Star Facet Length in RBC

[V_sh]

Hello friends,

I've just been curious about Star Length out of the GIA suggested range (45-65%) and did some search about it and found two nice articles, and I would be happy to know your opinions.

in the following charts, WLR stands for "Weighted Light Return" which indicates overall Brilliance
and DCLR stands for "Dispersed Colored Light Return" which is a metric for Fire








The following statements are about Diamonds with a Main Crown of 34-35 degrees


it's obvious that there is not a big loss if we go for a Star length out of the GIA accepted range; for example, a Star length of 65-70% and also 35-40% seems almost as good or better than a Star length of 50%

I also questioned that maybe the physical appearance of the diamond would be not well balanced if we go for something other than 45-65% which GIA said, but by some virtual designing I saw no big downside if we use shorter or longer stars (at least I think like that)


So, this would be great if you share your experiences with real and virtual diamonds with star lengths of 65-70% or 35-40%
any advantage/disadvantage related to the Upper Girdle/Star length you spotted?


I appreciate it if experienced friends participated in this topic, please
@Serg @Garry H (Cut Nut) @oldminer @0-0-0 @Karl_K


References:

https://www.gia.edu/doc/Modeling-the-Appearance-of-the-Round-Brilliant-Cut-Diamond-An-Analysis-of-Brilliance.pdf

https://www.gia.edu/doc/modeling-the-appearance-of-the-round-brilliant-cut-diamond.pdf

 

[Serg]

As far as I remember, GIA created the DCLR metric to assess the potential of a cut to create fire. However, this metric is almost independent of dispersion magnitude (if two cuts with identical proportions but made from materials with different dispersion are cut, they will give almost identical DCLR metric values, even if one of the materials has a zero dispersion value). The value of the metric mainly depends on the angles at which the rays exit the diamond. So, I recommend not using DCLR metric graphs for analyzing fire in diamonds.

 

[V_sh]

As far as I remember, GIA created the DCLR metric to assess the potential of a cut to create fire. However, this metric is almost independent of dispersion magnitude (if two cuts with identical proportions but made from materials with different dispersion are cut, they will give almost identical DCLR metric values, even if one of the materials has a zero dispersion value). The value of the metric mainly depends on the angles at which the rays exit the diamond. So, I recommend not using DCLR metric graphs for analyzing fire in diamonds.

Thanks for your precious advice Sergey, If DCLR has this issue then this can not be a scientific experiment.

I think you may suggest measuring fire intensity by ETAS, Right?

I became curious because also Cutwise reports proof that some long/short stars can perform well and we can consider them.

Did Octonus do any experiment about Star length?

 

[Serg]

Thanks for your precious advice Sergey, If DCLR has this issue then this can not be a scientific experiment.

I think you may suggest measuring fire intensity by ETAS, Right?

ETAS is not related to Fire. It is a tool for assessing the potential (probability) of a cut to capture a light source (redirect it into the observer's eye).

 

[Karl_K]

The greatest effect of star% is how it changes the upper girdle/halves angle relative to the CA angle and table size.
In simple terms:
The steeper the CA the lower the max star%.
Large tables also lower the max star%

With tables 58 and under............
Steep CA over 36 need shorter stars to reduce the upper girdle facets angle so they dont lose edge to edge brightness.
This is why many OEC were cut with 40-45% stars.
Over 35 degrees needs to stay around 50% and under for the same reason.
Shallow crowns under 34 can pick up some scintillation with long stars because the upper girdles are less flat.
In the 34-34.9 CA range many different star% will work.
Larger tables the star% needs to trend down some.

Some of my favorite diamonds ever where super tightly cut 41p/34CA/56T/60+% stars.

 

[Karl_K]

The problem with Garry's bad cut example CZ is over steep upper girdle angles. It demonstrates the problem perfectly.

 

[V_sh]

The greatest effect of star% is how it changes the upper girdle/halves angle relative to the CA angle and table size.
In simple terms:
The steeper the CA the lower the max star%.
Large tables also lower the max star%

With tables 58 and under............
Steep CA over 36 need shorter stars to reduce the upper girdle facets angle so they dont lose edge to edge brightness.
This is why many OEC were cut with 40-45% stars.
Over 35 degrees needs to stay around 50% and under for the same reason.
Shallow crowns under 34 can pick up some scintillation with long stars because the upper girdles are less flat.
In the 34-34.9 CA range many different star% will work.
Larger tables the star% needs to trend down some.

Some of my favorite diamonds ever where super tightly cut 41p/34CA/56T/60+% stars.

Thanks for your consideration and the information.

So in your experience more flat upper girdles (long or digged) can help increasing the Scintillation?

And you prefer shorter stars for deep crowns and longer stars for shallow crowns?

what would be your opinion about short stars like 35-40% ? Will they make the RBC look less balanced? Any decrease in light performance?

The problem with Garry's bad cut example CZ is over steep upper girdle angles. It demonstrates the problem perfectly.

thanks, Where can I read about these bad cut CZs?

 

[V_sh]

ETAS is not related to Fire. It is a tool for assessing the potential (probability) of a cut to capture a light source (redirect it into the observer's eye).

Thanks for the point you mentioned Sergey,

I meant color ETAS (CETAS) for measuring fire, I think dispersion scores of MSS project is measured by CETAS, Am I right?

 

[Karl_K]

Thanks for your consideration and the information.

So in your experience more flat upper girdles (long or digged) can help increasing the Scintillation?

opposite, with shallow crowns angles steeper/less flat uppers can increase scintillation.

And you prefer shorter stars for deep crowns and longer stars for shallow crowns?

Its more than a preference its science.

what would be your opinion about short stars like 35-40% ? Will they make the RBC look less balanced? Any decrease in light performance?

Depends on the CA and table size.
At some extremes it might be needed. 40% is not super rare in very steep crowned oec designs.

thanks, Where can I read about these bad cut CZs?

You can buy some.

Ideal and Average cut quality CZ comparison boxed set for the professional and retailer.

www.datlas.com

 

[V_sh]

opposite, with shallow crowns angles steeper/less flat uppers can increase scintillation.

Its more than a preference its science.


Depends on the CA and table size.
At some extremes it might be needed. 40% is not super rare in very steep crowned oec designs.



You can buy some.

Ideal and Average cut quality CZ comparison boxed set for the professional and retailer.

www.datlas.com

Thanks for your time and sharing your knowledge Karl, so valuable to me.

 

[Garry H (Cut Nut)]

Hello friends,

I've just been curious about Star Length out of the GIA suggested range (45-65%) and did some search about it and found two nice articles, and I would be happy to know your opinions.

in the following charts, WLR stands for "Weighted Light Return" which indicates overall Brilliance
and DCLR stands for "Dispersed Colored Light Return" which is a metric for Fire








The following statements are about Diamonds with a Main Crown of 34-35 degrees


it's obvious that there is not a big loss if we go for a Star length out of the GIA accepted range; for example, a Star length of 65-70% and also 35-40% seems almost as good or better than a Star length of 50%

I also questioned that maybe the physical appearance of the diamond would be not well balanced if we go for something other than 45-65% which GIA said, but by some virtual designing I saw no big downside if we use shorter or longer stars (at least I think like that)


So, this would be great if you share your experiences with real and virtual diamonds with star lengths of 65-70% or 35-40%
any advantage/disadvantage related to the Upper Girdle/Star length you spotted?


I appreciate it if experienced friends participated in this topic, please
@Serg @Garry H (Cut Nut) @oldminer @0-0-0 @Karl_K


References:

https://www.gia.edu/doc/Modeling-the-Appearance-of-the-Round-Brilliant-Cut-Diamond-An-Analysis-of-Brilliance.pdf

https://www.gia.edu/doc/modeling-the-appearance-of-the-round-brilliant-cut-diamond.pdf

There was very little useful or accurate information from either WLR or DCLR.

 

[Garry H (Cut Nut)]

Thanks for your consideration and the information.

So in your experience more flat upper girdles (long or digged) can help increasing the Scintillation?

And you prefer shorter stars for deep crowns and longer stars for shallow crowns?

what would be your opinion about short stars like 35-40% ? Will they make the RBC look less balanced? Any decrease in light performance?



thanks, Where can I read about these bad cut CZs?

I can not find an actual scan of a bad CZ at present, but this is the proportions sent to the cutter (Winks Thai guy who was very accurate).
The upper girdles are about 42.8 degrees - not supper steep.
The deep pavilion has more impact than the crown side angles because it doubles the impact.

 

[Garry H (Cut Nut)]

I found an actual scan. Shooting a ray through the upper girdle and dropping the pavilion angle from 44.31 to 42.31 results in nice light return and dispersion in both diamond and CZ with this model.
As shown is for CZ:

 

[Serg]

Thanks for the point you mentioned Sergey,

I meant color ETAS (CETAS) for measuring fire, I think dispersion scores of MSS project is measured by CETAS, Am I right?

I do not know a CETAS . We developed ETAS and DETAS( Dynamical ETAS).

 

[V_sh]

I found an actual scan. Shooting a ray through the upper girdle and dropping the pavilion angle from 44.31 to 42.31 results in nice light return and dispersion in both diamond and CZ with this model.
As shown is for CZ:

Thanks for sharing the information Garry,

It seems that the main problem is the main pavilion as you mentioned

what's your experience with long and short stars anyway? do you think they are completely useless?

 

[V_sh]

I do not know a CETAS . We developed ETAS and DETAS( Dynamical ETAS).

I appreciate your valuable patience Sergey,

So I'm going the completely wrong way, could you please generously enlighten me

What method do you suggest for measuring the fire/dispersion?

I'm looking for the method behind the "Disp" Score in the MSS project:

 

[Serg]

I appreciate your valuable patience Sergey,

So I'm going the completely wrong way, could you please generously enlighten me

What method do you suggest for measuring the fire/dispersion?

I'm looking for the method behind the "Disp" Score in the MSS project:

Each virtual facet is essentially a prism. Therefore, it can be attributed with Dispersion of the ray exiting this prism during reverse ray tracing, when the ray is emitted from the observer's eye. This dispersion has a strong correlation with the probability of observing a colored flash in this virtual facet if it reflects the source into the observer's eye (ETAS determines the probability of how many virtual facets simultaneously reflect the diamond). In other words, it is a measure of the conditional probability of a colored flash, given that there will be any flash at all. It should be noted that dispersion depends on the path of the ray, although the path of the ray is reversible. Thus, the dispersions of the prism in the directions of eye-light source and light source-eye are different. Therefore, this value does not precisely correlate with the conditional probability of a colored flash. Also, the probability of a colored flash depends on the angular size of the source. The larger the source, the lower the probability of seeing Fire. However, despite this, Dispersion is very convenient for developing cuts, as it, like ETAS and DETAS, describes the properties of the cut that are independent of the light sources. Meanwhile, Fire, LR, Brilliancy, Scintillation - always depend on the type of lighting used.

 

[V_sh]

Each virtual facet is essentially a prism. Therefore, it can be attributed with Dispersion of the ray exiting this prism during reverse ray tracing, when the ray is emitted from the observer's eye. This dispersion has a strong correlation with the probability of observing a colored flash in this virtual facet if it reflects the source into the observer's eye (ETAS determines the probability of how many virtual facets simultaneously reflect the diamond). In other words, it is a measure of the conditional probability of a colored flash, given that there will be any flash at all. It should be noted that dispersion depends on the path of the ray, although the path of the ray is reversible. Thus, the dispersions of the prism in the directions of eye-light source and light source-eye are different. Therefore, this value does not precisely correlate with the conditional probability of a colored flash. Also, the probability of a colored flash depends on the angular size of the source. The larger the source, the lower the probability of seeing Fire. However, despite this, Dispersion is very convenient for developing cuts, as it, like ETAS and DETAS, describes the properties of the cut that are independent of the light sources. Meanwhile, Fire, LR, Brilliancy, Scintillation - always depend on the type of lighting used.

Your amazing detailed answer is of great value to me and so helpful, thanks a lot.

then it seems that the Dispersion map is so valuable for designing, I hope you will make it available someday.

So I have to study ETAS, DETAS, and Ray Tracing techniques in detail as they are so valuable because of independence to the type of light source.

I appreciate sharing your wisdom.