Adding Symmetry to HCA scores

[gm89uk]

@Serg I applaud your dedication and thought process involved with every aspect of your analysis.

But this is a bit of a game changer for me. Cutwise cut performance metric appears to me to have poor correlation with HCA.

Your top ranking stone is: https://cutwise.com/diamond/29532?sp=47 with an HCA of 5.6. It's a 59/61.9 32.7/41.8 stone. I like the dynamic ASET that shows the area of leakage turn green on tilt, that may explain the discrepancy between static and dynamic scoring.

Just wondering:
1) There appears to be a weak correlation with carat size, the higher carat, the higher the cut performance. Do different carat sizes benefit from different proportions?

2) Does what you and your team's analysis undo what we all thought was a more sparkly diamond? There are plenty of 'standard ideal cut' tolks among the top rankers, but too many that are not to be ignored!

3) It almost appears that lack of optical symmetry can contribute to brilliance, fire and cut performance, particularly as carat size increase.

Tolk proportions are a safe bet, and easier to buy online. It seems some of the really well performing stones and the duds look really similar on paper.

Sorry for the thread jack, it may be better in a separate thread.

 

[Serg]

Btw. GIA grade is EX, and our cut grade is also EX. I have not any idea why an operator uploaded LAB cut grade VG, I have asked to check a reason and fix it.( most probably human mistake)

It was not a Cutwise operator mistake.
GIA have just changed cut grade. it was VG before 25 January 2019 , see old and new grading reports.
somebody recut diamonds( mass was 5.30ct , current mass is 5.29ct) to slightly reduce girdle valley thickness in 1-2 places.

 

[Serg]

@Serg I applaud your dedication and thought process involved with every aspect of your analysis.

But this is a bit of a game changer for me. Cutwise cut performance metric appears to me to have poor correlation with HCA.

Your top ranking stone is: https://cutwise.com/diamond/29532?sp=47 with an HCA of 5.6. It's a 59/61.9 32.7/41.8 stone. I like the dynamic ASET that shows the area of leakage turn green on tilt, that may explain the discrepancy between static and dynamic scoring.

.

Yes!, we ( I and Garry) were always disagree in such diamonds.
You could find many 5-10 years old PS discussion about such diamonds.
Garry names such diamond "dead ring".
I see huge fire and nice brilliancy in same diamonds.
These diamonds are defiantly are not good for Cyclops , but they are very nice for human observation in viewer distance range 30-50( may be more)cm.
Fire is great and there is not any black ring under table for stereo observer .

Cr32.7P41.8 optically is similar to Cr34.5P41.4
I prefer diamonds with slightly less pavilion and same crown like Cr34.5P41.2

Cr34.5P41.4 is very risky, Cr34.5P41.6 could have "Garry Dead Ring " even under stereo observation conditions.

btw. Main Cutwise destination is Hi-Vi( High Vibrancy ) fancy cuts .
We did these technology to help select and sell nice fancy cuts.
See for example Cushion cut variety.
https://cutwise.com/~J8zQ
RBC is very boring market which is very difficult to change. I do not like fight with AGS0, GIA EX, HCA, H&A, etc. I am working for Hi-Vi fancy cuts consumer market. RBC we use only as reference .

 

[Serg]

@Serg
Just wondering:
1) There appears to be a weak correlation with carat size, the higher carat, the higher the cut performance. Do different carat sizes benefit from different proportions?

we use 2 score types : Absolute and Relative( by area)
Absolute Fire strongly depends from Size.
Compare 2 RBC with similar proportions but with difference in mass in 2 times.
https://cutwise.com/compare/diamond-colorless?id[]=29494&id[]=24162
Bigger diamonds has absolute score 1.81, smaller diamond has 0.9, when the difference in relative Fire score is just 4%( 113% and 109%)

 

[blueMA]

we use 2 score types : Absolute and Relative( by area)
Absolute Fire strongly depends from Size.

This makes a perfect sense to me. It also correlates to real-world observation, so segregating and making available both data seem prudent, especially to compare different shapes/depth.

I kept the data of an interesting pair of diamonds I've observed in real life.
Table 63, depth 61.5, 35/41 HCA 4.1
Table 61, depth 61.9, 33/41.6 HCA 5.4

In all sense of knowledge under PS, these diamonds should've been rejects, but I must say they were absolutely stunning to look at in person. even compared to my super-sparkly H&A ring. I observed them through Idealscope and didn't even notice leakage.
@Serg 's posts here helps to finally make logical sense.

 

[Garry H (Cut Nut)]

[QUOTE="blueMA,
I kept the data of an interesting pair of diamonds I've observed in real life.
Table 63, depth 61.5, 35/41 HCA 4.1
Table 61, depth 61.9, 33/41.6 HCA 5.4
I observed them through Idealscope and didn't even notice leakage.
[/QUOTE]
I expect the stones were painted.

Yes, Sergey and me have had this debate for several years. His argument re stereo vision is good, but in my experience diamonds with those proportions look terrible when they get dirty (Sergey disagrees with me, we had that debate too).

AGS changed their rules and use ASET 32.5 degree obstruction in ASET (presumably because it enables them to show hearts patterns on their grading reports) which I think is nuts. 25cm / 8 inches may be the US military standard viewing distance but has not much to do with diamond enjoyment.

Add in the effect of stereo observation and there are 2 points I want to make:

1. as per Bruce Harding, the obstruction from one side of the head to the eye on the same side of the head is 8 to 14 degrees (approx depending on viewing distance) so 24 degree obstruction is even too much. I know there are stereo video issues Sergey, but I think it is an important factor. I think if you lowered the obstruction to say 15 degrees your result would become much closer to HCA.

2. As viewing distance become greater we become closer to virtual cyclops. I agree that each eye will still see different sparkles from different parts of a diamond in all but flood lighting (and more often in office lighting).

 

[Karl_K]

Pixel counters have a fixed location for lighting(location 1) and a fixed location for gathering that light(location 2).
Diamonds that gather light from location 1 and return it to location 2 the best are going to score higher on the machine.
How well that matches real world viewing is a huge question mark with pixel counters.
Is that diamond just better on the machine or is it better on the machine and in the real world????

 

[blueMA]

How well that matches real world viewing is a huge question mark with pixel counters.
Is that diamond just better on the machine or is it better on the machine and in the real world????

Interesting question indeed - because I've often wondered the same of the AGS method where the measurement criteria has been so directional and static. I've seen many LIVELY diamonds that perform so well in real life that wouldn't score so well under the AGS.

 

[Serg]

I expect the stones were painted.

Yes, Sergey and me have had this debate for several years. His argument re stereo vision is good, but in my experience diamonds with those proportions look terrible when they get dirty (Sergey disagrees with me, we had that debate too).

AGS changed their rules and use ASET 32.5 degree obstruction in ASET (presumably because it enables them to show hearts patterns on their grading reports) which I think is nuts. 25cm / 8 inches may be the US military standard viewing distance but has not much to do with diamond enjoyment.

Add in the effect of stereo observation and there are 2 points I want to make:

1. as per Bruce Harding, the obstruction from one side of the head to the eye on the same side of the head is 8 to 14 degrees (approx depending on viewing distance) so 24 degree obstruction is even too much. I know there are stereo video issues Sergey, but I think it is an important factor. I think if you lowered the obstruction to say 15 degrees your result would become much closer to HCA.

2. As viewing distance become greater we become closer to virtual cyclops. I agree that each eye will still see different sparkles from different parts of a diamond in all but flood lighting (and more often in office lighting).

Garry,
I am agree that dirty diamonds are ugly diamonds. It is not right penalty only one type proportions of round diamonds for this reasons. How have we penalty fancy cuts in such case?
there are 2 options :
a) we grade clean diamonds
b) we grade dirty diamonds
but we can not grade one diamonds in clean conditions and other diamonds in dirty conditions. we have to use same rules for all diamonds , it is main idea of our system

1) You missed my comments about low head obscuration. the human head obscuration is not only negative phenomena, it is very important part of Brilliancy . if one eye sees black facet and other eye sees same facet as white in same time then your brain sees great Brilliancy . it is come from stereoscopic rivalry .
if we use head obscuration 20 degree and less it is kill brilliancy in many diamonds.

I could agree with 24 degree but 15 degree is wrong conditions
2) Do you see "Dead ring" in diamonds P41.2Cr34.5 from 3m viewer distance? from 1m viewer distance? Sorry, I have not so fantastic vision.

 

[Garry H (Cut Nut)]

Now I am totally thread jacking my own thread.
Sergey, the camera lens is BLACK. Any way to replicate a human head colour / shade? I think the result would be quite different.
And even from a distance there is no way both eyes will see those dark stars around Tolkowsky and shallower.
And as we know - if one eye sees dark and the other sees bright then the metallic brilliance kicks in. And that is the real brilliance factor that is so hard to define.

Thinking about how one checks to see if the star symmetry pattern is present, absent a H&As or other scope, we form a 'tube' with our hand and look with one eye. No way can we see that star pattern with both eyes at close viewing distance.

 

[Serg]

Now I am totally thread jacking my own thread.
Sergey, the camera lens is BLACK. Any way to replicate a human head colour / shade? I think the result would be quite different.
And even from a distance there is no way both eyes will see those dark stars around Tolkowsky and shallower.
And as we know - if one eye sees dark and the other sees bright then the metallic brilliance kicks in. And that is the real brilliance factor that is so hard to define.

Thinking about how one checks to see if the star symmetry pattern is present, absent a H&As or other scope, we form a 'tube' with our hand and look with one eye. No way can we see that star pattern with both eyes at close viewing distance.

Garry,
Do you really believe that I have not know the difference between mono camera and human head obscurations ? I spent 20 years to add stereoscopic measurements to Diamond performance evaluation . We even developed stereo adapter for Vibox.( 10 years ago?)
Any approximation is approximation . It is truth for all available methods to measure diamond performance: FS, H&A, ASET, IS, BS, HCA , AGS, GIA ,.. and of course for Cutwise.
But only cutwise metrics use stereo movies to score diamond.( we have not yet stereo metrics for all type diamond beauty )
please check my early explanations here:
We compared our impression from Dibox stereo movies and real diamonds when we selected Dibox2.0 light environment included camera obscuration angle . We compared stereo with stereo!
Btw you also can see stereo movies for cutwise diamonds if you have 3D stereo TV. Just Please do it before you will try explain me the difference between camera and human Head obscurations . Better to see real facts before to do theoretical statements.

it is important to see Balance between similarity and difference.
You can do big mistake if just see a difference and ignore similarity.

And again,
1)Cutwise metrics is just approximation, but it is best available now approximation .
2) Cutwise metrics are good enough to create shortlist best for your diamonds from huge variety of different cuts.
You may compare Cushions, RBC, Ovals, Princess, Emeralds in same rule. Select short list then compare its with AGS0 round in any "real " light conditions if you wish.
3) Cutwise is way to improve Fancy cuts in short term. You will see more better fancy cuts a soon.
4) Cutwise is selection tool , instead rejections tools like HCA, ASET, IS,..

Final decision you always have to do by your eyes and by your brain because it have to match with your Taste.

 

[gm89uk]

Thanks @Serg for all your explanations.

What would be ideal would be if your system created a brilliancy or cut performance curve at several distances, and customers could select their viewing distance, based on whether they are shopping for an ering, necklace or studs.

Obviously this is unrealistic for now because it requires different hardware.

Isn't it possible to mathematically calculate the best viewing distance on a diamond (based on the scan of the diamond to calculate when its brilliance score is highest. From this you could make an 'obstruction score',a penalty for diamonds that have significantly reduced brilliancy at normal viewing conditions, something that is commonly asked about here.

Does cutwise equipment allow you to change the distance and obstruction or is that something is now set once the machinery has been created?

 

[Serg]

Thanks @Serg for all your explanations.

What would be ideal would be if your system created a brilliancy or cut performance curve at several distances, and customers could select their viewing distance, based on whether they are shopping for an ering, necklace or studs.

Obviously this is unrealistic for now because it requires different hardware.

Isn't it possible to mathematically calculate the best viewing distance on a diamond (based on the scan of the diamond to calculate when its brilliance score is highest. From this you could make an 'obstruction score',a penalty for diamonds that have significantly reduced brilliancy at normal viewing conditions, something that is commonly asked about here.

Does cutwise equipment allow you to change the distance and obstruction or is that something is now set once the machinery has been created?

It is very easy to reduce obscuration angle ( we do not need to change distance between camera and diamond) by adding extra ring in light bowl hole. ( we never did it, but I sure it is easy to do)
If somebody likes use several Office/fire light conditions with different obscuration angle we need replace ASET, H&A lights to Office/Fire light with different head obscurations.( otherwise he has to add/remove the ring for each diamond)
It is not difficult or expensive engineering task.
The problem is to explain to market a lot of different scores:
1) Absolute/Relative
2) Fire/Brilliancy
3) Short distance/Medium distance/Long distance
4) Symmetry Optical and 3D
5) Fluorescent, Color, Clarity
6) Milkiness
7) Mono, Stereo
8 ) Spread

For my opinion it is too much. If we will add different head obscuration angles it will even more scare retail and consumers .
May be latter. Market is not ready yet for such complexity .

 

[Serg]

It was not a Cutwise operator mistake.
GIA have just changed cut grade. it was VG before 25 January 2019 , see old and new grading reports.
somebody recut diamonds( mass was 5.30ct , current mass is 5.29ct) to slightly reduce girdle valley thickness in 1-2 places.

Final confirmation that grade had been changed after recut
please find:
models comparisons between 5.30ct and 5.29ct( after recut)
several crown halves had been repolished with angles changing 0.5-1degree

and Dibox ASET photo comparison with photoreal image for model 5.30ct

 

[sledge]

It is very easy to reduce obscuration angle ( we do not need to change distance between camera and diamond) by adding extra ring in light bowl hole. ( we never did it, but I sure it is easy to do)
If somebody likes use several Office/fire light conditions with different obscuration angle we need replace ASET, H&A lights to Office/Fire light with different head obscurations.( otherwise he has to add/remove the ring for each diamond)
It is not difficult or expensive engineering task.
The problem is to explain to market a lot of different scores:
1) Absolute/Relative
2) Fire/Brilliancy
3) Short distance/Medium distance/Long distance
4) Symmetry Optical and 3D
5) Fluorescent, Color, Clarity
6) Milkiness
7) Mono, Stereo
8 ) Spread

For my opinion it is too much. If we will add different head obscuration angles it will even more scare retail and consumers .
May be latter. Market is not ready yet for such complexity .

This is a sad reality.

We talk about the way technology can improve in diamonds but when I look around the most prevalent thing I see is the need for consumer education.

People of all intelligence levels and walks of life buy diamonds. A very small percentage of them have advanced analytical abilities to discern cut quality. Additionally most buyers won't put in the effort to learn the data either. Big, shiny and fits budget seems to be the current criteria.

It's my opinion that using advanced technology and presenting in a user friendly format is the key. I do think most buyers could comprehend and appreciate seeing that a diamond scores 9 in this lighting environment and 6 in another condition for an overall 8 score.

You have to keep the metrics easy, 1-10 for example and the descriptions just as easy. Maybe office, candlelight, sunny day, cloudy day, etc. Presentation has to also be clean and simple.

We can learn lots from history. I admired the works of Steve Jobs at Apple. He pushed his people to take complex data and make simple and easy. He then marketed and sold it based on ease, beauty and the cool factor.

That is the trick here as well, taking complex data and making it digestible to the hot shot engineer, janitor, old lady, young soldier, etc.

 

[OoohShiny]

This is a sad reality.

We talk about the way technology can improve in diamonds but when I look around the most prevalent thing I see is the need for consumer education.

People of all intelligence levels and walks of life buy diamonds. A very small percentage of them have advanced analytical abilities to discern cut quality. Additionally most buyers won't put in the effort to learn the data either. Big, shiny and fits budget seems to be the current criteria.

It's my opinion that using advanced technology and presenting in a user friendly format is the key. I do think most buyers could comprehend and appreciate seeing that a diamond scores 9 in this lighting environment and 6 in another condition for an overall 8 score.

You have to keep the metrics easy, 1-10 for example and the descriptions just as easy. Maybe office, candlelight, sunny day, cloudy day, etc. Presentation has to also be clean and simple.

We can learn lots from history. I admired the works of Steve Jobs at Apple. He pushed his people to take complex data and make simple and easy. He then marketed and sold it based on ease, beauty and the cool factor.

That is the trick here as well, taking complex data and making it digestible to the hot shot engineer, janitor, old lady, young soldier, etc.

Serg has lamented the difficulty of creating a simple language and tool for improving customer understanding and assessment of diamonds in a previous thread! I'm not sure if you've seen it but you might like to kickstart discussions again
https://www.pricescope.com/communit...th-high-brilliancy-fire-scintillation.240693/

 

[sledge]

Serg has lamented the difficulty of creating a simple language and tool for improving customer understanding and assessment of diamonds in a previous thread! I'm not sure if you've seen it but you might like to kickstart discussions again
https://www.pricescope.com/communit...th-high-brilliancy-fire-scintillation.240693/

Thank you for sharing. I haven't seen yet but will take a peek.

 

[blueMA]

Thank you for sharing. I haven't seen yet but will take a peek.

Of a special interest, you may stumble on @Miki Moto's post where she couldn't understand why she absolutely loves her Tiffany 3.15ct imperfect 57 table, 60.8 depth, 33.7 crown, 43.9 pav depth% (prob 41.3 pav angle) with HCA 3.3 compared to her superideal Hearts On Fire two diamonds (HCA 0.9 and 1.2) that she absolutely hated. I haven't seen her on PS as of recent but I think she'd find @Serg's posts here quite interesting.

To quote her - "I have often wondered why a non-super ideal would sparkle more than a super-ideal."

https://www.pricescope.com/communit...uper-ideal-diamond.174235/page-2#post-4324607

 

[Athena10X]

Garry,
Do you really believe that I have not know the difference between mono camera and human head obscurations ? I spent 20 years to add stereoscopic measurements to Diamond performance evaluation . We even developed stereo adapter for Vibox.( 10 years ago?)
Any approximation is approximation . It is truth for all available methods to measure diamond performance: FS, H&A, ASET, IS, BS, HCA , AGS, GIA ,.. and of course for Cutwise.
But only cutwise metrics use stereo movies to score diamond.( we have not yet stereo metrics for all type diamond beauty )
please check my early explanations here:
We compared our impression from Dibox stereo movies and real diamonds when we selected Dibox2.0 light environment included camera obscuration angle . We compared stereo with stereo!
Btw you also can see stereo movies for cutwise diamonds if you have 3D stereo TV. Just Please do it before you will try explain me the difference between camera and human Head obscurations . Better to see real facts before to do theoretical statements.

it is important to see Balance between similarity and difference.
You can do big mistake if just see a difference and ignore similarity.

And again,
1)Cutwise metrics is just approximation, but it is best available now approximation .
2) Cutwise metrics are good enough to create shortlist best for your diamonds from huge variety of different cuts.
You may compare Cushions, RBC, Ovals, Princess, Emeralds in same rule. Select short list then compare its with AGS0 round in any "real " light conditions if you wish.
3) Cutwise is way to improve Fancy cuts in short term. You will see more better fancy cuts a soon.
4) Cutwise is selection tool , instead rejections tools like HCA, ASET, IS,..

Final decision you always have to do by your eyes and by your brain because it have to match with your Taste.

If the DiBox 2.0 3D video could be made VR compatible, maybe retail stores could showcase diamonds in 3D via Oculus or Samsung Gear VR glasses. The consumer could select different lighting conditions in the VR environment and see the diamond’s estimated performance based on the metrics. Seeing is believing (especially something tangible) and more powerful then merely reading stats and trying to convert numbers into something visual. Maybe I’m thinking too far outside the box.

 

[blueMA]

If the DiBox 2.0 3D video could be made VR compatible, maybe retail stores could showcase diamonds in 3D via Oculus or Samsung Gear VR glasses. The consumer could select different lighting conditions in the VR environment and see the diamond’s estimated performance based on the metrics. Seeing is believing (especially something tangible) and more powerful then merely reading stats and trying to convert numbers into something visual. Maybe I’m thinking too far outside the box.

Actually love that VR idea. Diamond shopping in the future may become of such.

 

[OoohShiny]

If the DiBox 2.0 3D video could be made VR compatible, maybe retail stores could showcase diamonds in 3D via Oculus or Samsung Gear VR glasses. The consumer could select different lighting conditions in the VR environment and see the diamond’s estimated performance based on the metrics. Seeing is believing (especially something tangible) and more powerful then merely reading stats and trying to convert numbers into something visual. Maybe I’m thinking too far outside the box.

I'm guessing you mean diamonds in virtual inventory / held off-site, as surely one would look at the diamond itself if it was in the shop... lol

 

[Serg]

If the DiBox 2.0 3D video could be made VR compatible, maybe retail stores could showcase diamonds in 3D via Oculus or Samsung Gear VR glasses. The consumer could select different lighting conditions in the VR environment and see the diamond’s estimated performance based on the metrics. Seeing is believing (especially something tangible) and more powerful then merely reading stats and trying to convert numbers into something visual. Maybe I’m thinking too far outside the box.

Oculus and even HTC Vive pro have not good enough resolution for such applications yet.
3D Helmets with necessary resolution could become available in 2 years .
Google Cardboard with 4K phones is easiest now way to see Dibox stereo movies .

 

[gm89uk]

You can see the diamonds in 3D at home by clicking the 3D button. You make yourself crosseyed until you see 3 pictures from 2. Then relax your accommodation to see in 3D looking at the middle picture. It works but is hard to do.

http://www.stereoscopicsociety.org.uk/WordPress/resources-2/viewing-stereoscopic-images/

These are nice examples.

https://digital-photography-school.com/9-crazy-cross-eye-3d-photography-images-and-how-to-make-them/

cutwise has a similar feature.

 

[blueMA]

You can see the diamonds in 3D at home by clicking the 3D button. You make yourself crosseyed until you see 3 pictures from 2. Then relax your accommodation to see in 3D looking at the middle picture. It works but is hard to do.

http://www.stereoscopicsociety.org.uk/WordPress/resources-2/viewing-stereoscopic-images/

These are nice examples.

https://digital-photography-school.com/9-crazy-cross-eye-3d-photography-images-and-how-to-make-them/

cutwise has a similar feature.

That's pretty cool! I have a bit of a headache now!
It really helps to have a pen tip or an object in front to keep crosseyed because my wanting to focus on the image breaks the 3D otherwise.

 

[Athena10X]

I'm guessing you mean diamonds in virtual inventory / held off-site, as surely one would look at the diamond itself if it was in the shop... lol

Oculus and even HTC Vive pro have not good enough resolution for such applications yet.
3D Helmets with necessary resolution could become available in 2 years .
Google Cardboard with 4K phones is easiest now way to see Dibox stereo movies .

Yep, for virtual/online inventory. I have a pair of Google cardboard, just not a 4K phone. This would make for a fun project!

 

[OoohShiny]

That's pretty cool! I have a bit of a headache now!
It really helps to have a pen tip or an object in front to keep crosseyed because my wanting to focus on the image breaks the 3D otherwise.

It's basically the same as those Hidden Eye images that were all the rage in the late 90s lol

 

[blueMA]

It's basically the same as those Hidden Eye images that were all the rage in the late 90s lol

I had absolutely no trouble seeing those Hidden Eye images back in the 90s.

These 3D images needed a bit of work, following the instruction to start with a pen tip from the monitor towards me to the midway.
Now I can see the 3D so clearly it's quite amazing. I especially like this image for a practice.

 

[Karl_K]

Interesting question indeed - because I've often wondered the same of the AGS method where the measurement criteria has been so directional and static. I've seen many LIVELY diamonds that perform so well in real life that wouldn't score so well under the AGS.

One could certainly design a diamond that sucked in the real world using ASET principles.
It is easy to design a stone that beats a super-ideal MRB at any one given thing, designing a all around stone is what is hard.
For example, want a stone that rocks to the top in candle light and is round, cut an optimized OEC for that lighting.
It will blow away the ideal cut in that lighting, but in other lighting maybe not so much.
Everything is a compromise!

 

[blueMA]

Yep, for virtual/online inventory. I have a pair of Google cardboard, just not a 4K phone. This would make for a fun project!

I can see the 3D diamond comparison just cross eyed! Quite interesting.

 

[Serg]

I suggest to train your brain see stereo on DF( darkfield) movies firstly
like https://cutwise.com/diamond/333?sp=51

then Office light( brilliancy), then Fire.

it could be hard to start from Fire movies