Hi all, Long time reader, first time caller... I will admit that I'm moderately new to using Context Capture, but I have been finding myself spending quite a bit of time with it lately. I've got a few questions related to CCC and network processing/virtualization: 1. Who is the lord of Network processing? There must be one of you out there who are processing (like me) serious amounts of data. I've had to hack together a solution to get ccengine to function, but I'm finding it really difficult to understand how exactly the preferences are being set. As an example, running a windows10 client machine with ccengine as a startup will fail every single time - the userspace hasn't mapped out the network drive because you haven't used winlogon, which is the dependent action before ccengine can run. I figured out a way to make it work, but I'm not really happy with it. 2. Is anyone running Vulkan via virtualization? I have a bunch of ATI Instinct cards (M25) that I am running for these VM's, and things seem to be running just fine. My biggest curiosity is now - does CC use opengl *AND* vulkan at once? I've set Vulkan up, but after reading about it more I came to understand that Vulkan is responsible for CPU/GPU management optimization.. I think I am doing things right, but I would love to hear from others on how their setups went. 3. Is there any understanding of who has the fastest (read: best) Bentley cluster in the world? I am sure there are some quiet customers with big big big Bentley processing nodes, but I am wondering what kind of setup any of you have out there. My current setup is 10x 64 core / 400gb ram machines with 1x M25 Instinct cards running Vulkan per pod.. I'm usually running one pod, but once Azure gets their capacity restraints solved from the Batflu outbreak, we hope to have 5-6 pods running (50-60 machines) at once. I want to use docker to potentially spin up more, but the lack of graphics card support has me running regular VM's which honestly has been one of the biggest nightmares of 2020 for me. If I have any advice for anyone, its stay away from Azure if you are using Context Capture Center. It has been absolutely nothing but problems, and very very very expensive to operate. 4. Does Bentley care about Context Capture? I see interesting things coming down the pipeline. We are an EAP customer, and I'm just totally baffled by the lack of core functionality while they're adding things like tensorflow classifiers to detect objects in meshes. Frankly, we have been able to do that for a pretty good amount of time now and I don't consider it difficult to do. What I consider more difficult to do is get a mesh out of Bentley that doesnt suck without adding 1.2 million tie points. I've already come up with a solution to this problem, but I'm curious to know if Bentley is doing anything about it? It doesn't seem to me like they put much emphasis on understanding their user applications. I would think they would be talking file caching/acceleration .. they would be thinking about how their shared drive ideology affects the performance of their core applications, and they'd be thinking about how to use things like tensorflow to augment SfM algorithms to create their end users (cuistomers) a better product out of the box. Any insight would be much appreciated. I've found myself having to basically figure all of this out on my own - with a moderate level of success. I have various degrees of photos (most of the photos I'm getting at 20-30mp but I've recently started processing 75mp photos at 50mb each and it's starting to get painful doing AT on something that large. I'm considering downsampling, but the documentation doesn't say whether it'll downsample the textures on the final product as well. I've read the documentation up and down, I'm writing my own xml blocks and getting pretty good at custom AI tie points.. but damn, is there a secret I don't know about? Any Bentley ninja guidance would be more than greatly appreciated.
↧
Forum Post: Context Capture Center Vulkan and Network Processing
↧
Forum Post: RE: How to use thermal images in CC and DJI XT2 pose issue
Hm. New here, but that is about my question too and I'm afraid I don't see any answers to this thread :0
↧
↧
Forum Post: RE: Any plans to integrate with the new Ipad Pro?
Definitely an interesting product. The next iPhone is also rumored to use the lidar sensor. How do you see this incorporating into MicroStation. This is a much better as an extension to ContextCapture Mobile or maybe Navigator.
↧
Forum Post: RE: how dose python api "mergeBlocks((BlockVec)blocks)" work?
Got reply from Bentley.This is how to initialize Blockvec: # -------------------------------------------------------------------- # make block array # -------------------------------------------------------------------- blocks = ccmasterkernel.BlockVec() blocks.append(block1) blocks.append(block2)
↧
Forum Post: RE: Any plans to integrate with the new Ipad Pro?
[quote userid="332735" url="~/products/3d_imaging_and_point_cloud_software/f/contextcapture-descartes-pointools-forum/194561/any-plans-to-integrate-with-the-new-ipad-pro/577585"] How do you see this incorporating into MicroStation. [/quote] Oh, I think that most users would use their Ipad to capture an interior or an elevation of a building (See Greenscapes example above). There will be two types of info: photo info and depth/point infor provided by the LIDAR. It would be good to have some realtime indication of the coverage captured. The 6D_ai demo is impressive in the way it gives the user realtime feedback on whether and how much photo coverage he has. Ditto for the feedback on the mesh capture. The display of the meshes shown here could be enhanced to 'leave' the mesh overlaid in place when the user is moving about so that he can tell what bits of the chair for example hasn't been scanned. Not sure what the format ARKIt uses, but it would be good to be able to import the info directly into Mstn. Once there it should be possible to publish an i.model so that the info can be added to later? I imagine that the info be converted into Mstn's RealityMesh format, but it would be useful to have the component point cloud and photo splats(?) available, It would also be great to be able to access the video footage. [quote userid="332735" url="~/products/3d_imaging_and_point_cloud_software/f/contextcapture-descartes-pointools-forum/194561/any-plans-to-integrate-with-the-new-ipad-pro/577585"]This is a much better as an extension to ContextCapture Mobile or maybe Navigator. [/quote] Not really. These would be secondary functions / use cases. CC Mobile is currently viewing only. Before the info can be consumed by the typical CC Mobile user you will need processing (hopefully there will be a lot of tools available via ARKit Apple ecosystem) especially if stiching and geolocation is involved. It would be great if CC Mobile can be extended to also function as a data capture tool along the lines of Immersal which has extended a lot of the ideas underlying PhoyoSynth3d. I can see this kind of thing being very useful for building sites and asset owners. Picdata is already doing this with 2d photos. Combo? This would be a great platform / boost for Synchro Site and XR. Even Synchro Modeler would benefit if volume info can be extracted from the mesh info like Komatsu / Cesium's latest offering . Navigator? Sure, but federating data and doing clash detection is not particularly license / billing time intensive. How many subscriptions compared to Mstn and its verticals. Secondary market.
↧
↧
Forum Post: Can new coordinate systems be added in ContextCapture Editor "Update 5"?
I would need to add a new coordinate system in the ContextCapture editor library. It can be done? and if so, how? P.S. I have extrapolated from another program the coordinate system that I would like to use in .prj format.
↧
Forum Post: RE: Any plans to integrate with the new Ipad Pro?
Another startup S3D in the Scan-to-BIM space. Good illustration of how attractive it would be to be able to scan a space and immediately generate the context or background for your Mstn user to start working in context immediately. Lidar phones look like they are going to be mainstreaming soon. No need to buy one of the expensive Leica BLK360s or even BLK3D .
↧
File: Getting Started with the Context Capture SDK
↧
Wiki Page: Supported Data Centers
The Reality Modeling Cloud Services support 6 datacenters (see below) in strategic geographic locations.
↧
↧
Wiki Page: Select datacenter of new ProjectWise Project
Note: The following steps can only be performed by users who possess the “Administrator” or “Co-Administorator” user roles. 1) Go to https://connect.bentley.com/ 2) Click your enterprise's name in the Bentley CONNECT header to acces the enterprise portal: 3) Click the "User Management" tile: 4) Click the "Manage" button: 5) Use the "Search" field to filter the list of users: 6) Select the user and click "Modify Roles": 7) Toggle ON the "CONNECT Geo-Location" role in the list: *If the Role is not available, contact Bentley CONNECT Technical Support. 8) Click "Apply": Note: changes to roles may take up to 15 minutes to take effect. 9) The user will be allowed to set the new project's data center to one of the options supported by the Reality Modeling Cloud Service . Note: It is not possible to modify the Data Center Location of an existing ProjectWise Project.
↧
Forum Post: RE: Any plans to integrate with the new Ipad Pro?
Probably going to tie in with Apple's Indoor Survey app . Should make generating Digital Twins very accesible and mainstream. I wonder if the 3d meshed maps can also help with providing location precision indoors .
↧
Perimeter measurements in feet
Would like to show perimeter measurements in feet instead of yards?
↧
Forum Post: [Contextcapture U15] Python SDK matrixToYawPitchRoll
How to get yaw pitch roll values from pose using matrixToOmegaPhiKappa function? Documentation doesn't help much as YPR functions are for block export only and there is no direct function matrixToYawPitchRoll.
↧
↧
Wiki: ContextCapture | Descartes | Pointools Wiki | Orbit (Türkçe)
↧
Wiki Page: Reality Modeling Wiki
Please use the Table of Contents on the left to locate information about ContextCapture, Descartes, Pointools and Orbit 3DM.
↧
Wiki: ContextCapture | Descartes | Pointools | Orbit Wiki
↧
Wiki Page: Orbit Online License Activation
This page describes the concepts and requirements of Online License Activation applicable for Desktop and Server products. Online Activation Key and Concurrent Uses An Orbit product must be validated online via an Activation Key. An Activation Key is valid for a specified number of Concurrent Uses. Orbit manages all Activation Keys and Concurrent Uses online via the Orbit License Service. To use Orbit, outbound connectivity to the Orbit License Service https://activation.orbitgt.com/service on port 443 is required. In case you need to add a firewall rule, use the Server's Name and not its IP. The License Service is a scaled Cloud Service and has no fix IP. Trial and Demo Licenses Trial and Demo Licenses are available for Desktop products. Both are subject to license restrictions. My Orbit GT Customers are able to consult and manage their Activation Keys and Concurrent Uses via https://my.orbitgt.com . License Leases A successful activation validates product use for a well defined period, so-called “license lease lifetime”. Depending on the product type (see below), a certain license lease lifetime is applicable. By closing the application properly, the license lease is released immediately. This allows to restart the application right away on any hardware using the same license. Beware of the exception for the Orbit WebClient SDK (see below). Via this portal, specific individuals can manually end a license lease. This is required when the application is stopped in an irregular way. In such case, technically, the application was not able to successfully communicate the end-of-lease to the Orbit License Service. Any use is reported, abuse is seen as license violation. Get Access Access to the my.orbitgt.com portal is granted to customers only and managed via an existing Orbit GT Account. Only the Primary Contacts, being those individuals within the company that are responsible for the use of the Orbit products and licenses, have access. You don't know if you have an Orbit GT Account or forgot your password, try to login here https://orbitgt.com/login/ You don't have an Orbit GT Account, register here https://orbitgt.com/register/ You have an Orbit GT Account but you don't have permissions to access to this portal, request permission via sales@orbitgt.com. Product Activation Depending on the type of Orbit Product, a slightly different validation check is applied. Orbit Desktop Standalone Orbit Desktop Standalone products connect to Orbit Online License Service every time you start and stop the application. By closing the application properly, the license use is released immediately allowing you or any co-worker to start Orbit right away on a different machine using the same license. Online connectivity is required to start the application and to release your license use at closing the application. If you successfully started the application the license use is valid for 1 hour. When your license lease is about to expire the application reconnects automatically to the online license service to extend its license lease lifetime with another hour. After 3 failed attempts to automatically extend a valid license lease the application shutdown timer is displayed. Once you confirm acknowledge, the application will be blocked by the License Manager window in 5 minutes. This give you enough time to save your work or to re-validate your license manually. In case (1) the application is not closed properly or (2) no successful connection to the online license service could be established when stopping the application, your license will be released once its license lease lifetime expires. If the Orbit desktop can't be started due to an invalid license the License Manager Window displays the reason of activation failure and how to solve the issue. Orbit Desktop Clients Orbit Desktop Clients connect to their Orbit Server, see below, every time you start and stop the application. No online connectivity is required. By closing the application properly, the license use is released immediately allowing you or any co-worker to start Orbit right away on a different machine using the same license. If you successfully started the application the license use is valid for 30 minutes. When your license lease is about to expire the application reconnects automatically to its Orbit Server to extend its license lease lifetime with another 30 minutes. After 2 failed attempts to automatically extend a valid license lease the application shutdown timer is displayed. Once you confirm acknowledge, the application will be closed in 5 minutes. This give you enough time to save your work locally. In case (1) the client isn't closed properly or (2) no successful connection to the Orbit Server could be established when stopping the application, your license use will be released once its license lease lifetime expires. Orbit 3DM Publisher and 3DM Feature Extraction Server An Orbit Server operates the same as the above described Orbit Desktop Standalone, differences : The license lease lifetime is 24 hours. No acknowledge and shutdown timer can be used. If the license lease expires the Orbit Server license will become invalid. Orbit Servers connects to the Orbit Online License Service every time you start and stop the Orbit EOS Service correctly. A valid Orbit Server is able to activate a well-defined number of concurrent Orbit Desktop Clients. The Orbit Server is not connecting to the Online License Service to validate a connecting Orbit Desktop Client. Orbit 3D Mapping Viewer and Plugins The Html5-based 3D Mapping Viewer SDK and any Plugin provided by Orbit GT developed Plugin using the 3D Mapping Viewer SDK are free of charge and do not require online activation. From version 18.0 Plugins provided by Orbit GT using the Flash-based Mobile Mapping or Oblique Viewer SDK are free of charge. The Orbit 3DM Publisher version 18.0 or higher includes unlimited Activation Keys for all Plugins developed by Orbit GT that use the Flash-based Viewer SDK. Contact Sales at Orbit GT to obtain the configurations for Publisher 18.0 or higher to distribute a valid Activation Key for the Plugins you have developed.
↧
↧
Forum Post: RE: Holes on mesh when printed to PDF
Thanks Oto. I'm going to try to convert the 3MX to 3SM inside ContextCapture.
↧
Forum Post: Once I've installed Descartes Connect, which programs I can uninstall?
I just wanted to reinstall Descartes Connect to be part of the VR bundle (with CC) but turned I've downloaded too many old Microstation programs that take too much space on my computer. Which of the next (Descartes and Microstation) I may uninstall? Thnaks, Michael
↧
Wiki Page: My processing is failing due to insufficient memory despite having the tiling correctly adjusted
Product(s): ContextCapture Version(s): N\A Environment: N\A Area: Reconstruction Subarea: N\A Error message “Failed: insufficient memory. You may solve this issue by subdividing the reconstruction into smaller tiles. Please review your reconstruction settings.” In most of the cases, this error appears when the tiling is set incorrectly set, and the RAM estimation is too high for your computer In some critical cases , the processing might still be failing with an “insufficient memory” error message even if you have set the maximum RAM estimation under your computer’s capabilities. This is due to an incorrect estimation of the maximum RAM consumption for the tiles . This can happen when a lot of pictures are looking at the same area of the model. Such critical cases typically happen for assets such as telecommunication towers for which the data acquisition implies a lot of redundancy in the capture with all the photos focusing on the same area from all possible angles. Solution If you are facing a similar problem, we recommend lowering the tile size even more , until the process finally succeeds. If the problem persists, please submit a Service Request . Our team is working on improving the RAM estimation to avoid issues like this from occurring in the future. A fix should be available in ContextCapture Update 16.
↧