GIA measurments vs. Sarin measurements - variability?

[Venzen007]

From GIA -
6.83 x 6.81 x 4.71
69.2% Depth
61% Table.

Sarin measurements from Good Old Gold of the same stone.
6.85 x 6.82 x 4.70
68.9% Depth
61.3% Table

Every value is a bit different. What explains these differences? I would maybe expect a tolerance of +/- 0.01 mm, assuming they are using the same caliber of precision equipment. Are tolerances just more than that?

Is GIA also using Sarin?

 

[Johnbt]

In a word, rounding.

https://beyond4cs.com/cut/what-is-a-sarin-scan/when-would-it-be-useful

"Typically, proportions and angles provided in AGS reports have been found to be spot on and precise. On the other hand, GIA applies a slightly different methodology when displaying their measurements.

Due to the rounding and averaging of numbers, you might see variances when you compare their results to that of an actual Sarin scan."

 

[Venzen007]

I'd have considered rounding but for both GIA and the Sarin numbers both showing significant figures to the hundreth place, at least on several measurements.

 

[headlight]

This is really interesting!

 

[Karl_K]

What cut?

 

[Venzen007]

The one I recently bought, the asscher.

 

[Karl_K]

yea that's normal scanners suck at measuring step cuts.

 

[Venzen007]

So that's just the normal variance, then, assuming GIA is using a Sarin to get their measurements, too?

 

[Karl_K]

So that's just the normal variance, then, assuming GIA is using a Sarin to get their measurements, too?

Mostly these:
DiaMension® AXIOM
https://sarine.com/products/diamension-axiom/
But they might still have a few of these:
DiaMension® HD
https://sarine.com/products/diamension-hd/
GOG last I knew had the HD.

When specifically configured for step cuts they are not that bad but the default is set up for RB and not optimized for step cuts.

The table is hard for any scanner because the diamond is sitting on it and I suspect its calculated rather then measured in many scanners.

 

[Karl_K]

better image with RB.....

 

[oldminer]

All of the various manufacturers of diamond measuring scanners publish their tolerances for reliable measurement. None are perfect, but all measure close enough to work sufficiently well. The older units have wider error rates and broader tolerances. The newest versions are tighter, but there is simply no reason to make them more perfect. It is a matter of making the devices accurate enough while at the same time able to be purchased at a reasonable cost.

I own a digital micrometer with a high degree of resolution, .001mm, and often use it for exact depth measurement. Since the exact depth is basically the only measurement on most diamond which can be perfectly measured, having that bit of data seem of some potential importance in identification. Virtually no diamond has been measured with that sort of precision or repeatability with any scanner. If I need to be sure I am seeing the same stone at a later date having such a precise measurement adds a layer of assurance, but we don't need that accuracy to determine a cut grade.

 

[AV_]

I'd have considered rounding but for both GIA and the Sarin numbers both showing significant figures to the hundreth place, at least on several measurements.

Just bragging...

I read somewhere around here that the precision achievable in cutting is more than grading arbiters can measure - funny enough; but, who - where - how?

Invisible details have some kind of draw.

 

[Venzen007]

If I need to be sure I am seeing the same stone at a later date having such a precise measurement adds a layer of assurance, but we don't need that accuracy to determine a cut grade.

The underlying reason for my interest came out of a concern for being able to use these measurements for identification.

 

[AV_]

All the details together must be enough for ID, even if none are exact, or so I think.

 

[Karl_K]

Just bragging...

I read somewhere around here that the precision achievable in cutting is more than grading arbiters can measure - funny enough; but, who - where - how?

Invisible details have some kind of draw.

True at one time and at the normal settings probably today.
At the slowest highest resolution settings maybe not with an RB.

 

[Karl_K]

The resolution rating is the max the machine can in theory achieve, not the real world everyday balanced between speed and accuracy settings used in the labs.

 

[Karl_K]

All the details together must be enough for ID, even if none are exact, or so I think.

True in that there is a balance between scan time and accuracy that plays a large part in the outcome, with gia the need to symmetry grade by scan sets the accuracy standard higher then the gia gross rounding of all the numbers requires.

 

[oldminer]

Venzen007;
The lab report provided measurements along with apparent color, clarity, clarity plot, laser inscription, girdle thickness ranges, culet size, UV fluorescence strength, table %, crown angle all go into re-identification by an expert examiner. A consumer cannot hope to re-identify a diamond from measurements alone when even an expert wants and needs more data to be really certain. The measurements of a diamond are an essential and important re-identification tool, but trickery is not as rare as one would imagine.

Like Karl suggested, the labs don't necessarily use their scanners at their finest resolution on all diamonds. Even when they go slow and measure at their best, there is machine error that better contact type mechanical tools can pick up. The minor differences don't mean anything of importance to the current accepted grading process, but measurements on lab reports are not absolute in precision.

It may be that the few rarefied vendors of a Ultra-Ideal cuts found on Pricescope do use more precise tools and more precise measures at their diamond cutting shops to get the perfected results they are marketing. However, the measurements on AGS Lab reports which come with those stones do not reveal the proprietary and rather secretive additional labor and measurements which result in the cutting of these special diamonds.