Gradient Factor Presets by Manufacturer/Computer

[CandiveOz]

Suboptimal, yes, but since the goal was to add more conservatism based on diver's personal considerations... well, that's what it does.

Yes, I guess you can use the term “suboptimal” when it comes to different GF pairs. But then the question begs; why would you want a suboptimal decompression profile?

Just getting back to the objective of this post. While we won’t be able to determine what the optimal GF pairs are for Presets, I think we identified which pairs are probably suboptimal as highlighted in the red block in #24. As divers continue to use gradient factors, certain GF pairs will become common and hopefully a consensus will develop over time (and testing) to determine which GF pairs are appropriate for different types of diving. But I don’t think we are their yet.

That is why I believe this issue is important today because more and more manufacturers are offering dive computers with gradient factors and divers are purchasing these units without fully understanding the implications. In fact, most divers, I hate to say, base their purchase on form factor, colour screen, air integration and “coolness.” Little consideration is given to the algorithm. And most likely, these divers, will leave the GF setting on its default Preset with the rational that the dive supplier “knows best.” I’m not sure that this is the case and this is the reason why I wanted to highlight this issue in my original post. Hopefully we have all learned something. I certainly did.

 

[KenGordon]

why would you want a suboptimal decompression profile?

Fashion.

 

[dmaziuk]

Yes, I guess you can use the term “suboptimal” when it comes to different GF pairs. But then the question begs; why would you want a suboptimal decompression profile?

Why would you take a pony on a 20 m reef dive? Or two Shearwaters? People do weird stuff. You let them add more conservatism, some will add too much conservatism.

Buhlmann's M-value function is a straight line, tried and tested since Workman. As long as you modify it to a straight line to the "safe" side of the baseline, I really don't see how that can be "more risky". It's when you make it something that tries to balance a "gradient" of theoretical properties of a hypothetical bubble, then you gotta wonder how "untested" a conservatism factor of "-3" really is.

 

[CandiveOz]

As long as you modify it to a straight line to the "safe" side of the baseline, I really don't see how that can be "more risky".

I understand your position and mathematically it is justified. But remember this is just a model representing a physiological effect and therefore has its limits. What you are implying is that 20/80 = 80/80 = 95/80 (an inverse GF); essentially the decompression profile doesn’t matter because you get out of the water at 80% regardless. I don’t believe this is necessarily true. More studies and test dives would be necessary to convince me. In the meantime, I’ll fall back on my on my 20 minute argument I stated before; if a GF profile generates a longer deco run time, maybe a conventional deco with the extra time spent in the shallows is a better or more optimal option.

Prior to gradient factors, dive computers were pretty straight forward. They had a default setting plus one or two conservative factors. Most divers just strapped them on and jumped into the water. When manufacturers marketed their algorithms; be it DSAT, PZ+, RGBM or Buhlmann ZH-L8/16 they would boast that their algorithm was stringently tested with a substantial number of man dives etc. While gradient factors are based on Buhlmann which has been tested and used for decades, the actual gradient factor pairs themselves haven’t been. And that is the risk regardless of the mathematical validity.

 

[dmaziuk]

More studies and test dives would be necessary to convince me.

Yeah, good luck with that. Join the Navy: they're about the only ones doing any studying -- but they optimize for somewhat different parameters than us underwater tourists. So if you want their studies to be relevant to you: dive their profiles.

 

[NAM001]

Candiveoz

Thanks for the time and data you compiled. And the list of questions on one of the posts should certainly happen. Im going to throw a little poop in the water here based on some videos i have seen. There was a video,,, I think a shearwater,,, where the computer was presurized to a deep location like 100 ft and when they went to slowly vent to simulate a ascent they slipped and they went to something like 30-20 ft in a few seconds, before catching it and going back to deep and and restarting the ascent. When that happened from a near to NDL time frame, no alarms went off other than ascent rate. I confident that both gradient factors lo and high values were exceeded in the loss of pressure control incident. If that was a valid example of how the computer reacts then gf lo is totally moot until you hit NDL and gf Hi is moot also except for use in calculating NDL time. So it appears that , to me, the gf Lo is just there to fit the format of processing that is needed to do a deco ascent. And yes once your enter deco you are no longer rec diving. So from the rec world the only value that makes any diference appears to be (with that maker) is gf HI and that is solely for NDL calc. I say this because the high ascent rate and depth change triggered neither of the gf values. I believe that ,,,if in the video,,, the NDL was exceeded the outcome would have been much different. If that is true across the board with makers then setting a custom gf's serves no other function than controlling your allowed NDL time. So 40/85 is no different than 85/85 on ascents prior to NDL. and that 85/85 is no different than 100/100 prior to hitting NDL with the exception of ndl TIME CALCULATION.

tHOUGHTS?

 

[dmaziuk]

That's why the algorithm itself is a relatively minor part of a system design. You have to design for a user base and so on and so forth. Shearwater's target user base may be people whose gear will kill them in 4 minutes if they aren't careful, so what their computer does not do may be OK with them. There's a problem with selling a design like that to us clueless vacay divers: one day you may get class-actioned by some concerned mother in California.

 

[tursiops]

Candiveoz

Thanks for the time and data you compiled. And the list of questions on one of the posts should certainly happen. Im going to throw a little poop in the water here based on some videos i have seen. There was a video,,, I think a shearwater,,, where the computer was presurized to a deep location like 100 ft and when they went to slowly vent to simulate a ascent they slipped and they went to something like 30-20 ft in a few seconds, before catching it and going back to deep and and restarting the ascent. When that happened from a near to NDL time frame, no alarms went off other than ascent rate. I confident that both gradient factors lo and high values were exceeded in the loss of pressure control incident. If that was a valid example of how the computer reacts then gf lo is totally moot until you hit NDL and gf Hi is moot also except for use in calculating NDL time. So it appears that , to me, the gf Lo is just there to fit the format of processing that is needed to do a deco ascent. And yes once your enter deco you are no longer rec diving. So from the rec world the only value that makes any diference appears to be (with that maker) is gf HI and that is solely for NDL calc. I say this because the high ascent rate and depth change triggered neither of the gf values. I believe that ,,,if in the video,,, the NDL was exceeded the outcome would have been much different. If that is true across the board with makers then setting a custom gf's serves no other function than controlling your allowed NDL time. So 40/85 is no different than 85/85 on ascents prior to NDL. and that 85/85 is no different than 100/100 prior to hitting NDL with the exception of ndl TIME CALCULATION.

tHOUGHTS?

Why is this so complicated?
GFlo is irrelevant until you hit NDL. It could be anything.
If you inadvertently exceed NDL, however, then GFlo matters. Possibly a lot. So it needs to be selected with that in mind. It is, therefore, an important parameter in the computer setup, which if you are doing NDL diving you hope never to have to use. But, you might, so it needs to be chosen wisely, not arbitrarily.

 

[dmaziuk]

I understand your position and mathematically it is justified. But remember this is just a model representing a physiological effect and therefore has its limits. What you are implying is that 20/80 = 80/80 = 95/80 (an inverse GF); essentially the decompression profile doesn’t matter because you get out of the water at 80% regardless. I don’t believe this is necessarily true.

Do you also believe that multi-level profiles are dangerous? Because there is no difference between having a "deep" decompression stop and exploring a ledge at the same depth for teh same amount of time as part of your planned profile. If one has questionable physiological effects, then so must the other.

 

[NAM001]

Why is this so complicated?
GFlo is irrelevant until you hit NDL. It could be anything.
If you inadvertently exceed NDL, however, then GFlo matters. Possibly a lot. So it needs to be selected with that in mind. It is, therefore, an important parameter in the computer setup, which if you are doing NDL diving you hope never to have to use. But, you might, so it needs to be chosen wisely, not arbitrarily.

I dont think it is that complicated so long as any one that looks into this understands that when tech people talk about these things they often do not speek to their use in a a rec application. When talk about GF Lo occurs in a rec setting,,,,,, converstion immediatly leaves the rec world when GF LO = first stop is said. And often that happens with out warning. Then when talk about GF Hi starts,,,, it is usually spoken as the max percentage the diver wants to be exposed to. To most rec divers that means nothing either. Most divers in the rec community will never reach the limits where GF's come into play. Many do not even know that GF's are in their computer. Some do know how to fix their puter to let them stay down longer and do so with out any understanding what they are doing. I posed the question a while back, if there was a chart that compared various puters and their conservative settings. Some one in this thread made the painstaking effort to make such a table. THANKS TO WHOEVER DID THAT. That table is going in my dive book so I can allign my puter with others so we all are running on the same close calculated NDL. I wish there was a LAYMANS LEVEL video that talked about m values and gf's and how to use them in a REC setting and their affect on safety stops. Just a hint of what a M value is and a little about micro bubbles occuring before the m value and,,,, I know so many that would look up and say the light bulb came on about this. If something like this was onlyt vailable and presented in the most basic concepts. To be honest it should be a topic in the AOW course. You know a couple hours about haldine and ultrasound results and how that effected dive tables, conservative settings and the relation or affect of a safety stop. I have to say that this issue is part of the fall out of no linger teaching tables and its theory.