To mark our 30th anniversary, I invited former dGB Earth Sciences CTO and long-time OpendTect user Friso Brouwer to share a few words about his experience with the software. Here's what he had to say:
As a former employee and current user of their software, my friends at dGB Earth Sciences asked me to write something to celebrate their 30th birthday. They requested a short reflection on OpendTect—what I like about it.
The challenge? Keep it to just half a page. That’s a bit of a problem, of course, because there’s so much to like about OpendTect. But here, I’ll focus on one thing that has always made my work more enjoyable.
OpendTect allows the interpreter to work freely—not like a spoiled child, but in the sense that they’re not boxed in or constrained by the software. Achieving that in a software platform is less straightforward than it sounds. There’s always a trade-off between simplicity and flexibility. The simplest solution is a black box with a “go” button; the most flexible is letting users write everything from scratch—powerful, but time-consuming and complex. Interestingly, OpendTect allows for both. But where it truly shines is in striking the ideal balance between the two.
A great example of this is the attribute engine. It provides a set of basic building blocks and allows users to chain attributes together and optimize parameters. It’s an ideal setup for tackling complex problems and unleashing the creativity of the seismic interpreter.
Above is an image of an unsupervised classification on the Maui dataset (https://lnkd.in/eF85qi-r). While the key technology here is the UVQ classification neural network (a plugin), the best results only came after redesigning some of the input attributes in the attribute engine in a clever way. Having the ability to tweak the workflow made it possible to find a better, problem-specific solution. And being able to experiment and innovate like that is what makes geoscience work so enjoyable.
Happy birthday, dGB—and thanks for sharing the toys!