AGS Lab releases VPA software

[Garry H (Cut Nut)]

AGS Visual Proportion Analyser software is a little application that is free in a beta format.
You can download it by clicking the download button on the bottom of http://agslab.com/

It is a complex process. The manual that comes with it has the operating instructions on page 7 - copied here below to save you time:

How to Operate The Software
Copy the contents of the CD Rom to your desired location. Do not run the program from the CD. Open the folder entitled AGS_Visual_Performance_Analyzer. In order to process your own .stl or .srn files, place the files in the same directory of the VPA software. Double click the software AGS_Visual_Proportion_Analyzer.exe file and it will process the report files into a separate sub-folder called VPA_Images. Open that folder and enjoy looking at the images of diamonds cut to different degrees of precision.

Now on my computer the folders actually ended up back in the folder where the original .srn files came from. Not what I read in the instructions - but rather neat.

The charts produced show the angular and azimuth variation in simple images. It is an easy to follow system, and rather like GIA's recent announcement publishing the rules for Symmetry so that manufacturers know before sending a stone that they will achieve Excellent or at least Very Good - I guess this is also what AGS are doing.

Unfortunaltely it only works on Sarin srn files and .stl files.
On that matter there is an interesting article in Gems & Gemology by a team of GIA research scientists and cornel Uni folk that uses an ancient goniometer to measure the angles of stones to create calibration masters. They found that the scanner they compared had very large angular deviations from what they believe to be very accurate results from the Goniometer. They did not name the brand of scanner. The 5 stones they compared each had deviations of greater than 0.4 degrees. With that in mind one might need to be sure when using the AGS VPA software that its not GIGO.

I do not have time today for any more - but hopefully somone will post some of the display outputs.

 

[Karl_K]

very interesting, however I gave up doing this type of eval on scanner data because it is simply not accurate enough.
Helium set for the max accuracy mode might get fairly close but the common sarin scanner? no way.

Here are the images:
1:

2:

 

[Garry H (Cut Nut)]

very interesting, however I gave up doing this type of eval on scanner data because it is simply not accurate enough.
Helium set for the max accuracy mode might get fairly close but the common sarin scanner? no way.

Here are the images:
1:

2:

Karl I believe Sarin have a more accurate scanner now, and also Peter and Jason, aka Q, are happy to work with OctoNus to enable those proprietary formats like .gem .dmc and .mmd to be used also.

 

[Karl_K]

Karl I believe Sarin have a more accurate scanner now, and also Peter and Jason, aka Q, are happy to work with OctoNus to enable those proprietary formats like .gem .dmc and .mmd to be used also.

Hi Garry,
I am familiar with the output of the new Sarin scanner.
While it is a step up from the old one I still have large doubts about doing this type of evaluation using the data.
Very few places if any have totally switched to the new one so the majority of people using this tool will be using the old one.

Karl

 

[Rhino]

Very cool and some very useful information at a glance too. Here's a fun one of a more commonly cut diamond in our inventory.

In this first report you can very quickly see just how severe (or not) each crown and pavilion main deviate from the average angle as well as the upper and lower halves by the depth of color. You can see how badly the inconsistency is and more particularly in the upper halves and secondly the crown mains.

I hope they plan on allowing us to generate reports on all the shapes!

PS: Karl ... you have worked with Sarin's new scanner?

 

[Rhino]

And for the 2nd Report ... look at the twisting of the mains and the degrees of digging on both the upper and lower halves.

Excellent tool for judging cut quality, at least in the realm of craftsmanship. Garry do you know of any plans to be able to run this on cushions, asschers, princess?

 

[Garry H (Cut Nut)]

And for the 2nd Report ... look at the twisting of the mains and the degrees of digging on both the upper and lower halves.

Excellent tool for judging cut quality, at least in the realm of craftsmanship. Garry do you know of any plans to be able to run this on cushions, asschers, princess?

Afraid not Jonathon, but with a bit of luck Peter may join the discussion.

 

[diagem]

And for the 2nd Report ... look at the twisting of the mains and the degrees of digging on both the upper and lower halves.

Excellent tool for judging cut quality, at least in the realm of craftsmanship. Garry do you know of any plans to be able to run this on cushions, asschers, princess?

Afraid not Jonathon, but with a bit of luck Peter may join the discussion.

I am afraid that based on my experience..., The new Sarin HD "might" work for rounds but still not enough accurate for fancies.
But it will be a pleasure to hear Peter's view as well.

 

[Garry H (Cut Nut)]

And for the 2nd Report ... look at the twisting of the mains and the degrees of digging on both the upper and lower halves.

Excellent tool for judging cut quality, at least in the realm of craftsmanship. Garry do you know of any plans to be able to run this on cushions, asschers, princess?

Afraid not Jonathon, but with a bit of luck Peter may join the discussion.

I am afraid that based on my experience..., The new Sarin HD "might" work for rounds but still not enough accurate for fancies.
But it will be a pleasure to hear Peter's view as well.

I have heard that from other sources too DG

 

[Garry H (Cut Nut)]

Further commentary from Peter Yantzer from AGS laboratory:

"I see that some people don't think the machines can measure accurately enough to analyze a stone. The entire diamond industry all over the world runs on non-contact measuring devices. How can anyone suggest that they aren’t accurate enough? I think they may be missing the point. It doesn't matter what measuring device is used as long as it is reasonably accurate and repeatable and, perhaps most importantly, the operator takes care to clean the stone. Most devices are more than adequate to do that with the standard round brilliant cut.

For the AGS Visual Proportion Analyzer software the machines only need to be reasonably accurate in a relative way, not in an absolute way. We randomly and arbitrarily sampled 100,000 scans that were measured with different generations of non-contact measuring devices and the only problem we saw was in the measuring of stones that were dirty or not clean enough. It doesn’t matter if one machine says that a particular facet has an angle of 34.2 degrees and another measuring device says that the same facet is 34.0 degrees. Think relative.

There are no judgments with the software, except for a pass/fail on weight ratio, but some will try to force judgments on the accuracy. Which is not the point!

Please imagine for a moment:
• The images work beautifully as a selection or rejection tool and a quality control tool and a tool to prove or disprove.
• Manufacturers can use it to easily check the cutting precision of the goods coming out of their factories. And, manufacturers tend to have high end scanning devices.
• Retailers can use it to cherry pick their suppliers' inventories using scans that they request from their suppliers or creating scans from their own machines that are probably not as accurate as the manufacturers but more than sufficient nonetheless. I believe Sarin did very well at the AGS Conclave less than two weeks ago.
• Consumers will eventually become the engine that transforms the diamond industry. The VPA software is easy to learn, it’s FREE and it will go viral or close to it because it will be available on every free and shareware website on Earth, or at least a whole lot of them.

Having tried to imagine the above, please take a moment to think about that.

Imagine further:
Consumers ask retailers to provide them with a scan. The AGS Visual Proportion Analyzer is less than two weeks old. Do you doubt that there are companies that are already implementing or adding VPA images and analysis into their buying, their education, their selling and working every angle to give them a competitive advantage? They are now able to prove, rather than claim, to their customers that their diamonds or Elite line or Signature line or Ideal line is a superior stone. And, on the flip side they can use it to disprove competitors’ claims.

You may also be astounded about what you don't find when you Google combinations of words that involve diamond, software and free.

Lastly, imagine the power of VPA combined with an affordable copy of the AGS Performance Grading Software (PGS). Obviously, the key word is affordable. What a 1, 2 punch! Are you starting to believe, as I do, that this little, free software program is the catalyst that will change the diamond industry at every level? If you are selling a luxury product doesn’t it make sense to sell the most precise and highest performing diamond?


Heartily,
Peter

 

[Rhino]

Pete hit the nail squarely on the head. Thanks for posting this Garry. What he is saying is 110% true. Non contact measuring devices will always have a degree of tolerance and the software does exactly what it claims. An easy visual reference to the precision/craftsmanship of a diamond which is also key to its value as well. Hat's off to AGS.

 

[Rhino]

And for the 2nd Report ... look at the twisting of the mains and the degrees of digging on both the upper and lower halves.

Excellent tool for judging cut quality, at least in the realm of craftsmanship. Garry do you know of any plans to be able to run this on cushions, asschers, princess?

Afraid not Jonathon, but with a bit of luck Peter may join the discussion.

I am afraid that based on my experience..., The new Sarin HD "might" work for rounds but still not enough accurate for fancies.
But it will be a pleasure to hear Peter's view as well.

Greetings my friend. I believe if one is able to get an accurate enough scan to resolve all the facets it'll work just fine with fancies. In Sarin's machine, in particular, sometimes one needs to run the "shape wizard" if its a unknown or uncommon shape to accomplish this. We'll see but I hope this is implemented soon. I am all for tools that reward cutters who pay attention to detail (like you!) and don't just sloppily cut a product.

 

[Karl_K]

Further commentary from Peter Yantzer from AGS laboratory:

"I see that some people don't think the machines can measure accurately enough to analyze a stone. The entire diamond industry all over the world runs on non-contact measuring devices. How can anyone suggest that they aren’t accurate enough? I think they may be missing the point. It doesn't matter what measuring device is used as long as it is reasonably accurate and repeatable and, perhaps most importantly, the operator takes care to clean the stone. Most devices are more than adequate to do that with the standard round brilliant cut.

For the AGS Visual Proportion Analyzer software the machines only need to be reasonably accurate in a relative way, not in an absolute way. We randomly and arbitrarily sampled 100,000 scans that were measured with different generations of non-contact measuring devices and the only problem we saw was in the measuring of stones that were dirty or not clean enough. It doesn’t matter if one machine says that a particular facet has an angle of 34.2 degrees and another measuring device says that the same facet is 34.0 degrees. Think relative.

There are no judgments with the software, except for a pass/fail on weight ratio, but some will try to force judgments on the accuracy. Which is not the point!

Please imagine for a moment:
• The images work beautifully as a selection or rejection tool and a quality control tool and a tool to prove or disprove.
• Manufacturers can use it to easily check the cutting precision of the goods coming out of their factories. And, manufacturers tend to have high end scanning devices.
• Retailers can use it to cherry pick their suppliers' inventories using scans that they request from their suppliers or creating scans from their own machines that are probably not as accurate as the manufacturers but more than sufficient nonetheless. I believe Sarin did very well at the AGS Conclave less than two weeks ago.
• Consumers will eventually become the engine that transforms the diamond industry. The VPA software is easy to learn, it’s FREE and it will go viral or close to it because it will be available on every free and shareware website on Earth, or at least a whole lot of them.

Having tried to imagine the above, please take a moment to think about that.

Imagine further:
Consumers ask retailers to provide them with a scan. The AGS Visual Proportion Analyzer is less than two weeks old. Do you doubt that there are companies that are already implementing or adding VPA images and analysis into their buying, their education, their selling and working every angle to give them a competitive advantage? They are now able to prove, rather than claim, to their customers that their diamonds or Elite line or Signature line or Ideal line is a superior stone. And, on the flip side they can use it to disprove competitors’ claims.

You may also be astounded about what you don't find when you Google combinations of words that involve diamond, software and free.

Lastly, imagine the power of VPA combined with an affordable copy of the AGS Performance Grading Software (PGS). Obviously, the key word is affordable. What a 1, 2 punch! Are you starting to believe, as I do, that this little, free software program is the catalyst that will change the diamond industry at every level? If you are selling a luxury product doesn’t it make sense to sell the most precise and highest performing diamond?


Heartily,
Peter

I am really tempted to post one of my infamous strmrdr rants but will refrain for now.
To assume that the error is the same on measuring every facet to make it relative is somewhat delusional in my opinion.
There will be varies degrees of accuracy for each facet measured.

I do love the idea of this tool and rewarding tightly cut diamonds with high craftsmanship but I cant close my eyes to the accuracy of the data that goes into the images.

 

[Texas Leaguer]

This is exciting stuff. I love it when businesses act in alignment with their stated mission (consumer education and protection) and recognize that providing for the greater good will also benefit their bottom line in the long run. It's got to be tempting to keep to yourself technologies that have taken years (and mucho $$) to develop and are useful to you in a proprietary way. It takes vision to realize it is better to share it, and that it is even better to share it for free! It may not exactly be google earth, but it's pretty cool and potentially very useful for those of us in the diamond business.

My first question was also the impact on the output of standard instrumentation deviations. We will be analyzing it for use in our operations and will report back any insights we might gain. It will be interesting to test Peter's "Theory of Relativity"

 

[Karl_K]

My first question was also the impact on the output of standard instrumentation deviations. We will be analyzing it for use in our operations and will report back any insights we might gain. It will be interesting to test Peter's "Theory of Relativity"

rofl that cracked me up, guess I really am a geek.
I am looking forward to hearing about your results.
The tough thing is how do you verify the results, with a computer generated image you compare it to a real image.
If the scan data says a facet is 1 degree off on report 2 how do you verify that?

 

[Texas Leaguer]

Karl,
Hard to say what we will find or how to interpret and/or use what we do discover. Like testing any other tool, it may or may not be perfect for a particular use, but we also may find other uses for it. I believe it may have value for us to augment the other analytics we do on our inventory. Especially if the images turn out to be "relatively" consistent between scans and between different devices.

 

[Garry H (Cut Nut)]

The way to check scanner accuracy is with a goniometer. As I mentioned in the first post, GIA conducted one such study and Sergey and I have a patent for an even more accurate scanner partly based on some gonimoeter type principals.

GIA used an antique device similar to this one shown at number 2. http://en.wikipedia.org/wiki/Goniometer

There are more modern devices that claim to be more accurate such ashttp://www.newport.com/BG-Series-Goniometric-Cradles/584445/1033/info.aspx#tab_Specifications.

GIA's testing showed large errors in facet measurements from the unnamed scanner they tested, even for pavilion angles (which raises questions the accuracy of their goniometer and their testing methodology). In my own experience the crown facets are very hard for scanners to measure because - they are short and they are close to the stage of the scanner.

Peter and Q are also getting back to me on the question of Table Tilt vs optical axis of the stone being examined.

 

[japroffitt1]

I just ran a diamond I was considering and definitely helped me exclude it. The scan seems to have been bad, but the diamond was already shipped backed to my jewelers local office from their main office that has the Sarin so I was unwilling to take a risk on it. Very helpful tool!

 

[Serg]

I just ran a diamond I was considering and definitely helped me exclude it. The scan seems to have been bad, but the diamond was already shipped backed to my jewelers local office from their main office that has the Sarin so I was unwilling to take a risk on it. Very helpful tool!

Report has critical bugs. Pavilion main facets and lower halves had been mixed .

Diamond is not so bad as the report informs .

 

[Garry H (Cut Nut)]

I just ran a diamond I was considering and definitely helped me exclude it. The scan seems to have been bad, but the diamond was already shipped backed to my jewelers local office from their main office that has the Sarin so I was unwilling to take a risk on it. Very helpful tool!

If you send me the 3D Sarin file we will try to convert it to correct the scan error so you can see what the stone file really looks like Japroffitt1

 

[japroffitt1]

File is attached

 

[Garry H (Cut Nut)]

File is attached

Thanks - I had my laptop stolen and do not have the Pro version of DiamCalc licence yet - so can not do more detailed reporting, but here is a wire frame image - you can see a lot of problems.
It would also appear the stone was sitting on some dust so the errors may even be smaller for the crown than the report indicates. (Table Tilt).

 

[Garry H (Cut Nut)]

Peter Yantzer from AGSL asked me to pass the word that you can now download a complete unlocked version with rewritten instructions from our website - http://www.agslab.com - the download button is at the bottom of the page.

Regarding the current stone we are discussing Peter says:
"As I mentioned, the measurements are relative and the images are not exact to each other but in no way do they keep you from making an intelligent buy/don't buy, pass/fail or prove/disprove decision. I would hope that people would really take the time to actually understand what VPA is really about. VPA is about the images, not the numbers!

And, we are getting download requests from literally every corner of the earth. I couldn't be happier because of that."

 

[Serg]

Peter Yantzer from AGSL asked me to pass the word that you can now download a complete unlocked version with rewritten instructions from our website - http://www.agslab.com - the download button is at the bottom of the page.

Regarding the current stone we are discussing Peter says:
"As I mentioned, the measurements are relative and the images are not exact to each other but in no way do they keep you from making an intelligent buy/don't buy, pass/fail or prove/disprove decision. I would hope that people would really take the time to actually understand what VPA is really about. VPA is about the images, not the numbers!

And, we are getting download requests from literally every corner of the earth. I couldn't be happier because of that."

re:VPA is about the images, not the numbers!

as results on image "Halves-Facet direction" we see wrong information Image(direction precision) and wrong numbers 8-11 degree error in Average .

this mistake ( misleading information) comes from PVA.
other misleading information about variation between main facet angles ( crown and pavilion) most probably comes from bad quality Sarin 3D model. I do not believe what such Table Tilt present on real diamond

So 2 big and critical mistakes( one from scann , other from PVA) give completely wrong appearance about quality of this particular diamond

 

[Garry H (Cut Nut)]

Andrey is going to attach the files shortly:
Hi Garry:

Thanks for offering to post the .stl files on the Pricescope thread. Along with the .stl files you can post the paragraphs below if you want.

These scans are of the same AGSL calibration stone and have been generated over the last 3 years using 8 different measuring devices - 3 Sarin Diamension machines in standard and most accurate mode, two separate scans from 1 Octonus Helium machine and 4 Sarin HD machines.

The file names tell you which machine measured what. I think that if you look at the VPA images you will see that the images give you very valuable information even though the results are not exactly the same. This is the relative nature of the VPA software that I mentioned earlier.

I believe that you will agree that you can make meaningful buy/don't buy, prove/disprove, pass/fail or accept/reject decisions using VPA. Keep in mind that GIGO ( garbage in = garbage out ) definitely applies here. So don't infer anything from a bad scan. As I see it, the ultimate responsibility for scan accuracy will eventually rest with the manufacturer because a wholesaler, retailer or consumer will reject a stone with a bad scan. And, a wholesaler, retailer or consumer doesn't care to try and learn that a bad scan might be able to be adjusted and make a decision from the adjustment. I believe they will all want accurate scans.

Please go to www.agslab.com and download the fully unlocked version of the Visual Proportion Analysis software and run the attached .stl files. The download button is at the bottom of the first web page. You can then decide for yourself if the VPA software does or does not provide an amazing amount of valuable information.

The software is free to anyone. If you are a consumer you can also download the software and try it out.


[ PSAdmin: Added STL files per request for downloads. ]

 

[Serg]

re:Keep in mind that GIGO ( garbage in = garbage out ) definitely applies here.

1)How can consumer define that scan GI is?

2)some software increase garbage other reduce garbage.

for example software that use light performance metrics reduce garbage in ( for example table tilt)
in same any software ,that use directly facet angles with reference to 3D model table , increase garbage out

 

[Karl_K]

Serg can you open sample 13 in DC. Scan error or possible parsing error in DC?
edit: sample 20 looks questionable to me also.

 

[Serg]

Serg can you open sample 13 in DC. Scan error or possible parsing error in DC?
edit: sample 20 looks questionable to me also.

Hi Karl,
I will check it in few hours latter ( I have not Hasp Keys with me )

There is one more source of error which reduce visual model precision . It is export in to STL from Polyhedron or ASCII formates and then import from STL into Polyhedron or ASCII formates.

STL can use only triangles facets. we have to split facets with 4 and more vertexes . take for example 4 vertexes main crown facets.
after split in into 2 triangles and rounding( each operation in computer has limited precision ) vertexes coordinates then 4 vertexes are not in one plane more. so if you then you do import from STL with same or even worse precision ( criteria of concavity ) you receive additional ( false) facets.
To avoid such false facets you have to do import always with precision worse than export.
we can reduce number of such problems but we can not avoid its at all

Best way is to avoid using stl formats for diamond 3D models

 

[Serg]

I just ran a diamond I was considering and definitely helped me exclude it. The scan seems to have been bad, but the diamond was already shipped backed to my jewelers local office from their main office that has the Sarin so I was unwilling to take a risk on it. Very helpful tool!

Hi japroffitt1,

Could You please send me the address of your jeweler?
I want ask him to scan this diamond on any other scanner .

 

[Garry H (Cut Nut)]

My post 28 May 2012 21:55 a few above now has all the .srn files thanks to Andrey