2. Dabarti Capture Results.

Before I dive into the process, here are 2 renders. First one with captured fabric textures. Input photographs were taken with mobile phone. Second is Rusty old metal Captured using Sony A7RII.

3. Process.

For basic capturing process you will need:

• Digital Camera.
• Light source.
• Spherical reflective object for calibration. 
• Optional but highly recommended: Remote shutter or intervalometer.
• Dabarti Capture Software (Demo or Pro)

Read More

In this post I want to cover the proper way of bug reporting as I believe it’s very important for both the users and developers.

I base this on my experience with various companies over the last years, and approach I describe here helped me greatly with quick resolution of most of the issues. Of course there are companies that don’t care about consumers and best bug reports won’t help.

Good bug report should be reproducible and clearly described. What I do is:

1. Describe the issue. Best if it’s short but clear. “Something is broken” is NO-NO.
2. Provide screenshots or record video of actual issue. Because when reporting while in production you don’t want to waste your valuable time, I recommend using ShareX – “Screen capture, file sharing and productivity tool”. It can grab screenshots, let you paint over them and upload it automatically to one of many destinations. When it’s done it will give you a link ready to be shared with anyone. Example Video 
3. Provide the scene files on top of this it’s worth providing reproduction steps. It will speed up the bug fixing process. In 3ds Max you can archive whole scene with textures into single zip file and than quickly upload it with ShareX to your secure destination.
4. Include Software versions and hardware specs. 

By doing this you make your life easier in long-term and help other users in the process. Be a hero, report bugs properly.

I’ve practiced that over last years with Chaos Group and V-Ray RT GPU. It helped greatly as I’m to spot some issues in life production situations. Even the best developers won’t envision all the creative ways their software will be used and your feedback is very important.

Also if you’re not sure if what you’re experiencing is a bug/issue or just lack of experience it’s still worth asking support. Just remember to google / bing / forums search for answers first.

Photometric stereo.

Edit: Since releasing this article, I’ve spent some time to create easy to use processing tool for non-technical artists that would like to try this technique – https://dabarti.com/capture .

A while ago I did small research project with the aim of developing best way to acquire accurate normal bump maps from photos. In this post I’m going to describe how I did it.

Initial idea was to use Agisoft and photogrammetry, this seemed as the most straight forward approach, but while it produced some promising results I found that detail of the textures was not nearly good enough. After going though some papers on that subject I struck gold by stumbling upon technique called Photometric Stereo.

Great explanation of what it is and how it works is found on Wikipedia.  What I love about this idea the most is that I could potentially take photos from single point of view and generate normal map with the quality limited only by the lens and camera sensor. Think about this like “reverse rendering”. If you know the color of the given part of the object and direction of the light, you can basically find out the surface normal.

Read More

Check it out here or on our Youtube channel.

Read More

Something very good is about to be added to V-Ray. It’s called Adaptive Lights and it speeds up rendering huge amounts of lights. Still under R&D but it’s already very impressive. And It makes me very happy as for quite some time I’ve been unimpressed by the speed of rendering lots of V-Ray Lights. As usual Chaos Group team listened and created something simply amazing.

I’m excited. Check out why:

For this image render times went down from 43:37 to 10:29 :

For this one render times went down from: 53:59 to 7:31

I trust numbers and like to test a lot on variation of scenes. So I took shots we did in the past and calculated how much faster it would be on average compared to Probabilistic Lights or Full approach:

https://docs.google.com/spreadsheets/d/1TeluZJFuUcycV6FG7RPwMW_0hSc8w-q_cU6SHXOsWeI/edit?usp=sharing

My personal thoughts based on those tests:

1. Probabilistic lights are not universal solution for scenes containing huge amounts of lights.
2. It’s smart, it adapts to the conditions of scenes and users don’t have to worry about playing with settings as much as they had with Probabilistic lights.
3. New approach is more universal and compared to Probabilistic Lights it will be a lot faster.
4. More lights = bigger the difference. It will render very fast regardless of number of lights.
5. It’s good to complain that something is slow ;)…

If you have access to nightly builds, you can test it out already (use maxscript code for that “renderers.current.options_probabilisticLights=2” ). This is still under research and development but it’s supposed to be included in next service pack.

If you would like to learn more about adaptive lights I recommend you to visit this ChaosGroup Lab blog post .