Can anyone share feedback on Sarin''''s Cutgrade Analyzer?

[Regular Guy]

Although there seems to be a fair amount of discussion and controversy here concerning the use of one appliance vs. another on the evaluation of a diamond''s light performance, no one seems to question very much Sarin''s positive command for handling the task of measuring with good precision a diamond''s assorted angles.

So...and perhaps it is either new or expensive...but since I''ve not seen it mentioned here, what can we assume about the apparatus they describe here, where they describe that: "after accurately scanning and measuring your diamond with DiaVision™ (you can) use CGA to evaluate and grade the diamond’s light performance parameters – brilliancy, fire, and scintillation."

Has anyone here used it? Doesn''t anyone want to use it? Are there some issues about it that explain its lack of mention here?

Thanks in advance.

 

[JohnQuixote]

Ira,

The feature is available but not used much. "Light performance" evaluation is viewed as dubious even in machines like BS and Isee that do direct assessment. With Sarin’s given error any assessment is potentially off to begin with.

You know there is also a H&A function in DiaVision that works with a licensed DiamCalc component but the problem again is that the scans aren’t accurate enough (hence Helium being developed). If they were accurate enough we could rely less on live photos and use .srn files - although Sarin would have to upgrade their grading of the H&A... Brian spoke with them about the problem with their example of “perfect” hearts – the people at Sarin told him he was the first to comment on it not being perfect.

Additionally there are subtle factors currently being studied which may not yet be accounted for by any scanning devices. More on this as we get into '05.

 

[Regular Guy]

Hey, John, thanks for replying. So, if I understand you correctly, the "valuing" of the light performance is done "one-off," by way of interpretation of the angles just measured. It does not in any way take advantage that the diamond could still be right there, and analyzed directly? I think that''s what you''re saying.

From the text at the web site, is any interpretive framework teachable to it, would you think? For example, presume you have an established ideal cut. According to the sarin site, with this apparatus...

..."You can compare your diamond’s light performance to any reference stone of your choice, or to out-of-the-box reference stones corresponding to the best known cuts according to worldwide gem labs...."

Any guess as to what is being asserted...what system for brilliance, fire & scintilation is being applied either by default, or after a reference stone is provided?

 

[mdx]

Date: 1/31/2005 10:38:55 AM
Author: JohnQuixote

Ira,

The feature is available but not used much. ''Light performance'' evaluation is viewed as dubious even in machines like BS and Isee that do direct assessment. With Sarin’s given error any assessment is potentially off to begin with.

You know there is also a H&A function in DiaVision that works with a licensed DiamCalc component but the problem again is that the scans aren’t accurate enough (hence Helium being developed). If they were accurate enough we could rely less on live photos and use .srn files - although Sarin would have to upgrade their grading of the H&A... Brian spoke with them about the problem with their example of “perfect” hearts – the people at Sarin told him he was the first to comment on it not being perfect.

Additionally there are subtle factors currently being studied which may not yet be accounted for by any scanning devices. More on this as we get into ''05.

John

I think the Ogi system suffers the same problem as the others although I think this one works on ray tracing

Johan

 

[JohnQuixote]

Sarin does not have added mechanics for light performance assessment: It takes the proportions results we are all aware of and imports them into a licensed DiamCalc component to create a virtual model of the diamond. Then it uses that virtual model to estimate light performance. The problem is that the proportions measurements are already not consistent, so meaningful predictions and comparisons are not possible in an absolute sense.

Does that answer your questions, Ira?

 

[JohnQuixote]

Date: 1/31/2005 3:39

6 PM
Author: mdx
I think the Ogi system suffers the same problem as the others although I think this one works on ray tracing

Johan

Would that ray tracing be cousin to Sarin''s virtual modeling? Or does OGI actually have added mechanics for this, Johan?

 

[valeria101]

Date: 1/31/2005 1:3:25 AM
Author:Regular Guy

Although there seems to be a fair amount of discussion and controversy here concerning the use of one appliance vs. another on the evaluation of a diamond's light performance, no one seems to question very much Sarin's positive command ....

Well, I cannot agree 100% (just 90%

). There were several instances of Sarin data compard with AGS or alternative reports of proportions, with results falling all over the place. This is just the latest example, there are others on record.

Until present I fealt rather confortable with the idea just because these numbers are destined for the HCA and that tool is smart enough to give a wide bracket for these swings (that 2-or-below versus everything else threshold).


However... how about this little argument:

If HCA scores make a reliable envelope for Sarin & AGS proportions readings, that really means that those diamonds are cut with much more precission than Sarin&All can actually measure. So at some point on the production chain, more accurate numbers could be recorded than labs and sellers might obtain afterwards. If this is true... well, why is there any need to reinvent the wheel and produce faulty (and expensive) measurements in lieu of some better ones that never got recorded

 

[Regular Guy]

Date: 1/31/2005 5:23:51 PM
Author: JohnQuixote



Does that answer your questions, Ira?

Well, I was going to say..

which is it that gets things started really...is it the chicken or the egg?

But, Ana beat me to the punch.

Actually, I think she's got a good question that I hope someone will look at.

I've got another...in fact, John...see one more post from me

 

[mdx]

Date: 1/31/2005 5:26:51 PM
Author: JohnQuixote

Date: 1/31/2005 3:39

6 PM
Author: mdx
I think the Ogi system suffers the same problem as the others although I think this one works on ray tracing

Johan

Would that ray tracing be cousin to Sarin''s virtual modeling? Or does OGI actually have added mechanics for this, Johan?

John the devise is called a Firetrace and seems to have a scanning function. I think Gary (Cut Nut) mentioned to me that he had seen it at the Vagas show and it was based on a ray trace model.
I would assume it scans a 3d image and applies a software driven ray trace simulation.
Johan

 

[Regular Guy]

Looks like the Ogi option is represented here.

Not sure I''m entirely sensitive to the variability in differences suggested in output from the "direct measurement" approaches, vs those done my inference. Seems like even those doing direct measurement...may extrapolate in some measure...at least if analysis is included. (In contrast, things like IS, for my untrained eyes, are just not that easy to read and discern.) Isn''t measurement error always relatively operative?

 

[Rhino]

Date: 1/31/2005 5:23:51 PM
Author: JohnQuixote

Sarin does not have added mechanics for light performance assessment: It takes the proportions results we are all aware of and imports them into a licensed DiamCalc component to create a virtual model of the diamond. Then it uses that virtual model to estimate light performance. The problem is that the proportions measurements are already not consistent, so meaningful predictions and comparisons are not possible in an absolute sense.

Does that answer your questions, Ira?

Hi John,

You are mistaken my friend and confusing DiamCalc import/plugin with a new Cut Grade Analyzer that Sarin is incorporating into their software. Insofar as I know I am the only beta-tester of the CGA software unless they forwarded you guys a copy.

Regarding its results, it''s too early to say at this point and the software I was forwarded is a bit buggy. It''s supposed to be based on ray trace analysis similar to DiamCalc. I keep in close touch with the guys at Sarin so I''ll see if there is a more recent version of the software which is less buggy and if so I''ll post some results for ya''ll to check out.

Peace,

 

[Garry H (Cut Nut)]

1. Sarin are not pushing it hard - it is from an outside organization
2. Ogi are trying harder but have the same sort of problems
3. Sarin can not do even simple visualization software which is why they buy it from OctoNus (The realistic view DiamCalcSarin HASP for $500 from Sarin)
4. The type of ray tracing is slow and simplistic - I believe neither ogi or sarins products acount for even simple reflection polarization
5. it appears only DTI and OctoNus are in this race
6. GIA is even ahead of S & O in ray tracing

 

[Rhino]

I can believe that bro.

 

[JohnQuixote]

Date: 1/31/2005 7:29:10 PM
Author: Rhino

You are mistaken my friend and confusing DiamCalc import/plugin with a new Cut Grade Analyzer that Sarin is incorporating into their software. Insofar as I know I am the only beta-tester of the CGA software unless they forwarded you guys a copy.

Regarding its results, it's too early to say at this point and the software I was forwarded is a bit buggy. It's supposed to be based on ray trace analysis similar to DiamCalc. I keep in close touch with the guys at Sarin so I'll see if there is a more recent version of the software which is less buggy and if so I'll post some results for ya'll to check out.

Peace,

Greetings RhinoKnight.

With the exception of not knowing about the beta software, I do not believe I am mistaken, my friendly mammal.

Ira asked if Sarin is able to “take advantage that the diamond could still be right there, and analyzed directly?” My answer clarified that there are no direct assessment components on Sarin a la B’Scope, Isee, Imagem, etc (software notwithstanding).

Let me modify my statement thusly…

“Sarin does not have added mechanics for light performance assessment: It takes the proportions results we are all aware of and imports them into a licensed DiamCalc component new CGA component to create a virtual model of the diamond. Then it uses that virtual model to estimate light performance. The problem is that the proportions measurements are already not consistent, so meaningful predictions and comparisons are not possible in an absolute sense.”

…does this clarification assuage misstatement? Or is there more?

I’m not familiar with the beta CGA software, but the foundation of my statement is that Sarin is taking it’s own measurements and then using additional software - whether their own, DiamCalc’s, or JCPenney’s to do the CGA, yes?

Succinctly put, Sarin’s measurements are not exact enough for this. There is enough proof in the puddings of threads flavored with Sarins and grading reports being different.

(cough Helium! cough)

 

[Serg]

1)Sarin has license for distributing of cut-down variant of GA (calculations of BLR OctoNus were deleted by initiative of Sarin). Actually this software is used
only for photorealistic visualization in structural lighting like IS or H&A

2)CutGrade which Sarin and OGI are developed by no means concerns of OctoNus

3)RayTracing is always included to the base of photorealistic images and calculations of BLR somehow. But realization of adequate BLRs is more complex task than realization and test of BLR. In particular, we create function ETAS and necessary for it ConeTracing more than 3 years

4)The creation of adequate BLR is not enough to create laboratory CutGrading System. But this is very helpful tool for cutters if they want to develop new cuts or to improve current cuts. This year we will release new sets of BLR for cutters and possibly helpful tools for creation of cuts.

 

[JohnQuixote]

Thank you, Serg.

For clarification, BLRs include (1) positive perceived responses such as brilliance, fire, contrast and scint, (2) negative responses such as light leakage, “fish-eye” and “nail-head” and (3) cut coefficients like ETAS - which can overlap areas 1 and 2.

In the future - if there were a globally accepted set of BLR for human observation of diamonds - then grading would be possible… But only with more accurate scanning than Sarin can now offer, correct?

 

[Serg]

Date: 2/1/2005 10

6:54 AM
Author: JohnQuixote


Thank you, Serg.

For clarification, BLRs include (1) positive perceived responses such as brilliance, fire, contrast and scint, (2) negative responses such as light leakage, “fish-eye” and “nail-head” and (3) cut coefficients like ETAS - which can overlap areas 1 and 2.

In the future - if there were a globally accepted set of BLR for human observation of diamonds - then grading would be possible… But only with more accurate scanning than Sarin can now offer, correct?

John,

re:2) negative responses such as light leakage, “fish-eye” and “nail-head

We combined “nail-head” and light leakage to one BLR ( BlackZones) now.

re: ETAS - which can overlap areas 1 and 2.

We use negative/positive , white/color ETASes now.

re:But only with more accurate scanning than Sarin can now offer, correct?

Answer strongly depends from tasks CutGrading.
For example I think perhaps You can receive more correct result for LR by good Sarin model and our software than by BS .

 

[RockDoc]

I feel that the ray tracing is not up to snuff..

Why?

Read the thread where Marty Haske is discussiing the Octanus ray tracing methods ( the most potent information from Marty is from page 10 on in that post.)

http://www.gis.net/~adamas/giacut.html lots of very informative articles on many subjects

First, ray tracing currently is done in reverse, which is "risky" for accurate results.

Secondly, there is no standard for the lighting and angles for the illumination.

Thirdly, it considers non absorbent data where the subject stone needs to consider the absortion factor.

Marrty has spent years on this subject and with color science research, and is in my opinion far more knowledgeable than many people are.

Read about it here: https://www.pricescope.com/community/threads/ags-new-cut-grade-system-early-2005.19268/page-10

I have had long discussions on this with him, and want to make it clear on the forum so that he is acknoledged ror his contributions. We do share information but HE is the one that is the ultimate expert in this field.

To do ray tracing in a mode that is really accurate, it requires a lot of data be set up, and to do that would require a lot of time for the analysis, not to mention a computer that is more powerful than what is out there currently.

Perhaps this will change with the 64 bit software and chips now being introduced..

There is nothing out there that is PERFECT.... its approaching, but not here yet.

I believe there is severe variances from producing a virtual image as compared to the actual inspection of the diamond. In this venue, Gemex I believe the best way to measure light return.. but even with them the B Scope has a little ways to go. Gemex has imaged 700,000 diamonds and is ahead of everyone else using direct measurement of the light return.

Gemex is a very dedicated company, working on getting the best possible analysis of light return. They are constantly working to improve their B Scope ratings and method of analysis.

There is a lot of work that AGS has done...but a lot of this work is sort of in flux...

Rockdoc

 

[JohnQuixote]

Date: 2/1/2005 11:57:16 AM
Author: Serg


John,

Answer strongly depends from tasks CutGrading.
For example I think perhaps You can receive more correct result for LR by good Sarin model and our software than by BS .

Your answer regarding overall light return measure is not surprising, since BLR takes numerous illumination scenarios into account but BS does not. BS also has an admitted 5% error for (whatever) it is measuring.

In this sense, I agree that the future of ray tracing will yield more objective and meaningful measure than any current technology that uses atypical lighting scenarios to generate skewed results.

This year we will release new sets of BLR for cutters and possibly helpful tools for creation of cuts.

…And this will help in the development of a globally accepted set of BLR for human observation in general? (will get expert/consumer feedback in this stage?)

Thank you for the update. Can you explain more about the new ETAS categories (negative/positive , white/color) ?

 

[JohnQuixote]

Date: 2/1/2005 12:55

9 PM
Author: RockDoc
I feel that the ray tracing is not up to snuff..

Why?

Read the thread where Marty Haske is discussiing the Octanus ray tracing methods ( the most potent information from Marty is from page 10 on in that post.)

http://www.gis.net/~adamas/giacut.html lots of very informative articles on many subjects

First, ray tracing currently is done in reverse, which is 'risky' for accurate results.

Secondly, there is no standard for the lighting and angles for the illumination.

Thirdly, it considers non absorbent data where the subject stone needs to consider the absortion factor.

Marrty has spent years on this subject and with color science research, and is in my opinion far more knowledgeable than many people are.

I have had long discussions on this with him, and want to make it clear on the forum so that he is acknoledged ror his contributions. We do share information but HE is the one that is the ultimate expert in this field.

To do ray tracing in a mode that is really accurate, it requires a lot of data be set up, and to do that would require a lot of time for the analysis, not to mention a computer that is more powerful than what is out there currently.

Perhaps this will change with the 64 bit software and chips now being introduced..

There is nothing out there that is PERFECT.... its approaching, but not here yet.

I believe there is severe variances from producing a virtual image as compared to the actual inspection of the diamond. In this venue, Gemex I believe the best way to measure light return.. but even with them the B Scope has a little ways to go. Gemex has imaged 700,000 diamonds and is ahead of everyone else using direct measurement of the light return.

Gemex is a very dedicated company, working on getting the best possible analysis of light return. They are constantly working to improve their B Scope ratings and method of analysis.

There is a lot of work that AGS has done...but a lot of this work is sort of in flux...

Rockdoc

Hello Bill

To be fair to Serg and the technology he is developing I think it's also appropriate to point out Marty’s position on BrillianceScope.

Date: 11/27/2004 3:52:38 PM
Author: adamasgem

1) The BS scope tests diamonds 'for light return' in a lighting environment the diamond will never see again, and doesn't see in nature, so what is it telling the consumer..

2) Knowing the way the BS scope lighting works, you can probably design a 'cut' to maximize its returned 'performance' metric.

Not having played with one extensively, but having seen it, I'm not impressed. I've also had some discussions with Randy Wagner about the BS scope, and told him my feelings regarding it. Maybe they have improved it since the last time I took a look at JCK.

This Marty quote can be read in context in this thread discussing GemEx. Garry also nicely illustrates ETAS/ray tracing versus how B'Scope works in the same thread.

Bill, I strongly agree with your endorsement of Marty's expertise, but I have to side with his opinion of BS.

Not only does it evaluate diamonds in unnatural conditions – it tends to assign higher scores to a type of facet arrangement that does not, in our experience, result in a character of beauty with optimum visual balance.

Michael Cowing has done extensive work on aspects of diamond beauty. His insight on direct assessment is as follows: “To be meaningful, measurements of diamond beauty should be made in the typical illumination circumstances in which human judgment of that beauty is made. Measurements of beauty in atypical lighting can give high scores to cuts that have lower scores in typical illumination.” (from his aticle “Describing Diamond Beauty”).

Unless B-Scope type technology changes considerably and somehow manages to take into account the numerous variables in a diamond’s daily panorama of illumination such as obscuration/contrast brilliance, multiple (some possibly distant) light sources and a reasonable range of tilt through hundreds of illumination scenarios it will be hard for any of the direct assessment devices like GemEx’s to have the potential that 3D computer modeling has for the future.

I'm not saying they can’t eventually build the perfect mousetrap. I just believe the obstacles are significant.

 

[RockDoc]

Hey John

Give Marty a call.... with Gemex's updated camera and software, he is coming around to potentially changing his opinion, the quote on his website is rather old.

I have recently had extensive conversation with both Marty and Randy, and have stirred them up together.... so we'll see what develops.

He and Randy can talk about things that I can't, since I am a computer grape.. the computer part is not one of my forte's.
My forte is in legal research and doing what an expert witness does in the legal arena, as you are aware.

In fact, ask Dave Atlas and Marty about the course in Mediation that we all took together...Like bringing people together..


But results from the stones I have seen and looked at the results has been helpful to me and to Gemex, and Marty is potentially coming around on this.

What's your opinion on reverse ray tracing... and absorbancy issues?


Todd posted a very good post about his experience with the New BScope.

Did you see that one.

Rockdoc

 

[Garry H (Cut Nut)]

Effective Total Angular Size or ETAS is a complex idea to get your head around.
(Would you have thought Sergey would think of something easy?)

Here is an explanation:

A diamond is an optical instrument that redirects light from illumination sources and re-distributes it into the space surrounding it. A diamonds facet can be imagined to be virtual window into more facets and back into the surroundings, but of course it is more complex than this too, because sometimes we see 50% out of a facet, and 50% of our feild of view is reflected. Imagine looking into a fish in a pond - you can choose to see the reflection off the water, or you can focus on the fish underwater.

As a diamonds facets split any light beam into several smaller (secondary) beams; their effective total angular size (ETAS) may be greater or smaller than that of the primary beam.

Increasing the number of a diamonds facets can result in an increase or a decrease of the ETAS. If the number of facets is small enough, ETAS grows with the number. But ETAS begins to decrease when the size of virtual facets becomes so small that our eye looses it’s capacity to resolve the brightness of the primary source from that of secondary sources. This also explains why the size of a diamond needs to be included in any cut grading solutions that account for human perceptions of visual appearance.

 

[JohnQuixote]

Date: 2/1/2005 7:33:26 PM
Author: RockDoc
Hey John

Give Marty a call.... with Gemex's updated camera and software, he is coming around to potentially changing his opinion, the quote on his website is rather old.

I have recently had extensive conversation with both Marty and Randy, and have stirred them up together.... so we'll see what develops.

He and Randy can talk about things that I can't, since I am a computer grape.. the computer part is not one of my forte's.

My forte is in legal research and doing what an expert witness does in the legal arena, as you are aware.

In fact, ask Dave Atlas and Marty about the course in Mediation that we all took together...Like bringing people together..

But results from the stones I have seen and looked at the results has been helpful to me and to Gemex, and Marty is potentially coming around on this.

What's your opinion on reverse ray tracing... and absorbancy issues?

Todd posted a very good post about his experience with the New BScope. Did you see that one.

Rockdoc

Bill,

That quote I cited from Marty is only 9 weeks old.

I find it hard to believe he has converted, since he’s been pretty staunch in the past… Say, did you use some hypnosis on him and Dave during your mediation course? Or just fancy lawyer-talk?

As for ray-tracing, the task that lies ahead is complex. My understanding of the problem with most ray tracing software is that it considers the “eye” as a tiny point aperture. This is simplistic, since pupil adaptation and our stereoscopic vision - among other things - make the eyes geometrically complex…but not insolvably so (consider science has already replicated the anatomy for high-end cameras and binocular technology). The issue at hand is that most software has also treated light as discrete rays or simple parallel beams. This is where ETAS, what Garry described above, is so important. Calculating ETAS through a range of tilt is the only viable model I have yet seen that has any potential for estimating scintillation.

Of course, neither direct-assessment models or computer assessment models take the subjectivity factor into account, but that’s a discussion for another thread.

I looked at Todd’s post and thought it was well stated. I am not opposed to the use of BrillianceScope among those who understand its limitations - pros and consumers alike. I consider it useful for distinguishing good LR from poor LR over a broad range of cut quality. However, when making comparisons between equivalent makes of the finest diamonds the results are spurious and inconsistent. Further, I have fundamental issues with the aesthetic relevancy and analytic accuracy of what it purports to measure…Relevancy due to subjectivity and environment, and inaccuracy due to failure to account for contrast qualities of brilliance and ETAS.

That said, like many others I find the results interesting, if not especially meaningful. I do like the photos. My problem is with those who treat its results as absolute and ‘sell’ consumers on the idea that it will paint the entire landscape of a diamond’s beauty, which it will not.

 

[RockDoc]

Hi John

Let me clarify Marty hasn't changed his position.... but he is open to the new developments in B Scope....he is very scholarly in keeping his mind open.

I didn't want to infer that his mind is changed yet, and if so then hopefully this will clarify my former post.

As of yet, there is only qualitative results out there based on the limitations of technology, but a lot of folks are working towards a solution to report more and develop a standard.

Both Marty and Serg have done a lot of work and invested a ton of time. Both are to be commended.

In one way or another we are all working to improve education of the trade and consumers...The methods used can certainly vary but having an opinion helps each other when done amicably.

Just wanted to make my former posting a little clearer so readers don't get the wrong idea.

Rockdoc