shape
carat
color
clarity

AGS less strict on colour?

You are vastly under estimating how hard it is getting GIA to change a grade on a recheck or regrade even when you think they are very far off.

Is the stone linked to the original grade even if sent back in again fresh?
 
Is the stone linked to the original grade even if sent back in again fresh?
They say they can do so even if the number is removed or it has had minor work done. It is also something like 1/2 price to send the original report in with it.
My belief is they can with more recent stones but if you sent one from say 10 years ago they probably could not if it didn't have an inscription.
 
There doesn't seem to be much consensus at all. I mean we don't even really know if GIA is using machines for colour grading yet. Seriously.
I believe this was addressed earlier in this thread - primarily at GIA NYC and Carlsbad locations, and only ever for melee sizes. And for the remaining millions of stones they see per year - https://www.gia.edu/doc/Coloring-Grading-D-to-Z-Diamonds-at-the-GIA-Laboratory.pdf
https://www.gia.edu/gia-news-press/gia-ibm-transform-diamond-grading
https://www.ibm.com/case-studies/gemological-institute-of-america/
 
Last edited:
I believe this was addressed earlier in this thread - primarily at GIA NYC and Carlsbad locations, and only ever for melee sizes. And for the remaining millions of stones they see per year - https://www.gia.edu/doc/Coloring-Grading-D-to-Z-Diamonds-at-the-GIA-Laboratory.pdf
https://www.gia.edu/gia-news-press/gia-ibm-transform-diamond-grading
https://www.ibm.com/case-studies/gemological-institute-of-america/

Texas leaguer said they they did some by machine but we don't have confirmation on what carat size it goes up to.

There are some articles where GIA suggest most colour grading is automated. If that really is the case then we can put the human element as a big factor to bed.

The IBM stuff looks interesting.
 
There are some articles where GIA suggest most colour grading is automated. If that really is the case then we can put the human element as a big factor to bed.
Then you open a huge can of worms of the machine element and the conversion to the old system.
The hardest parts are raising a error on stones it can not accurately grade for what ever reason and taking the results and interpret them on the existing color scale.
That last part is a real issue because the results for 5 stones all say G grade may look nothing like each other from the raw data. They may have different tone/colors or even color zoning or a colored inclusion or a ton of other factors. Even if they are clean and dont have embedded metal from tweezers which a human grader should catch quickly.
 
The end game is that I buy a stone sight unseen and when it arrives it is what I was expecting. Nothing more, nothing less but definitely not 2 colour grades lower which can easily happen when balancing the 4C's and cross shopping labs. I've adjusted my expectations but I could have saved some time had I paid heed to earlier advice on this very board on the subject.

That's not the message seemingly being conveyed based on your adamant replies to everyone participating in this thread.
You stated in another thread that you're abroad, so you won't be able to see any diamond before you buy. Your concern is warranted since you should absolutely get validation, regardless of the grading lab for any particular diamond, that you will be getting what you expect for your money.
My final advice: Because you are worried about color grading being inaccurate for the AGS diamond that you have narrowed your list down to, then perhaps ask WF to provide you a set of side by side pictures and videos of that AGS graded diamond to multiple GIA diamonds of the same color grade...they may have them in stock, since their Premium Select line are GIA graded.
If not, then my advice would be to simply stick to what you are familiar and comfortable with: GIA graded diamonds or snatch up that CBI you are also interested in.
You're going to drive yourself insane with analysis paralysis, otherwise.
Cheers!
 
That's not the message seemingly being conveyed based on your adamant replies to everyone participating in this thread.

If at the end of this some new information ends up being shared, facts confirmed or myths put to bed that's a good thing for everyone.

Cheers
 
Then you open a huge can of worms of the machine element and the conversion to the old system.
The hardest parts are raising a error on stones it can not accurately grade for what ever reason and taking the results and interpret them on the existing color scale.
That last part is a real issue because the results for 5 stones all say G grade may look nothing like each other from the raw data. They may have different tone/colors or even color zoning or a colored inclusion or a ton of other factors. Even if they are clean and dont have embedded metal from tweezers which a human grader should catch quickly.

If they're using some sort of AI or computer learning then I imagine humans step in for edge cases which then inform the analysis for the next time. It would seem to me much less complicated than clarity grading at least
 
If they're using some sort of AI or computer learning then I imagine humans step in for edge cases which then inform the analysis for the next time. It would seem to me much less complicated than clarity grading at least

No brainer. I expect fewe humans grade diamnds under 1ct at GIA. They grade probably 5 times more diamonds under 1 carat than over.
There are 111M under 1ct listed on RapNet now compared to 333k over 1ct.
The price difference for color increases dramatically as carat weight increases (because the grading systems are totally messed up - as diamonds get bigger clarity and color grades get softer).

Buyers are less likely to challenge a grade on a smaller diamond because of the cost and time, so its a case of "who cares". Also if they try to resubmit a smaller stone for another try they first have to polish off the mandatory free laser inscription. (You have to pay for laser inscription over 1 ct)
 
GET 3 FREE HCA RESULTS JOIN THE FORUM. ASK FOR HELP
Top