AGS new cut grade system early 2005

[Serg]

Date: 11/2/2004 11:21:52 AM
Author: adamasgem
Someone correct me if i am wrong, but in my previous post on total angle of acceptance TOA, it seems that the larger the stone, the wider the range of angles of rays with respect to the table the viewer will ''process'' as being part of the stone.

If we fix the TOA (TOA_fixed), as is suggested in the GIA article, and then try to correlate with observations with a varying ''real'' TOA (TOA_ith), then if

TOA_fixed < TOA_ith,

we can''t possibly get any ''good'' correlation of the theoretical TOA_fixed model with the observed.

You are right. Result for big diamond should be quite different from result for small diamond.

To read IDCC proceedings.

Did you remember our discussion what optimal quantity of facet depends from diamond size ?

I will absent on PS next 7-10 days

 

[adamasgem]

Date: 11/2/2004 11:41

4 AM
Author: Serg

Date: 11/2/2004 11:21:52 AM
Author: adamasgem
Someone correct me if i am wrong, but in my previous post on total angle of acceptance TOA, it seems that the larger the stone, the wider the range of angles of rays with respect to the table the viewer will 'process' as being part of the stone.

If we fix the TOA (TOA_fixed), as is suggested in the GIA article, and then try to correlate with observations with a varying 'real' TOA (TOA_ith), then if

TOA_fixed < TOA_ith,

we can't possibly get any 'good' correlation of the theoretical TOA_fixed model with the observed.

You are right. Result for big diamond should be quite different from result for small diamond.

To read IDCC proceedings.

Did you remember our discussion what optimal quantity of facet depends from diamond size ?

I will absent on PS next 7-10 days

For any of the techies out there, please chime in..

What we are saying is that their is a interrelationship between TOA (Total Observation Angle) and stone diameter, viewer's distance, and Pupil diameter.

Well, you can plug in the numbers you want for TOA and calculate a pupil diameter for the viewer's position you specify, BUT, when you plug in the pupil diameter into standard equations (Moon& Spencer 1944 and DeGroot & Gebhard 1952 in Wyszecki & Stiles) that relate the pupil diameter to illuminance, you find some numerical limitations imposed by the log_10 function inherent the two sets of equations giving 2.41mm
Now this suggests that the critical factor that we don't know is the luminance in the viewing hemisphere which strongly suggests the FIXED pupil diameter we should be using,

which then defines what the real TOA is for the viewers position and the fixed diameter of each stone. We just can't arbitrarily pick a TOA to use.

Now if you look at the pictures of the GIA viewing hemispheres they used, with differing blacked out areas,
you can bet your bottom dollar that they used the same FIXED bulb setup to illuminate those hemispheres, and given the light sucking ability of the blacked out areas in the hemisphere, the lumninance varied inside each hemisphere, effecting the pupil diameter.

So while the viewer's position didn't change, each hemisphere setup whould have a different mean pupil diameter among all viewers. I'll bet a buck that's another reason why their observed data didn't correlate
with their models, because they probably assumed a fixed TOA, where in reality each setup had a different TOA for the same diamond.

Oh well... we will probably never know what the truth is, and the trade will be force fed a little flawed research once again..

Not on my watch..

 

[Garry H (Cut Nut)]

Translation to mere mortal English:

When GIA turned all their lights on (fluoro and LEDS) in their illumination box, your pupils get smaller, and what you see is different - you see the brightness OK, but not as much fire. This is partly because the brightness (glare reflected, and bright sparkles of mainly white scintillation) washes out our ability to detect the inherently less intense rainbow colors of fire. And partly because our pupils are narrower (Marty''s FOV = field of view) because that is what happens when the light is bright. If our pupils are wide open, as in a dim restaurant, we can see a braoder spectrum of the dimmer fire because more can get inside onto the rods in the back of our eyeballs.
Because GIA have not said how much light they used - Marty and Sergey can not test the GIA software - and this is not good in the world of science, where peer review is an important concept.
We would like to invite GIA to join this discussion and provide some answers please.

 

[adamasgem]

Date: 11/3/2004 5:37:15 AM
Author: Garry H (Cut Nut)
Translation to mere mortal English:

When GIA turned all their lights on (fluoro and LEDS) in their illumination box, your pupils get smaller, and what you see is different - you see the brightness OK, but not as much fire. This is partly because the brightness (glare reflected, and bright sparkles of mainly white scintillation) washes out our ability to detect the inherently less intense rainbow colors of fire. And partly because our pupils are narrower (Marty''s FOV = field of view) because that is what happens when the light is bright. If our pupils are wide open, as in a dim restaurant, we can see a braoder spectrum of the dimmer fire because more can get inside onto the rods in the back of our eyeballs.
Because GIA have not said how much light they used - Marty and Sergey can not test the GIA software - and this is not good in the world of science, where peer review is an important concept.
We would like to invite GIA to join this discussion and provide some answers please.

Good translation Gary. Additionally if one looks at the hemispheres in Figure 7 of the new 2004 article, you will realize that GIA was effectively dimming the lights also, and I''ll bet that their theoretical model didn''t account for that in terms of the Pupil radius, however much that may be. We don''t know the basic illmination intensity to relate to pupil size.

Irrespective of all this, what about a model that mimiced that average vieweing environment the consumer will see everyday after they purchase their diamond. Why did they concentrate on "trade" environments for a "metric". NO MENTION OF THAT IN THE ARTICLE.

It would be interesting to see relative performance differences between "trade" and "everday consumer" environments. Isn''t this what it is supposed to be about, or was the $$ cart before the horse?

 

[Garry H (Cut Nut)]

Trade re consumers:
1. Remeber the Fluoro article - the non trade people were all over the place and the information was useless.
2. Trade needs this study "sold" to them so it is logical to include and win them over (or GIA may a. loose grading business and b. loose donations)
3. trade people can discern small distinctions by virtue of experiance (we would do the same)

 

[adamasgem]

Date: 11/3/2004 4:13:21 PM
Author: Garry H (Cut Nut)
Trade re consumers:
1. Remeber the Fluoro article - the non trade people were all over the place and the information was useless.
2. Trade needs this study ''sold'' to them so it is logical to include and win them over (or GIA may a. loose grading business and b. loose donations)
3. trade people can discern small distinctions by virtue of experiance (we would do the same)

1) You know I''m on record regarding the farce of the GIA Fluorescence article and the change in the grading standards there
2) The "trade" will buy anything that makes junk appear better than it is
3) That''s why some people come to independent people they trust to help them in the purchase. My experiance with internet purchasers is that they were turned off by retail sales hype and pricing, along with general lack of information given on the retail level.

 

[Garry H (Cut Nut)]

Now now Marty. Leave the politics to the pollies.

The fact remains that 95% of diamonds are sold in B&M stores and those people need some help. Any help that GIA provides consumers is a step in the good direction. I agree that we interested nerds have a role to play (we had a politician in Australia - Don Chipp - who started a minority party that held the balance in the senate - he coined the phrase "Keep the Bastards Honest").

But as our mummies told us - we catch more flies with honey

So far I can see the gIA cut study has headed in generally the right direction. Of course their standard will be looser than AGS''s, but AGS has a niche position to maintain. I can see some possible problems with the GIA study too - but none that look too terrible. take stone RD07 for instance. A shocker at first glance, but it has poor light return and this makes the potential for fire much greater. It has long splintery lower girdles - this gives it many small virtual facets - now this would work for larger stones - and it was one of the largest in their study (.76)- but if they used a 0.44ct (smallest) then I doubt they would have conferred second top grade on it.

(T53% c36.5 p41.4 with lower girdles 90)

 

[strmrdr]

I disagree Garry if gia wants to prove they arent industry puppets then they had better get it right the first time.
Iv said for a while now it that it looks like they went in with the results they wanted and instead of doing real studies they are coming up with junk to try and support what they want it too.
That is is no way shape or form good for consumers.

 

[Garry H (Cut Nut)]

Storm have you read this latest report in G&G?

If not then you have no right to make such a comment.
If you have then we welcome your comments and your review.

But unless you have a valid and pertinent criticsm, please resist the temptation to cut down tall poppies and attack large institutions.

 

[strmrdr]

Date: 11/4/2004 1:36:36 AM
Author: Garry H (Cut Nut)
Storm have you read this latest report in G&G?


If not then you have no right to make such a comment.

If you have then we welcome your comments and your review.


But unless you have a valid and pertinent criticsm, please resist the temptation to cut down tall poppies and attack large institutions.

link it and Ill read it.
Because everything Iv read online has given me the impression im right and Marty is confirming it.
You have confirmed it yourself in this thread and others.
The bottom line is that there is something rotton about it the question is what..

 

[adamasgem]

Date: 11/3/2004 9:29:45 PM
Author: Garry H (Cut Nut)
Now now Marty. Leave the politics to the pollies.

The fact remains that 95% of diamonds are sold in B&M stores and those people need some help. Any help that GIA provides consumers is a step in the good direction. I agree that we interested nerds have a role to play (we had a politician in Australia - Don Chipp - who started a minority party that held the balance in the senate - he coined the phrase ''Keep the Bastards Honest'').

But as our mummies told us - we catch more flies with honey

So far I can see the gIA cut study has headed in generally the right direction. Of course their standard will be looser than AGS''s, but AGS has a niche position to maintain. I can see some possible problems with the GIA study too - but none that look too terrible. take stone RD07 for instance. A shocker at first glance, but it has poor light return and this makes the potential for fire much greater. It has long splintery lower girdles - this gives it many small virtual facets - now this would work for larger stones - and it was one of the largest in their study (.76)- but if they used a 0.44ct (smallest) then I doubt they would have conferred second top grade on it.

(T53% c36.5 p41.4 with lower girdles 90)

Gary, Unfortunately $$ and politics have interjected themselves into the GIA study from the start, in my opinion. I don''t think that we should support misleading information, period. Unfortunately the good work of some people gets misdirected with preset agenda. "Here is the answer, give me the rhetoric to back it up" is the tone I read.

Regarding your statment "flies with honey", it should be "flies with money"

I agree with keeping the bastards honest.

 

[Garry H (Cut Nut)]

it will not be online for some time I expect

 

[Garry H (Cut Nut)]

Marty did you see this link www.gia.edu/gemsandgemology/361/ig__gi_data_depository.cfm

It has quite a bit of additional detail (still not all we would like - but more )

 

[adamasgem]

Date: 11/5/2004 10:11:51 PM
Author: Garry H (Cut Nut)
Marty did you see this link www.gia.edu/gemsandgemology/361/ig__gi_data_depository.cfm

It has quite a bit of additional detail (still not all we would like - but more )

Yah Gary I had looked at it before, it descibes the hemispheres they tested. What is more interesting is their picture Figure 7 showing the setup and someone using one of their hemispheres. What is apparent is that their modle Figure 10 is inconsistent with the picture in Figure 7 unless they had the stone on a black background.

Nevertheless, Figure 7 tells me that they are dealing with a photopic viewer with a small pupil size, probably under 3mm, but we don''t know the lux in the hemisphere. If my memory serves me right, the 1.5 degree "model" would be about right if all the stones were 6.5mm with a 4.1mm pupil at a 8" viewing distance, which ain''t the case.

 

[adamasgem]

The "problem" with a metric based on a limited "viewer" perspective is that it has to assume a fixed size stone, other wise one has to change the viewers position (height above the stone or pupil size) to get the same perspective, as a function of the size of the stone.

If the TOA is changed, because of the inherent mathematical non linearity of the problem, there is no guarantee that the relative performance from one stone cut design to another will remain the same. Stone #2 may become "better" than stone #1 and vice versa depending on size and viewer perspective (TOA).


Nevertheless, the TOA observed for each stone in the GIA study was NOT GOVERNED, in my opinion, by the 3deg or otherwise specified holes in their viewing hemispheres.

 

[adamasgem]

ght intensity for a light box tray using a Minolta Chrometer II I have, and found that the intensity at 8 inches or so was on the order of 120 foot candles give or take, which would result in Pupil diameters on the order of 2mm using the DeGroot formulation for about 350 cd/m^2.

So I calculated the TOA (double TOA to get GIA solid angle) and found, based on diameters extrapolated from the weights given for the GIA test stones, and a fixed viewing distance of 8", for a fixed pupil diameter of 2mm, the following variations in TOA.


I believe that the pupil diameter for GIA''s hemispheres will remain fairly constant with a particular viewer, as it appears that the primary driving force for their observations will be the light tray background that creates the scattered illumination in the hemisphere.


Each viewer of course, will have a different Pupil diameter for the same viewing conditions, sometimes dependent on age


I suspect that the stone was placed on a black tray in the hemisphere, at least I hope so.

Interesting data, I might say.. Comments anyone?

 

[Garry H (Cut Nut)]

Pretty sure it would be the grey tray Marty - they did say they found that better - probably too much contrast between the tray and the diamond?

Could you please label the axes' of your chart. It is hard to comment without knowing what it is.




(It is time we had a concerted campaign to save you from Kerryistic behaviour Marty - remember the KISS principle?)

 

[adamasgem]

Date: 11/6/2004 4:47:19 PM
Author: Garry H (Cut Nut)

Pretty sure it would be the grey tray Marty - they did say they found that better - probably too much contrast between the tray and the diamond?

Could you please label the axes'' of your chart. It is hard to comment without knowing what it is.






(It is time we had a concerted campaign to save you from Kerryistic behaviour Marty - remember the KISS principle?)

Gary I labeled the x axis as corresponding to the GIA models RD01 to RD46


The y axis is obviously then the TOA (Total Observation Angle) as I defined it previoulsy, sort of 1/2 of GIA 3 degree solid angle..


PLEEEEZE don''t accuse me of being John Kerry like..

That''s baaaaad..


I don''t flip flop like that

, even if I am from MA now, and used to know him in the 80''s when he first won the senate seat and was "dating" my next door neighbor who is now a judge.

 

[Garry H (Cut Nut)]

I have not got a clue what your x axis is Marty?

It is not the stone #

It is not the stone weight

It is not the stone diameter

 

[adamasgem]

Cutting and pasting didn''t work. Here are my model definitions.. The graph of TOA shows TOA calculated for models 60 to 104

 

[adamasgem]

RE: Pupil Diameter..

There a few online references regarding pupil diameter, other than in Wyszecki & Stiles..

One is from..http://www.blackwell-synergy.com/openurl?genre=article&sid=nlm:pubmed&issn=0275-5408&date=1995&volume=15&issue=6&spage=601

For field diameters of up to 25°, replotting the data in
terms of corneal flux density (i.e. the product of luminance and subtended
area) causes an approximate convergence onto a single function described
by D = 7.75 -5.75 [(F/846)^0.41/((F/846)^0.41 + 2)] [I added the missing negative sign and missing parenthasis]

where D is the pupil diameter (mm) and F is the corneal flux density (cd/m2 deg2).

This is consistent with the DeGroot formulation in Wyszecki & Stiles

IBM''s web site shows data from a 1920 article (Reeves) giving much larger pupil diameters vs illumination, which I believe is overshadowed by recent publications.

http://www.pc.ibm.com/ww/healthycomputing/vdt13eyec.html

 

[adamasgem]

I ran both the AGS 33 degree solid head angle and GIA 46 degree solid head angle with the cosine squared weighting function to get an idea of the relative differences in discrimiation from this type of metric. Pete Yanzter kindly suppiled me with DiamondCalc DMC files for the new GIA RD01-46 definitions, and had run DC metrics on those stones.

In comparing the data, normalized to RD01, it is obvious that a limited perspective viewpoint creates larger apparent differences between diamonds than a simple cosine squared metric, HOWEVER, one can note from the DiamondCalc data, that as the viewers angles changes from overhead to 30 degrees, there is a eduction in the apparent differences between stones.

Now, the DC 30 tilt model, I assume is ONE perspective, but what I realized and theorized is that 30 degree tilt metric (or AGS''s 15 degree observation point) will also change, even in a symmetric diamond, if the diamond is rotated about the table normal (the viewer is still at 30 degrees), how much it changes will have to be calculated, but it will show an 8 fold symmetry in the metric''s oscillations

This argues that a static viewpoint model may not be appropriate. In an assymetric stone, a metric from a viewers perspective not centered over the stones table, will probably have wider variations, as the stone is rotated about the table normal. (are you looking at the 34.0 or 34.5 degree main on the same stone)

 

[adamasgem]

Regarding Static Viewpoits..

I ran 1 million rays and binned the data in 1 degree increments for tilt (0 to 90 degrees) and one degree increments in azimuth (rotation around table to culet axis), with optical symmetry considerations I grouped used modulo 90 degree grouping, as I can do an out of round stone..

Now these data effectivly are for a situation for a +/- 0.5 degree TOA, and one can boxcar smooth 3 data points and get the 3 degree solid angle that GIA used..

Note the 10% peak to peak variation in the metric depending on the viewpoint azimuth.. The next post will look at the data averaged over all Azimuths, versus the tilt (look angle), with the same grouping. Some may not like the implications of the results I post here, as to static look angle representation of a diamonds performance, but the data is the data..

 

[adamasgem]

Here is the data sets averaged over all azimuths and plotted versus the tilt (or look angle), again the data is for a +/- 0.5 degree angle grouping.. Boxcar smoothing of 3 data points should emulate pretty well the conical grouping of backward ray traacing.

I think these data seriously call into question the used of static viewpoints.

As a note, to lessen the run time, I limited the number of internal facet interactions to 12, which I believe is probably consistent with the backward ray trace methodologies used. This resulted in an AVERAGE of 6 to 8% of the rays energy remaining when the forward ray trace terminated..

Comments anyone??

Especially note the LARGE shifts in energy metric at around 30 degrees.. These dramatic shifts occur at positions which are probably related to crown and pavilion main angles, although not entirely so..

 

[Serg]

Marty,

Just read your last posts.

Are you sure in your last plot?

Could you describe difference of condition between last two plots. Its are not clear for me.

 

[adamasgem]

Sergey.. All I did is bin the data by the 1 degree quantized spherical coordinates (phi,theta) of the exiting ray projection on a hemisphere.. (phi, theta), where phi is the azimuth.

Surprised the hell out of me as well when I can the first stone. I could easily rationalize the validity of the data averaged over all tilts versus Azimuth, but the magnitudes of the data versus tilt were unsuspected by me.

The second plot is the average data in a one degree annular ring at a given theta (or tilt angle with respect to the table

It is all the same data set..

I can send you an Excell file of the 1 degree grouped data for one of the runs..

I ran 1000000 rays at 12 facet interactions so in the most densely populated bins there were up to a 1000 samples in a bin.

 

[adamasgem]

On another note, regarding the GIA hemisphere tests, I think the "effective pupil" size is governed by the solid angle viewing port hole (3 degrees or whatever) and the hemisphere's radius (3,6, or 8 inches).

If you note the picture in the G&G article of the hemispheres they had some fairly sizable viewing ports, and since both the eye and the head move around in actuality (dither, technically) in thier cycloptic viewer, you average within the restricted view, so the actual Pupil diameter doesn't seem to be material to their metric modeling, but the stone size still does to compute a metric.

It certainly wasn't clear from the G&G article, but it actually makes sense in a way, although physically I think the TOA should be increased as not only is the head and eye moving, but the stone is not static.

They don't state what baseline viewing distance (hemisphere radius) is used for their brightness metric with a 3 degree "hole", but for an 8 inch radius and 6mm stone it works out to about a 2.34 degree TOA.

 

[Serg]

Date: 11/6/2004 2:21:11 PM
Author: adamasgem

I measured the light intensity for a light box tray using a Minolta Chrometer II I have, and found that the intensity at 8 inches or so was on the order of 120 foot candles give or take, which would result in Pupil diameters on the order of 2mm using the DeGroot formulation for about 350 cd/m^2.




So I calculated the TOA (double TOA to get GIA solid angle) and found, based on diameters extrapolated from the weights given for the GIA test stones, and a fixed viewing distance of 8'', for a fixed pupil diameter of 2mm, the following variations in TOA.




I believe that the pupil diameter for GIA''s hemispheres will remain fairly constant with a particular viewer, as it appears that the primary driving force for their observations will be the light tray background that creates the scattered illumination in the hemisphere.




Each viewer of course, will have a different Pupil diameter for the same viewing conditions, sometimes dependent on age




I suspect that the stone was placed on a black tray in the hemisphere, at least I hope so.

Interesting data, I might say.. Comments anyone?

fluctuation is +_10% only. On the basis of this plot all RD01 - RD46 should be similar, but its are quite different in real life .

 

[Serg]

Date: 11/10/2004 11

4:27 PM
Author: adamasgem

I ran both the AGS 33 degree solid head angle and GIA 46 degree solid head angle with the cosine squared weighting function to get an idea of the relative differences in discrimiation from this type of metric. Pete Yanzter kindly suppiled me with DiamondCalc DMC files for the new GIA RD01-46 definitions, and had run DC metrics on those stones.

In comparing the data, normalized to RD01, it is obvious that a limited perspective viewpoint creates larger apparent differences between diamonds than a simple cosine squared metric, HOWEVER, one can note from the DiamondCalc data, that as the viewers angles changes from overhead to 30 degrees, there is a eduction in the apparent differences between stones.

Now, the DC 30 tilt model, I assume is ONE perspective, but what I realized and theorized is that 30 degree tilt metric (or AGS''s 15 degree observation point) will also change, even in a symmetric diamond, if the diamond is rotated about the table normal (the viewer is still at 30 degrees), how much it changes will have to be calculated, but it will show an 8 fold symmetry in the metric''s oscillations

This argues that a static viewpoint model may not be appropriate. In an assymetric stone, a metric from a viewers perspective not centered over the stones table, will probably have wider variations, as the stone is rotated about the table normal. (are you looking at the 34.0 or 34.5 degree main on the same stone)

MOst interesting stones are RD 27 and RD 45( Biggest difference. )

For RD 27 there is very big difference even for new and old GIA metric.

 

[adamasgem]

Date: 11/21/2004 7:11:29 AM
Author: Serg

Date: 11/6/2004 2:21:11 PM
Author: adamasgem


I measured the light intensity for a light box tray using a Minolta Chrometer II I have, and found that the intensity at 8 inches or so was on the order of 120 foot candles give or take, which would result in Pupil diameters on the order of 2mm using the DeGroot formulation for about 350 cd/m^2.&nbsp;






So I calculated the TOA (double TOA to get GIA solid angle) and found, based on diameters extrapolated from the weights given for the GIA test stones, and a fixed viewing distance of 8'', for a fixed pupil diameter of 2mm, the following variations in TOA.






I believe that the pupil diameter for GIA''s hemispheres will remain fairly constant with a particular viewer, as it appears that the primary driving force for their observations will be the light tray background that creates the scattered illumination in the hemisphere.






Each viewer of course, will have a different Pupil diameter for the same viewing conditions, sometimes dependent on age






I suspect that the stone was placed on a black tray in the hemisphere, at least I hope so.

Interesting data, I might say.. Comments anyone?

fluctuation is +_10% only. On the basis of this plot all RD01 - RD46 should be similar, but its are quite different in real life .

Sergey.. The plot and commentary you are referring to deal with the total observation angle based on the stone size and pupil size, NOT the resulting metric..