AGS new cut grade system early 2005

[adamasgem]

----------------
On 9/29/2004 12:52:01 PM Serg wrote:

I hope this diamond has same parameters except upper girdle facets.----------------

Sergey, Thanks. Here is a composite of the two pics, unfortunately I can't make an overlay. There appears to be a lot more going then just what you noted regarding the lower part of the hemisphere

 

[adamasgem]

----------------
On 9/28/2004 1:18:42 PM Serg wrote:

----------------
re: What I am trying to figure out is in my pics is the why for the better the optical symmetry, the purer the hues with more broadflash.

I think You could see this phenomena only If you use a lot of small sources light at the same time .

Asymmetrical diamond produce much more possibility to see small light source . This and because size of image of light source could more then size of virtual facet are reason why in one point you could see light from different light sources.
If you use normal quantity of light source this phenomena is very rare( impossible)----------------

Go stand under a tree, not so impossible. The problem is if you limit your analyses to a small, fixed number of light sources then you potentially alias your result. Regarding "Asymmetrical diamond produce much more possibility to see small light source", well I agree that assymetrical diamonds will create smaller patches of fire, but wouldn't you agree that the smaller patches are less resolvable to the eye, the pinfire opal

 

[Serg]

Re:" well I agree that assymetrical diamonds will create smaller patches of fire, but wouldn't you agree that the smaller patches are less resolvable to the eye, the pinfire opal"

Not only size, color and brightness of flashes but also quantity of flashes are important for positive human perception of a diamond. Average quantity of simultaneous flashes and duration of a period of time when you don't see any flash are very important for person too.

Certainly, size and color are important but time and quantity play not less important role in perception of diamond beauty.

If you find for *8 diamond of 1ct that flashes have perfect color and enough size then you have possibility to increase number of facets for the stone with diameter in 2 times more than previous. In such way you increase probability to see the flash without worsening of quality of this flash in comparison with 1ct *8 diamond.

Of course, some consumers during transition from 1ct to 8ct will prefer to save increasing of flash size instead of additional increasing of probability to see more flashes. But try to check the same transition from 8ct to 64ct conformably to these consumers. Here the wish to see more flashes surely will arise. Indirect indication that this wish will arise near 10ct is the practice not to use classical cut for such big stones. Developers make new cuts with greater number of facets for such big stones. 13ct *8 diamond is not the rule on the market. It is exception.

The reason of this phenomenon is far from attempts to get big yield, although the aspirasion (wish) to save the weight for big stones surely is present and sometimes this aspirasion leads to very ugly results (note: usually cutting with less number of facets).

Scintillation brings even more contribution to demand to increase the number of facets because the color of flash is not important for Scintillation, but the number of flashes, their distribution and changing are very important.

Draw your attention to the number of virtual facets on the crown of Princess cut. What times their size is less than the size of virtual facets of round cut? But this is don't prevent you to enjoy wounderfull Scintillation of 1-2ct Princesses?

I am not the enemy of big facets. I am the follower of right balance between the size of diamond and sizes of virtual facets. Most likely the cut must have two types of facets to open full potential of the diamond: big and small facets. These facets of two types sholud be evenly distributed on the surface. Big facets are necessary to see PERIODICALLY (from time to time) big color flashes (Fire); small facets are necessary to see CONSTANTLY some moving flashes (Scintillation).


Marty,
and now what do you think how I should answer on your question: YES or NO?


P.S. I hope this English is more understood for anybody because this text is translated not by me.

 

[adamasgem]

----------------
On 9/30/2004 12:28:22 PM Serg wrote:

Re:' well I agree that assymetrical diamonds will create smaller patches of fire, but wouldn't you agree that the smaller patches are less resolvable to the eye, the pinfire opal'

Not only size, color and brightness of flashes but also quantity of flashes are important for positive human perception of a diamond. Average quantity of simultaneous flashes and duration of a period of time when you don't see any flash are very important for person too.


I am not the enemy of big facets. I am the follower of right balance between the size of diamond and sizes of virtual facets. Most likely the cut must have two types of facets to open full potential of the diamond: big and small facets. These facets of two types sholud be evenly distributed on the surface. Big facets are necessary to see PERIODICALLY (from time to time) big color flashes (Fire); small facets are necessary to see CONSTANTLY some moving flashes (Scintillation).


Marty,
and now what do you think how I should answer on your question: YES or NO?

Well, In my opinion, for the same size stone, SMALL=LESS RESOLVABLE, and you get small with lack of optical symmetry, what is the sense of more scintilation if you can't see it at normal viewing distances

P.S. I hope this English is more understood for anybody because this text is translated not by me. ----------------
The English is just fine, as is your work in DiamondCalc

 

[adamasgem]

----------------
On 9/30/2004 12:28:22 PM Serg wrote:

Re:' well I agree that assymetrical diamonds will create smaller patches of fire, but wouldn't you agree that the smaller patches are less resolvable to the eye, the pinfire opal'

Not only size, color and brightness of flashes but also quantity of flashes are important for positive human perception of a diamond. Average quantity of simultaneous flashes and duration of a period of time when you don't see any flash are very important for person too.

blockquote>

Sergey To better understand the problem, I modified the SAS2000 3D Raytrace software to compute and display (in any perspective), chromatic flares. The sample attached is for the Tolkowsky cut, 3% girdle, as GIA used. The picture I show would be that of an observer looking down on the projected hemisphere, but I can also show profile views. I used 8 fold symmetry to generate this pic but since real stones are assymetric, I can not use the symmetry aspect if so desired.

In playing with my software, I had the following observations:

1) The NUMBER of samples used effects the chromatic flare pictures, but more importantly, the position of the samples with respect to facets SIGNIFICANTLY effects the "picture" and any attendent analyses. This is probably the most important criterion for assessing the relative goodness of a particular cut.

2) I've used a parallel wavefront on the entire stone, with illuminant C for these pics, but can do it for virtually any source type. What I noticed in the GIA Fire analyses, was that they used spot size as one of their criteria for DCLR using real pictures in their 16' hemisphere, but I believe that the spot size is directly related to the collimation of the wavefront and relative intensity to the illumination effects, which I can also model.

3) Because of the input position's sensitivity to generating chromatic flare,
I question (without any real justification) the accuracy of the use of "beams" or "cones" of light rather than "rays", in assessing relative performance from one cut. I haven't convinced myself one way or another.
However, in the Monte Carlo analyses I perform, I wonder how many "rays" are enough. I guess I'll have to do convergence studies to see how fast the "answers" settle down, like I did for the WLR, and maybe play with the backwards ray tracing technique. I looked at that before but was uncertain on how many interactions to go back.

4) Of course, one of the PRIMARY problems is how to weight the results for the human perspective and Field Of View (distance, scotopic or photopic viewer, rod and cone sensitivities, etc, which I can experiment with in the SAS2000 software, It will be inetresting to compare results, like I have with you in the past, as others are reluctant to "share" detailed comparisons (They don't like me throwing a monkey wrench into their gear works).

It is an interesting problem, that GIA has spent the last 15 years or so playing with and spent untold millions as well. As a one man band, given what I've been able to do in a relatively short period of time since I decided to attack the problem, I can only wonder how efficient some are.

How many man years has your group spent in developing the DiamondCalc Software, may I ask?

 

[Serg]

Marty,

At first I suggest to reach an approximate coincidence of pictures and only after that we can discuss conclusions.

I need to know sizes of used by you hemisphere and diamond to model. If size of hemisphere is greatly more than size of diamond then "Samples" are too small for used by you light source.

Below I publish data for diamond: Pv=40.75; Cr=34.5; D=6.00ìì ;T=57; GirdleBasel=3%, Star=50%, LF=75%( GIALF=73%). Radius of hemisphere 1000 mm. Distance from diamond to light source 600 mm. Size of light source 4 mm (if I don't decrease sizes of light source to model yours variant of light source then "Samples" are very small and are visible bad. May be you use smaller size of hemisphere?)

I is most easily to check primary scale and parameters of cut due to primary reflections (without refraction).

That's why I publish pictures for quantity of inner reflections: 0, 4, 12.

 

[Serg]

4

 

[Serg]

12

 

[Serg]

----------------
On 10/5/2004 2:03:46 PM adamasgem wrote:

----------------
On 9/30/2004 12:28:22 PM Serg wrote:

Re:' well I agree that assymetrical diamonds will create smaller patches of fire, but wouldn't you agree that the smaller patches are less resolvable to the eye, the pinfire opal'



blockquote>



It is an interesting problem, that GIA has spent the last 15 years or so playing with and spent untold millions as well. As a one man band, given what I've been able to do in a relatively short period of time since I decided to attack the problem, I can only wonder how efficient some are.

How many man years has your group spent in developing the DiamondCalc Software, may I ask?


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

Project DC. It took 8 person-year, perhaps 7 or 10 (this period is without calculation of testing with real objects. Photos and preparing to photography took very much time).

Half of year we spent on Beam tracing, 2 years on ConeTracing, 30% - GUI, near 30% we spent on development (with testing) different BLR taking into account features of human perception. + a lot of time we spent to support various stereo-eyes in vain. Unfortunately good stereo-eyes with good drivers didn't became popular.

 

[adamasgem]

----------------
On 10/6/2004 9:17:29 AM Serg wrote:

Marty,

At first I suggest to reach an approximate coincidence of pictures and only after that we can discuss conclusions.I hope our differing methodologies will allow that comparison

I need to know sizes of used by you hemisphere and diamond to model. If size of hemisphere is greatly more than size of diamond then 'Samples' are too small for used by you light source.
I can change size of "hemisphere" used for projection, but it is all relative

Below I publish data for diamond: Pv=40.75; Cr=34.5; D=6.00ìì ;T=57; GirdleBasel=3%, Star=50%, LF=75%( GIALF=73%). Radius of hemisphere 1000 mm. Distance from diamond to light source 600 mm. Size of light source 4 mm (if I don't decrease sizes of light source to model yours variant of light source then 'Samples' are very small and are visible bad. May be you use smaller size of hemisphere?)
GIA and I saturate the stone with parallel rays meaning every point can get a ray, although I have looked at(and can change size) of a "gemprint" using a 1mm spot size. I use a 6.53mm diameter stone although I can rerun with a 6 mm size and a 4mm spot

I is most easily to check primary scale and parameters of cut due to primary reflections (without refraction).
Forgive me, I don't understand

That's why I publish pictures for quantity of inner reflections: 0, 4, 12.----
Here is the problem, the pattern we see is a function of where on the stone the incident ray hits, so somehow we may have to correlate that. Let me run your stone with a 4mm spot and see what I get
------------

 

[adamasgem]

----------------
On 10/6/2004 9:40:11 AM Serg wrote:

----------------
On 10/5/2004 2:03:46 PM adamasgem wrote:

----------------
On 9/30/2004 12:28:22 PM Serg wrote:

Re:


It is an interesting problem, that GIA has spent the last 15 years or so playing with and spent untold millions as well. As a one man band, given what I've been able to do in a relatively short period of time since I decided to attack the problem, I can only wonder how efficient some are.

How many man years has your group spent in developing the DiamondCalc Software, may I ask?


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

Project DC. It took 8 person-year, perhaps 7 or 10 (this period is without calculation of testing with real objects. Photos and preparing to photography took very much time).

Interesting, if DiamondCalc took 8 man years (in the US that would be over $1,000,000 in engineering time), then I wonder how long it would take to duplicate my entire SAS2000 Spectrophotometer Analysis Software package, and what would that cost be? Sort of a business valuation problem

I haven't really played with the backward ray tracing yet, let me think about that problem.


Half of year we spent on Beam tracing, 2 years on ConeTracing, 30% - GUI, near 30% we spent on development (with testing) different BLR taking into account features of human perception. + a lot of time we spent to support various stereo-eyes in vain. Unfortunately good stereo-eyes with good drivers didn't became popular.

The stereo eye thing would be neat, but as you imply might be overkill, and people wouldn't appreciate the amount of work that went into it.
----------------

 

[Serg]

Re:GIA and I saturate the stone with parallel rays meaning every point can get a ray, although I have looked at(and can change size) of a "gemprint" using a 1mm spot size. I use a 6.53mm diameter stone although I can rerun with a 6 mm size and a 4mm spot


Marty,

We do not use parallel rays.

Each point of diamond get a light for any size light source!

 

[Garry H (Cut Nut)]

RE - The stereo eye thing would be neat, but as you imply might be overkill, and people wouldn't appreciate the amount of work that went into it.

What Sergey refers to is glasses that are made of LCD crystals that alternate dark and light at about 50 hertz. The computer screen sends a signal that alternates the image and shuts / blackens one lens, and then the other so that you can see full color stereoscopic on the computer screen. (far better than the red blue glasses)

It is really cool - but the technology and programming are both very expensive - perhaps if it ever becomes main stream and made as a standard feature on Plasma TV's - it might become more than a really cool expensive novelty.

Hope that was a good explanation Sergey

 

[adamasgem]

----------------
On 10/7/2004 1:22:45 AM Serg wrote:

Re:GIA and I saturate the stone with parallel rays meaning every point can get a ray, although I have looked at(and can change size) of a 'gemprint' using a 1mm spot size. I use a 6.53mm diameter stone although I can rerun with a 6 mm size and a 4mm spot


Marty,

We do not use parallel rays.

Each point of diamond get a light for any size light source!----------------

OK, Great. That is the correct way of doing it, rather than the parallel rays which alias the results. You have to have a lot of rays using monte carlo to do that, but I'll generate something with a 10 degree observer to try it out. Then to get the pictures without showing every possible flare do you limit the Field of View (FOV) and only plot rays coming from the limited FOV.

What is your "observer definition.. I can do that by Pupil Size and Distance from stone. I can also modify the intensity by the scotopic or photopic sensitivities. Do you also use those sensitivities?

 

[adamasgem]

----------------
On 10/7/2004 3:41:35 AM Garry H (Cut Nut) wrote:

RE - The stereo eye thing would be neat, but as you imply might be overkill, and people wouldn't appreciate the amount of work that went into it.

What Sergey refers to is glasses that are made of LCD crystals that alternate dark and light at about 50 hertz. The computer screen sends a signal that alternates the image and shuts / blackens one lens, and then the other so that you can see full color stereoscopic on the computer screen. (far better than the red blue glasses)

It is really cool - but the technology and programming are both very expensive - perhaps if it ever becomes main stream and made as a standard feature on Plasma TV's - it might become more than a really cool expensive novelty.

Hope that was a good explanation Sergey

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

NEATO but a "little" expensive for our industry

 

[adamasgem]

----------------
On 10/7/2004 1:22:45 AM Serg wrote:

Re:GIA and I saturate the stone with parallel rays meaning every point can get a ray, although I have looked at(and can change size) of a 'gemprint' using a 1mm spot size. I use a 6.53mm diameter stone although I can rerun with a 6 mm size and a 4mm spot


Marty,

We do not use parallel rays.

Each point of diamond get a light for any size light source!----------------

OK Sergey, I'm totally confused..

1) You previously stated you used a 4mm spot on a 6mm stone.. I take that to mean that ONLY the center 4mm of the stone is illuminated.. Is that a correct interpretation?

2) When I think of a "spot" illumination with regard to Fire I think of a "parallel" wavefront like GIA said they modeled in their 16" hemisphere physical model..

I do believe Fire should be NOT be evaluated in this manner, but Fire from any and ALL randomly oriented INCIDENT rays should evaluated.

3) For qualitative display or comparison one needs to limit the number and define the position and angle of incidence of rays used to generate chromatic flares, beacuse you are going to get "random" pictures if you don't do that..


I have the capability to limit the position of INPUT rays to the centroid of each facet on the crown. That eliminates one variable in any comparison.

Then we have to consider the direction of the incident ray with respect to the normal of the facet hit..


You showed 4 pics of flares labeled 1, 4 and 12.. Does that mean, only 1 4 and 12 internal relections??

 

[Serg]

re:1) You previously stated you used a 4mm spot on a 6mm stone.. I take that to mean that ONLY the center 4mm of the stone is illuminated.. Is that a correct interpretation?

No. If you use 100mm bulb , is it means that only 100mm your room is illuminated?

re:You showed 4 pics of flares labeled 1, 4 and 12.. Does that mean, only 1 4 and 12 internal relections??

0, 4, 12 internal reflections.

0 is means external reflection

 

[adamasgem]

----------------
On 10/7/2004 2:31:54 PM Serg wrote:

re:1) You previously stated you used a 4mm spot on a 6mm stone.. I take that to mean that ONLY the center 4mm of the stone is illuminated.. Is that a correct interpretation?

No. If you use 100mm bulb , is it means that only 100mm your room is illuminated?

OK, I think I get it now. I reread your original post
4mm diameter light source 600 mm away from 6mm stone

re:You showed 4 pics of flares labeled 1, 4 and 12.. Does that mean, only 1 4 and 12 internal relections??

0, 4, 12 internal reflections.

0 is means external reflection----------------

Thanks Sergey

 

[adamasgem]

Sergey

1) 300 Random Input rays, flares plotted using 8 flod symmetry
2) 10nm wavelength separation 400 to 700 nm
3) 12 internal relection cutoff, all potential flares generated by each ray exiting crown
4) MSU Defined 6mm diamond
5) 4mm spot illumination 600mm from diamond
6)Cummulative intensity using RGB(100,100,100) = white , RGB(255,255,255) is pure white
7) no Scotopic/Photopic viewer intensity attenuation

 

[adamasgem]

I used a reduced jpeg quality to save the previous pic, I'll do a new simulation with the same conditions and save it as a gif as I think the jpeg altered the colors too much

 

[adamasgem]

Here is another 300 sample simulation using a gif format..

 

[Garry H (Cut Nut)]

Marty Sergey used 0 4 and 12.
0 = did not refract - only reflections.

I am curious as to why your flare patterns are not 8 fold symmetrical?

What's with the cyclonic swirl?

 

[adamasgem]

----------------
On 10/7/2004 9:26:06 PM Garry H (Cut Nut) wrote:

Marty Sergey used 0 4 and 12.
0 = did not refract - only reflections.

I am curious as to why your flare patterns are not 8 fold symmetrical?

What's with the cyclonic swirl?----------------

Gary. They are 8 Fold symmetric, that is the way I plot them, but since I've only plotted the effects of 300 rays I may not catch the mirror image. By 8 fold symmetric I maen I just rotate each point by 45 degrees. I could also flip the initial (x,y,z) point to (x,-y,z) (mirror image) and rotate that by 45 degree increments(45,90,135,180,225,270 and 315 degree rotations, that would be legitimate also, good idea Gary, quick fix

 

[Garry H (Cut Nut)]

Marty, use English please.
Simplifying what your words will achieve 2 things:
1. more people will read the thread.
2. more people from other technical fields and experiance may contribute.

what I was trying to explain is shown in these two cut and pastes. the one on the right is from 1/3rd way out from the center.
If the stone model is symmetrical, then the flashes should be symmetrical.

 

[adamasgem]

----------------
On 10/8/2004 12:48:25 AM Garry H (Cut Nut) wrote:

Marty, use English please.
Simplifying what your words will achieve 2 things:
1. more people will read the thread.
2. more people from other technical fields and experiance may contribute.

what I was trying to explain is shown in these two cut and pastes. the one on the right is from 1/3rd way out from the center.
If the stone model is symmetrical, then the flashes should be symmetrical.

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

I know what you meant Gary.

Perhaps I'm talking in not the Queen's English. Is this picture better with what I call an 8 fold mirror symmetry

 

[adamasgem]

I might note that an 8 fold Mirror or 8 fold symmetry is ONLY applicable for a SYMMETRIC illumination assumption on the diamond. The minute the spot (or illimination) is off center or assymetric all flare symmetry assumptions go out the window.

 

[Garry H (Cut Nut)]

This is Sergey's 12 and your most recent Marty.
Want to explain the differences?

 

[Serg]

Marty,
We should compare our results step by step.

Please send you image with first reflection only.

 

[adamasgem]

----------------
On 10/8/2004 3:09:27 AM Serg wrote:

Marty,
We should compare our results step by step.

Please send you image with first reflection only.----------------

Do you mean glare, If you do there is no chromatic flare from glare.
As to your meaning of "first reflection", the results are dependent on the point of incidence on the stone.
I have an option in which I can limit the point of incidence to the centroid of any facet, then the angle of incidence variation will be small, unless you want to fix the source illumination at (x,y,z) = (0,0,600mm) which would also eliminate the angle of incidence variation. That would be an ideal solution, although it would not test the table. I use a 0.5% culet also.

Also, when I compute, using double precision arithmatic, the angles of the facet planes with respect to the table plane, based on the computed facet plane coefficients (A,B,C,D; where A*x + B*y +C*z +D =0), I get variations in the 5th or 6th decimal place from what it desired, i.e. 34.5 degree crown angle may be 34.9999996 for 4 mains and 35.000001 for the other four.. have you checked that in your simulation. I don't know how much that effect the result but it is virtually impossible to overcome.

I think this public cross checking is a very good process, although some might get lost, there are a lot out here who purchase diamonds that have very good technical education and will enjoy it; my clients for diamond evaluation seem to be very educated, as the vast majority come from the internet.

 

[Serg]

reo you mean glare, If you do there is no chromatic flare from glare.

Main flashes after 4 "refractions" could have much bigger intensity. Yes, I ask about first reflection( 0 refraction), and first reflection should gray( although reflectivity factor depends from wave-length .

)

re:As to your meaning of "first reflection", the results are dependent on the point of incidence on the stone.

Please use same algorithm for this test.

Why do you want use special algorithm for first reflection?