Non-scientific Comparison of Dive Computer Conservatism

[padudle]

Hello All,

Recently I purchased a new dive computer. On the first few dives I compared it with some of my old computors for sensor accuracy and NDL.

New Computer: Oceanic Veo 2.0 set to the DSAT algorith, air and no conservatism factor.

Previous Computer: Cressi Archimede set to air and no conservatism factor. This is my trusted reference computer. I have several hundred dives on this computer with no decompression problems.

Third Computer: Suunto Mosquito set to 24% nitrox and no conservatism factor. I knew from previous experience that this Suunto is super conservative. I set it to 24% nitrox to try to make it at least usable with the other computers.

On the first day, I carried the new Oceanic and the Cressi. At the end of the second dive the Cressi showed less than 3 minutes NDL and the Oceanic something like 15 minutes. I obeyed the Cressi without entering deco.

On the second day, I carried the Oceanic, the Cressi and added the Suunto. At the end of the second dive the Suunto showed 11 minutes of deco obligation. The Cressi showed 6 minutes of deco obligation and the Oceanic was still happily showing about 15 minutes of NDL. I did the extra hang to clear the Cressi and disregarded the Suunto as usual.

For the rest of the long vacation I dived nearly every day and carried only the Oceanic and obeyed the limits of that computer. I suffered no ill health effects.

Conclusions: There is a wide variation in NDL provided by different dive computer models. I really enjoyed the extra limits provided by a less conservative computer.

Final notes: When switching dive computers, I think it is good practice to compare the new computer with your old computer. In this way you can get a feel for the differences in conservatism. If the new computer is less conservative you can consider dialing in a conservatism factor. In my case, I decided to enjoy the extra bottom times offered by my new computer. Diving conditions were very mild and I have no health problems. Please don't call the PADI police.

Pete

 

[redacted]

Third Computer: Suunto Mosquito set to 24% nitrox and no conservatism factor. I knew from previous experience that this Suunto is super conservative. I set it to 24% nitrox to try to make it at least usable with the other computers.
Pete

Toilet bowl, Goodwill or Ebay - it really needs to go.

 

[Water Boy76]

Yes I do this too. My Oceanic <DSAT> is the most liberal when compared to my backup Hollis DG02 <Buhlman>. I usually have about 12 Min left on the Oceanic when the Hollis goes into Deco. I consider it my safety notification. If the dive is worth an extra ten minutes I'll use them. If it's not, I'll ascend and save the NDL for the next drop.

 

[tbone1004]

best to have a computer that has a predictable algorithm and factors that you can adjust and know what you're adjusting instead of arbitrary numbers.... VPM, GF, etc. Computers that have algorithms that you can't model scare the crap out of me.....

 

[padudle]

best to have a computer that has a predictable algorithm and factors that you can adjust and know what you're adjusting instead of arbitrary numbers.... VPM, GF, etc. Computers that have algorithms that you can't model scare the crap out of me.....

Tbone, Can you elaborate a little. What do you mean by model? Do you have computer codes for specific mathematical models?

It would be nice if all dive computer manufacturers were required to publish their deco algorithms in open-source format. Maybe I should do a Kickstarter project for an open-source dive computer.

 

[tbone1004]

the ability to do dive plans and know what the computer is going to do prior to the dive. The computers you've listed above all have "fudge factors" that can penalize you for things that you don't know. Quick ascents, sawtooth profiles, short SIT, etc and without knowing the algorithm you can't accurately predict what your deco obligations are going to be, aside from the fact that while they will do deco, none of them were really designed as a "decompression computer". If you get something like the Petrel, Seabears, Liquivisions, etc. that are running unadulterated decompression algorithms, you can actually plan your dives using something like vplanner and know what is actually going to happen. Conservatism can then be experimented with on land knowing what your tissue loadings are going to be like.

 

[Steve_C]

I have noticed that my Sunnto Zoop is much slower to give credit for ascending a bit than the Aeris (Oceanic). On dive 1 the NDLs are not too far apart but as soon as I go up a few feet the Aeris starts giving me more NDL. I have to go up a lot further with the Sunnto and then the credit is less. Almost like the adjustments for the Suunto are being done on a much coarser depth grid. It is not just the algorithm at a given depth but how frequent the updates and how they are done as the depth changes that can make a big difference.

 

[Water Boy76]

I live in a diving world where your diving profile can't be properly planned. You start with a plan to go to a certain depth, then the weather turns nasty so you stay shallow, or the weather is better so you go deeper, or somebody got thier nitrox fill too high, or the planned spot showing no fish so you dive elsewhere, or there's boats on your spot so you don't dive the planned spot, or the viz is terrible so you go deeper, etc....

Planning is useful and necessary for technical divers. But when your plans always change like mine not very useful.

Not sure why someone would be afraid to dive computers with proven Algorithm's? There's not epidemic of bent divers that I'm aware of.

 

[dmaziuk]

It would be nice if all dive computer manufacturers were required to publish their deco algorithms in open-source format.

[br]
You could google and find the code for at least one. Plus the planner in (open source) subversion.

[br]
The issue is that bubble formation models in particular are rocket science and a) are hard to understand and b) require a whole lot of compute power to run. Because of the latter what dive computers actually run is some sort of a "close enough" approximation. Because of the former nobody can tell how close it is to the real thing or create a dive plan that will match the computer's readings. It's not than important for recreational divers who don't go into deco and are usually limited by their air rather than ndl anyway, esp. if you play nice, end your dives in the shallows and take hour+ surface intervals. Dives that require detailed planning are a different story.

 

[RonR]

It’s true that most computers touting bubble models run an approximation of those models created by adding bubble factors on top of a Haldane/ Buhlmann calculation. This is computationally much simpler, and not a bad thing at all. However several computers exist that do run full bubble calculations in real time- the Atomic Cobalt kicks in “full” RGBM for dives deeper than 150’, VPM firmware is available for Liquivision and Shearwater computers, and full RGBM at least for Liquivision. There are probably several others I’m just not familiar with.

But even if you have the code, the algorithm is only part of the story. The integration of an algorithm into a real time system, one that handles all of the unexpected things a diver can do (rapid ascents- how much and for how long?, partially completed stops, re-descents after entering deco, literally hundreds of possibilities) has a great deal to do with how a computer functions. Most of these things are not defined by any algorithm, but are up to the computer developer. Desktop simulators don’t run as a real time system, and can present an idealized version of an algorithm. Actual computers must fit the algorithm to real time events. How that is done is about 98% of the work of algorithm implementation, and not surprisingly varies from one manufacturer to another. That, and different decisions about conservatism factors, are why even computers running the "same" algorithm will give different profiles. -Ron