Here you can find information about what is new in HP Carbon version 1.11.10.
This version is available for download from the Octonus official web site. |
To improve the model building process, we implemented a mechanism to include contours with light dazzle from the model building. It results in improving of model building quality.
Some examples:
| HPC 1.9.2 | HPC 1.11.10 |
|---|---|
|
|
| HPC 1.9.2 | HPC 1.11.10 |
|---|---|
|
|
| HPC 1.9.2 | HPC 1.11.10 |
|---|---|
|
|
|
|
Sometimes clients faced following message during attempt of scan&build models containing contours with light dazzle:
Model was not built for such cases.
Now contours with light dazzle are corrected during scanning. For example, compare contours for the same photo with light dazzle below:
Before | After |
|---|---|
|
|
|
|
|
|
Introducing the new allocation algorithms: "23. Single", "23. Next" and "23. FixedForm" (please refer to page Algorithms comparison to understand meanings of algorithms):
The table provides the average algorithm execution times on modern rough models, measured in seconds. These results were obtained using an Intel Core i7-7700K processor.
New algorithms find the best cost-effective solution considering the specified price list. However, user needs may differ.
Upcoming adjustments:
Performance-driven solutions
- Why? Sale price depends on performance, not just diamond size
- What? The algorithms will provide solutions with best performance in addition to the maximum price solutions,
More diverse solutions
- Why? Some customers rely on external price lists for evaluation, inaccessible to algorithms
- What? Algorithms will provide more diverse solutions in terms of cut and grade, if the user desires, even if it requires additional time.
Continuous improvements for user-centric solutions
- Addressing customers needs
- Providing personalized solutions for enhanced decision-making
- Optimizing algorithm efficiency for improved productivity
While new global allocation algorithms prioritize finding the most price-effective solution based on a given price list, it's important to recognize that user requirements are diverse.
The value of a particular diamond to a company depends on various factors beyond its carat price alone, including:
The algorithms are unable to consider all these factors when searching for the optimal solution.
We offer users the ability to customize modern algorithms for obtaining more diverse solutions in terms of cut, grade, and clarity. So users can access more choices and improve decision-making.
Of course, it's important to note that additional solutions searching might require a slightly longer processing time. However, you have the ability to customize the balance between algorithmic speed and solution diversity.
Our latest update introduces three distinct levels of diversity: cut diversity, grade diversity, and clarity diversity. You have the flexibility to choose which diversity levels to use based on your specific needs and preferences. Upon installing this update, default values will be set, with cut diversity at 20%, grade diversity at 10%, and clarity diversity at 5%. Rest assured, your current settings will be saved even if you exit the program, allowing you to continue next time with your customized levels of diversity. In the old algorithms, diversity levels are inactive and not used. We recommend you to explore and use this new feature wisely.
The values refer to three parameters: Cut, Grade, and Clarity. When you assign higher values to any of these parameters, you will receive a greater number of diverse solutions based on that particular parameter. However, keep in mind that increasing the values will also result in longer algorithm processing time.
While searching for solutions, these three parameters have an impact on the price of the solution, along with the weight. The algorithms aim to explore all possible combinations of these parameters. The algorithms optimize their search to find the best outcome with minimal computational efforts. At any given time, there exists a globally best solution, referred to as GBS. If the predicted final price of "Cut," "Grade," and "Clarity" parameters specific combination is lower than the GBS price, calculations using that parameter combination are halted to save time. Therefore, only solutions that have the potential to surpass the already found GBS are computed until the end. This behavior corresponds to a Diversity level of 0%.
Due to this time optimization, you may receive only one combination without any choice. Use Diversity to obtain a greater number of options
If you set Cut Diversity to N%, after finding GBS, the algorithms initiate searches for solutions with a price higher than [GBS_Price * (100% - N%)] for each diamond cut. The search doesn't start from scratch but utilizes the intermediate results obtained during the search for GBS. Consequently, for each Cut, a solution will be obtained if it differs from the best solution by less than N%.
The same principle applies to Grade and Clarity Diversity as well.
If you set Clarity Diversity to 100%, you will get the best solution in every Clarity, but it may take a long time
You can view the Diversity levels values in the Alloc field tooltip of the solution.
It's important to note, Diversity levels have a significant effect on the speed of the algorithm. The default settings of Diversity levels slow down the algorithm by approximately two times when working with 10 cuts. To indicate the potential productivity impact, Diversity scales are logarithmic and have color differentiation. It is not recommended to use values above 30%. If you always choose the most expensive solution, we recommend setting the Diversity levels to 0% and regularly updating the price lists to ensure accurate targets for the algorithms.
A new cut Emerald_3x4_GRID SmartCut was added.
Emerald_3x4_GRID is an upgrade of the Emerald_3x4 SmartCut via a GRID Performanceware appraiser.
The GRID Performanceware appraiser contains grades based on Cutwise Brilliance metric for approx. 600000 Pa1-Pa2-Pa3-Pa4-Ca1-Ca2-Ca3 Emerald slopes combinations.
Emerald_3x4_GRID has 3 preconfigured Profiles: Max_Mass, Commercial, and High_Performance.
The previous Emerald_3x4 SmartCut was based on the rectangular appraiser and SweetLine. It could produce solutions with large black zones in some cases.
Emerald_3x4_GRID could filter these bad parameter combinations using GRID Performanceware appraiser.
The Emerald_3x4_GRID cut optimization forms are updated and encrypted. These forms are now exclusively accessible by the allocation algorithms within HP Carbon.
For example, Emerald_Demo5 project
Typically Emerald_3x4_GRID gives solutions with approx. 2.5% smaller Carat Weight, but 5-10% higher Brilliance.
The global allocation by for DLL-cuts Princess.2S-Princess.5S was modified in the HPCarbon version 1.11.10. Previously, the main focus was on preserving the height of the chevrons, whereas the latest version ensures similar widths of the chevrons.
| Previous | Current |
|---|---|
|
|
Furthermore, both the starting form and allocation cut forms was updated.
А new cut Oval_WBT_C32_G64_P24_2023_MAY_18 is added.
This cut is based on the previous Oval_WBT_C32_G64_P24_2017_JUN_21 cut with following changes: the appraiser was re-designed to seamlessly integrate with the standard Smart Cut optimization process. Global allocation algorithms focus on the Absolute part of the appraiser and a comprehensive set of internal parameters, while SmartRecut algorithms operate on the main set of parameters, including both Absolute and Relative parts.
Finally, the Ultra_High_Performance, High_Performance, Commercial and Max_Mass profiles was specifically configured to yield desired results for different targets.
In diamond production, at each technological stage, manufacturers work with a diamond represented by a scan model and a virtual model of the plan they aim to achieve at the current stage. As a result of processing, an intermediate iteration occurs. After completing the physical processing of a stone at each stage, the intermediate iteration is scanned.
It is necessary to assess the processing of the intermediate iteration, and the following questions arise:
1. What should the intermediate iteration model be compared to?
2. How can the quality of the intermediate iteration processing be assessed?
3. Stick to the plan or make changes?
To address these issues, a 'preplan' concept has been developed - a predictive model expected upon completion of each technological stage. The preplan represents a virtual model of the diamond-cutting process.
Moreover, a specialized tool is required to transfer accumulated data through a technological pipeline into a new scan. The accumulated data includes the original plan, solutions, inclusions, cavities, and previous scans.
At technological stages of diamond processing, such as Sawing, Table Polishing, Bruting, Blocking, and Polishing, the seamless transfer of accumulated data from previous projects is crucial when obtaining a new scan. A distinct preplan is created at every stage.
Importing previous data into a new model is executed through a model-matching technology called Recognition. To facilitate this process, the utilization of a preplan is introduced to the Recognition.
Notably, the Recognition algorithm has become more reliable and accurate in the latest version of HP Carbon due to information about the expected result.
| HPC 1.9.2 | HPC 1.11.9 |
|---|---|
|
|
In the HP Carbon 1.11.10, a new type of preplan for Sawing has been introduced.
This enhancement allows users to create a Sawing preplan for a diamond in solution, comprising two diamonds and a sawing layer.
For precise angle adjustments using the Compass, it is essential to have well-polished reference facets with angles within the Compass's operational range of up to 5 degrees. This eliminates the need for additional utilization of dop clicks and fine-tuning. During the blocking procedure, these reference facets are created using the conventional Dry polishing technique.
Additionally, we require a new procedure to execute the first reference facet and establish the facet sequence.
We recommend following the outlined pipeline for Autostop.
It is essential to perform the "Correction of Actual Model," which involves three key steps: scanning, Recognition, and setting the newly created scan as the Main Scan.
Initially, the operator establishes processing orders separately for Crown and Pavilion within the Facet Types.
Ensure minimal risk of cutter error by recommending a secure choice for the 1st facets in the processing orders for Crown and Pavilion.
If the cutting depth of the 1st facet in a processing order for Crown or Pavilion is less than 20 microns, it is considered unsafe. In response, an evaluation is carried out on the symmetrical facet to the first one within the same facet group of the processing order. The identified facet is subsequently reassigned as the new initial facet.
The confirmation of the safety of the 1st facet is now integrated into the Facet Types. This validation becomes active during the saving process that occurs after the creation or modification of the processing orders.
Before | After |
|---|---|
|
|
Firstly, a cutter has to create a preplan (Simplified Plan) for blocking of the target Plan both the Crown and Pavilion. This process is carried out based on the facet processing orders in the "Generate Next Stage" panel, specifically in the Blocking mode.
Please ensure that the option "Align facet azimuths with the dop click" is specified, as it is turned on by default. Additionally, specify the depth allowance with the default minimum set at 100 microns.
The resulting preplan model takes into account the specified depth allowance.
Steps 4 to 7 for the Crown must utilize the preplan as the Galahad plan.
Create a faceting stage for 1st facet for Crown
Please choose the "First facet" option in the "Facet processing mode" control.
Create the appropriate faceting stage for the 1st facet in the Facet Types processing orders for the Crown by clicking the "Generate Stage" button in the "Generate Next Stage" panel. Please save the file to store recent changes in the Faceting Stages.
Make the 1st facet for Crown
Manufacture of the 1st facet in the Facet Types processing orders for Crown. The initial facet should be created by a cutter using visual judgment.
Then make a HP scan of semipolished diamond with 1st facets on Crown.
Repeat the step 4.2 for Pavilion.
Please select the "Pavilion" option in the "Facet Types processing order".
To achieve the desired azimuth and slope of the first facet for the Crown, adjustments will be made with depth allowance in Compass Autostop. This process requires utilizing the Recognition of the Scan data from the previous stage, resulting in a model with the accurate azimuth and slope of the first facet.
Making all facets in the processing order for Crown, except the first one, of the preplan.
Please select the "Exclude first facet" option in the "Facet processing mode" control.
Generate the stage and save the file.
Finalizing the model from the previous stage based on the initial Plan without making any adjustments to the facet azimuths during the processing stage.
Please select the initial Plan, activate the "Do not align facet azimuths with the dop click" option, and ensure that the option "All facets" in the "Facet processing mode" control is specified. Create the faceting stage and save the file which will be used in Compass Autostop.
An addition has been made to present the number of dop clicks from the first facet in the processing sequence (absolute), as well as from the preceding facet (relative).
Given that the dop handle for setting the azimuth can rotate in both clockwise and counter-clockwise directions, an indicative sign has been added to the count of DOP clicks: "+" signifies an increase in azimuths, while "-" denotes a decrease.
The calculations regarding the number of dop clicks have been conducted under the premise that the switch between facets by azimuth can be accomplished with a minimal count of clicks, taking into account the direction of handle rotation.
Furthermore, the display of the processing stage is introduced, showing whether it is Polished or Blocking.
I3D report incorporates the "Global Maximum" line (BlueLine) into the "Remaining Depth" mode within the I3D Report, displaying the distance to the furthest areas of a scan as a whole.
is eliminated.
Functionality has been implemented to prevent inclusions from entering the interior of a diamond.
An inclusion can enter the diamond if it is located too close to the diamond model. To prevent this, the use of a Safe inclusion buffer is proposed. The Safe inclusion buffer is a copy of the inclusion model displaced by a certain distance (Safe distance) towards the relevant diamond (active diamond). To check if the diamond is at a safe distance from the inclusions, the following steps need to be taken. For each inclusion located near the diamond, create a Safe inclusion buffer and initiate Bound Swim. The resulting solution will be considered "safe", i.e. all inclusions are further away from it than the Safe distance.
Two methods for selecting inclusions to create the Safe distance have been implemented: manual and automatic.
To create a Safe buffer for a specific inclusion, select the "Create Safe buffer for active diamond" option from the inclusion's context menu.
Choose the desired Safe distance.
As a result, the inclusion will now have a Safe buffer.
In this method, all inclusions located at a distance from the diamond equal to 2 * Safe distance are automatically selected to create Safe buffers.
Select the option "Create Safe buffers for inclusions near the active diamond" from the menu in the panel header.
Choose the desired Safe distance.
As a result, multiple inclusions will now have Safe buffers. Here are two examples of such inclusions.
Additionally, we have incorporated a convenient option for removing previously created Safe buffers all at once.
When working with inclusions, the main focus is on the inclusions that are located within the scan. On the contrary, inclusions located outside the scan can be disruptive, and the operator may want to remove them. For this purpose, an automatic removal of inclusions located outside the scan has been implemented.
To remove inclusions, select the menu item "Remove inclusions outside the scan". All inclusions located outside the main scan will be removed.
The image shows the inclusions before and after removal.
| Before | After |
|---|---|
|
|
Now you can delete multiple inclusions simultaneously:
In the Inclusion filter, fixed the message if both check boxes (inside and outside) are unchecked. Was "filter failed", now "filter turned off".
The filter worked wrong in case of Sandwich (shell was wrongly hidden).
Enforced crash guards in the Solution List and the method selector in Scan & Build mode.
Fixed: The width and length calculations for specific Heart shapes, along with their orientations in the reports (diamond_82007.ox2z):
| HPC 1.9.2 | HPC 1.11.10 |
|---|---|
|
|
Removed the need for confirming file replacements and introduced the full installation mode in the Report Templates installer.
| HPC 1.9.2 | HPC 1.11.10 |
|---|---|
|
|
