AGS 0 -- what exactly is required?

[tenbinko]

I understand that the AGS uses a 3-D system and that a 3D model of the diamond is first created and a cut grade assigned afterwards... but HOW is it assigned, does anybody know? I've attached a file with a cutting guideline chart... but a guideline is only a guide, right. My guess is that if a diamond's got the specs that place it in an AGS "0" box, it's a candidate for a 0 grade, but a candidate only?

Then, does the grader look at the ASET to determine if the cut grade is truly 0, or perhaps 1, or even lower?

Surely, it's not as simple as using newer (post-2005, 3-D-based) proportion charts and considering symmetry, polish and such to yield a cut grade and then giving an ASET image derived from the 3-D model only as extra info for the consumer?

 

[Karl_K]

http://www.agslab.com/members/content/docs/Complete_Explanation_of_AGS_Cut_System.pdf

 

[Karl_K]

The simple answer is the 3d scan is fed into a computer program and it spits out the grade.
http://www.americangemsociety.org/performancegradingso

 

[yssie]

Perhaps someone with more knowledge (Karl/Paul/Wink/Rhino?) will be able to answer your questions more thoroughly, but my 2c -

1. You're right - the proportions guideline is a guide only. A stone can have perfect proportions by the chart but not get the 0 grade, and a stone can miss 0 by the chart and earn the grade.

2. PGS: http://www.americangemsociety.org/performancegradingso
PGS is a ray tracing programme. The stone is scanned and PGS output determines the cut grade - so the ultimate cut grade depends entirely on how well the stone is represented by the scan. This means that you, the consumer, need to factor in a couple of things not on the report -
A) Scanner error and/or inefficiencies have direct effects on accuracy. I *believe* AGSL uses helium in the labs but mostly Sarin for routine grading, helium is both more precise and more accurately captures slight wonk (imprecise facet meets, etc.)

You can find listed precision of various scanners on Sarin's site http://www.sarin.com/diamensionhd.asp
Most are listed w/ error of +/- 0.1 or 0.2deg (angular). Rather puts quibbling over an average of 40.8 vs 40.9 into perspective.

B) Since cut grade is based entirely on a scan of the stone's faceting which doesn't factor in internal characteristics, cut grade doesn't account for effects of inclusions and body colour on type and quantity of light return. Those generated ASET pics printed on some of the DQDs are derived from the same scan so obviously are subject to the same inconsistencies - that's why we always say get an actual pic. I've found that variations are often more notable w/ princesses than rounds.

AGSL won't grade fracture filled stones at all.



ETA Oops, slow typing and Karl already answered

 

[yssie]

Here's a GOG page comparing Sarin & Helium http://www.goodoldgold.com/technologies/diamondscanners/helium/

 

[Karl_K]

Here's a GOG page comparing Sarin & Helium http://www.goodoldgold.com/technologies/diamondscanners/helium/

There is a new sarin machine out since then, sarin HD it is closer to helium than the old model at least for rounds if used in the most accurate mode.

 

[Karl_K]

Perhaps someone with more knowledge (Karl/Paul/Wink/Rhino?) will be able to answer your questions more thoroughly, but my 2c -

I am cracking up here, good job :} You know your stuffs.

 

[diagem]

A) Scanner error and/or inefficiencies have direct effects on accuracy. I *believe* AGSL uses helium in the labs but mostly Sarin for routine grading, helium is both more precise and more accurately captures slight wonk (imprecise facet meets, etc.)

You can find listed precision of various scanners on Sarin's site http://www.sarin.com/diamensionhd.asp
Most are listed w/ error of +/- 0.1 or 0.2deg (angular). Rather puts quibbling over an average of 40.8 vs 40.9 into perspective.

Only one point I would like to comment on (Yssie, you are doing great job), the angular margin of error on the Sarin scanner is 0.3-deg (e.g. 0.15 +/- each side) and definitely not 0.1 as listed on their website. I have no experience with Helium but it would be hard for me to imagine higher accuracy & repeatability ( repeatability is sometimes more important than a single margin of error.)

 

[yssie]

Karl, I'm preening

DiaGem - if the Lord of the Cutters says so, well, so it must be...

I can definitely understand why repeatability is a vital real-world consideration. You just acknowledge that there is going to be error you can't identify or quantify, but if you can specify with some confidence what the variance including that unknown error is expected to be, or what the trends are, you've gone a long way toward removing those variables even without correcting for them! In that system.

 

[diagem]

Karl, I'm preening

DiaGem - if the Lord of the Cutters says so, well, so it must be...

I can definitely understand why repeatability is a vital real-world consideration. You just acknowledge that there is going to be error you can't identify or quantify, but if you can specify with some confidence what the variance including that unknown error is expected to be, or what the trends are, you've gone a long way toward removing those variables even without correcting for them! In that system.

Yssie, I am definitely no Lord..., but thanks.

I didn't acknowledged "I" can't identify or quantify the errors.
On the contrary, I must identify these errors to complete the cuts aimed for optimal accuracy and precision.
What I am saying is I wish the margin of error would stand at angular 0.01deg, that would make my life as a cutter to those precision levels much easier. But honestly I can't see perfect 3D measurements happening so soon.

 

[Karl_K]

I'm going to disagree with my buddy Yoram just a little. :}
Helium when set for max accuracy is more accurate, repeatable and creates better models than even the newer sairin HD.
The scan time however is 3-4x longer at that setting than sarin which is why AGS does not use them for routine grading.
There is a trade off on any optical scanner - time vs accuracy.

 

[Karl_K]

I didn't acknowledged "I" can't identify or quantify the errors.
On the contrary, I must identify these errors to complete the cuts aimed for optimal accuracy and precision.
What I am saying is I wish the margin of error would stand at angular 0.01deg, that would make my life as a cutter to those precision levels much easier. But honestly I can't see perfect 3D measurements happening so soon.

Amen to that.
.01 would make your life a lot easier.

 

[tenbinko]

Thanks, Karl and Yssie. I have that AGS cut system pdf, actually, and I read it a few times (so embarrassed to admit that... because I still didn't know at the end how the cut grade was determined exactly, but I guess that's because I didn't read the last part of the document very carefully, where I thought it got more technical and referred to other types of cuts as well, and initially, I just wanted an in-depth article on proportions on rounds and learn more about the ASET).

So, as Yssie mentioned, it could potentially be a bit inaccurate due to inclusions and scanning errors, etc. That's why an Idealscope, actual ASET, or GCAL optical brilliance image or something like that always trumps the computer-generated ASET on the report, eh?

But BTW, I thought I read somewhere AGS uses the Helium scanner... if it uses the Sarin, then its measurents aren't that much more accurate than GIA's, right, since GIA's lab uses Sarin, I believe. But of course, it rounds too much.