User tips for DiamCalc

[Garry H (Cut Nut)]

Regular Guy, that is on http://www.ideal-scope.com/Using%20DC_files/frame.htm

But since doing this we have noted that sometimes the table size estimation is wrong because of paralax errors related to camera lens proximity.
Follow the instructions, and after you have completed everything, go back and enter the known table size if you have it.

 

[Garry H (Cut Nut)]

To open a Gem Adviser file in DiamCalc (e.g. from a vendor website), click on the Gem adviser file or image.
A dialogue box will ask you to save it and where to put it.
Before saving it, change the file ending from .gem to .dmc.
Then open the file using OPen> (top left tool bar).

If you want to open an already filed .gem file in DiamCalc rather than Gem Adviser, go to the correct folder and Open > with the same top left tool bar control. You will only see .dmc files displayed. Hit the option in the lower Box option for All files (*.*)

Find the file you want as a .gem file and open it. It will work. You can also change the file type by renaming it, or when you have it opened in DiamCalc, simply resave it.

 

[Regular Guy]

New question, or set of questions...

When using Diamond Calc, can you help me understand the expected relationship between:

- a) Cut quality score
- b) IS image (I would have thought (a) & (b) to be similar, except for testing just done)
- c) HCA score
- d) cut type, i.e., TIC, FIC, BIC

Sorry, maybe too big a question.

I wanted to see how DC would interpret this phenomena of the below 1 on HCA Storm's been talking about, and first, sought to compare it to a solid "1" HCA ranked score. I went to the search by cut database, accepted all the defaults (.9 - 1.5, etc) and said OK, then sort twice on HCA scores, to get a sort of low to hight, and at the top of page 2, found a stone, modeled it in DC, and got this (see attached GEM file, showing very nice light return of .99, .98 stereo):

 

[Regular Guy]

(continued)

...then, to model another one, I go back to the sorted database, pick the diamond scoring .4 on the HCA at the top of page 1, also conforming to Storm''s conjecture, again model it in DC, and it''s attached as a Gem file. Here, the light return is not as good...but not shockingly bad....: .99 LR mono, and now .95 stereo.

 

[Regular Guy]

...for add some perspective, I throw in the data for my wife e-ring, with an HCA score of .9, an FIC, and see how the data presents; the GEM file is attached, and new questions arise for me. Do the Cut Quality scores of .88 mono and .91 stereo kind of disappoint? I look at the IS image, and it looks kind of good. Is it possible the FICs should be evaluated differently somehow, such that the cut quality presentation becomes less relevant?

 

[Regular Guy]

...finally, turning again to the default sorted options, I turn again to the last diamond presented, in the sort organized lowest to highest for HCA scores (with the worst still being very good, but about 3.6 still), and the gem file is presented. Now here, the cut quality score is a good bit better than my .9 FIC, with .96 for light return, both mono & stereo. But...the IS more clearly disappoints, and may be more consistent with the HCA score.

Four entries ago, above, I review the questions on the table, here. I am trying to understand what info I should be evaluating, helpfully, from DC, and looking for meaningful relationships between the number presentation of cut quality scores, the IS view, their relationship to HCA scores, and if the type (TIC, FIC, etc), may be a confound for the DC presentation of info.

Thanks in advance!

 

[Garry H (Cut Nut)]

Ira:

1. read the AGS and GIA threads - understand the issue of stereo scopic vision

2. DiamCalc does light return with the stone rocked thru 30 degreess (in a knight type movement). You can open the dialogue box and see the results for table only - and also whole top of the stone.

3. Sergey removed leakage calculator a few years back because there is not such a strong correlation with IS type leakage and light return. again - read the AGS and GIA threads - understand the issue of stereo scopic vision

 

[Garry H (Cut Nut)]

Contrast is useful - but it does not yet have a "distribution" measure for human preference.

There is no scintillation and no fire yet.

High brightness and high fire are rare - maybe even impossible - to be combined

 

[pricescope]

There are some discrepancies between DC and HCA scores as well as IS images. I do not think it is worth trying to find connection or disagreement between them (I''ll try to explain below why).

Although each of these tools (as well as any other tool available today) are helpful, they probably can only be used as an aid in making preliminary screening. In order to get the most of them, it helps to understand how each of them works.

I-S shows and magnifies the effect of the "light leakage". It was shown many years ago that diamonds with a lot of light leakage look bad. Therefore, people decided that any light leakage is bad. I-S magnifies the effect.

The good:

1. provides standardized viewing environment
2. easy to spot and reject stones with a lot of light leakage
3. many diamonds with little leakage has a great chance to look great
4. can be used on fancies as well

The bad (it was discovered/realized a couple of years ago):

1. Small or limited leakage can not be too bad - diamonds with some leakage can still look great, localized leakage can help by increasing the contrast
2. Evaluate diamonds with mono (one-eye) vision
3. static, table up position only
4. shallow diamonds do not sow leakage but can appear dark because of the head/observer obstruction effect

HCA. It was designed based on both I-S and DC data. It used to penalized diamonds for the light leakage and also was the first method to add spread as one of the desirability factor. Some corrections were made later to penalize darker shallow stones and to ease penalties for some limited light leakage.

The good:

1. fast and free tool to weed out known bad performer
2. gave an opportunity to avoid old AGS0 steep/deep.
3. find a good candidates outside of old AGS0

The bad:

1. Scales/ratings are subjective. e.g. diamond with 0.5 score might not be nicer than the one with 2.0.
2. relies on the measurement data that might not be accurate enough
3. doesn''t take girdle, minor facets, and symmetry into equation
4. problems related to I-S
5. cannot be applied for fancies

DiamCalc is the work in progress and a great tool for modelling and study of the diamond cut.

The good:

1. Helps to predict light performance and appearance for both mono and stereo vision as well as in static and tilting positions.
2. allows to simulate diamonds in motion
3. With certain skills, one can learn to predict diamond appearance for different combinations of proportions.
4. allows to import 3-D diamond models.

The bad (not really bad but limitations of using it as a grading tool, which is never actually was claimed

):

1. Light Return score cannot be used for grading diamond beauty (again it was never meant/claimed to be)
2. Requires learning curve to use properly
3. I cannot think about anything else because DC is designed for studying rather than grading.

HCA supposed to become obsolete when more reliable grading system will become available (AGS, GIA, or MSU/Octonus). My point is not to hook on the particular results, scores of any of these tools. They might have poor correlation between each other for certain proportions. It is fine as long as we understand how to interpret these results.

Hope this makes sense

 

[belle]

it makes perfect sense

thank you leonid for giving an overview of these tools. i hope people will read your post and understand the benefits and limitations for each before using them to make any concrete determinations of diamond performance.

 

[valeria101]

Date: 8/29/2005 2:05:12 PM
Author: Pricescope


DiamCalc is the work in progress and a great tool for modelling and study of the diamond cut.

The good:

- allows to import 3-D diamond models.

The bad (not really bad but limitations of using it as a grading tool, which is never actually was claimed):


- I cannot think about anything else because DC is designed for studying rather than grading.

Quick question about the 'good' cited above:


How good do these imported files look? E.g. for 3D Sarin, is it necessary to make manual adjustment in order to make the DC image look like the stone? Or... is it enough to import the data and see a reliable wireframe model.




About the 'bad' cited.

If only DC had better graphics, it could be allot more. Seriously. Why not

 

[adamasgem]

Date: 11/26/2005 4:20:39 AM
Author: valeria101

Date: 8/29/2005 2:05:12 PM
Author: Pricescope


DiamCalc is the work in progress and a great tool for modelling and study of the diamond cut.

The good:

- allows to import 3-D diamond models.

The bad (not really bad but limitations of using it as a grading tool, which is never actually was claimed):



- I cannot think about anything else because DC is designed for studying rather than grading.

Quick question about the ''good'' cited above:


How good do these imported files look? E.g. for 3D Sarin, is it necessary to make manual adjustment in order to make the DC image look like the stone? Or... is it enough to import the data and see a reliable wireframe model.




About the ''bad'' cited.

If only DC had better graphics, it could be allot more. Seriously. Why not

The graphics are quite good, BUT in general, it is for an NON ABSORBING diamond, and look quite much better if you use a correct absorption data. I''ve actually asked Sergey to pur in a in input file. I am going to add the capability to generate an estimated absorption file from my SAS2000 data. It is sort of a kludge as we are dealing with cut stones but using n=my 3D foreward ray trace I''ve made some estimates of the average pathlength I see in the method and illumination I use.

DiamoncCac acn be VERY USEFULL, and I have long recommended it, although I am a novice in its use, but I played with it after your post, and I do recommend it to my clients, and have had the methodology to do the absorption in the SAS2000 for a long time. AGS, I believe used it in the development of their system, and it is MUCH QUICKER (factors of 100) than the forward raytracing I used to verify what GIA was saying in their Brilliance article. Sergey''s ETAS and my Chromatic flare agree quite well, I believe, and Sergey is the only developer who is open about what he is doing, and we had a VERY informative technical thread last year on PriceScope..

I''ll try to supply some representative absroption files to DiamondCalc based on my masters and analysis. those, coupled with accurate Helium data, will give pretty nice visualizations I believe, but the time to do it will increase. It will also depend on the illumination envirionment you model.nn the chromatic flare (FIRE) depends on the color of the diamond, as well as the virtual image and the fineness of the simulation parameters you select.

i would prefer to see 5 or 10 nm inputs rather then the 20nm internals Sergey chose (to better capture the diamond 415nm absorption), but that would substantially increase the simulation time involved in the processing.

 

[valeria101]

If this thread is meant to collect all odd bits about DC use...

This is yet another.

Bruce Harding has promised an piece about bow tie (on this thread). Which reminds me - bow tie hardly shows in DC graphics. It does in IS and Asset setups (or anything built like them), but not the standard lighting settings. I have not tried to adjust them to show bow-tie any better than already done.

For example, I would expected to get one by changing the azimuths of pavilion facets (not on the ''oval'' model, but on a stretched ''brilliant'' model that gives me that option). And... not much happens.

Perhaps I just do not hate enough this bow-tie fault to see it. Anyone tried the same?

 

[valeria101]

Date: 11/26/2005 9:51:02 AM
Author: adamasgem


The graphics are quite good, BUT in general, it is for an NON ABSORBING diamond, and look quite much better if you use a correct absorption data.

I don''t want to sound ungrateful... your detailed reply to my poor post has bee around for a while and I should have answered.
However, after some attempts to understand what exactly changes in the graphic representation from one procedure to another, I have to admit deafeat.

By ''better graphics'' I mererly thought of... something impresionistic rather than technical.

 

[Garry H (Cut Nut)]

To make your own lighting models in DiamCalc you should follow this procedure:

Open the Lighting control box by >Options>advanced>Lighting Configuration OR by >Options>Lighting>Customize

Then with any type of lighting turned on, eg jewellery shop, Expand the Lighting Configuration and play with lights, move them around by clicking + and altering size and direction etc. Click apply as you go to see the effects.

You can see what the lighting looks like by going back to >Options>Advanced>Lighting Map where if you click a couple of buttons you can see where the lights are in variuos hemisphere views.

 

[Garry H (Cut Nut)]

Now you must save your modeled lighting for later use.
When you do this with the lower left button you must also tick "Simulation Parameters" - this will be automatically ticked in the next release.

And to open the lighting - 1. you must have some sort of light turned on - not a wire frame.
2. open up the lighting control box and click >Load and open the lighting model you saved.