No Two Dive Computers Are Alike

Dive computers have replaced decompression tables in most types of diving since they came onto the market some 30 years ago. According to Divers Alert Network (DAN), during that same time, the overall incidence of decompression sickness (DCS), at least in sport divers, hasn't changed. While data dispels the old worries that abandoning tables for computers would result in increased DCS, they indicate that computers are no panacea. And questions still linger about how safe dive computers are, especially when comparing brands -- or even models within the same line -- to each other.

There are dozens of dive computer models on the market, and they differ in design, quality of manufacturing, and, in particular, which decompression algorithm they use. But the manufacturers generally don't disclose publically information about their algorithms, their operational use, or their DCS risk. According to Petar DeNoble, senior research director at DAN, there are at least two reasons for this: Dive computers aren't regulated by any official organization, and validating decompression safety is complicated and expensive. So while manufacturers do test dive computers, they don't have to generate all the data necessary to support claims that their computers control or reduce the risk of DCS.

Generally, most sport divers tend to prefer computers with more conservative algorithms because they're not into personal risk-taking. However, without the detailed knowledge of how each dive computer manages decompression, it's hard to know which computers are more conservative than others. And as a new study shows, some dive computer algorithms even decrease or increase their conservatism during dives, depending on the diver's depth and duration.

To most sport divers, computer algorithms are mysterious, complicated, and impossible to comprehend fully. So we rely on the good will of the computer manufacturers and presume their computers will give us a safe dive. Some divers want to know more, however, so at the risk of boring our readers with too many details about decompression profiles, let us summarize a study in the journal Diving and Hyperbaric Medicine by dive algorithm expert Martin Sayer, head of the U.K. National Facility for Scientific Diving.

Sayer and his team tested 43 individual sport-diving computers using square profiles, which are more suitable for scientific testing, and are often followed by technical and military divers, who make decompression stops. Although sport divers tend to make multilevel, no-decompression dives, square profile results are important in order to understand more about one's own computer or when scouting around for a new dive computer. Tested three times at their default settings, they were compressed in a recompression chamber to five simulated depths -- 50 feet, 65 feet, 100 feet, 130 feet and 165 feet. At each depth, they recorded when each computer registered "no decompression," and then when various decompression times were indicated or exceeded.

In studying the differences between 0-, 5-, 10-, 20- and 30-minute decompression intervals, they discovered considerable variation in the times recorded for all the depth/decompression combinations, and the largest differences in permitted times didn't always belong to the same computer unit. For example, the average maximum no-decompression stop time for 50 feet was 69.7 minutes for all computers, but it ranged from 60.7 minutes (for the Apeks Quantum) to 83.7 minutes (for the Oceanic Atom 2). That's a 23-minute difference in the time one can spend at 50 feet! The average maximum no-decompression stop time for 100 feet was 18.4 minutes, but the times ranged from 15.7 minutes (Uemis SDA), to 22 minutes (Oceanic Datamask Hud).

The computers tested generally tended to be more conservative than standard USN decompression tables at depths shallower than 100 feet (and particularly at 65 feet), but less conservative than the tables between 100 and 165 feet. The differences were not always consistent between a manufacturer's computer models. In some comparisons, there were large differences at shallower depths, but then those differences weren't evident at deeper depths.

Some specific examples: For no-decompression dives, the Oceanic Veo 250 gave no-decompression times that were less than the average at 50 feet and 65 feet, but above the average at 100 and 130 feet. The Mares Nemo Sport was among the more conservative computers at 50 feet, but was the least conservative at 130 feet. Similar anomalies were found in decompression dives. The Uemis SDA gave longer than mean times at 50 and 65 feet, but shorter than mean times for the deeper depths.

There were some general trends within the major brands -- Mares computers tended to be the most conservative overall, followed by Uwatec and Suunto. In both no-deco and decompression tests, most Oceanic computers gave the longest times. Overall, results from models made by the same manufacturer were reasonably close,; unsurprising, since they use the same form of algorithm for their family of computers. But each seems to use a different model, which they often modifying without publishing the criteria they use for those modifications.

For example, Oceanic uses a modified version of the standard Haldane model (which assumes that all gas is dissolved into the tissues, and remains dissolved as long as maximum tissue tensions are not exceeded). Suunto uses its reduced gradient bubble model (RGBM), designed to protect a recreational diver from the effects of micro-bubble buildup. Uwatec uses versions of the ZH-L8 ADT, their own eight-tissue algorithm (it divides your body into eight "compartments" and mathematically follows the uptake and release of nitrogen in each area.). Mares uses its Mares-Wienke RGBM, a Haldanian model with some extra safety factors. The differences between models weren't always consistent across the depth range, so significant theoretical similarities must exist.

Sawyer ends the study by stating, "In a computer-driven era, it remains disappointing that dive management decisions . . . continue to be based largely on subjective assessment. This will remain an issue until there is an accepted 'gold standard' for decompression modeling. [Otherwise,] it will remain difficult for there to be any consistent approach to the manufacture of decompression computers."

Undercurrent columnist John Bantin, a veteran dive equipment tester, looked over Sawyer's study and notes that some results are out of date due to manufacturers' changes to algorithms since then. "What he says, especially about Oceanic computers, was correct once, yet he fails to mention that for at least three years these computers have been supplied with dual algorithms. This was caused in no small part by a campaign by myself and other dive writers that the DSAT algorithm Oceanic originally employed was not suitable for decompression-stop diving (or no-stop diving deeper than 100 feet, for that matter). Oceanic therefore introduced the Pelagic+ algorithm as an alternative algorithm available with virtually all its products.

"Alas, the default setting for Oceanic computers is still DSAT algorithm, and dive store staff fail to point out that the Pelagic+ algorithm is much more suitable for serious diving (and akin to the algorithms used by the other manufacturers). Buyers are more interested in peripheral functions they can understand than the invisible algorithm that might keep them safe from injury. So if you use a modern Oceanic computer, I advise you to set it for Pelagic+. Incidentally, today, Suunto, Mares, Cressi computers and many other brands now use the algorithm developed by Bruce Wienke, with very similar results."

Sawyer admits some of the computers he tested were out of date, due to limited research funds that couldn't buy the latest and greatest models. "However, the samples we have tested and are continuing to test are representative of what may still be in use today," he told Undercurrent. "Divers just don't throw away a computer when a new model comes out; if they do, some sell them."

And while there are also new versions of the computers being released, it's with minimal notification to the consumer, says Sawyer. For example, compare the no-decompression stops listed in the October 2008 manual and the August 2011 versions of the Suunto Vyper Air manual -- Sawyer's team saw significant differences, particularly in the personal factor and altitude settings. "The implication is that the two versions must be alluding to different versions of the decompression algorithms, or at least how these are modified with personal or altitude settings," he says. "This suggests that Suunto released different versions of the Vyper Air computer, but with no differences in the model name, so divers would get a different performance than expected if they were changing their settings based on the 2008 manual but had a 2011 version of the computer. Suunto could have allowed for this by using version numbers of the computers, but didn't. So although we cite computer names/models [in our study], they can still be different versions of the same unit being tested.

"Overall, the points we are making are no different from those made by John: that computer manufacturers are not good at informing the consumer of what the basic/default operations of their computers are, or when and where changes to the algorithms are being made. That means many divers may not be using their computers to their full potential."

While this study was aimed at the use of dive computers by occupational divers, Sawyer is now attempting to examine the effects on recreational divers by doing a series of multi-level dives over a number of days but using repeated standardized "recreational" dive profiles. He expects to report those results later this year.

(On a separate note, there were 28 battery changes and 19 computer failures during the trials. Some data-download failures were due to low battery power, others occurred when the downloaded data didn't equate to the dive profile. It's unclear whether the download errors were representative of real-time problems that could affect a diver's ability to get valid information, and thus, abort a dive. But if you assume it could, then, the researchers say this equates to a battery change or failure every 37 hours or 54 hours of diving, respectively.)

"Decompression management by 43 models of dive computer: single square-wave exposures to between 15 and 50 msw," by M.D.J. Sayer, E. Azzopardi and A. Sieber; Diving and Hyperbaric Medicine, vol. 44, pgs. 193-201.

- - Vanessa Richardson