AGS New Cutting Guideline Charts for Princess Cuts

[JohnQuixote]

Date: 11/21/2004 5:40:42 AM
Author: Garry H (Cut Nut)

John were you drinking when you posted this - not sure if you were doing a Star Trek joke - or being serious?

Well, tie me kangaroo down, sport...No way!

Drinky + Posty = poor spelling/grammar/rambling and my dear old mum would find it (she googles) and slap me down over long-distance.

*Mom, if you're reading this, don't listen to Garry.

Okay, seriously, let me reference your prior post:

Date: 11/19/2004 9

5:47 PM
Author: Garry H (Cut Nut)
I agree with Rhino - but yes - they will use the same approach - the ASET scope will show more green in the returned light.

<BR class="ibbquote"
So...At IDCC the Sasian/Yantzer/Caudill discussion of angular spectrum (Proceedings page 16-17) used practical examples for RB. The color coding (fig 2) represents angular rays arriving from H to 45 degrees as Green, 45 to 75 degrees as red and 75 to vertical as blue.

As to my question: Referencing AGS "Making The Cut," page 19, the example of a Princess (confirming ASET with DiamCalc) shows contrast of a typical princess cut. There is an abundance of green in the returned light - as you noted - and I was wondering if, as with the IDCC presentation's example of a RB, the green angular rays are those arriving from H-45. Or am I on a different page?

PS - I am now drinking as I post this, but it's all Starbucks, baby.

 

[oldminer]

Well, I have been reading as listening. This is an exciting time for the diamond trade, isn't it? While AGS and GIA go into the future with predictive results based on careful and complete measurements of each diamond, I am going, again, in a somewhat different direction. I have signed an agreement with Imagem, Inc. and taken delivery of one of their grading devices which we hope to soon have up and running. Intead of "predicting" and being dependent on a huge number of measured variables, Imagem technology simply and elegantly measures light return, scintillation, and intensitydirectly from each diamond placed into the device. It does not make use or ray or beam tracing and has been very thoroughly tested over a number of years. A direct reading just makes more sense.

The GIA went into this arena with a huge databsae of existing diamonds that they had graded over the years. They chose to use this information and make predictions of light return in a very complex manner. AGS seems to have gone in this direction, too, although with a somewhat different twist.

Imagem, Inc. came along a couple years after GIA decided on its course of action and was able to work with direct light measurement technology that was new to the market and not there when GIA made up its mind. I think the real future of light performance grading will be with directly measuring behavior, not predicting it.

We hope to soon be offering direct light measuring reports that can be added onto existing GIA, AGS, EGL and any other lab report for a very nominal amount. It will initially be only for round diamonds, but we soon hope to have the other shapes ready for prime time. This approach, while different, should prove totally compatible with GIA and AGS systems, if their measurements and predictions are really working well. If not, then Imagem may be a better methodology. Right now, I can only tell you that it will be a challenging period to see which system or systems give the trade and consumers the best information and the best service.

 

[moosewendy]

How does imagem compare with the gemex brilliantscope? Is it similar technology?

 

[Garry H (Cut Nut)]

Here is a diagram of the ASET lighting model.
Light that comes from the lens direction is blue (15 degrees each side of top dead center = 30 total - Ideal-scope is 24 total)
Light from 15 to 45 is red (30 to 90 of the 180 hemisphere)
the remaining 45 degrees each side (to the horizon) is green

 

[Garry H (Cut Nut)]

Sir John, the answer is yes - that is the lighting angles for ASET.

Dave it wil be interesting to compare results of the Brilliancescope and ISEE2 to Imagem. Have you done it?
BTW - I must check and see if there is back light - I do not think there is - so the right side image would be wrong

 

[oldminer]

Yes, it will be of interest to see how Brilliancescope and ISEE2 compare to Imagem. I have truthfully not made a comparison to date. No doubt, this will be done extensively by many interested parties.

Imagem MEASURES. It does not make its grade based on comparison to a visual database. It counts bright pixels, dark pixels, anddetects patterns and quantifies them. I don''t believe BScope or ISee2 do their work in a similar manner, but maybe I am mistaken.

 

[moosewendy]

The brilliancescope counts pixels, but I don''t think it detects patterns. I think it does a straight count, which is one of the problems with the technology, particularly with fancies. I guess the biggest question would be under what kind of lighting conditions "return" is measured. Does it simulate "store" lighting, direct sunlight, etc.

 

[oldminer]

Ah, everyone wants to know about how the lighting model is constructed. And it is my belief that Imagem will not release this information as it is their intellectual property. However, over time, the method of collection will not be the main question, but how well it works compared to other systems and/o what our human peerception tells us. So long as the end result appears correct and is based on quantifiable and repeatable results, I don''t forsee a major problem. Diamonds do look different in different lighting. Of course. They look black in a dark room, too. The way light reacts to the cut of the diamond creates the light return that can be measured. I believe Imagem has employed a lighting model that will provide real world results and not used some model that does not equate to what our eyes might see. The basis for much of Imagem is the science of imitating human-like results with technology.

 

[moosewendy]

Dave, that''s the key. It needs to mimic human perception, not "improve" on it. If the results are an accurate predictor of what people would choose if given the opportunity to view 2 diamonds side by side, then the technology is a winner.

 

[oldminer]

Yes, I agree, that "mimic" is key not "improve". Imagem, Inc reads these postings and is very sensitive to the needs of future and potential clients and consumers. The idea of repeatable, honest results is undoubtedly the key to all efforts, GIA, AGS, MSU and Imagem. We have already had plenty of less capable attempts including my own AGA Cut Class system. I never thought my system was the ultimate tool, but it still has merit in protecting people from foolish purchases and in quality assurance. I have developed a very refined AGA Cut Class version to use with Imagem so that we get results on par with AGS or GIA. I sure don''t want to confuse people with conflicting results unless there is a very good reason to create a problem.

 

[JohnQuixote]

Date: 11/22/2004 2:42:40 PM
Author: Garry H (Cut Nut)
Sir John, the answer is yes - that is the lighting angles for ASET.

Dave it wil be interesting to compare results of the Brilliancescope and ISEE2 to Imagem. Have you done it?
BTW - I must check and see if there is back light - I do not think there is - so the right side image would be wrong

Thanks Garry.

Dave, I know you are treating this Imagem angle with care, but I continue to have concerns about any device which purports to quantify overall performance using a limited, artificial lighting scenario. A diamond's panorama of illumination is so highly variable that unless you give separate reports for "in sunlight," "under spotlights," "in mostly diffuse light," "in medium candlelight," "under 60w light bulb" etc. it seems that a single report will be prejudicial. Brilliance Scope is famous for punishing stones cut for an abundance of dispersion – but many people think those diamonds are just as beautiful (or more so) than those with narrow pavilion mains that BS rewards.

Further, contrast brilliance/obscuration/and pedestrian, real-world lighting conditions make a lighting chamber formula even more unrealistic to me… Now we need to have reports for “in sunlight with 2 viewers and guy wearing black shirt 5 feet away,” or “in candlelight with white cat sleeping on lap and glass of merlot 16 inches to the horizontal.”

I believe IdealScope to be a tangible indicator of light leakage without prejudice. Proposed ray-tracing and 3D modeling will, at least, treat each diamond equally through the paces. A static lighting chamber is just too limited to take all possible illumination scenarios *AND* human taste relative to those scenarios into consideration in my opinion.

You know someone has to say it

 

[strmrdr]

Date: 11/22/2004 3:45:41 PM
Author: moosewendy
If the results are an accurate predictor of what people would choose if given the opportunity to view 2 diamonds side by side, then the technology is a winner.

That is an impossible task.
You could have 20 people in a room and have them go thru the same 30 diamonds and get 20 different answers on which is best.

Or put Gary H and David from diamondsbylauren in a room with 500 diamonds and have them select the 100 best ones.
Id be shocked if even one diamond was on both lists.


The best any tool can show under these conditions using this criteria this diamond scored better.

The problem is that most of them are saying put the diamond in this box bingo this one is better and not telling the conditions and criteria.
Pretty trustworthy huh?

 

[Garry H (Cut Nut)]

Date: 11/23/2004 12:56:45 AM
Author: strmrdr

Date: 11/22/2004 3:45:41 PM
Author: moosewendy
If the results are an accurate predictor of what people would choose if given the opportunity to view 2 diamonds side by side, then the technology is a winner.

That is an impossible task.
You could have 20 people in a room and have them go thru the same 30 diamonds and get 20 different answers on which is best.

Or put Gary H and David from diamondsbylauren in a room with 500 diamonds and have them select the 100 best ones.
Id be shocked if even one diamond was on both lists.


The best any tool can show under these conditions using this criteria this diamond scored better.

The problem is that most of them are saying put the diamond in this box bingo this one is better and not telling the conditions and criteria.
Pretty trustworthy huh?

If the selesction criteria was "good" and "bad" storm, and you had to divide the pile into 2 equal halves of 250 stones - what difference do you think there would be in the disputed differences?

Do you think we would each aggree on the top 200 and worst 200 - and dispute 100?
Want to put some #''s on it?

 

[valeria101]

500 makes me dizzy only thinking of the sheer amount of work! I'd hate all diamonds in the world way before hitting #100 !

I would not worry much about individual choice strm... these tools all say about the same simple thing: brilliant diamonds are nicer to look at than dull ones. Sounds right, given that these thingies have no color, and are rather small for what most mortals consider paying... what else is left but brilliance? They'd better glitter their diamond soul out!

The rest is just numbers... where "dull" begins, and so forth is just room for marketing, IMO.

 

[Garry H (Cut Nut)]

Ana i often sort +2,000 plus stones a day when on a buying trip.
If color is low, then color first.
But usually just idealscope them all, then color and clarity followed by a final ideal-scope.
I usually end up with less than 200 ''buys'' if they are larger, and maybe 500 at best if they are smaller high quality goods.

 

[valeria101]

Date: 11/23/2004 3:28:45 AM
Author: Garry H (Cut Nut)
Ana i often sort +2,000 plus stones a day when on a buying trip. [...]I usually end up with less than 200 ''buys'' ...

... that''s sorting err... about four pounds of rice grain by grain!


Is this anywhere near standard practice? Someone must be ending up with the "other" 1800 pieces...

 

[strmrdr]

Date: 11/23/2004 2:54:57 AM
Author: valeria101

I would not worry much about individual choice strm... these tools all say about the same simple thing: brilliant diamonds are nicer to look at than dull ones. Sounds right, given that these thingies have no color, and are rather small for what most mortals consider paying... what else is left but brilliance? They''d better glitter their diamond soul out!

The rest is just numbers... where ''dull'' begins, and so forth is just room for marketing, IMO.

brilliant vs dull under what conditions val?
Saying its brilliant without saying under what conditions isnt very informative is it?

 

[strmrdr]

Date: 11/23/2004 2:43:16 AM
Author: Garry H (Cut Nut)

Date: 11/23/2004 12:56:45 AM

Author: strmrdr

If the selesction criteria was 'good' and 'bad' storm, and you had to divide the pile into 2 equal halves of 250 stones - what difference do you think there would be in the disputed differences?


Do you think we would each aggree on the top 200 and worst 200 - and dispute 100?

Want to put some #'s on it?

At the 250 point depending on the mix of diamonds there would be some overlap likely.
At that point you both would have likely gone outside your taste in diamonds in selecting them at that point.

 

[JohnQuixote]

Date: 11/23/2004 2:54:57 AM
Author: valeria101
500 makes me dizzy only thinking of the sheer amount of work! I''d hate all diamonds in the world way before hitting #100 !

I would not worry much about individual choice strm... these tools all say about the same simple thing: brilliant diamonds are nicer to look at than dull ones. Sounds right, given that these thingies have no color, and are rather small for what most mortals consider paying... what else is left but brilliance? They''d better glitter their diamond soul out!

The rest is just numbers... where ''dull'' begins, and so forth is just room for marketing, IMO.

Ana - There is some truth to that. For these tools to say "Brilliant are better than dull," is fine. Arguably, HCA has proven just as good for that purpose.

I can agree that these light-chamber devices are useful for separating good light return versus poor return over a range of average to good cut quality, but when making comparisons between equivalent makes of the best cuts they are inconsistent and inaccurate (let''s not get started on "scintillation" measure

), as well as having the relevancy issues I alluded to above.

Best,

 

[valeria101]

Date: 11/23/2004 10:51:41 AM
Author: JohnQuixote

I can agree that these light-chamber devices are useful for separating good light return versus poor return over a range of average to good cut quality, but when making comparisons between equivalent makes of the best cuts they are inconsistent and inaccurate

I understand this. Surely a measure of variation (e.g. "scintillation" defined by pixel counting) can only be times less precise than the basic measurements. For this reason alone it deserves being left to rest for now...

 

[valeria101]

Date: 11/23/2004 9:42:42 AM
Author: strmrdr

brilliant vs dull under what conditions val?

I am not taking the challenge... my previous post (the one you are referring to) contained just a qualitative statement: a position of principle that in my mind has no connecton with modeling technicalities that far.

Down to technicalities, one can always select a critical position to represent a range (= decide which lighting still relevant to human vision makes diamonds look worst, and then pick up those that manage to look decent even then - a crutical path approach of sorts). Not that it is done (the picture of such a Reversed - Bscope would look like crap!), but most of the time this is possible.

Besides, choosing lighting conditions it is obviously a matter of taste (i.e. do you want an "evening cut" ?) and choice (as is shaped the current contest of proprietary sets of such conditions). Matters of atste are not the most critical... IMO.