Thanks very much for building this great website and your helpful contributions elsewhere. I notice on Reddit you mention blue hue a lot. So my question is, in your experience, do blue hue diamonds receive a higher letter grade for equivalent color saturation (CIE C*) in comparison to yellow hue diamonds? This could be the case if a grader has a Cape series master set and is grading “whiteness”. On the other hand, at least GCAL is explicit that they grade by saturation and try to ignore hue. So what would you say is the deal on letter grade vs C* vs hue angle?
From what I’ve seen, I don’t think blue receives a higher color grade (at least not compared to the variance within each grade for IGI). Although I’m not completely sure or have any real data to back it up.
For naturals, there is a separate set of master stones for brown (https://www.gia.edu/doc/Coloring-Grading-D-to-Z-Diamonds-at-the-GIA-Laboratory.pdf). Quoting: “Some observers try to grade just as they would yellow diamonds, and only look for saturation differences (the “amount” of yellow), which can result in an incorrectly high determination compared to laboratory grading. If yellow master stones are the only ones available, the observer should assess the overall depth of color and equate it to the overall depth of the yellow master stone.”
Do you have a link to the GCAL description? Grading saturation only seems contrary to what GIA has described to avoid what you’re suggesting about “incorrectly high determination.”
Thank you for linking these videos – I will be sure to make my way through the series.
Perhaps I should have also quoted the preceding sentence in the GIA report: “When doing so, the observer must remember to assess the overall depth of color— the combined effect of tone (lightness to darkness) and saturation (strength or weakness) of a color (King
et al., 1994).” That is consistent with the GIA fancy color description (https://www.gia.edu/doc/ColDiaChartBklt.pdf). On pages 8-10 of the PDF (pages 5-7 by page number) they show various examples in the HSL color space. On page 8, they say that colors show their strongest saturations at different lightness levels. So my interpretation was that “overall depth of color” should account for saturation+lightness as opposed to saturation alone when the hue is different.
That stands in contrast to what is being said in the video you’ve linked arond 28:30 in, where GCAL says that they’re looking at saturation only.
This is somewhat beyond my knowledge of color spaces and the algorithms used for automated color grading by something like the Sarine machines. Should you use saturation only or saturation+lightness and which one is a better model that matches human perception of diamond color? Interesting topic, but beyond my knowledge.
I don’t shy away from blue because I think they’re overgraded for color. I shy away from blue because I think most people who buy a lab diamond want it to look similar to a natural diamond (i.e. not raise any questions about whether it’s lab or natural). A yellow hue is most similar to the typical natural diamond. Yellow HPHT is the same price as blue HPHT, so I recommend yellow over blue unless unless there’s a stated preference for blue.
Thanks again for your comments. To tell you the truth, I’m not completely sure how to interpret the GIA observations. Color spaces like HSL suffer from not being perceptually uniform or orthogonal, because they’re based on RGB primaries rather than the CIE XYZ tristimulus values. So if they used that space rather than the similar but perceptually uniform CIEL*C*h°, this could easily account for why the saturation and lightness are correlated. They don’t mention explicitly which color space they refer to in this otherwise very useful document (thank you for the link). Perhaps none in particular, if they’re still eye grading for the most part. There are certainly some oddities of color perception (e.g. dark, not saturated orange is perceived as brown) that aren’t fully captured by measurements, even when the measurement is faithful to the human vision system.
It’s fascinating to me how much the perception of a diamond will vary based on cut and polish quality, or even just size, given identical material characteristics. Although grading is standardized, it perhaps isn’t always representative.
Thanks for the clarification on your reasoning for not preferring blue HPHT. I’m somewhat surprised that blues and yellows are available at the same price. High color grades are not so straightforward to make in yellow hues as you need a lot of/very efficient nitrogen getter (easy to end up with severe inclusions as these are metals like Al or Ti), whereas blue is fairly straightforward in that you use less getter and add enough boron to compensate whatever nitrogen you have left over. That being said, a little excess boron makes it very blue, so getting the balance exactly right may be difficult.
The GCAL methodology is described here: https://www.youtube.com/watch?v=6TqqvhELqb8 (nice series of webinars, by the way).
From my data, IGI has the highest repeatability of all labs, suggesting they’re grading by machine (probably just C*). I didn’t know that GIA has separate yellow and brown master sets; thanks for that information. By my reading, their advice is to avoid grading by “yellowness” and instead look at “non-whiteness”, which is consistent with GCAL’s approach, I think. But as blue can look “whiter than white”, I wondered if blue nuance can get a grade bonus compared to the yellows and browns when an inexperienced grader grades it by eye, and that’s why you shy away from it? Grading strictly by C* should avoid that problem, so I don’t necessarily expect to see it in IGI results. GIA data would be very interesting, assuming they don’t also have a blue master set.