Hi Daniele,
you mentioned that after installation there are files in C:\Program Files\Capturing Reality\RealityCapture>
Can you check if they are all needed files to run RealityCapture (compare it with the ordinary installed application)?
Also, what is the reason to run it is such way?
Also, our developers were informed about this and we will discuss, what could be wrong there.
The reason behind is an attempt to scale our workflow better.
Currently we use expensive machines with GPUs and have a python script which runs every step in RC, from align to export, managing the input parameters, retrying if necessary, etc.
However, since it is running on a single machine that means we are using expensive GPU but not utilizing them. According to our research we need GPU only during the reconstruct-unwrap-texture.
Also, as soon as project is finished it is deleted from the machine so we can’t retrigger any particular step later on.
Instead, we were thinking of running it in k8s, with 2 pools: GPU and CPU-only machines, with different specs. Then, using any workflow tool (f.e. Argo) we can schedule, orchestrate and manage our workflows having the best resource utilization, so at the end we expect it to become faster and cheaper.
Hi Daniele, we had just call with our developers, and as the RC is not starting at your side, there could be some issues, which we don’t know about.
There are some conditions, which need to be satisfied to run RC on docker. You need to have a proper container for Windows, also the computer, where it is running needs to have the same Windows version, you need to have a GPU with CUDA on that machine, also some kinds of modifications, etc.
There are some links they sent me:
We will try to investigate this issue more.
Hello!
Is it possible to run RC on Windows Server 2022?
I know that there are no Windows 11/10 Docker images, there are only images for Windows Server.
Hi, it should be possible:
https://dev.epicgames.com/community/learning/knowledge-base/Wj7B/capturing-reality-realitycapture-os-and-hardware-requirements#:~:text=Add%20to%20Bookmarks-,RealityCapture%20requires%3A,-64bit%20machine%20with
Hey!
After some time I managed to run RC in container.
RealityCapture.exe -headless -addFolder C:\CalculateTextured3DModel\Images -align -save C:\CalculateTextured3DModel\Images\save"
Works fine and creates new save folder and creates fine in it.
However, when I add -calculateTexture the folder save never appears. Do you know why the commands like -align and -setReconstructionRegionAuto seem to work, whereas -calculateTexture fails?
Do you have created model before the texturing?
Hey!
Do you mean calculateNormalModel step?
Hello - I’m trying to do the same thing, but I cannot get it working. I would only need the alignment step to work.
In my case, I’m installing Reality Capture 1.5.1 on a Windows container (base image mcr.microsoft.com/windows:ltsc2019 ) . I’m trying to run RealityCapture in powershell. When I launch any RealityCapture command, I get a crash dump written to C:\Users\ContainerAdministrator\AppData\Local\Temp\.
Could you share what steps you did here? Which base image did you use? Did you need to do some steps in relation to CUDA or enabling vendor specific graphic APis as discussed in the blog posts linked above?
Hi, yes I meant that.
In general the new folder should be created also after the calculate texture command.
How does look the modified command line?
Hey!
I don’t know if I can help you with your case, but I could definitely share the steps I did in order to run RC.
FROM mcr.microsoft.com/windows/server:ltsc2022
COPY "RealityCapture-1.4.1.117425.msi" "/"
RUN /RealityCapture-1.4.1.117425.msi /quiet
So I use ltsc2022, copy the installer, install RC and then simply run cli commands.
Did you need to do some steps in relation to CUDA or enabling vendor specific graphic APis as discussed in the blog posts linked above?
I did nothing related to this.