Dive computers and DCS

[Ojai Diver]

Can somebody educate me on the premises at work here?

Given the unforgiving nature of DCS, at what point does flouting a computer's conservatism (or seeking a less conservative computer) become a recipe for getting bent?

I'm not trying to be critical of anybody here, just trying to understand.

Isn't some safety being lost when one seeks more bottom time, greater depth, or shorter SITs through a less conservative computer?

Or is it that the more conservative computers are somehow flawed in their algorithms?

 

[lowviz]

@Ojai Diver: That is a good question and should be a new thread (IMHO).

You may wish to report your own post and ask to have it turned into a new thread. (mods, feel free to remove this post)

 

[rhwestfall]

@Ojai Diver: That is a good question and should be a new thread (IMHO).

You may wish to report your own post and ask to have it turned into a new thread. (mods, feel free to remove this post)

Agree! This may become one of the most interesting discussions in some time.

 

[Freewillow]

Can somebody educate me on the premises at work here?

Given the unforgiving nature of DCS, at what point does flouting a computer's conservatism (or seeking a less conservative computer) become a recipe for getting bent?

I'm not trying to be critical of anybody here, just trying to understand.

Isn't some safety being lost when one seeks more bottom time, greater depth, or shorter SITs through a less conservative computer?

Or is it that the more conservative computers are somehow flawed in their algorithms?

Whaw, this will provide us with 20+ additional pages.

In a way, you are right, the more liberal the algo, the more risk one has to develop DCS. But this is pure logic. The matter of "getting bent" is so personal, there are so many factors involved that it is difficult to determine which algo and modifiations of it will suit you.

While the Cressi Leonardo is a very good computor for beginner or occasional divers, the others - Buhlmann based - behaves differently for more intensive dives.

The problem can be of two orders.

If one multiply the amount of consecutives dives and shorten the surface interval, the Cressi Leonardo goes faster into deco mode.

I am used to "ride" the NDL and even enter into Deco up to a TTS ( time to surface ) of 15/20 minutes. This means 10/15 minutes of MANDATORY stop. Buhlmann algo computors tend to start the deco countdown at depth shallower than 30 feet while the Leonardo starts much higher in the water column - arount 10 feet. This has for consequence that in multilevel dives, starting with a deco obligation, you can "look around" the coral at 20/15 feet while you cannot do that with a computor like the Cressi Leonardo. If you have one, you need to go immediatelly to your deco stop @ 18 or 10 feet.

One is not really -(+/-) - more liberal than the other. Simply the algo is not the same. So the "advise" is different.

 

[scubadada]

Can somebody educate me on the premises at work here?

Given the unforgiving nature of DCS, at what point does flouting a computer's conservatism (or seeking a less conservative computer) become a recipe for getting bent?

I'm not trying to be critical of anybody here, just trying to understand.

Isn't some safety being lost when one seeks more bottom time, greater depth, or shorter SITs through a less conservative computer?

Or is it that the more conservative computers are somehow flawed in their algorithms?

All commercially available decompression algorithms are "safe" for recreational diving. All are more conservative than the present US Navy tables. There is no data to support that any of the available algorithms are more or less safe when used appropriately. More liberal algorithms are just that, more liberal. More conservative algorithm are also just that, more conservative. Best way to avoid DCS is to not dive. Choice of a decompression algorithm is a personal one and may be affected by individual risk factors. A liberal algorithm can always be dived more conservatively. There's nothing you can do with a conservative algorithm. Decompression diving is another topic and currently, quite contentious (see Deep Stops)

 

[stuartv]

Can somebody educate me on the premises at work here?

Given the unforgiving nature of DCS, at what point does flouting a computer's conservatism (or seeking a less conservative computer) become a recipe for getting bent?

I'm not trying to be critical of anybody here, just trying to understand.

Isn't some safety being lost when one seeks more bottom time, greater depth, or shorter SITs through a less conservative computer?

Or is it that the more conservative computers are somehow flawed in their algorithms?

At some point beyond where any of the mainstream, current, recreational computers are today.

In the many threads on this subject, here on SB, that I have read, I have yet to see anyone ever post any kind of evidence that even the most liberal recreational computers available today are so liberal that they have any SIGNIFICANT risk of getting a person bent.

The most liberal recreational computers I know of are the ones that offer the DSAT algorithm. This is many Aeris and Oceanic models plus, I believe, the Sherwood Amphos. I have yet to see anyone post any evidence, or even claim, that any of these computers are so liberal that they post any more risk for getting the user bent than any of the less conservative computers.

Do I think they actually ARE more likely to get you bent? Yes, I think they probably are. But, I think the increase in risk is so small and the safety cushion they build into all of them is so big that I would simply class the liberal computers as "Safe" and the more conservative computers "Safer". Some, I might even call "Unreasonably safer". As in, just no need at all to be THAT conservative.

From what I can tell, a lot of the most vocal people in decrying these liberal computers as "too liberal" are people who have never been bent. So, they have no basis for saying computer A is TOO liberal or unsafe versus saying the computer B is simply way more conservative than it needs to be. But, they say it anyway because they've been using computer B for years and never gotten bent, so they somehow conclude that that means it's the right amount safe and anything more liberal is a hazard. All the while they could have been driving 45 in a 55 MPH zone and just never realized it.

The conservative computers are not flawed. They just give the diver a bigger safety margin than the more liberal computers.

 

[stuartv]

the more liberal the algo, the more risk one has to develop DCS.

I do not believe this is categorically true.

There seems to be mounting evidence that bubble model algorithms, like RGBM or VPM-B, allow the diver more slow tissue loading for the same total ascent time as the model expressed in the Buhlmann implementation, and that this provides a higher risk of DCS.

So, it is POSSIBLE that a more liberal algorithm may still yield less risk for DCS, even though allowing a longer NDL, by virtue of the way it accounts for time spent at different depths during a dive.

This is a complex subject and I don't think anyone can really explain what I'm trying to say in just a few sentences. The point is that, unless you are comparing 2 computers that use the same algorithm and one simply uses parameters (i.e. details of implementation) that result in longer NDLs, I don't think you can make the general statement that a computer that gives shorter NDLs (which is what I take to be implied when you say "the more liberal algorithm") gives less risk of DCS. That will be true some of the time, but I don't think it's always true.

If you're comparing two computers that both implement the Buhlmann ZHL-16C algorithm and one uses Gradient Factors of 30/70 and the other uses GF 60/95, then, yes, I think you could categorically say that the former is going to yield less risk of DCS.

But, if one computer implements Buhlmann and the other implements RGBM, then even though one may usually give shorter NDLs than the other, it still could be that sometimes it will get the diver bent where the other wouldn't have.

 

[scubadada]

Cochran is actually the most liberal of the available algorithms, though, you can argue that Buhlmann with GF can sometimes be more liberal. Single dives do not tell the whole story. Multiple dive data such as generated by ScubaLab can be very enlightening
http://ads.bonniercorp.com/scuba/PDF/ScubaLab-Computer-Test-September-2014-data.pdf
Their 2016 data is also available off their website as an Excel file.

 

[dmaziuk]

Can somebody educate me on the premises at work here?

Given the unforgiving nature of DCS, at what point does flouting a computer's conservatism (or seeking a less conservative computer) become a recipe for getting bent?

 

[rhwestfall]

Cochran is actually the most liberal of the available algorithms, though, you can argue that Buhlmann with GF can sometimes be more liberal. Single dives do not tell the whole story. Multiple dive data such as generated by ScubaLab can be very enlightening
http://ads.bonniercorp.com/scuba/PDF/ScubaLab-Computer-Test-September-2014-data.pdf
Their 2016 data is also available off their website as an Excel file.


View attachment 382906

that is because the Cochran algorithm is a "Navy Schedule" (IIRC, they are a contract vendor/supplier for the US Armed Forces and other countries too), and from there you can assign some conservation to it.... I dive one of their computers (EMC-20) as a "lost deco gas, get me home somehow" back-up to a Petrel.