AGS new cut grade system early 2005

[Garry H (Cut Nut)]

4 times less?

Correction (edited).
4 times for surface area intensity - but dispersion is near linear, so 2 times.
Maybe 2.5times less? dispersion is not purely linear?

 

[adamasgem]

----------------
On 10/17/2004 12:09:48 PM Serg wrote:

Marty,

Presume we have two color flashes on sphere with same power.

Presume the dispersion angle of first flash is 1 degree, the dispersion angle of second flash is 2 degree.

Second flash has intensity in two times less then first flash.

Do you agree?----------------

If the flare is composed of the same wavelength interval lets say 400 to 500nm, then the AVERAGE intensity for the second flare is 1/2 that of the first flare, but its length is double, so without any weighting for exit angle they would appear equal in a GIA type metric.

We have a few basic issues to consider in metrics for either brilliance or fire
1) Do we use the same environment for Fire as we do for brilliance, the only reason a "spot" seems to be used for Fire is that you can "see", via chromatic flare plots, the "flares"
2) The unit intensity variation for the "hemisphere", should we use the classic cloud covered sky and its D65 wavelength distribution or a "blue sky" model with head occlusion.
3) or a series of "spots" with a defined color temperature and intensity level with respect to the background, or do we treat each differently. Numerically, one effect may get lost in another.
4) The effects of internal absorption, because to date, no one has published anything regarding that. I'll post something later for the effects of a type IIA spectra inregard to the GIA type brilliance metric

In regard to the "viewer"

5) The physiological ability to resolve the differences in the flares, which depends, of course on the "background" and "surround" you are dealing with, human visual color adaptation, a nonlinear process, from what I've done (implemented) and studied in regard to human color vision.

Since we are not dealing per-se with phenomena like Fresnel diffraction at a slit or through a hole, in terms of smearing of the energy (Airy disk effect), I've been treating each wavelength as a "point" of energy (zero "area" in effect, the flare only having a length variable. Can you suggest another

I guess the first thing one might due is to modify any metric by the wavelength sensitivity of the human eye, for either a scotopic or photopic viewer.

 

[Serg]

----------------
On 10/18/2004 9:46:19 AM adamasgem wrote:

----------------
On 10/17/2004 12:09:48 PM Serg wrote:

Marty,

Presume we have two color flashes on sphere with same power.

Presume the dispersion angle of first flash is 1 degree, the dispersion angle of second flash is 2 degree.

Second flash has intensity in two times less then first flash.

Do you agree?----------------

If the flare is composed of the same wavelength interval lets say 400 to 500nm, then the AVERAGE intensity for the second flare is 1/2 that of the first flare, but its length is double, so without any weighting for exit angle they would appear equal in a GIA type metric.




----------------

But you do not take into 1/2(decrease) intensity . You take into Fresnel law only. Fresnel coefficients are same for same wave in both cases.
Are am right?

 

[adamasgem]

----------------
On 10/18/2004 10:30:33 AM Serg wrote:

----------------
On 10/18/2004 9:46:19 AM adamasgem wrote:

----------------
On 10/17/2004 12:09:48 PM Serg wrote:

Marty,

Presume we have two color flashes on sphere with same power.

Presume the dispersion angle of first flash is 1 degree, the dispersion angle of second flash is 2 degree.

Second flash has intensity in two times less then first flash.

Do you agree?----------------

If the flare is composed of the same wavelength interval lets say 400 to 500nm, then the AVERAGE intensity for the second flare is 1/2 that of the first flare, but its length is double, so without any weighting for exit angle they would appear equal in a GIA type metric.




----------------

But you do not take into 1/2(decrease) intensity . You take into Fresnel law only. Fresnel coefficients are same for same wave in both cases.
Are am right?
----------------

I look at it simplistically from the standpoint that I have unit exiting vector V0 (vector) at intensity I0, V1 at intensity I1 and V2 at intensity I2. Lets say V0 is ray at 400nm , V1 at 410nm and ray v2 is at 420nm

vector V2-V0 may be twice as long as V1-V0 and the magnitudes I0, I1, and I2 are what they are, based on the incindent intensity at 400, 410, 420 nm and the pathelength history. SIMPLISTICALLY, The metric I've computed is the integral of the intensity I_k from V0 to Vn weighted by the cosine squared observability function. Fresnel reflectivity makes I_k whatever it is.

 

[Serg]

----------------
On 10/19/2004 10:01:03 AM adamasgem wrote:

SIMPLISTICALLY, The metric I've computed is the integral of the intensity I_k from V0 to Vn weighted by the cosine squared observability function. Fresnel reflectivity makes I_k whatever it is.



----------------

I am disagree.

Your metric is the integral of the Power(Energy/time) by area "weighted by the cosine squared observability function" .
Power(Energy/time) is the integral of the intensity by area .

Fresnel law is for Power(Energy/time), but not for Intensity.

Intensity=Power/area

Post was edited.

 

[oldminer]

Well guys, this thread has been totally and thoroughly "hijacked" at this point. It could be that you two are the only two people here who know what either is talking about. You've lost most of us. I imagine that doesn't really matter to either of you, because you are immersed in your train of thoughts, but it seems to be now somewhat "off-target". I'm being kind.




I think the public who come to Pricescope for meaningful advice might gain an awareness from this thread that all the details of the near future cut grading scenario have some unresolved issues that are at relatively high level of science and technological knowledge. We all want consumers to have the right information and we also want to build their confidence. Could the information contained in this thread build up consumer confidence in our good intentions to do it right? Might consumers get the feeling that all is not so well in the near future of diamond grading?




Maybe it would be a good time to tell consumers if the overall game plan will help them or hurt them, in plain and simple language. Give them a few reasons, ones they can comprehend, to defend your position. From this, we might all benefit....THANKS.

 

[Serg]

David,

My discussion with Marty devoted to problem of verification the result any Cut Study. I am sure, it is very important problem for diamond industry
Labs use absence the verification system for advancement unready Cut grading system to market.
Only open discussion could change current situation. there is very danger to build consumer confidence are keeping back problems.
More than two people are reading and understanding this thread.
I try find the new persons for joint work and disputes. Joint dispute is very helpful. What is why I am work in PS.

May be Yurii and Garry could help with "in plain and simple language". It is impossible job for me. Sorry.

Marty,
I will strongly support works for developing verification standard . Thanks for your help.

 

[oldminer]

Sergey: I do realize the language barrier is, in itself, a major hurdle for you to overcome. I personally appreciate all I have read here, in the limited capacity I have myself, to understand only a small part of what is being discussed. The post by me above was not a criticism of what your interests are, but only of the appearance here of such difficult information which only a very few can begin to appreciate. Even fewer would claim to understand in full.




Marty can attest to the fact that I have encouraged him in his past research and have done my little part by purchasing the DiamCalc program to help support your underfunded efforts. I fear the real powers in the diamond business want to sweep many of the issues revealed here under the rug rather than address them. Since they are so powerful, that will probably happen regardless of what is right.




Marty, how about posting a "short" recap of any conclusions you can draw from what has been discussed here?


Sergey, how about the same from you, or from Yuri or Garry, as you suggested?




I think all readers here would want to get an understanding of what you two agree upon, what is disputed, and what the AGS and GIA are proposing that has both of you concerned. I think these three things are what the messages back and forth have been about, in the most general sense.

 

[denverappraiser]

I, for one, have been following this thread with great enthusiasm although I too find it a little difficult at times. I have no problem struggling through it for as long as these guys are willing to keep at it.

Neil Beaty, GG ISA
Independent Appraisals in Denver

 

[Iiro]

I am one of those who still read and try to understand this. I like Sergeys style with one question, issue, at a time. Martys writing much harder to follow, there is so much tehc in one post, and a lot of time it does not address to question at hand.

 

[oldminer]

DenverA and Iiro: Both of you are sort of where I am in all of this. We are interested, but the thread has lost its focus for several reasons. I thought possibly we could get the main particpants to regroup and organize for those of us in the wings.

 

[adamasgem]

----------------
On 10/19/2004 10:52:42 AM Serg wrote:

----------------
On 10/19/2004 10:01:03 AM adamasgem wrote:

SIMPLISTICALLY, The metric I've computed is the integral of the intensity I_k from V0 to Vn weighted by the cosine squared observability function. Fresnel reflectivity makes I_k whatever it is.



----------------

I am disagree.

Your metric is the integral of the Power(Energy/time) by area 'weighted by the cosine squared observability function' .
Power(Energy/time) is the integral of the intensity by area .

Fresnel law is for Power(Energy/time), but not for Intensity.

Intensity=Power/area

Post was edited.----------------

I'm refering to the Freznel reflectivity equations for I/Io, which determine what % of the internal ray (my "Intensity") is refracted out verses what is internally reflected at an interaction.

 

[adamasgem]

----------------
On 10/19/2004 2:24:08 PM oldminer wrote:


Sergey: I do realize the language barrier is, in itself, a major hurdle for you to overcome. I personally appreciate all I have read here, in the limited capacity I have myself, to understand only a small part of what is being discussed. The post by me above was not a criticism of what your interests are, but only of the appearance here of such difficult information which only a very few can begin to appreciate. Even fewer would claim to understand in full.


Marty can attest to the fact that I have encouraged him in his past research and have done my little part by purchasing the DiamCalc program to help support your underfunded efforts. I fear the real powers in the diamond business want to sweep many of the issues revealed here under the rug rather than address them. Since they are so powerful, that will probably happen regardless of what is right.


Marty, how about posting a 'short' recap of any conclusions you can draw from what has been discussed here?

Sergey, how about the same from you, or from Yuri or Garry, as you suggested?


I think all readers here would want to get an understanding of what you two agree upon, what is disputed, and what the AGS and GIA are proposing that has both of you concerned. I think these three things are what the messages back and forth have been about, in the most general sense.
----------------

David.. So far Sergey an I have shown that two differing methodologies are in agreement as to reflection patterns (glare effect) so that we can try to define and resolve "metrics" for Fire and Brilliance.

What I have repeatedly tried to point out for the consumer, is that "metrics" which may or may not have been proposed so far by Labs to ascertain what is a good cut versus what may not be as good are highly dependent on the environments chosen to create metrics. In otherwords, it is possible to tailor environments to meet the objective.

Sergey's methodology is a lot faster than the strict Monte Carlo analysis I conduct, and makes certain assumptions which can be verified by doing a "full blown" analyses like I try to do. Both of us (not to exclude anyone else) has to make certain critical assumptions, which may or may not effects our results in a material way (like doing 1000000 rays versus 10000 rays). Only by OPENLY discussing what we are doing and how we are doing it, can people understand that what we are trying to do, rather than what has evolved into
opinions by some organizations, which are largely based on the "trust us, we are G-Ds" methodology, who can do no wrong.

I realize that this may appear laborious to some, in part to the communication/language barriers, but what other groups who purport to be "industry leaders" are doing this. Most don't accept criticism, nor open their research up to peer review of any meaningful nature. Communication is largely one way, you tell them what you know, but they can't or won't tell you. Everything is proprietary because of the big bucks involved with selling their "paper".

Sergey is the ONLY one who will engage in a technical discourse of this nature.

Bear with us, and we will get to the meat of the problem, in that we will be able to independently and jointly review what is published by others, what makes sense and what is pure consumer BS, and maybe eventually be able to tell the consumer, with a reasonable degree of certainty, what are sound facts and what is hyperbole.

Some in the industry probably won't like our comments

 

[adamasgem]

----------------
On 10/19/2004 2:40:52 PM Iiro wrote:

I am one of those who still read and try to understand this. I like Sergeys style with one question, issue, at a time. Martys writing much harder to follow, there is so much tehc in one post, and a lot of time it does not address to question at hand.



----------------

Sometimes we have to discuss differences in our interpretation to get to what is the "question", bear with us.. I'm an engineer by training, I don't know what Sergey's basis is, but I suppose and optics/physics major, and we can talk two different languaages sometimes, and I don't mean Russian/English. Physics majors talk in wavenumbers, engineers in nanometers, and others in electron volts, all mean the same thing, but translating can be a pain.

 

[rfath]

I, too, am having a great time slogging through the thread - giving google a fabulous workout to figure out the technicalities mentioned here.

Then again, I also got to talk with Marty in person after he got done appraising my ring... it makes more sense to me with the gestures and emphasis as well as being able to see what he's talking about in real life with his software and hardware in front of us.

~Rachel

 

[Serg]

----------------
On 10/19/2004 2:51:32 PM adamasgem wrote:

----------------
On 10/19/2004 10:52:42 AM Serg wrote:

----------------
On 10/19/2004 10:01:03 AM adamasgem wrote:

SIMPLISTICALLY, The metric I've computed is the integral of the intensity I_k from V0 to Vn weighted by the cosine squared observability function. Fresnel reflectivity makes I_k whatever it is.



----------------

I am disagree.

Your metric is the integral of the Power(Energy/time) by area 'weighted by the cosine squared observability function' .
Power(Energy/time) is the integral of the intensity by area .

Fresnel law is for Power(Energy/time), but not for Intensity.

Intensity=Power/area

Post was edited.----------------

I'm refering to the Freznel reflectivity equations for I/Io, which determine what % of the internal ray (my 'Intensity') is refracted out verses what is internally reflected at an interaction.


----------------

Fine.
We main question now what is Intensity?

You use Intensity like Power.
I use Intensity like illuminance ( Power/area) . Like brightness for human.

 

[adamasgem]

----------------
On 10/19/2004 3:30:55 PM Serg wrote:

----------------
----------------

Fine.
We main question now what is Intensity?

You use Intensity like Power.
I use Intensity like illuminance ( Power/area) . Like brightness for human.
----------------

I understand now..

If you look at a spectral distribution for, say D65, I define the peak intensity (illuminance) as 1.0, the intensity at lets say 700nm may be 0.8 (or what ever).

Now I can't understand how one defines "area" like GIA did in their DCLR report, as what is the "area" of a single ray of light..???

 

[Serg]

From GIA Fire Article p.183 Fall 2001:
"The VFI diagrams display a variety of properties that can be combined into a metric, such as the total number and relative brightness of colored spots, and the lengths and angular distribution of colored streaks made up of these spots. The metric we derived- dispersed colored light return, or DCLR .... That is. DCLR is sum over all colored streaks. of the sum over all colors( sampled every 10 nm), of the size(area) of each colored streak multiplied times the "smoothed" brightness( intensity) of each spot along the streak, times an exit-angle weighting factor...
"
We could discuss only this definition. May be GIA realization this metric by software is quite different from this description. I do not know. May be GIA realization like Marty realization. May be GIA realization better or worse than GIA definition.

I can not discuss GIA realization, I have not enough information for this work.
But I know a lot reasons why GIA definition is very bad for conception Fire.
Some explanation you could read http://www.cutstudy.com/cut/english/grading1/6.htm

new point : We can not use size of spots on sphere for grading quality Flash in human eye. Some times bigger spot on sphere will produce smaller flash in eye. Brightness on sphere and brightness in eye have not direct correlation. ( Sometimes its have inverse correlation)
old point: Metric should find good balance between quality and quantity of Flashes. What is better one big flash or two small....? When one big flash is better than two small? When two small flashes are better than one big flash?

 

[Garry H (Cut Nut)]

Dave and everyone else (except the 2 nerds - will talk to you later

)
This picture shows the difference between power and intensity.
Of 100% of the light hitting the stone, 17% is reflected, 83% goes into the stone and (barring absorption) comes out somewhere; this is power.

The Intensity of the beam refracted out the crown is going to be brighter or more concentrated to an obsevrer than the original light source (if they were to see the entire flash of color from one view). This is one of diamonds amazing features - the ability to make a sparkle that is brighter than the original source. (Sergey and Marty - even if you find errors in this simple desrciption - email me, dont confuse).

Now regarding the hijacking of the thread - most of the content of this thread is in a similar topic. This was the wrong thread (as has been discussed a few pages back) and I have asked Marty to use english, and he has improved, but Marty as you can see others are watching, so keep it as simple as possible please. Sergey is trying very hard to use good simple english (and doing very well indeed).

Sergey's point is that many people read these threads who because of the organizations and positions - they can but lurk. We know this because we have private communications. There is also a hidden forum on Pricescope for the participants of the Diamond Cut Conference - it is pretty dead at present and we are considering inviting all those who are interested and have something to contribute (consumers too) to participate.

Now What is this all about?
Sergey believes that it is important that organisations intending to perform cut grading be subject to peer review. e.g. when GIA did release something about their polarisation methods in the WLR study - Sergey, Vladimir and others posted here and pointed out a basic mistake in their approach. They subsequently admited this and changed what they wrote on the website. There is no reason to think therefore that what they wrote in G&G in Fall 1998 was completely valid.

So what Sergey is doing does have a purpose - those little specs on imaginary spheres are useful - they come from those rays in the models I posted here. There are many different assumptions and many possible reasons why 1 approach may be more useful than another.

Finally, AGS and GIA have established some ways to grade the cut quality of a diamond. Great. But!
Those methods will not make it possible to design new cuts with different facet patterns than those that they have a grading system for. Sergey's vision is for a software based system that can take a scan of any rough diamond and create a one off cut just for that piece of rough diamond.
Imagine one off art cuts!!
And they will look amazing!!

Now that is some way off - and at present the holding devices and polishing technology could not cope anyway - but the future we see is just mind bogglingly beautiful

 

[strmrdr]

Let me see if I get this :}
Garry you intensity diagram is confusing it may lead someone to believe that more light is coming out than going in which is not the case.
It is just gathering it in from a larger area and concentrating it in a smaller area.
Your text is correct in calling it concentrated.

Now here is where the problem comes in:
If you measure the light output at the 129% marked rays exit point it will be lower than an equal sized entrance point if that enterence point is the only illumination source: ie the rest of the diamond is masked off, which is what your power diagram is showing.
But the diamond is concentrating the light from a larger incoming area and sending it out a smaller area.
So in affect its acting as funnel for light.
Funnel == A large wide low pressure flow comes in and a smaller higher pressure flow comes out.

Diamond: A large low intensity beam comes in and a smaller higher intensity beam comes out.

In both cases the total power out is less than the input power but the apparent power/intensity is higher.

Ps. The analogy isn’t perfect because the loses and gains are different but as a concept does it sound good?

 

[Serg]

strmrdr,

1)This diagrams right for specific light only.( plane wave front, simple ray tracing.) You can not use such light in real life.
2) The one of goal diamond to collect light from hemisphere and send this light to zone of eyes. You need add more funnels to your analogy. A lot of funnels collect light and redirect in to two funnels. Power in eyes funnels could much bigger than in any primary funnels.

 

[strmrdr]

----------------
On 10/20/2004 5:07:45 AM Serg wrote:

strmrdr,


1)This diagrams right for specific light only.( plane wave front, simple ray tracing.) You can not use such light in real life.

2) The one of goal diamond to collect light from hemisphere and send this light to zone of eyes. You need add more funnels to your analogy. A lot of funnels collect light and redirect in to two funnels. Power in eyes funnels could much bigger than in any primary funnels.----------------

Ok I understand the adding funnels part.

Light that cant be used in real life is confusing when discussing the "looks" of a diamond which is very real life. That is where you lose me I dont have the background to look at it the way you do.
Is non-real world light even relevant to diamonds?

 

[Garry H (Cut Nut)]

Is non real world ever relevant?

This is the problem - unless we can make the problems bite sized, we can not solve them, and then we can not unleash this powerful capacity to design based on the shape of the rough rather than the expectations based on previous diamonds.

Isn't Sergey's english so wonderfully concise? He gets right to the bare bones.

The sharp end of the funnels are our eyes. The pupils are bigger in a restaruant than in Walmart, so we see more fire. Sergey's displays on the surface of the sphere can account for that. Then it is an analysis deal to decide if a diamond is more likely to show fire and strong sparkles etc.
The science is capabale of deciding if a diamond is good or bad.

 

[adamasgem]

----------------
On 10/20/2004 4:04:39 AM Serg wrote:

From GIA Fire Article p.183 Fall 2001:
'The VFI diagrams display a variety of properties that can be combined into a metric, such as the total number and relative brightness of colored spots, and the lengths and angular distribution of colored streaks made up of these spots. The metric we derived- dispersed colored light return, or DCLR .... That is. DCLR is sum over all colored streaks. of the sum over all colors( sampled every 10 nm), of the size(area) of each colored streak multiplied times the 'smoothed' brightness( intensity) of each spot along the streak, times an exit-angle weighting factor...
'
We could discuss only this definition. May be GIA realization this metric by software is quite different from this description. I do not know. May be GIA realization like Marty realization. May be GIA realization better or worse than GIA definition.

That is the $64,000.00 Question

I can not discuss GIA realization, I have not enough information for this work. I took my best crack at it

But I know a lot reasons why GIA definition is very bad for conception Fire.
Some explanation you could read http://www.cutstudy.com/cut/english/grading1/6.htm

I'll go back and review that tonight

new point : We can not use size of spots on sphere for grading quality Flash in human eye. Some times bigger spot on sphere will produce smaller flash in eye. Brightness on sphere and brightness in eye have not direct correlation. There is our friend non linearity ( Sometimes its have inverse correlation)

You are right, in part, I believe. "Relative" brightness of course effects observability/resovlevability, as does "size". The "size" (or area)of the spots in any physical or theoretical projection depend on the radius of the hemisphere. The human, because of his ability only to "see" rays directed toward him, ony sees a fraction of the hemisphere at a time. GIA, quite appropriately I believe, weights the flares by the cosine squared of the exit angle to take into account that the primary viewing position is looking down at the diamond.

old point: Metric should find good balance between quality and quantity of Flashes. What is better one big flash or two small....? When one big flash is better than two small? When two small flashes are better than one big flash?----------------

 

[adamasgem]

----------------
On 10/20/2004 4:11:00 AM Garry H (Cut Nut) wrote:

Dave and everyone else (except the 2 nerds - will talk to you later

)

Now regarding the hijacking of the thread - "

"Garry, If you say I "hijacked" this thread one more time, I'm going to tell you to Buzz off. You keep on trying to compare apples and oranges and to make a play on trying to commenting on what couldn't or shouldn't be compared because AGS and GIA and MSU did THREE DIFFERENT THINGS, and they are going to get THREE DIFFERENT ANSWERS, and the THREE DIFFERENT THINGS THEY DID were not well defined at all


----------------

 

[strmrdr]

----------------
On 10/20/2004 6:45:54 AM Garry H (Cut Nut) wrote:

Is non real world ever relevant?


This is the problem - unless we can make the problems bite sized, we can not solve them, and then we can not unleash this powerful capacity to design based on the shape of the rough rather than the expectations based on previous diamonds.


Isn't Sergey's english so wonderfully concise? He gets right to the bare bones.


The sharp end of the funnels are our eyes. The pupils are bigger in a restaruant than in Walmart, so we see more fire. Sergey's displays on the surface of the sphere can account for that. Then it is an analysis deal to decide if a diamond is more likely to show fire and strong sparkles etc.

The science is capabale of deciding if a diamond is good or bad. ----------------

It can be if it corelates with the real world conditions to some degree.
Is it a step backwards not to use real world conditions or as close as possbile?

Yes I find Sergey fairly to understand when he isnt shooting over my head.

Now back to funnels the sharp ends of the funnels arent the eyes the eyes are funnels that are 10 inches or 16 inches or however far depending on whose method you want to use away from the sharp ends of the funnels that are the exit points on the diamond.
correct?

 

[Serg]

----------------
On 10/20/2004 5:22:57 AM strmrdr wrote:




Light that cant be used in real life is confusing when discussing the 'looks' of a diamond which is very real life. That is where you lose me I dont have the background to look at it the way you do.
Is non-real world light even relevant to diamonds?



----------------

Real source is 3-dimensional and has spherical wave front and continuous spectrum. The ray using in the raytracing usually is one-dimensional. In DiamCalc we use 3-dimensional beam but with discontinuous spectrum and plane wave front (any real beam has divergence angle). Such properties as continuous spectrum and non-zero divergence of the beam also influence to changing of intensity in considered case. Therefore neither ray nor parallel beam can't model all transformations which are taking place with lighting source.
For example, intensity for non-divergent beam doesn't depend on distance, intensity of divergent beam is decreasing with the distance.
There is limit of application for any model of real universe .

In the context of this limit using of model can be successful, outside of this limit using of the model can't be successful.
If we speak about intensity of the light from the diamond near surface, then we can neglect of divergence and discontinuous spectrum, and parallel beam gives good model for remote source of the lighting.
If we consider intensity of the light far from the diamond, then this model can't give correct conception about intensity even for remote source of the light source.

Also it is important to understand that intensity of the light from the diamond and image brightness on the retina are two big differences.

For example: let's consider lampposts lighting the road. The brightness of the bulbs which you can see on these lampposts is the same and doesn't depend on the distance to the lamppost. Contribution to the illumination for the place where you are locating is very different for different lampposts. Actually nearest lamppost is most important for illumination, because influence of other lampposts is decreasing proportionally to the squared distance.

Therefore if we speak about flashes on the sphere, then its instrument can be used only for analysis of probability to view the flash (or other similar characteristics). Brightness and size of the flash viewable by the human is determined entirely another laws and in another virtual space.

All the following things are equally important for estimation of perception by human of Fire, Scintillation: the frequency of flashes (this is generalization of its quantity), its brightness, saturation (for Fire), size. Also there are another important characteristics, for example evenness of its location.

 

[Garry H (Cut Nut)]

Marty re my comments about "hijacking" I was responding to Dave's comments:

Well guys, this thread has been totally and thoroughly "hijacked" at this point. It could be that you two are the only two people here who know what either is talking about. You've lost most of us. I imagine that doesn't really matter to either of you, because you are immersed in your train of thoughts, but it seems to be now somewhat "off-target". I'm being kind.

But you point out Marty that this is relevant because AGS have used computer modelling as part of their cut grading system, and therefore
You keep on trying to compare apples and oranges and to make a play on trying to commenting on what couldn't or shouldn't be compared because AGS and GIA and MSU did THREE DIFFERENT THINGS, and they are going to get THREE DIFFERENT ANSWERS, and the THREE DIFFERENT THINGS THEY DID were not well defined at all

But hey - here is the funny part - AGS used DiamCalc

 

[adamasgem]

----------------
On 10/20/2004 4:45:21 PM Garry H (Cut Nut) wrote:

Marty re my comments about 'hijacking' I was responding to Dave's comments:

Well guys, this thread has been totally and thoroughly 'hijacked' at this point. It could be that you two are the only two people here who know what either is talking about. You've lost most of us. I imagine that doesn't really matter to either of you, because you are immersed in your train of thoughts, but it seems to be now somewhat 'off-target'. I'm being kind.

But you point out Marty that this is relevant because AGS have used computer modelling as part of their cut grading system, and therefore
You keep on trying to compare apples and oranges and to make a play on trying to commenting on what couldn't or shouldn't be compared because AGS and GIA and MSU did THREE DIFFERENT THINGS, and they are going to get THREE DIFFERENT ANSWERS, and the THREE DIFFERENT THINGS THEY DID were not well defined at all

But hey - here is the funny part - AGS used DiamCalc

----------------

It's not so funny Gary, In fact a long time back I had recommended that they do, DiamondCal is an excellant visualization tool. They also used another ray trace product. The issues lie importantly in what environment is being modeled, and in defining and discussing that environment

 

[adamasgem]

Well.. The GIA article is out now.

A lot of backtracking, and some "strange" changes.. Seems that they are now
relying on their series of grader "observations" and poor statistics to "prove" what they wanted to "prove". I haven't digested it all yet because it seems a little difficult to stomach..

They've backtracked on their "brilliance" article and now define "brightness" by using a new hemispherical lighting model in a static environment

For "Brightness" Metric

1) Static Viewer over the table with a 3 degree solid angle (1.5 degree cone angle) total Field Of View

2) Hemispherical lighting with central 46 degree solid angle (23 degree cone) blacked out. So illumination is ONLY coming in from angles from +/- 23 to 90 degrees of the hemisphere, and with no intensity variation.

3) Black hole at girdle plane

4) Metric now includes the effects of glare

They haven't apparently changed their "Fire" inputs, a collimated overhead beam of light with a cosine squared weighting

So, in effect, what they have done is create a model where the angular ray distribution to define the INPUT to their metric is a function of the stones' cutting angles.

As far from a realistic input model as one can get, just because their empirical "observations" didn't "fit" previous theoretical results using Pearson Product Moment correlation tests and Cronbach alpha (Boy I'm going to have to get that reference).

Never mind using accepted illumination models from science like a cloud cover sky with intensity falloff, or the Petrovsky blue sky model (with head occlusion)

I participated in one of their observation tests in Tucson a few years back, and their "brightness" lighting environment sure doesn't fit what I saw..


They do however, have a nice hologram on the front cover of G&G.