AGS 000 and ImaGem results

[JohnQuixote]

Date: 6/12/2006 5:46:24 PM
Author: Paul-Antwerp

2. The H&A-view of a princess will show you its symmetry. Do not look for arrows, or hearts, look for a symmetrical built-up. Symmetry is always the key to light performance. I am sure that non-symmetrical light performance exists, but cutting planned asymmetry is extremely difficult and unlikely.

To further this point, while we shouldn’t expect to see hearts, arrows or bunnies, we can expect that aiming mirrors more precisely at each other will have a positive result... Strm, at the AGS presentation Peter stressed that, in emerald cuts, pointing the corners at each other improves performance.

Chalk another one up for the optical symmetry pundits (we’re the ones in the white hats).

 

[JohnQuixote]

Date: 6/12/2006 5:27:03 PM
Author: strmrdr

This thread is giving me flashbacks.
When im feeling better I will hunt down the earlier princess thread where this was discussed with the b-scope.
Measuring light return in a fixed position doesnt mean as much with fancies.

Exactly what I was thinking.
I hear it every day from my friendly neighborhood cutter, actually.

People have tried to interpret fancy performance based on that kind of info in the past.
(No! Bad people. No biscuit.)

I know Brian was advising against it long, long ago.

 

[JohnQuixote]

Found the thread I had in mind.

Brian made these comments back in 2003, advising against princess light performance judgments using just proportions info / ideal-scope.

Date: 12/1/2003 2:55:26 AM
Author: BrianTheCutter

Light leakage evaluation should be combined with a live examination, especially for fancies, and that is what we do. One should remember diamonds are not worn in a static position and the human eyes look at the diamond from different angels which will change the way one would see the stone in a static upright position...

Symmetry is not always a guarantee of the best beauty in fancies - small amounts of light leakage
are good for contrast and character. Ask Serg at MSU I think he would agree!

Original thread here (a blast from the past!)

More people appear to understand this now. I’ll forward this conversation to Brian; it will make his day to see the great discussion in this thread.

 

[researcher]

Date: 6/15/2006 1:33:01 AM
Author: JohnQuixote

Date: 6/12/2006 8:39:42 AM

Author:oldminer


Maybe someone with more experience with the AGS 000 princess cuts will have a differing point of view or different information they wish to share. I'd sure be glad to know why the light return is not judged face-up as the primary grading characteristic. All opinions appreciated.

Great thread Dave.


Rounds are the most optimized shape for light return. As such the static view is appropriate. On the other hand, in fancies the look changes more in dynamic situations.


IndieJones voiced this nicely from a consumer perspective.


Researcher, I know you’re an experienced princess enthusiast. Do you feel different angles change the look of yours dramatically? If so, does your single, face-up view still represent the stone's best qualities to you?

I love how you said "enthusiast" as a description for me--I am that!

I'm obviously not an expert, so please bear with me as I massacre the proper diamond terminology! Also, please remember that these are just my personal observations of my stone (I've included the virtual ASET as a reference).

Yes, different angles and different lighting does dramatically change the look of my stone. I can't say for certain what I like best, but I can describe the differences.

When viewed straight on (hand held out straight at eye level in front of me) in bright overhead light I am literally blinded by the reflection off my table and can see lots of pinfire on the crown. When tilted, I get broad flashes of color and white light and my stone seems to reflect the light throughout the entire stone.

When viewed straight on indoors with more diffuse lighting, my stone shows more scintillation than when tilted. However, it appears more brilliant in this lighting when it's tilted.

In candlelight my stone has lots of pinpoint fire when viewed straight on, and broader flashes of color in addition to the pinpoint fire when tilted just slightly.

Interestingly, the stones in my wedding band tend to look best straight on or with a very slight tilt, but when tilted at say 30 or more degrees I can see dead spots near the corners and the brilliance is not as good

Also, with the stones in my wedding band the fire does not seem to change from pinpoint to broad flashes for the most part. I have to tilt the band just right to catch the broad flashes of light whereas with my e-ring I can move my hand in any direction and get the broad flashes.

Finally, when I use the light that comes with the ideal scope to backlight my rings I think they look best head on, but just slightly (it may be because I like the optical symmetry and contrast in the stones though).

In sum, yes, I do notice differences in the performance of my stone(s) when viewed at different angles, but it's hard to say if the head on view is best. I can say for certain my stone looks best to me when it's moving instead of static, but it performs too differently when viewed straight on and tilted to say which is better--especially in different lighting.

I just asked my hubby his opinion and he said he prefers the stone when it's tilted and we're in bright light because he likes when I accidentally blind people with the reflection off the table

To each their own!

 

[researcher]

Now, I have a question. How is it that the cutters decide on the numbers/formula for their princess cuts (and the AGS charts) when, as a consumer, I''m told there''s no way to judge a princess cut by the numbers? This question has plagued me from the start. I understand that our inability to go by the numbers is due in part to the limited knowledge we''re given regarding the crown and pavilion angles, but that can be remedied by providing those numbers. If that''s not enough information and there isn''t a perfect tool for evaluating princess cuts, how can AGS accurately grade their performance? What are the employees at AGS being taught about princess cuts and their performance, and why can''t we the consumers be given the same information?

Finally, is there a way to at least somewhat predict the light performance of various princess cuts? For instance, does the number of pavilion chevrons impact the brilliance of the stone? Does the number of facets impact the fire such that some stones will exhibit broader flashes while others will exhibit more pinfire in most lighting conditions? It would be extremely useful for consumers to have more information similar to the advice given on PS that, "as table size increases, the chances of getting a good performer decrease". Such helpful hints are desperately needed when considering stones on the internet, especially when we''re told the tools available don''t begin to tell the whole story.

I hope I''m making at least a little sense in this post. Not only is it really late, but my head always spins when these princess cut discussions arise. I can''t tell you enough how much I enjoy them though--even when the discussions are way over my head

 

[oldminer]

The thread is now encompassing a large range of good information. It would be a challenge to organize it, but there is valuable information here in several posts.

I can't even begin to address some of the thoughts expressed. I have spent a long time in this subject, but certainly do not have all the answers. No one does. I liked the analogy of a diamond being a sculpture which should be appreciated from various angles. Once we buy a diamond, we do look at it from all angles and appreciate it. I believe that in choosing a diamond one must screen out the poorer ones first and this can be done face-up. The ImaGem process screens out those stones which an astute person would not select if they had expert eyes. Most consumers don't know what to look for, so having diamonds pre-graded with such a technology tool would make the process much safer. Dealers realizing the need consumers have for better cut, would then cut more diamonds of higher performance to satisfy demand. In the end, we'd have mostly well cut diamonds in the better quality ranges. I see this as a plus for consumers and the trade, in general.

Surely, one can enjoy the look of fancy shaped diamonds by moving and tilting them to their heart's content. However, I'd suggest that one can grade their Light Behavior and find the best looking group of stones by only examining them face-up. No one in their right mind would cut a diamond with a poor face up view and an superb tilted view. It would be financially a disaster and who would buy such a stone. No doubt, some of the finest diamonds have excellent tilted views, but they should also enjoy an have an equally fine or better face-up view. What we feel is that diamonds which face-up with lower performance than tilted do not belong in the top category.

If some group wants to make their top category include more stones by allowing somewhat less face-up performance with higher tilted performance would widen the category. That would favor cutters, but not be beneficial to consumers. I believe this is what is taking place and don't think it is a good move for the trade. Consumers are far less trusting and uneducated today than 50 years ago. From what I see, we need an honest system that discriminates performance in rather tight categories. The best looking diamonds are found in the Excellent + category yet we see no degradation in visual performance for Excellent, Very Good+ and Very Good. This follows the general standards of grading diamonds. Think of the IF category for clarity. We see no degradation with our eyes at VVS1, VVS2, VS1 and usually VS2. In Color, D is tops, but our eyes don't really detect E,F or sometimes G in the face-up position of well cut stones.

Not everyone will opt for an Excellent+, just as they won't opt for IF or D. If price related to visual quality, then the price for the IF D would be the same as VS1 F. This is not the case at all. We feel the same ought to apply to cut grading / Light Behavior. Our discrimination and grading should be reasonably fine, rather than coarse.

Again, the more variables you toss into the soup of grading Light Performance, such as tilting, the more error prone and worthless it becomes. If you want honesty and an understandable system, you won't include tilting into the grade. That does not mean you cannot say to someone, tilt the stone and appreciate its extraordinary beauty in any direction, but it should not be how a gemological grade is set.

Hope I am making sense to those of you patient enough to read this long posting..... I do appreciate all the courtesy here and the excellent give and take.

 

[strmrdr]

Look at these images for a bit.
The ones on the ends score better on B-scope and will on other pixel counters.
BUT
in smaller sizes a majority of consumers prefere the middle one.
once they hit much over 2ct the other looks take over.

 

[Rhino]

Just catching up here today.

Good response Paul. I''d like to add some thoughts.

Date: 6/12/2006 5:46:24 PM
Author: Paul-Antwerp
I wished that we had a tool or even an idea about one tool. It would make education and resulting sales a lot easier.

At this point in time, one can only use a series of tools, while making sure not to put too much weight in one single tool:

1. As a start, getting stones with an AGS-grade on light performance is fantastic. One does not get these grades by accident. Therefore, do not expect to find them in non-AGS-graded stones. It will be like finding a needle in a haysteack (spelling?).

Ain''t that the truth.

2. The H&A-view of a princess will show you its symmetry. Do not look for arrows, or hearts, look for a symmetrical built-up. Symmetry is always the key to light performance. I am sure that non-symmetrical light performance exists, but cutting planned asymmetry is extremely difficult and unlikely.

True. The only addendum I would add to your statement is "Symmetry combined with proper proportions is always the key to light performance". I''ve seen some highly symmetrical stones that were duds.

3. In the Idealscope, look for a nice contrast-pattern, without aiming for the total absence of light leakage.

4. In the ASET-scope, look for that same pattern, and a good division of reds, greens and blues.

5. When assessing a Sarin- or Helium-measurement, make sure to get all the figures, and forget about the averages. In princess-cuts, cutters are used to ''following the rough'' and ''cutting for weight''. A P1-angle of 64° might mean two angles of 66° and two of 62°. If possible, you should check all the angles, starting with the pavilion, since this is the foundation of light performance.

6. Do not forget the ultimate tool: your eyes.

And an AMEN to that. In response to Neil''s question I think the ultimate tool of course is the eyes, however for mapping purposes, a mutilcolored reflector such as the ASET is a great mapping/logging tool. Better yet, a Helium scan with accompanying Gem Advisor model and you have your ASET, every facet measurement, etc. My .02c.

Peace,

 

[Rhino]

Hi Dave,

I''m sure folks have answered what I''m about to so I''m just paging along here.

Date: 6/14/2006 12:58:11 PM
Author: oldminer

Dr. Aggarwal has replied as follows:

I am not sure what is the definition of ''direct'' and ''diffused'' light used by ''Rhino''.
The lighting environments used by ImaGem and Brillianscope are different in many ways. The angle of incidence of the light with respect to the table with ImaGem is as broad or more broad than the angles used in Gemex. To establish statistical correlation, one must have a representative and a large sample. A true study comparing the performance of different systems on light behavior has yet to be conducted.

For clarification.

Direct light means there is nothing between the light source and the object it is striking. Simple example. Standing outside in the sunlight, with a direct path between the sun and the diamond it is illuminating is "direct light". It doesn''t necessarilly have to be the sun ... it can be any type of direct lighting that is not filtered or diffused.

Diffused. Say the sun goes behind a cloud. That direct light from the sun is now being diffused through a cloud. It can also be referred to as "filtered light". Diffuse light contributes to an entirely different appearance in diamond. One in which the elements of contrast are more plainly observed.

A general rule of thumb with regards to these lighting environments ...

Direct light will generally induce the optical characteristic of fire within diamond. The stronger the light source the more intense the reflections will be.

Diffuse light will generally induce the optical characteristic of brightness within diamond. Same principals as above.

Scintillation can be observed in all lighting environments. The nature of that scintillation will change depending on the lighting.

I made a video on this subject Dave which demosntrates all these things that I know you''d be interested to see and about all of this is covered in our newly revised cut tutorial.

Regarding this statement...

The lighting environments used by ImaGem and Brillianscope are different in many ways. The angle of incidence of the light with respect to the table with ImaGem is as broad or more broad than the angles used in Gemex. To establish statistical correlation, one must have a representative and a large sample. A true study comparing the performance of different systems on light behavior has yet to be conducted.

Regarding the lighting environment ... is the Imagem''s tests done in direct or diffuse lighting?
If they are different in many ways, exactly how are they different?

Regarding the statement "The angle of incidence of the light with respect to the table with ImaGem is as broad or more broad than the angles used in Gemex."

I have given the exact spectrum with which the Bscope is taking its measurements from. This answer from Dr. Aggarwal is very generalized. "Is as broad or more broad" doesn''t tell me anything. If Dr. Aggarwal understands things like "the angular spectrum" such as employed by AGS through the studies of scientist Dr. Jose Sasian, this is the answer I''m looking for. Over what angular spectrum does the Imagem concentrate light on to get its readings? Dave, this isn''t hard to figure out. If you drop me a phone call I can tell you how to do it.

Date: 6/14/2006 12:58:11 PM
Author: oldminer

What can be demonstrated using ImaGem''s technology is that light performance
of AGS 0 stones varies significantly on the measures of light behavior
developed by ImaGem for Princess cut. Furthermore, the differences in the
light behavior of AGS 0 Princess cut are large enough that they can be
easily perceived in their images. These are important findings and require
careful review of the cut grading system developed by AGS and the impact
direct measurement of light behavior is likely to have on the cutting,
grading, selling and valuation of Princess cut stones. For example, are
there non-zero AGS stones which have better light performance than AGS 0 cut
stones? One might pose several other questons.

We are finding this too but with regards to direct lighting tests. Oftentimes appearance in diffuse light will corellate but not always as there are stones that can have great fire but not great brightness due to poor contrast.

Date: 6/14/2006 12:58:11 PM
Author: oldminer

The direct, diffused lighting question is answered in this way.

ImaGem uses direct, diffused lighting.
Gemex uses direct, non-diffused lighting.

Lighting types may be indirect or direct. Each of these two types may be non-diffused or diffused.

If a question about this remains, please let me hear from you.

Sure. I have to chuckle when I read this answer Dave... Imagem uses direct, diffused lighting.

It sounds like the good Doctor is trying to cover all his bases in this answer but is failing to realize that lighting is generally either direct or diffuse. Ie. The BrillianceScope is doing its readings with a view to direct light. The 5 primary positions. The 6th light view is not taken into account for the measurements but is a type of diffused light view. Ie. Direct light is hitting a white diffuser dome and the light is being filted onto the diamond in that view.

The 2 questions I''d really like answers to are.

Is the Imagem taking its measurement by direct lighting or diffuse lighting?
What is the angular spectrum with which it is taking its measurements from?

I appreciate the help and cooperation Dave and for looking into these matters. Please express my gratitude to Dr. Aggarwal for his help in trying to answer too. All of us, at least in this community are interested to know these things.

Kind regards,
Jonathan

 

[Rhino]

Date: 6/14/2006 4:42:26 PM
Author: oldminer
Paul: We have a high precision 12 inch ruler and are measuring nothing exceeding 10 inches. I think we have the tool for the job. The entire process has been controlled so that no diamond can possibly measure greater than the ability of the device to discriminate performance. Performance does not totally equate to every individual''s human preference of what is most beautiful, but we believe that all the most beautifully performing stones will score in the upper category of the system proposed. Some consumers may select a diamond from a lower category, but this may be not only personal preference, but a budgetary or weight preference. Whatever they wish to purchase, knowing all the facts, is fine with all of us. I would think engineers will probably go for the highest performance numbers, but most people who are less number conscious will use their eyes, wallets and other faculties to make choices more in moderation yet still in the top two or three categories.

Nothing to date with thousands of test stones measured indicate any problem with this ability to measure brilliancy, sparkle and intensity as defined by ImaGem. I would further warn people reading our little discussion that we personally get along nicely and that this is a discussion of importance, not a minor fight.

I am far more concerned about the world of diamonds being upset with utilizing measuring devices that don''t measure all that well, then rounding off these somewhat faulty results and then giving total reliance on a final prediction based on poor science, poor measures and quasi-visual results to confirm the whole mess.

Hi Dave,

From the stones I have sent to you for testing (the 3 princess cuts from last years discussion), there was direct corellation between Bscope and Imagem results. This leads me to believe that the Imagem is doing its testing in direct light.

To date, the only digital light analyzer I''ve encountered which takes measurements in diffuse light is the Isee2. However, late last year Isee2 made changes to their machine to now, not only measure light in diffuse lighting for brightness but also added the ability to measure fire. My plate has been full this year so I have not had the chance to carefully study the changes that went into the recent developments of the Isee2, but from what I have been able to see and compare so far it appears the Imagem is taking its measurements in direct light as opposed to diffuse. Since I don''t have the machine or have graphics to see inside of it, this is only speculation on my part but would appreciate clarification and confirmation of that.

What I am going to do, is see if I can drum up a stone that has very good/excellent fire but fair/poor brightness and send it over to you for testing to see how it fairs. That''ll also help determine the strengths/limitations of the technology. Too bad you can''t send it here for a week or 2 so I can play with it.

Peace,

 

[Rhino]

Date: 6/14/2006 8:40:39 PM
Author: He Scores

RE: Light study of face up views.

It''s been my contention all along regarding ''look grades'' that the diamond is a piece of sculpture. Name me one other piece of sculpture in the world of art that is viewed in only one position, in a certain lighting environment.

Diamonds of all shapes are viewed by the wearer in all sorts of directions.

Frankly, I felt most of the studies limiting the observer or observing instrument to the ''face up'' view was.....limiting.


Bill Bray
Diamond Cutter

Good point Bill. Which is why I always prefer the visual assessment over any technology that limits views.

 

[RockDoc]

I''ve looked at the results of Imagem which corresspond to a numberical basis, but unlike the Brilliance Scope I haven''t yet seen any graphical imaging such as the B Scope is capable of producing.

All of us that have the B Scope Analyzer don''t totally rely on the blue bar ratings, but rather the images.

I do agree that machines do not have human eyes, nor the resolution of the human eye, as per your statement, however I am unaware if the Imagem does produce imaging that can be analyzed as well, and if this is reported to consumers in the report.

If it is just numbers, where there is a basis of how and what the numbers mean is really relevant. It appears that divulging the technology succinctly, is missing from what you and Dr. Aggarwal disclose. Can you be a lot more specific. so Imagem isn''t considered to be a black box mystery item?

You mentioned a grading range of from "good" to Excellent. If good is the lowest grading a stone gets, what about the real duds, that are more properly described as Fair, Poor, Low etc.? Do that have a place in your resultant gradings?

When we do see a "dead" stone, saying its "good" indicates that the result could be misinterpreted. I will grant you that the industry is certainly on a path to cutting livlier goods and better proportioned stones as consumers become more aware, so maybe we are not seeing real "dogs" for testing, but I do know they are out there. It appears to me that a rating of just GOOD, isn''t descriptive enough for such goods.

Can you post some of the graphical images here and is there a report issued by the IMAGEM that does show them? While I certainly have not looked at all the reports on your site it is capable of, I am wondering if a more detailed report other than the small "credit card" sized one exists.

As for the direct/diffused lighting subject, the B Scope does show a diffused light image which is the image 6 on their report.

Rockdoc

 

[Rhino]

Date: 6/14/2006 4:07:22 PM
Author: researcher
I just wanted to say I''m learning a TON from this thread. Please keep up the discussion! It''s nice to have the true experts considering the consumers'' perspective and tools for performance measurement

Thanks for chiming in researcher. It is after all your perspective that counts most of course. The goal of all these manufacturers of technologies is to better help communicate which diamonds will appear brightest/most fiery to your eyes. It''s not the case in every single instance but from our personal experience using them it does happen to be the case in most circumstances. As Paul pointed out earlier in this thread heavy emphasis should not be placed on any single technology but the overall results of all the data that can be acquired. If you have any specific questions about a technology used that I''m familiar with feel free to ask and I''ll do my best to help.

Kind regards,

 

[Rhino]

Date: 6/15/2006 1:50:05 AM
Author: JohnQuixote

Date: 6/12/2006 5:46:24 PM
Author: Paul-Antwerp

2. The H&A-view of a princess will show you its symmetry. Do not look for arrows, or hearts, look for a symmetrical built-up. Symmetry is always the key to light performance. I am sure that non-symmetrical light performance exists, but cutting planned asymmetry is extremely difficult and unlikely.

To further this point, while we shouldn’t expect to see hearts, arrows or bunnies, we can expect that aiming mirrors more precisely at each other will have a positive result... Strm, at the AGS presentation Peter stressed that, in emerald cuts, pointing the corners at each other improves performance.

Chalk another one up for the optical symmetry pundits (we’re the ones in the white hats).

Thanks for sharing this John. Now ... I could''ve sworn I saw bunnies in the last Asscher we inspected.

 

[Rhino]

Date: 6/15/2006 2:06:21 AM
Author: JohnQuixote
Found the thread I had in mind.

Brian made these comments back in 2003, advising against princess light performance judgments using just proportions info / ideal-scope.

Date: 12/1/2003 2:55:26 AM
Author: BrianTheCutter

Light leakage evaluation should be combined with a live examination, especially for fancies, and that is what we do. One should remember diamonds are not worn in a static position and the human eyes look at the diamond from different angels which will change the way one would see the stone in a static upright position...

Symmetry is not always a guarantee of the best beauty in fancies - small amounts of light leakage
are good for contrast and character. Ask Serg at MSU I think he would agree!

Original thread here (a blast from the past!)

More people appear to understand this now. I’ll forward this conversation to Brian; it will make his day to see the great discussion in this thread.

I am finding this as I''m exploring certain combinations of steep/deep rounds as well which are stones I had once not considered.

 

[Rhino]

Awesome ASET image researcher!

Date: 6/15/2006 5:34:14 AM
Author: researcher
Now, I have a question. How is it that the cutters decide on the numbers/formula for their princess cuts (and the AGS charts) when, as a consumer, I''m told there''s no way to judge a princess cut by the numbers? This question has plagued me from the start. I understand that our inability to go by the numbers is due in part to the limited knowledge we''re given regarding the crown and pavilion angles, but that can be remedied by providing those numbers. If that''s not enough information and there isn''t a perfect tool for evaluating princess cuts, how can AGS accurately grade their performance? What are the employees at AGS being taught about princess cuts and their performance, and why can''t we the consumers be given the same information?

For starters, most Sarin''s will not give all the numbers necessary to make an accurate judgement.

A 3d file or better yet a Helium scan would be needed to accurately assess a princess cut by numbers but even then the expert examining the numbers would have to know how those numbers translate to face up appearance. Most people in this industry have not studied the numbers in princess cuts enough to be able to do this. This has been one of the primary studies we''ve been doing over the course of time and I am familiar with many proportion combinations in princess cuts and how they''ll face up but there are many more I''m not familiar with. I do know the cherry sets though and that''s what''s important to me.

A much easier task, than learning proportion combinations is acquiring a 3D scan of the stone and importing it into DiamCalc. In the software we can virtually see the stone and how it will perform in various environments. This requires a study of the light views it offers, even making some custom light views if necessary, and corelating how these views relate to real world environments. Gem Advisor files generated from 3D models is extremely valuable data IMO.

In fact, it is these 3d models that AGS generates (I believe they are using the Sarin for this, although they do own a Helium as well) and loads into their own ray trace software which basically generates the grade. This is how they are accurately grading performance and determing the light performance grades.

There are certain limitations to reflector based technologies (including ASET) and ray trace software which have been adopted into the AGS system which some view as a weakness because not always do the results corellate to human observation. I would say they do in most circumstances but not all. No system, IMO is flawless.

Finally, is there a way to at least somewhat predict the light performance of various princess cuts?

With a live analysis there are a few options.

1. Send the stone to AGS.

2. Acquire a 3D model of the stone allowing an expert to examine the proportions and symmetry and using that model in conjuction with DiamCalc. With that you can view and/or share the images shown via Gem Advisor including its ASET and Red Reflector results. A knowledgeable gemologist familiar with ASET results can interpret, pretty easily how the stone will appear. Red reflectors by themselves are ok but don''t always tell the whole story.

3. BrillianceScope and/or Imagem analysis is excellent information to attain coupled with a visual examination for confirmation.

4. Have an appraiser and/or AGS jeweler with the AGS PGS software run the model to determine AGS light performance score.

For instance, does the number of pavilion chevrons impact the brilliance of the stone? Does the number of facets impact the fire such that some stones will exhibit broader flashes while others will exhibit more pinfire in most lighting conditions?

Absolutely. It''s interesting you should ask this because we just recorded footage of this type of comparison comparing various flavors of AGS Ideal princess cuts. One of those comparisons being the difference in chevrons on the pavilion. I wasn''t going to publish this until I finished another project I''m currently working on covering the differences in appearance between GIA Ex vs GIA VG''s but later on when I get up to the store perhaps I can snap off a comparison shot for ya and show you how they look in a static position. Currently, within the realm of AGS Ideal princess cut there are more stones available of the 3/4 chevron type than there is the 2 chevron type. Both are beautiful yet vary in their appearance. The fewer chevrons contribute to a braoder flash look while the 3/4 chevron variety contribute to higher pin flash.

It would be extremely useful for consumers to have more information similar to the advice given on PS that, ''as table size increases, the chances of getting a good performer decrease''. Such helpful hints are desperately needed when considering stones on the internet, especially when we''re told the tools available don''t begin to tell the whole story.

Agreed.

I hope I''m making at least a little sense in this post. Not only is it really late, but my head always spins when these princess cut discussions arise. I can''t tell you enough how much I enjoy them though--even when the discussions are way over my head

Your post makes 100% sense researcher. I hope my answers are worded such that you can easily understand them. If you need clarification on anything don''t be afraid to ask.

Kind regards,

 

[oldminer]

Rhino, Imagem uses direct diffused lighting. Dr. Aggarwal does not wish to give angles of incidence at the present time, but anyone who has seen the Verigem can readily deduce a lot about lighting angles. Obviously, if the face up view is what is essential, then the angle does not need to be as broad as to mimic a tilted stone. Dr. A. certainly knows that lighting is either diffused or non-diffused. He also knows that either can be direct or non-direct. We have told you that the lighting is direct diffused. No kidding, that's the story with the lighting. It isn't rocket science and he is not hiding the fact. BrillianceScope is using direct, non-diffused lighting. That's the difference.... A diamond with excellent fire and poor brightness would be of interest for testing. Glad to do it. I sort of doubt it would be a knockout pretty stone.

"In fact, it is these 3d models that AGS generates (I believe they are using the Sarin for this, although they do own a Helium as well) and loads into their own ray trace software which basically generates the grade. This is how they are accurately grading performance and determining the light performance grades. "
How can you say these predictions are "accurate"? The measures for Sarin are not accurate, precise or even repeatable to a large extent. The Helium is better, but the larger the diamond, the more error it creates. It is not using the best camera for the purpose. Helium is a lot better than the current Sarin.....No problem with that.
xxxxxxxxxxxxxxxxxxx



RockDoc, you need to read the definitions of Brilliance, Sparkle and that ImaGem uses. They are quite simple measures and the machine is not a black box because the pixel measures and grey scale measures of pixels are hardly mystery items for science. The grades associated with overall Light Behavior run from Excellent + to Commercial. A total of 8 potential grades, so GOOD is #5 or the 8. We are finely dividing the upper ranges, as we all currently do with clarity and color. The lower ranges are far wider since it matters so much less.
xxxxxxxxxxxxxxxxxxxxxxxxxxx

Stmrdr, I sure don't know how you know what Imagem will do. There is a lot of programming behind the scenes to make the end result sensible. Others reading here may actually believe you "know" what Imagem will read on your three Asscher cut examples. We don't know ourselves and have not tested this. I sort of think you may be speculating.

Others reading this thread may give your statements credence without being aware of the lack of any proof. Admittedly, logic suggests things, but logic is not always correct guidance. The testing is the proof.

Do you have a series of round, princess or marquise shaped diamonds for us to test for you? I'll give you the dealer discount price and even pay 1/2 your air-fare (coach, of course). Lunch and dinner included. too. It would be excellent to meet and know you better. Do I have courage?

xxxxxxxxxxxxxxxxxxxxxxxx

Everyone: The eyes are the best way to select a final choice. They are not the most efficient way to measure performance or screen large numbers of diamonds. The trade is rushing to adopt predictive tools based on very questionable physical measurements and theoretical results. It may be good to take a deep breath and look at the possible alternatives. Just because the big guys have rushed into their brand of new technology does not make them correct. They are large, powerful entities. We have seen how they can be corrupted by other powerful players. Not often, we hope, but we know it can occur because it has occurred. Quoting the dictates of GIA, AGS or Gemex won't automatically give you the best possible choices although they are well positioned. As time goes on there will be more to discuss, no doubt.

I intend to have a Verigem unit with me at the GIA Symposium. If anyone attending has stones to test run, I would be glad to demonstrate it. I'll post my cell phone number near the time of the event in August so I can be contacted a the Symposium. Of course, my email and messaging will be working at the Hyatt.

 

[strmrdr]

Dave, yep speculation but you gave the math on how it gets each score and unless you aren''t being truthful about the math its easy to figure out what it will score in relative terms.
Reverse engineering the light grade part of it is a no-brainer which is why he wont be specific on the lighting model.
The hard part is access to enough diamonds to run thru it to make it meaningful.

 

[oldminer]

Storm:

People who know me would not say anything much about me not being truthful. A little geeky and weird, but honest to a fault, more than fair to nearly everyone, and open-minded. Since you don''t really know me, I won''t take offense, but your suspicion is misplaced.

Anyone who looks inside the top glass of a Verigem sees the entire lighting model for themselves. There is nothing all that tricky going on with the lighting. The part of the ImaGem system that is "tricky" or a better word would be "complex" is the programming (intellectual property), the Imagestatistics, the human perception studies and the thousands of diamonds that were processed by a third party investor of great competance. Jonathan has seen the machine and several others here have seen it. Some have peered inside from the top and seen the lighting. No one was amazed or shocked. It does not take a strange lighting formula to make diamonds perform. That''s one of the characteristics of diamonds, its easy to make them pretty.

Well, I tried to be convincing. Possibly someone else will be more understanding. I''d like to convince you too at some point. It is good to be a devil''s advocate in moderation.

 

[strmrdr]

Hey Dave,
Glad you didnt take offence..
I see now I worded that badly I wasnt implying you werent truthful.

What I was trying to say that is if the calculations are done as you have passed them on as being done then it is pretty easy to figure out how they will score relative to one another.

I agree the database is the valuable part of the imagem program.

Im curious convince me of what?
That imagem gives useful information or that imagem scores give all the answers?
For rounds the first part may be true but the second will never be true no one machine or technology can give all the answers.

My position is that with asschers and most likely princess cuts pixel counters will never be able to rank them.
I believe with rounds they can and do so to some extent.

 

[oldminer]

"That imagem gives useful information or that imagem scores give all the answers?"

ImaGem would never claim to give all the answers. All tools for categorization of beauty or appearance are laid over average data on human perception. ImaGem would claim that the information provided is sufficient to make smart choices and that their ranking system correlates well with human graders while never tiring or becoming inconsistent as humans often do. Since we don't claim to determine maximum beauty, but only determine performance, Light Behavior, the consumer and the trade can still use their own judgment about what is most beautiful. We sure can eliminate the ones that don't belong in the beauty contest with our eyes or by machine.

Technology evolves and improves to fit our needs. When cars came to be in the early 1900's, folks said, "wait a minute, how will you get over a fence with such a vehicle. Our horses are far better."
Well, cars still can't jump fences, but where are the horses on the roads today? Technology won out in transportation by car over the horse. When telephones were introduced people said they were worthless and telegrams and letters worked just fine. The first phones were installed in Wall St and gave stock brokers a huge edge over their old fashioned competitors who didn't adopt them. In short order, the entire stock market system became wired. When was the last time you wrote a letter when you could use email or the cell phone? Letters may be more personal and romantic but most correspondence is for other reasons. Remember the days of perfumed letters? But, technology is beating the heck out of typewriters and hand written letters.

I expect technology will overcome its possible present shortcomings in deterining the best group of diamonds of any shape. We know a lot about human perception and are making machines ever more smart in this regard. If the human mind can tell us which princess cut or which Asscher cut looks the best, then we can train devices to do it with higher accuracy and consistency that people can. Maybe we won't do it tonight, but it will happen in a very short time, not years.

 

[Rhino]

Good morning Dave. Totally concur with your comments below. I agree ... the face up appearance should be #1 priority since that is the view seen most. One thing to bear in mind however with AGS'' system is it is not discounting the face up view at the expense of the tilt metrics. Ie. if a stone scores good on the tilt metrics but suffers in the face up view, it''ll take the hit. If anyone has heard different please chime in but this is my understanding.

Peace,
Jon

Date: 6/15/2006 8:02:00 AM
Author: oldminer
The thread is now encompassing a large range of good information. It would be a challenge to organize it, but there is valuable information here in several posts.

I can''t even begin to address some of the thoughts expressed. I have spent a long time in this subject, but certainly do not have all the answers. No one does. I liked the analogy of a diamond being a sculpture which should be appreciated from various angles. Once we buy a diamond, we do look at it from all angles and appreciate it. I believe that in choosing a diamond one must screen out the poorer ones first and this can be done face-up. The ImaGem process screens out those stones which an astute person would not select if they had expert eyes. Most consumers don''t know what to look for, so having diamonds pre-graded with such a technology tool would make the process much safer. Dealers realizing the need consumers have for better cut, would then cut more diamonds of higher performance to satisfy demand. In the end, we''d have mostly well cut diamonds in the better quaility ranges. I see this as a plus for consumers and the trade, in general.

Surely, one can enjoy the look of fancy shaped diamonds by moving and tilting them to their heart''s content. However, I''d suggest that one can grade their Light Behavior and find the best looking group of stones by only examining them face-up. No one in their right mind would cut a diamond with a poor face up view and an superb tilted view. It would be financially a disaster and who would buy such a stone. No doubt, some of the finest diamonds have excellent tilted views, but they should also enjoy an have an equally fine or better face-up view. What we feel is that diamonds which face-up with lower performance than tilted do not belong in the top category.

If some group wants to make their top category include more stones by allowing somewhat less face-up performance with higher tilted performance would widen the category. That would favor cutters, but not be beneficial to consumers. I believe this is what is taking place and don''t think it is a good move for the trade. Consumers are far less trusting and uneducated today than 50 years ago. From what I see, we need an honest system that discriminates performance in rather tight categories. The best looking diamonds are found in the Excellent + category yet we see no degradation in visual performance for Excellent, Very Good+ and Very Good. This follows the general standards of grading diamonds. Think of the IF category for clarity. We see no degradation with our eyes at VVS1, VVS2, VS1 and usually VS2. In Color, D is tops, but our eyes don''t really detect E,F or sometimes G in the face-up position of well cut stones.

Not everyone will opt for an Excellent+, just as they won''t opt for IF or D. If price related to visual quality, then the price for the IF D would be the same as VS1 F. This is not the case at all. We feel the same ought to apply to cut grading / Light Behavior. Our discrimination and grading should be reasonably fine, rather than coarse.

Again, the more variables you toss into the soup of grading Light Performance, such as tilting, the more error prone and worthless it becomes. If you want honesty and an understandable system, you won''t include tilitng into the grade. That does not mean you cannot say to someone, tilt the stone and appreciate its extraordinary beauty in any direction, but it should not be how a gemological grade is set.

Hope I am making sense to those of you patient enough to read this long posting..... I do appreciate all the courtesy here and the excellent give and take.

 

[Rhino]

Hey Rock,

Just typing out loud here ...

Date: 6/15/2006 1:06:08 PM
Author: RockDoc

I've looked at the results of Imagem which corresspond to a numberical basis, but unlike the Brilliance Scope I haven't yet seen any graphical imaging such as the B Scope is capable of producing.

All of us that have the B Scope Analyzer don't totally rely on the blue bar ratings, but rather the images.

True. Which is why, as I look at the overall package provided by both machines, favor the Bscope. As far as I have been able to tell the Imagem takes its reading from one light view. Similar to the original LambdaSpec machine used by DPL labs which was the predecessor to the BrillianceScope. Even at that I am finding corellations between the numerical results between the 2.

I do agree that machines do not have human eyes, nor the resolution of the human eye, as per your statement, however I am unaware if the Imagem does produce imaging that can be analyzed as well, and if this is reported to consumers in the report.

Last I saw the Imagem produces a monchrome image which is displayed on the ID card generated from the machines software. Perhaps Dave can post an example or I can scan one in.

If it is just numbers, where there is a basis of how and what the numbers mean is really relevant. It appears that divulging the technology succinctly, is missing from what you and Dr. Aggarwal disclose. Can you be a lot more specific. so Imagem isn't considered to be a black box mystery item?

The answers to my questions from Dr. Aggrawal would be appreciated in helping to better understand the technology as there are question marks surrounding it. Dave, please don't take our questions/suggestions in the wrong spirit. We are for a better and more full understanding of exactly what it is the Imagem is doing. Ie. how it is testing it's stones and from what angular spectrum it is doing so. There may be advantages to it over the BrillianceScope but without the answers to our questions we can only, at best, speculate.

You mentioned a grading range of from 'good' to Excellent. If good is the lowest grading a stone gets, what about the real duds, that are more properly described as Fair, Poor, Low etc.? Do that have a place in your resultant gradings?

Good question.

When we do see a 'dead' stone, saying its 'good' indicates that the result could be misinterpreted. I will grant you that the industry is certainly on a path to cutting livlier goods and better proportioned stones as consumers become more aware, so maybe we are not seeing real 'dogs' for testing, but I do know they are out there. It appears to me that a rating of just GOOD, isn't descriptive enough for such goods.

Can you post some of the graphical images here and is there a report issued by the IMAGEM that does show them? While I certainly have not looked at all the reports on your site it is capable of, I am wondering if a more detailed report other than the small 'credit card' sized one exists.

Good questions! Hopefully Dr. A and Dave will accommodate.

BTW, if there are any questions you have as well Dave, please feel free to fire away. I don't want you to feel, in any sense ganged up on or overwhelmed by our questions but with Bill, strm and I ... we're 3 techy geeks when it comes to this stuff.

Ex/Ex/Ex

 

[strmrdr]

Date: 6/16/2006 7:44:17 AM
Author: oldminer
''That imagem gives useful information or that imagem scores give all the answers?''


ImaGem would never claim to give all the answers. All tools for categorization of beauty or appearance are laid over average data on human perception. ImaGem would claim that the information provided is sufficient to make smart choices and that their ranking system correlates well with human graders while never tiring or becoming inconsistent as humans often do. Since we don''t claim to determine maximum beauty, but only determine performance, Light Behavior, the consumer and the trade can still use their own judgment about what is most beautiful. We sure can eliminate the ones that don''t belong in the beauty contest with our eyes or by machine.


Technology evolves and improves to fit our needs. When cars came to be in the early 1900''s, folks said, ''wait a minute, how will you get over a fence with such a vehicle. Our horses are far better.''

Well, cars still can''t jump fences, but where are the horses on the roads today? Technology won out in transportation by car over the horse. When telephones were introduced people said they were worthless and telegrams and letters worked just fine. The first phones were installed in Wall St and gave stock brokers a huge edge over their old fashioned competitors who didn''t adopt them. In short order, the entire stock market system became wired. When was the last time you wrote a letter when you could use email or the cell phone? Letters may be more personal and romantic but most correspondence is for other reasons. Remember the days of perfumed letters? But, technology is beating the heck out of typewriters and hand written letters.


I expect technology will overcome its possible present shortcomings in deterining the best group of diamonds of any shape. We know a lot about human perception and are making machines ever more smart in this regard. If the human mind can tell us which princess cut or which Asscher cut looks the best, then we can train devices to do it with higher accuracy and consistency that people can. Maybe we won''t do it tonight, but it will happen in a very short time, not years.

Interesting Dave I cant find anything I strongly disagree with in it.
I do think great machine looks grading is far off.
DiamCalc/helium and AGS virtual aset cames the closest right now I think.

Performance grading is available today from several machines however.
Its into this catagory that Imagem and B-scope fall.
But performance and looks are 2 different things.

 

[Rhino]

Hi Dave,

Just catching up here. Thoughts below...

Date: 6/15/2006 2:35:52 PM
Author: oldminer

Rhino, Imagem uses direct diffused lighting.

As I have been accommodating in explaining to Dr. A what I mean by both direct or diffuse lighting, would he be so kind as to explain what he means when he says direct diffused lighting and perhaps provide an example as I have done? That would be appreciated.

Dr. Aggarwal does not wish to give angles of incidence at the present time, but anyone who has seen the Verigem can radily deduce a lot about lighting angles. Obviously, if the face up view is what is essential, then the angle does not need to be as broad as to mimic a tilted stone. Dr. A. certainly knows that lighting is either diffused or non-diffused. He also knows that either can be direct or non-direct. We have told you that the lighting is direct diffused. No kidding, that''s the story with the lighting. It isn''t rocket science and he is not hiding the fact.

It may not be your intention Dave, but if Dr. A understands there is a difference between diffuse lighting and direct lighting, then turns around and says it''s direct diffused ... there are folks here, including myself that see this as skirting the issue and not being straight forward with his answer. He needs to explain what he means and be as forthcoming with the information and cooperation that we are doing here. It seems he doesn''t want to be as accommodating.

If he can at least explain what he means by "direct diffused lighting" and provide a real world example then perhaps we can get a better understanding of what he is talking about. I''m not trying to be difficult. All we are looking for is clarification Dave.

Since Dr. A refuses to share angles of incidence with us it''s also not fair of him to say "The angle of incidence of the light with respect to the table with ImaGem is as broad or more broad than the angles used in Gemex." If he is not willing so share facts about his technology, it''s not entirely fair to make these claims because with no answers ... we just don''t know.

As I have expressed, it is at this point speculation of whether this statement is true or not.

''In fact, it is these 3d models that AGS generates (I believe they are using the Sarin for this, although they do own a Helium as well) and loads into their own ray trace software which basically generates the grade. This is how they are accurately grading performance and determing the light performance grades. ''
How can you say these predictions are ''accurate''? The measures for Sarin are not accurate, precise or even repeatable to a large extent. The Helium is better, but the larger the diamond, the more error it creates. It is not using the best camera for the purpose. Helium is a lot better than the current Sarin.....No problem with that.
xxxxxxxxxxxxxxxxxxx

Good question. Let me revise that answer ... It''s as accurate as AGS'' PGS software says it is.

This actually points to a contention I personally have with the AGS Cut Grading system. I appreciate the science of reflector based technology and as you know have been photographing under these devices for years but from certain examples I have seen and learned from regarding the weakness of reflectors and how they corellate to human observation, I have found some of these limitations have been adopted into the AGS system regarding certain thresholds of shallow angled diamonds and other facet arrangments which do not face up as well as others. I''ve stated in other threads ... AGS is being conservative regarding the critical analysis of steep/deep combos and is a change they''ve made from the 96 cut grade system which I believe is for the better, however I am finding that GIA has being more conservative on the shallow angled combos. If one understands their respective approaches to cut grading they''d understand why.

RockDoc, you need to read the definitions of Brilliance, Sparkle and that ImaGem uses. They are quite simple measures and the machine is not a black box because the pixel measures and grey scale measures of pixels are hardly mystery items for science. The grades associated with overall Light Behavior run from Excellent + to Commercial. A total of 8 potential grades, so GOOD is #5 or the 8. We are finely dividng the upper ranges, as we all currently do with clarity and color. The lower ranges are far wider since it matters so much less.
xxxxxxxxxxxxxxxxxxxxxxxxxxx

I think Rock''s point here is that the lowest score shouldn''t be called "GOOD". "Good" may lead a person to believe that "good" is just fine when in fact it represents the lowest grade the technology assigns and is in fact a pooper.

Date: 6/15/2006 2:35:52 PM
Author: oldminer

Stmrdr, I sure don''t know how you know what Imagem will do. There is a lot of programming behind the scenes to make the end result sensible. Others reading here may actually believe you ''know'' what Imagem will read on your three Asscher cut examples. We don''t know ourselves and have not tested this. I sort of think you may be speculating.

Others reading this thread may give your statements credence without being aware of the lack of any proof. Admittedly, logic suggests things, but logic is not always correct guidance. The testing is the proof.

Do you have a series of round, princess or marquise shaped diamonds for us to test for you? I''ll give you the dealer discount price and even pay 1/2 your air-fare (coach, of course). Lunch and dinner included. too. It would be excellent to meet and know you better. Do I have courage?

xxxxxxxxxxxxxxxxxxxxxxxx

Half air fair?!?! Go for it strm!

Everyone: The eyes are the best way to select a final choice. They are not the most efficient way to measure performance or screen large numbers of diamonds. The trade is rushing to adopt predictive tools based on very questionable physical measurements and theoretical results. It may be good to take a deep breath and look at the possible alternatives. Just because the big guys have rushed into their brand of new technology does not make them correct. They are large, powerful entities. We have seen how they can be corrupted by other powerful players. Not often, we hope, but we know it can occur bcause it has occurred. Quoting the dictates of GIA, AGS or Gemex won''t automatically give you the best possible choices although they are well positioned. As time goes on there will be more to discuss, no doubt.

I intend to have a Verigem unit with me at the GIA Symposium. If anyone attending has stones to test run, I would be glad to demonstrate it. I''ll post my cell phone number near the time of the event in August so I can be contacted a the Symposium. Of course, my email and messaging will be working at the Hyatt.

I''ll cya at the show Dave. Hope to get some more answers.

Regards,

 

[jasontb]

Who is going to get me a guest pass for the GIA Symposium?

 

[Rhino]

In relation to the topic of this thread and inspired by "researchers" questions I thought I''d like to share what footage we have taken so far with regards to various AGS Ideal Princess cuts in the following video file. This video is not yet complete but 2 of the comparisons relate directly to questions researcher has asked with regards to

1. Differing appearances between AGS Ideal Princess cuts.
2. How the amount of chevrons on the pavilion alter the appearance.
3. How even stones with teh same amount of chevrons can be impacted by proportion combinations.

http://www.goodoldgold.com/videos/AGSIDEALPRINCESS.wmv

In this short clip the following comparisons are made.

1. The first comparison is of 2 Princess cuts that we have found differs in light performance yet both get the AGS Ideal grade for "light performance". The stone on the left is a GIA stone we hand selected but on the PGS software gets ideal for light performance. The stone on the right in the first comparison is an AGS graded Ideal. Both of these stones have the same number of chevrons on the pavilion (4 chevrons plus the mains), both have Ideal polish & symmetry. Here the BrillianceScope results concur with the eyes and since I''ve found corellations between Bscope and Imagem I would suspect would also rank accordingly on the Imagem as well. I''m pretty confident about that. In this post and the next are the Bscope results of each. In this post is the results of the stone on the left. Next post will be the stone on the right.

2. In the 2nd comparison we''re examining a 4 chevron princess alongside a 2 chevrong princess. Both ideal light performance yet differ in appearance.

Hope you enjoy this researcher and anyone else following along here. I am finishing another project but when that''s done plan on finishing up the completion of this princess comparisons of AGS ideals (for light performance).

Kind regards,

 

[Rhino]

Here are the Bscope results to the stone on the right in the first comparison.

 

[oldminer]

"As I have been accommodating in explaining to Dr. A what I mean by both direct or diffuse lighting, would he be so kind as to explain what he means when he says direct diffused lighting and perhaps provide an example as I have done? That would be appreciated."


Direct, non-diffused light is where the light comes from the source and falls directly onto the object. (Gemex)
Direct, diffused is where light passes through a filter or softener screen and falls directly on the object (ImaGem)

Non-Direct light is when light is reflected off of a mirror or a surface, such as an integrating sphere before it contacts the object.
Non-Direct diffused light has a filter or softener screen between the light and the object in addition to its being reflected.
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

When you seen the lighting ring move down in Gemex it is creating 5 differing angles of lighting. ImaGem simply uses quite the same total angles but makes it happen all at one time with a broad, direct-diffused light source at a distance that allows it to happen. This is a vital difference between the two units in spite of similar outputs. We also believe the use of diffused light is crucial to being able to measure what needs to be measured in a meaningful way.

xxxxxxxxxxxxxxxxxxx

We believe the predictive tools are excellent for cutters to use as a guide, but for jewelers and consumers who want to know with certainty how their one unique diamond performs, then one must measure the performance on that stone versus predicting by measuring parameters. Parametric screening works. I have been providing a crude but useful tool for all shapes since the late 1980's and have had my share of abuse from those who used to say "Parameters don't work.". Now, much of what we hear is "Parameters work", but of course we now measure more parameters with better precision. I have said they work all along, but are not the way to make a final judgment when there are better tools Whenever I have a diamond to weigh and have a scale, I put it on the scale. If I don't have a scale I measure the stone and use a calculator, specific gravity and a formula to ESTIMATE the weight. Anyone knows which is preferred and most accurate.... What I see is a new style of scale that will quickly weight the performance of light with the confidence one gets from a digital scale when one needs a carat weight.

 

[jasontb]

Light from a source reflected off a mirror at an object is non-direct? Why? Isn''t it the same as if the light source were pointed directly at the object?