Does Photogrammetry Work For Archviz? Avoid These Mistakes For Good Results

Over the last few weeks I set out on a journey to find a photogrammetry workflow that is reasonable for my archviz needs. Photogrammetry is an enticing technology, and the idea of it is very appealing. The idea that you can get extremely realistic 3d models simply from pictures seems too good to be true. Can it replace our typical poly modeling workflows? Is it reasonable to do, even considering time and cost? What is the proper workflow?

These are all questions I set out to answer, and in the video you can follow along with me and find the answers. In the video, you will see the workflow that I landed on. It is free, and it is something anyone can experiment with. However, I also encountered a lot of pitfalls. You can learn from my mistakes and see what not to do!

These are the parts of the process:

1. Taking really good photos - This is the hardest part, and I discuss a lot in the video about what to do and what not to do

2. Use Meshroom (a free, open source, photogrammetry software) to analyze the photos and generate the 3d model - I will show you how to do this. It's easy!

3. Use 3d software to clean up and render your models - I used 3ds Max for this, specifically the retopology tools, and Vray to render. There are many 3d software that will do just fine. I see a lot of people using Blender for this part.

In the end, I will give my analysis of photogrammetry for archviz overall, and share my thoughts about how useful I think it is. And then I want to hear from you about your experiences with photogrammetry. Has it worked well for you? What is your workflow? Let me know in the comments.