not drawing any conclusion other than the basis for your conclusion is without foundation. We could argue that the solasfera scan might be a little wonky; that the standardization of our photo may vary from the simulation; that the angular spectrum which I have come to expect may not be appropriate for its configuration, blah, blah, blah… But this is a text book example of why I said what I said earlier in this thread and warned of the potential for abuse of the AGS PGS software, naturally I did not have my friend Rhino in mind when I made that statement; I was not aware that he was attempting to use the software to evaluate solasfera cut diamonds and imagine that he is doing so for comparative purposes of solasfera to solasfera cut diamonds only because as stated, it would be absurd to attempt to use the software to draw a comparison between a square solasfera and a princess cut diamond since the metrics would have to be entirely different for the differences stated above.

This is an opportunity for us to learn that a lot of different looks exist and we can “snap” any one of them in an attempt to support whatever argument we want to make: - 30blue; 40blue; 30blue/white; 40blue/white; all of the forward and reverse ray-tracing views as well; as yet-unincorporated views for small, medium, large and cosine20 scintillation events. All of this and we have not even discussed the other reflector topics which I could raise to further make this point. All of these possibilities when considered together tell a story; and it’s a useful one. However even that story, as seen by the AGSL, is only used for placing diamonds which it has had the opportunity to grade and scan using the equipment which the AGS PGS was designed specifically to be used with, is only for placing a diamond inside a broad cut grade, not to attempt to dissect minutia like you are attempting to do.

Todd I pointed you to the .gem files, I have accurately snapped the images with the most obstruction in them and therefore the ''worst'' position over the 30 degree tilt range. If you disagree just view the whole video of each and choose your own snapshot.



ASET 30 is a cone with 75 - 90 degrees and 90 - 105 degrees being considered blocked by the viewers head and obstructed (ASET Blue). This is the same cone of obstruction as in the white background ASET scopes where the images were taken.

It would be more misleading to choose ASET 40 as then a comparison with the photographed images would not be equivalent.

I believe that I understand the intent of your argument, but to be it in local terms, that dog won’t hunt... And I think that there are a lot of people who have spent a lot of time researching diamonds and believe that they are capable of making minor distinctions about diamonds from minute details as provided by images being evaluated on their computer monitor at an average quality of 72 dots per inch. But I’ve been in the business of buying diamonds since 1985 and I have seen thousands and thousands of diamonds cross my desk. And that is the point really… There is no substitute for the experience of seeing thousands of diamonds and having the opportunity of comparing what I see with the scans created by the equipment available to us. The same premise holds true for attempting to determine the extent and visibility, color, tone, etc. for a clarity image as viewed from a computer monitor as compared to actually being able to examine the inclusions as seen through a gem scope using various degrees of magnification and soapy water. Any gemologist experienced with these things will verify that what I am saying is true.

Author Mark Twayne once said “Lies, damned lies, and statistics” to describe the persuasive power of numbers, particularly the use of statistics to bolster weak arguments and the tendency of people to disparage statistics that do not support their position. “Lies, damned lies and statistics” seems like a statement that could easily be applied to the use of unauthorized, inaccurate and inappropriate use of the AGS PGS to attempt to designate a light performance rating for diamonds not graded by the AGSL and for which the AGS obviously never intended their software to be used… But we’d have to change it to “Lies, damned lies and computer generated images”.

But to address the real question at hand CCL… No, I am not trying to lead novices into thinking anything other than the fact that this subject is quite complex. Are you?

Todd,

Please invite Peter Yantzer here I will do the same.

To enter a proper discussion and a proper educational thread you need to explain alll the light performance metrics for the PGS software, which tilt positions are considered and what the weightings for each tilt position.

While some of this information on how each of these metrics are calculated and positions can found here the formulas for weighting are not. You have already created confusion on this when you said Dispersion is not being considered I beleive this to be false.

Rather than argue over my examples lets take a closer look at how the PGS calculates the performance cut grade. Anything to share like slides perhaps from the symposium?

In response to:
Todd what I find most disagreeable with this thread is that you are rejecting all comparison tools and leading consumers to only the "proven" branded lines. While I don''t generally have a problem with your statements which often promote directly or indirectly Infinity Diamonds (that you happen to sell), I do however feel this thread crosses a line especially since there are so many factual inaccuracies and assumptions (see below) made in your statements.

I''d like to say:
CCL, I''m not the one who attempted to make a comparison between a Solasfera square brilliant and an Infinity princess cut diamond... that was you! I simply pointed out that your comparison of the two diamonds was an inaccurate comparison because the two diamonds have distinctly different facet structures. The only similarity being shape. It was also you who attempted to make a direct comparison between the GOG Princess Cut Diamond, the Solasfera and the Infnity when you said: "This is the same for a Solasfera and the GOG Signature Princess cuts both of which feature more edge to edge zones of red than the Infinity brand over this entire tilt range." The fact that you seem so intent on turning this thread into a comparison between the GOG Princess; the Solasfera which is also sold by GOG and the Infinity princess which is not leads me to wonder whether you''re schilling for anybody, CCL?

You were quick to point out that I sell the Infinity brand of diamonds in what I can only imagine is an attempt to discredit my expertise, however anybody who actually reads this thread will note that I *NEVER* mentioned Infinity *anything* in my initial response, nor did I refer to any brand of diamond specifically - that''s all been you; at which point you attempted to draw a direct comparison between specific diamond brands... If not to sway people directly towards one brand over another, I don''t know why you would attempt to do so, other than assuming that you are being compensated to do so... the only time I mentioned Infinity was in response to your comparison of the Solasfera and Infinity cut diamonds, so if you want to be offended by something, be offended at your attempt to push one brand over another. The only preference which I stated was my personal preference for AGS grading as it specifically applies to princess cut diamonds, no brand mentioned and Infinity is not the only cutter who sends princess cut diamonds to the AGS for grading.

In response to:
The examples chosen were both of well cut and characterized Princess cuts, I am well aware the one has two chevrons (Infinity) while the other has three (Solasfera). But this is secondary and the examples chosen were only to to show why the faceup is a good overall indicator of the important tilted positions as well. Do you not see how little difference slight tilting makes to where these two diamonds gather light from? The greatest difference is small areas show contrast(obstruction) which is favoured in some positions over others.

I''d like to say:
CCL the difference between the facet design of the Infinity princess cut diamond and the Solasfera square brilliant are remarkably different... well beyond the number of chevron facets. The increased number of facets in the design of the Solasfera will result in a higher number of virtual facets, which subsequently will result in smaller flashes of light - this will appeal to some people, just as some people prefer the look of an 81 facet round to the look of a traditional 57 facet round (58 if you include the culet as a facet) and there is nothing wrong with that because it is simply a matter of personal taste and preference... at no time have I said "the Infinity princess cut diamond is superior to the Solasfera" so please do not put words in my mouth. All I''m saying is that the difference in the design of the two diamonds makes the comparison inaccurate.

In response to:
I will invite Peter to this thread I hope you will do the same.
Regardless, the AGS research on the correlation between ASET and real world diamond performance published here and elsewhere stands on its own and came before the PGS.

I''d like to say:
Say what?!?! Of course the article predates the release of the PGS, it has been stated many times that the ASET concept is a work in progress and that what has been discussed thus far is only the tip of the iceberg. What is the relevance of this statement?

In response to:
You have confused so many different things here.
1) Rhino did not produce the Solasfera nor is he their only vendor.
Solasfera is a brand name and a supplier in its own capacity.
2) The Solasfera has attained AGS 0 on grading reports already see the DQD for this diamond
3) I have no idea what you mean by the AGSL does not have a metric for the Solasfera?

I''d like to say:
1. Of course Rhino did not produce the Solasfera, I did not say that he did... I referred to the diamonds which he scanned as "his Solasfera" because they are in fact "his" after all, they are part of "his" inventory... I suppose that technically I should have referred to it as "the Solasfera" but I never imagined that anybody would get so caught up in the use of a single pronoun.

2. I never said that the Solasfera could not achieve an AGS Ideal 0 grade; I merely stated that most of their diamonds are sent to the GIA for grading.

I would like to point out however that the diamond which you used as an example of the Solasfera attaining an AGS 0 overall cut grade is actually NOT a Solasfera, but rather for one of an AGS Ideal princess cut diamond (note the traditional princess cut facet pattern indicated on the plotting diagram) which is advertised on the web site as a "2.1ct I VVS2 GOG Signature Princess Cut."

3. You''re being intentionally obtuse, right? Well, just in case you''re not... AGS would need to design a separate metric for a diamond like the Solasfera since they place tremendous weight on facet-structure and virtual facet implications. The fact that formula-B happened to earn AGS-0 when the selector in PGS was set to princess means nothing. Heck, I think that the Solasfera is nice and maybe it would be graded "0" if AGS designed a metric for it, I''m not saying that it wouldn''t... but it is not appropriate to deem it a princess simply because it is square, any more than it''s appropriate to consider the Leo or Star 129 a traditional RB and thus any output created by the PGS while set on the Princess cut platform is irrelevant.

I truly wish that Peter Yantzer was permitted to engage in casual conversation on the forum, it is my understanding that AGS Policy prohibits him from doing so

 

[Todd Gray]

Apple... Meet Orange.

 

[Lorelei]

Hi Phade,

I just wondered how you were getting on? Have you made any progress?