Has anyone had issues with Brian Gavin diamonds? Whiteflash

[drk14]

Hi,
Yes. see results of multifocus for sappfire

@Serg , thanks for taking the time to respond to my question, and for the interesting links.

The sapphire example brings what appears to be real facets into focus in a single image, and the feldspar example brings real inclusions(?) at different focal depths into focus in a single image.

What I'm curious about is whether this technique can be be used to bring virtual facets from disparate focal planes into focus in the same image, especially for cases in which the pathlength is long compared to the physical depth of the stone?

 

[Serg]

@Serg , thanks for taking the time to respond to my question, and for the interesting links.

The sapphire example brings what appears to be real facets into focus in a single image, and the feldspar example brings real inclusions(?) at different focal depths into focus in a single image.

What I'm curious about is whether this technique can be be used to bring virtual facets from disparate focal planes into focus in the same image, especially for cases in which the pathlength is long compared to the physical depth of the stone?

1) yes, its are real gematite inclusions in feldspar.
2) Stacking is necessary when you need high resolution and big depth of focus in same time.
High resolution is not necessary to see virtual facets. to see virtual facet for long path You may just reduce aperture and use low magnification( short focal length). it is much more cheap approach . what is your final goal?

 

[drk14]

You may just reduce aperture and use low magnification( short focal length). it is much more cheap approach . what is your final goal?

@Serg -- I have noticed that in all online photos and videos of "crushed ice" type facets, these virtual facets appear blurry/mushy. In some cases this may be due to degradation of the image when it is reflected a large number of times by imperfect mirrors, but I suspect that in most cases, the blurriness is simply due to the fact that the depth of field is not sufficient to accommodate both the real facets and the virtual facets -- i.e., if the stone's physical outline and real facets are in focus (as they should be), then the virtual facets with long pathlength are out of focus. I would like to know if it is possible to photograph diamonds that has "crushed ice" facets in a way that does not automatically make the far-away VFs appear like "mush".

If this could be achieved simply by stopping down the aperture, how come no major online retailer uses this method to present diamonds that are prone to "mush" (e.g., elongated cuts like oval, mq, pear)?

 

[Serg]

Drk14,
see some tests from 2015

1) Standard Dibox settings

2) distance +40mm ( less magnification and resolution) F13

3) Distance +40mm, F16


Stacking in movies is time consuming, you need spend in 3-10 times more minutes to take single movie . I suppose it is too expensive for current online retail business models . Also many sellers are not happy to use high quality, high resolution images because they clearly shows invisible for naked eye inclusions and cut defects . Most buyers mainly see diamond drawbacks in images instead nice overall appearance .
Stacking movies will show more inclusions, cut asymmetry , etc. So sellers thinks that they will spend more time( money) but receive less sales . It is not technical problem more.

 

[Serg]

Ps. you may see more Fire on photos with higher F( smaller aperture). Even green flashes show up.

 

[Serg]

on Multifocus photo ( stacking) SI 1 "crushed ice" diamond may looks as I1 . It is very misleading when
some VS looks as SI, SI as I.

 

[OoohShiny]

Excellent info, @Serg, thank you for posting

 

[drk14]

@Serg -- thanks for taking the time to post your insights and image examples.

Drk14,
see some tests from 2015

To be clear, those (bottom two images) were examples of a single image (not a multi-stack) acquired with reduced aperture, correct? I feel like the "mush" was cleared up (brought into focus) in some areas, but not so much in other areas (e.g., the top and bottom third of the table).

Stacking movies will show more inclusions, cut asymmetry , etc. So sellers thinks that they will spend more time( money) but receive less sales .

I don't know how big a role PS prosumer recommendations plays in the total sales of some of vendors who are popular here (e..g, JA, BN, etc.), nor how big of a share of their sales are made up of fancies, but I do know that 90-100% of ovals, marquise, pears, and radiants in their virtual inventories are not recommended to buyers who come to PS for advice. If this niche is "small potatoes" in the big scheme of things, then I see your point. On the other hand, if a 2x-5x increase in the sale of crushed-ice fancies to PS users would be significant, then it might be worthwhile for those vendors to make available tools that allow us to better evaluate these stones.

on Multifocus photo ( stacking) SI 1 "crushed ice" diamond may looks as I1 .

Would love to see some examples!

 

[Serg]

Drk,
yes , last 3 photos are not multi focus images

 

[Serg]

@Serg



I don't know how big a role PS prosumer recommendations plays in the total sales of some of vendors who are popular here (e..g, JA, BN, etc.), nor how big of a share of their sales are made up of fancies, but I do know that 90-100% of ovals, marquise, pears, and radiants in their virtual inventories are not recommended to buyers who come to PS for advice. If this niche is "small potatoes" in the big scheme of things, then I see your point. On the other hand, if a 2x-5x increase in the sale of crushed-ice fancies to PS users would be significant, then it might be worthwhile for those vendors to make available tools that allow us to better evaluate these stones.


Would love to see some examples!

For my point of view most PS prosumer recommendations are helpless and vain specially for fancy cuts.
In best case in beginning they ask ASET, IS images (which again can not to be used as only one tool to estimate Fancy cuts optical performance ) and then " advice" : "you have to look diamonds yourself and take that you like". Most of such PS prosumers do not even see the difference between ASET as selected tool for Standard Round cut and ASET as rejected tool for fancy cuts.

 

[drk14]

For my point of view most PS prosumer recommendations are helpless and vain specially for fancy cuts.
In best case in beginning they ask ASET, IS images (which again can not to be used as only one tool to estimate Fancy cuts optical performance ) and then " advice" : "you have to look diamonds yourself and take that you like". Most of such PS prosumers do not even see the difference between ASET as selected tool for Standard Round cut and ASET as rejected tool for fancy cuts.

This is sort of my point. Part of the reason PS prosumer are much better skilled at evaluating MRBs than fancy cuts is that there are few suitable tools available to PS users for assessing non-round cuts.
One result of this is that many shoppers are encouraged to shop for fancies in person, or to have one of a small number of vendors make personal recommendations (e.g., GOG or VC are often recommended for this purpose). I think that larger vendors such as JA, BN, ED lose many sales of fancies on PS. So, they should have a financial incentive to provide new imaging tools (e.g., multifocus stacks) that could allow a PS prosumer to be more confident in identifying well-performing fancy cuts.

Now, if it turns out that fancy cuts only represent, say, 10% of these vendors' sales, and that PS recommendations only account for, say, 5% of overall sales, then they would only stand to gain at most 0.5% by providing PS prosumers with better tools. In this case, I agree that the cost-benefit analysis may not be in favor of making available additional imaging tools. But I have no idea what the real numbers are.

 

[Rockdiamond]

This is sort of my point. Part of the reason PS prosumer are much better skilled at evaluating MRBs than fancy cuts is that there are few suitable tools available to PS users for assessing non-round cuts.
One result of this is that many shoppers are encouraged to shop for fancies in person, or to have one of a small number of vendors make personal recommendations (e.g., GOG or VC are often recommended for this purpose). I think that larger vendors such as JA, BN, ED lose many sales of fancies on PS. So, they should have a financial incentive to provide new imaging tools (e.g., multifocus stacks) that could allow a PS prosumer to be more confident in identifying well-performing fancy cuts.

Now, if it turns out that fancy cuts only represent, say, 10% of these vendors' sales, and that PS recommendations only account for, say, 5% of overall sales, then they would only stand to gain at most 0.5% by providing PS prosumers with better tools. In this case, I agree that the cost-benefit analysis may not be in favor of making available additional imaging tools. But I have no idea what the real numbers are.

What a great post!
From my perspective, the Serg's company makes the highest quality photographic equipment for diamonds on the market- by a mile.
He was a pioneer in the revolving images you see today.
The issue isn't how the quality of the image.
They even developed a special rocking device to show Scintillation.
Today's diamond market is by and large populated with images that must be done as quickly as possible.
Scintillation performance is based on factors that are simply not possible to accurately capture in a picture or rotating video.

 

[Texas Leaguer]

This is sort of my point. Part of the reason PS prosumer are much better skilled at evaluating MRBs than fancy cuts is that there are few suitable tools available to PS users for assessing non-round cuts.
One result of this is that many shoppers are encouraged to shop for fancies in person, or to have one of a small number of vendors make personal recommendations (e.g., GOG or VC are often recommended for this purpose). I think that larger vendors such as JA, BN, ED lose many sales of fancies on PS. So, they should have a financial incentive to provide new imaging tools (e.g., multifocus stacks) that could allow a PS prosumer to be more confident in identifying well-performing fancy cuts.

Now, if it turns out that fancy cuts only represent, say, 10% of these vendors' sales, and that PS recommendations only account for, say, 5% of overall sales, then they would only stand to gain at most 0.5% by providing PS prosumers with better tools. In this case, I agree that the cost-benefit analysis may not be in favor of making available additional imaging tools. But I have no idea what the real numbers are.

drk,
I think you make an astute point and here is why. Even though historical sales percentages suggest that the cost/benefit ratio may not be favorable, one has to recognize that online shopping for diamonds is growing rapidly. Remote buyers need effective tools. Someone who develops or adapts tools effectively for the purpose of selling fancy shape diamonds can potentially tap into a sizable emerging market. The fancy pie might be much smaller than the round pie, but if you get a much bigger piece of that pie, it could be very rewarding.

 

[drk14]

Bryan,
Thanks for your comment.

Someone who develops or adapts tools effectively for the purpose of selling fancy shape diamonds can potentially tap into a sizable emerging market.

Can I take this statement as confirmation that my premise (viz., that current online sales of fancy cuts is significantly below potential), is essentially correct?

 

[Texas Leaguer]

Bryan,
Thanks for your comment.



Can I take this statement as confirmation that my premise (viz., that current online sales of fancy cuts is significantly below potential), is essentially correct?

My instincts tell me that without question fancy sales are under-performing online. And primarily for the reasons that you have identified.

Perhaps the work Serg is doing will eventually find application here.

While AGSL has developed LP grading standards for a range of fancy shapes, and the potential to develop metrics for ANY fancy shape, it has so far not been widely embraced by the trade. There are reasons for this other than merit.

 

[Rockdiamond]

Return rates on Fancy Shapes by the "big boys" are astronomical.
I get what you're alluding to about the AGSL Bryan- and sure, part of the lack of acceptance is based on monetary reasons.
But IMO a larger deterrent is how restrictive the AGSL cut grades are.

 

[drk14]

There are reasons for this other than merit.

Care to expand on this? Extrapolating from David's comment above, is the problem that too many fancy cuts would not pass LP grading, creating a disincentive for adopting such metrics?

If so, is this because a large fraction of fancies are indeed not well cut (by any reasonable measure of light performance), or is it because proposed LP metrics do not correlate sufficiently well with consumer perception of fancy cut performance?

 

[Rockdiamond]

Hi Drk,
Based on your avatar, I assume you're a marquise lover....
What's a reasonable measure of light performance for a Marquise?
I ask because there really is no way to pin down what's "best"
A 2:1 LxW Ratio is considered "ideal"- but yet some people love them a bit chubbier. Are they wrong?
The very term "light performance" as defined by AGSL misses this point entirely IMO
And in my opinion, that's what's "wrong" with the AGSL way of grading make.
I'm sure that we can find plenty of examples of cutters saving weight on fancy shapes- and such cutters might be upset if AGSL imposed "standards"
But there's also plenty of cutters who want to cut great fancy shapes using their own design/ideas.
In a sense, the difficulty in selling them, as you're asking about, is what makes them so attractive.
A round is going to be...well, round.
But fancy shapes can vary wildly within the same shape. Variety is indeed the spice of life, no?

 

[Texas Leaguer]

Care to expand on this? Extrapolating from David's comment above, is the problem that too many fancy cuts would not pass LP grading, creating a disincentive for adopting such metrics?

If so, is this because a large fraction of fancies are indeed not well cut (by any reasonable measure of light performance), or is it because proposed LP metrics do not correlate sufficiently well with consumer perception of fancy cut performance?

It's multiple reasons including those you have mentioned above. I think the correlation is probably there with regards to AGSL grading of fancies. I really only have sufficient experience with the princess cuts, but I have seen a few ovals, cushions and emerald cuts that made AGS 0 and they will knock your socks off.

One of the problems is that AGSL is sort of treated by the trade as a 'one trick pony' in the sense that main goal of a cutter sending a stone to AGS vs GIA is to get the Ideal grade. The mindset never developed in the market that a AGS1 or 2 would be really nice make. Rather, it seems like they are viewed as cutting failures. Therefore, the relative correlations have never been established in the market. Meanwhile, GIA goes merrily along NOT providing any concrete information about cut quality. And most manufacturers take full advantage of the 'flexibility' this gives them.

The fact that making AGS Ideal is very demanding on the cutter and expensive to produce, and because falling even a tiny bit short earns them a certificate that is not properly valued in the market, few manufacturers are willing to go in that direction. And who could blame them!

 

[drk14]

Based on your avatar, I assume you're a marquise lover....
What's a reasonable measure of light performance for a Marquise?

Avatar is my wife's mq. My interest in fancy cuts (and the problem of evaluating them) is based in part on my experiences shopping for her stone.

What is light performance? In my opinion, it is the ability of the diamond's sparkles to mesmerize... It's a combination of physical (optical) effects and psychological responses. If there's nothing about the dynamic pattern of light reflections that makes you involuntarily gaze at the stone , then the attractiveness of the proportions and shape outline is irrelevant.

Can light performance be measured objectively? The physical effects can be predicted and quantified given an accurate representation of the diamond and lighting environment, and sufficient computational power. The psychological responses are less tractable but can be approximated by phenomenological models and validated experimentally. Such models will always have limitations (i.e., they will work well in some situations, but be less useful in other situations).

Do any currently available LP metrics solve the above problem successfully? Seems to me like the tools available for rounds are doing a pretty good job, but if there are counterexamples I will not dispute those assertions.

 

[Rockdiamond]

Thanks DRK,
Trying to correlate the methodology that works reasonably well for rounds to fancy shapes is a pointless exercise in my opinion. Too many variables in Fancy Shapes.

Can light performance be measured objectively? The physical effects can be predicted and quantified given an accurate representation of the diamond and lighting environment, and sufficient computational power.

Not really, if we're using reality, as opposed to computer modeling.
We certainly can predict certain factors fairly well, given fixed lighting and a fixed relationship between pupil and diamond.
But in real life, these aspects are never stable. Makes the results far less meaningful.
Add in the very real differences in psychological responses, and we're pretty much starting from scratch when it comes to fancy shape cut grading.

I really only have sufficient experience with the princess cuts, but I have seen a few ovals, cushions and emerald cuts that made AGS 0 and they will knock your socks off.

I'm sure the stones you saw were amazing Bryan- and I have seen countless examples of stones that would not make AGSL0 cut grade on princess and other fancy shapes that will also blow your sox off....

 

[drk14]

Bryan,
Thanks for further explaining the quandary of AGSL grading. A solution for fancies may be to develop tools that allows one to differentiate the relative performance of any two stones, without assigning an absolute grade to either (e.g., imaging techniques that don't attempt to quantify LP).

I have seen a few ovals, cushions and emerald cuts that made AGS 0 and they will knock your socks off.

Are there any videos available of AGS 0 ovals? I'm curious whether they tend primarily to be ovals that are dominated by larger VFs ("chunky" faceting) throughout the diamond, or alternatively, whether the AGS LP metric for ovals is able to also identify crushed-ice type ovals that "will knock your socks off".

 

[drk14]

Not really, if we're using reality, as opposed to computer modeling.
We certainly can predict certain factors fairly well, given fixed lighting and a fixed relationship between pupil and diamond.
But in real life, these aspects are never stable. Makes the results far less meaningful.

David,
If you re-read what I wrote, you'll find that we don't actually disagree. My assumptions are stated, albeit parsimoniously. To prove a conditional statement false, you have to agree to the antecedent ("given an accurate representation of the diamond and lighting environment, and sufficient computational power") and show that the consequent ("physical effects can be predicted and quantified") does not follow from the premise.

When I say "accurate representation", I do mean realistic. And this may include complex lighting environments as well as moving parts.

And for the record, I have not suggested that LP metrics that were developed for rounds can also be applied to fancies.

 

[Rockdiamond]

Thank you for the explanation DRK, although the logic is not direct to me.
Can't I disagree with the antecedent, which would then nullify results- or at least put them in a better context?
Although I do believe we do agree in general....
We agree realistic pics could help...
But in terms of realism, there's a HUGE issue- the actual size of diamonds.
To examine the details of a diamond we need to use magnificaiton- either in real life, or in pics.
But in real life viewing, aspects like scintillation make up a large part of what attracts us to a small object like a diamond, and you just can't accurately picture it in a close up picture.
Compared to contrast/brightness which can be far more easily documented.

Data to be presented in a metric ignores the reality that humans have a big part in the art of diamonds.
The largest diamond sellers today are all about removing the human element.
As is the AGSL cut grade.

about the oval:
I believe that the AGSL Oval cut grade is for "Oval Brilliant" as opposed to "Oval Modified Brilliant"
The second type of stone would typically have a more "crushed ice" sort of LP

 

[drk14]

Can't I disagree with the antecedent, which would then nullify results

It's a quirk of formal logic that if the antecedent of a conditional statement is false, then the conditional statement is automatically true, independent of whether the consequent is true or false. The following statement is true: "If the moon is made of cheese, then the Earth must be flat."

The point of my original statement was that realistic prediction of the physical behavior of diamonds is within the realm of the possible (i.e., I was not claiming that this is problem has already been completely solved).

 

[Serg]

Bryan,
Thanks for further explaining the quandary of AGSL grading. A solution for fancies may be to develop tools that allows one to differentiate the relative performance of any two stones, without assigning an absolute grade to either (e.g., imaging techniques that don't attempt to quantify LP).
".

Introduction of relative performance without absolute grade would very misleading for diamond industry.
Suppose you introduce it. then somebody take AGS0 round cuts as reference and create relative performance for any diamond and start use such relative performance to sort all diamonds in absolute scale . But in relative nonlinear metric B-C is not equal to ( B-A)-(C-A).
so you may receive that B is better than C in direct comparison and in same time that B is worse than C in relative comparison with A as reference.

 

[drk14]

@Serg -- I understand your point, but what I was suggesting was to develop non-quantitative tools (e.g., some form of imaging) to allow relative comparisons without assigning any numbers to the difference in performance.

 

[Rockdiamond]

They have it DRK
The ViBox has that capability. You can do a rocking video of two stones and the software places the images next to each other and synchronizes the rocking motion.
The quality of the images is stunning.
The newer crop of rotating cameras simplified a lot of the aspects on which the ViBox offers total control.
Such as lighting. F stop. White balance.
The ViBox offers total control. But the unit is the size of a refrigerator.
And it takes a few moments to set up each diamond- after you've spent a few months practicing.
The new crop of cameras have a more simple device for holding the diamond- a vacuum. The ViBox uses a complex device- it takes more skill, and practice to master. It also is better at the task.
I have owned the ViBox- and it is an amazing price of technology.
But quick and simple wins the day in production line shops.

 

[Texas Leaguer]

I think the essence of David's post is a good one. The R&D is there, and the equipment is there to make high quality visual comparisons (to share remotely). But the cost and learning curve is still prohibitive for industry-wide adoption.

However, individual companies can deploy this technology to showcase and promote their own goods. And that's where it has to start. Then competitors start competing and soon there is a standard against which consumers can compare across different vendors.

If Octonus or AGSL could develop a decent rejection tool like HCA for fancies, then there would be a lot more viability for fancy sales in the online space. That could be a pathway to more sophisticated video and/or laboratory analysis. However, parametric grading is too simplistic to work well for fancies.

Presently, fancy buyers are pretty much relegated to trusting their own untrained eye or taking the word of a vendor with a financial interest in selling them.

 

[Karl_K]

If Octonus or AGSL could develop a decent rejection tool like HCA for fancies, then there would be a lot more viability for fancy sales in the online space.

Not going to happen because numbers don't mean jack with most fancy cuts.
Which is the biggest failure of the agsl SE/EC cut grading.
Well after that it doesn't work because it does not and can not grade patterns and life.
You can follow the formula in the cutting guides to the letter and cut a crappy stone.