Black Box vs. Glass Box: A Diver's Guide to Decompression Algorithms (ADT MB vs. Gradient Factors)

As you progress in your diving journey, you start asking deeper questions. You move beyond “how long can I stay down?” to “why that long, specifically?” You begin to look at your dive computer not just as a gauge, but as a computational engine running a specific safety philosophy. And on advanced computers like the SCUBAPRO G3, you’re often presented with a fundamental choice: which philosophy do you want to follow?

This isn’t just a minor setting tweak. It’s a choice between two profoundly different approaches to managing decompression risk. On one side, you have adaptive algorithms, which we’ll call the “Black Box” approach. On the other, you have user-configurable models like Gradient Factors, the “Glass Box” approach. Understanding the difference is crucial for any diver looking to take their knowledge and safety to the next level.

The Common Ground: Professor Bühlmann’s ZHL-16

First, let’s be clear: these two approaches aren’t alien to each other. They both stand on the shoulders of a giant: Professor Albert A. Bühlmann and his ZHL-16 decompression model. This model represents the body as 16 theoretical “tissue compartments,” from fast-saturating blood to slow-saturating cartilage. The computer constantly calculates the theoretical nitrogen pressure (or “tension”) in each. The model also defines a maximum tolerated pressure for each tissue at any given depth, known as the “M-value.” Exceed this M-value, and the risk of bubble formation (DCS) becomes unacceptable.

Both ADT MB and GF algorithms use this fundamental 16-compartment engine. The difference lies in how they decide how close you’re allowed to get to those M-value limits.

The “Black Box” Approach: The ZHL-16 ADT MB

Think of the Adaptive (ADT) Microbubble (MB) algorithm as a sophisticated expert system. It’s a “black box” not because it’s unreliable, but because its decision-making process is proprietary and largely automatic. The manufacturer has encoded their extensive research and best practices into the logic.

• Adaptive (ADT): This algorithm doesn’t just look at your depth and time. It can factor in other variables that suggest increased decompression stress. For instance, if you’re working hard (inferred from increased air consumption, or directly from a heart rate monitor), the algorithm might become more conservative, shortening your NDL. Cold water, detected by the temperature sensor, can also trigger a more cautious profile.
• Microbubble (MB): This aspect of the model goes a step further. It acknowledges that tiny, asymptomatic “silent” bubbles can form on any dive. The MB logic aims to manage these microbubbles, preventing them from growing into problematic sizes. This often results in more conservative profiles for repetitive dives or dives with rapid ascents, where bubble “seeds” are more likely to be present.

The Philosophy: You are trusting the manufacturer’s expertise. You’re using a system designed to automatically give you a safer profile when it detects conditions that fall outside the ideal. It’s convenient, smart, and for the vast majority of recreational dives, it’s an excellent, safety-oriented approach.

[Image of a diagram showing depth/time input leading to an ‘Adaptive Algorithm Black Box’ which outputs a safe NDL]

The “Glass Box” Approach: The ZHL-16 GF (Gradient Factors)

If the ADT MB is an expert system, Gradient Factors (GF) turn you into the system administrator. This is a “glass box” because the mechanism is transparent, and you have direct control over the dials. Instead of relying on a hidden adaptive logic, you explicitly tell the computer your personal risk tolerance.

This is done by setting two numbers, GF Low and GF High (e.g., 30/70).

• GF Low (e.g., 30%): This controls the “deep end” of your decompression. It sets the maximum gas tension you’ll allow upon leaving the deepest part of your dive. A lower number forces you to make your first decompression stops deeper, starting to off-gas much earlier in your ascent. It’s a very conservative way to manage bubble formation from the start.
• GF High (e.g., 70%): This controls the “shallow end.” It sets the maximum gas tension you’ll allow when you finally surface. A lower number here means you want to be more “clean” of nitrogen at the end of your dive, requiring longer, shallower stops.

The Philosophy: You are taking direct control of your safety margins. You are telling the computer, “I don’t want to just stay below the M-value; I want to stay below 30% of its available margin in the deep, and 70% of its margin at the surface.” This allows an experienced diver to tailor their profile based on factors the computer can’t know: personal fatigue, hydration, or the specific demands of a complex dive plan.

[Image of a graph showing the M-value line, with the GF Low/High line drawn underneath it, creating a more conservative ceiling]

Head-to-Head: When to Trust the Box vs. When to Build Your Own

The Weight of Control: Capability Comes with Responsibility

The power of Gradient Factors comes with a significant caveat: the responsibility for a safe profile shifts from the manufacturer squarely onto your shoulders. Choosing the ZHL-16 GF algorithm is a declaration that you understand decompression theory well enough to make your own safety decisions. Setting GFs that are too aggressive (e.g., 90/95) can create a profile far less safe than the computer’s default settings. Conversely, setting them too conservatively without reason can lead to unnecessarily long decompression times.

Conclusion: Which Diver Are You?

The choice between an adaptive algorithm and Gradient Factors is a reflection of your diving experience, knowledge, and philosophy. There is no single “best” answer.

The “Black Box” ADT MB is a brilliant tool for the diver who wants a state-of-the-art, automated safety net that intelligently adapts to observable stress. The “Glass Box” GF is the essential tool for the technical diver or the experienced recreational diver who has the knowledge and discipline to manage their own safety margins with precision.

By offering both, a computer like the G3 doesn’t just give you a device; it gives you a choice. Your task is to gain the knowledge to make that choice wisely.