Dive computers... SO many choices!

[stuartv]

I'm not sure what the advantage is of setting all of your low values in the 20-
40 range, except for keeping people very far away from the M-values.

Let me be more direct. As far as I know GF Lo was specifically put in to let the user produce ascent plans that roughly match the deep stops that bubble models used to produce. The value of those deeper stops is now at least somewhat discredited. So:

Is there any study anywhere that has produced data to support the notion that there is any reason to even still HAVE a GF Lo? I.e. why not just use the same number for GF Lo and GF Hi? I know there are opinions and that's not what I'm asking for. I have asked a couple of experts in the field this question and they could not give me a reason. So, I'm asking you people that seem to really know what's up. Can you cite any study that shows a positive benefit to having a GF Lo that is actually lower than the GF Hi?

I'm just a hobbyist, not a deco scientist. My "research" is completely amateur. But, what I've been able to read and understand suggests to me that there is no reason to use a lower GF Lo value than what you use for GF Hi. E.g. Use GF 80/80 or 70/70 or 60/60, rather than 50/80 or 30/70 or whatever.

Is there any actual data that would support the statement that this is a bad idea? Like I said, my research is amateur and I'm not connected to the serious research circles on this, so me not finding out the reason why it's a bad idea doesn't mean it isn't one.

 

[stuartv]

Uh? Why do I need an example? What I have said is correct. I think maybe I am not getting my point across. Would be more than happy to explain it verbally though... lol

Well, you said that NDL is determined by the intercept of the GF line with ambient pressure. And then you said that is affected in a small way by GF Lo. Those two statements are contradictory.

 

[LandonL]

It is affected by GF Lo, ONLY when and if you hit the GF-line in planning software, not in practice using a computer.

That is only because your NDL is being recalculated for the change in your dive in the software based on some ascent rate assumptions.

 

[stuartv]

It is affected by GF Lo, ONLY when and if you hit the GF-line.

That is only because your NDL is being recalculated for the change in your dive.

I understand what you're saying. So, let me ask you this. Suppose you do a dive with GF 30/90. On your ascent, you get to 20' and your leading compartment pressure gradient is at 29%. That means you are going to exceed 30% before you get to 10' if you continue your ascent.

Do you prescribe a mandatory deco stop at 20' in that case?

 

[dmaziuk]

If we really really wanted to argue semantics, technically every dive is a decompression dive, but we don't run around arguing this point with everyone, and correcting new divers

And maybe we should. Maybe we also should tell everyone that there are always bubbles and that it's Buhlmann's model that's technically "mistaken" in that it it tries to prevent bubbles from forming...

 

[KenGordon]

I do agree with trying to keep things simple for the sake of being practical, but technically that isn't what he explicitly stated, maybe this is what Ken is focusing on? Am I correct Ken?

I think stuartv was claiming that what he said implies that the NDL has absolutely nothing to do with the GF lo.

So I guess if you really wanted to have an academic argument that is somewhat pointless, the GF Low does have some effect on the NDL mathematically, but practically these stops are so short and cleared by arrival, you can ignore them (which is what some of the computer models do - why stop for 10 seconds 30 feet above you when your tissue will be clear by the time you get there). Given this fact, is the stop really relevant if you have cleared it before arrival? Which is exactly what is being said in the expanded explanation from Shearwater.

That will depend on the ascent rate. One of the points of a NDL dive is to be able to ascend immediately to the surface. So any ceiling would seem to preclude that, especially in the case of messing up buoyancy, having a lift go wrong or an accidentally dropped weight belt. Also particularly low GF lo are more likely to see this happen with moderate ascent rates.

It isn't really academic. You, for example, are about to chose some GF numbers as your preset choices. If you follow the fashion and have low GF lo values you will have to chose how to handle this.

When your computer is in the water how does it determine what stops are due? Does it first check if an ascent can be made to the surface limited by GF high and only if that is not possible generate a interpolated GF stop?

 

[KenGordon]

As a point of clarification, isn't this untrue for the lo gradient? The deep stop studies showed higher levels of DCS when people were trying to do something closer to a reduced bubble model, which is what the lo gradient does with smaller numbers.

"Safe" for the lo gradient is a bit contradictory I believe:

- A low number (0-30) approximates RGBM or deep stops, which is known to not work and increase DCS
.

Not really, because the dive gets longer. Use a low lo value and any extra on gassing deep will keep you in the water longer at the last stop so as not to exceed the gf hi value.

This is different to the NEDU study where they were doing the stops deeper but the total time in the water was constant. With GF a 45/85 dive will have longer last stops as well as deeper stops than an 85/85 dive, so the surfacing risk ought to be similar. On the other hand the slower compartments will still be more saturated than the shallow profile but no so badly as in the NEDU case.

 

[KenGordon]

Lastly, this is a lot of debate about how to do something that Baker's algorithm wasn't really even intended for. I could be wrong, but I believe his algorithm is/was specifically intended to calculate ascents for decompression dives. It was not designed or intended to be used to calculate NDLs. So, how his implementation works when you try to use it for NDL dives may not really be the "best" way to use his algorithm or the best way to calculate NDLs. In other words, even though his implementation always calculates the first stop using GF Lo, that doesn't mean that that is the best way to calculate an NDL. I personally think the way Shearwater and others are doing it (using only GF Hi) is a better way of calculating an NDL.

So you agree that GF is a bit rubbish for an computer the main use of which is NDL diving?

"Here have this computer. It has two knobs. One isn't connected to anything and the other one... hmmm we can't really recommend a number for that. Look in this top hat. It has a selection of numbers to choose...."

 

[dmaziuk]

Not really, because the dive gets longer. Use a low lo value and any extra on gassing deep will keep you in the water longer at the last stop so as not to exceed the gf hi value.

This is different to the NEDU study where they were doing the stops deeper but the total time in the water was constant. With GF a 45/85 dive will have longer last stops as well as deeper stops than an 85/85 dive, so the surfacing risk ought to be similar. On the other hand the slower compartments will still be more saturated than the shallow profile but no so badly as in the NEDU case.

I think the salient point is that if your target demo is rec users who often don't plan their gas for longer stops (<raises hand/>), then having a gf lo that generates a stop, and then gf hi that pads the shallow one accordingly, sounds like a rather unwise design decision.

 

[LandonL]

As a point of clarification, isn't this untrue for the lo gradient? The deep stop studies showed higher levels of DCS when people were trying to do something closer to a reduced bubble model,which is what the lo gradient does with smaller numbers.

I made some assumptions above I maybe shouldn't have, like picking reasonable numbers and not diving 0/0, and considering on gassing at deep stops for slow tissues... anyway.... yes picking ridiculously low numbers would increase risk more than mitigate it in my view...

This issue is very complex though.... I am going to assume you are talking about the NEDU dive data that Dr. Doolette. If not, then maybe reading his excellent summary of the research will give you some insight. It is in the linked thread below. If you want my personal opinion, yeah I am not big on deep stops, and I think its clear if using the same gas throughout its better to just do a shallow decompression profile, which can be extrapolated to recreational diving as well. Multi-gas switches? It is hard to say for certain, and I would be very interested to see controlled studies that addressed this.

Deep Stops Increases DCS

For the hi gradient, obviously the lower the number the better - it just impacts time/ascent rate.

So, what makes sense for low conservatism? That's why I thought something like (40-60) / (lowest possible) would be best. I'm not sure what the advantage is of setting all of your low values in the 20-40 range, except for keeping people very far away from the M-values.

The problem with diving very low GF values, is you are pushing your decompression profile closer to the Ambient Pressure line, which greatly extends your dive time, and given the NEDU data for single gas deep dives, it is going to result in on-gassing of some of the slow tissue compartments at deep stops. It gets more complex when you throw in multiple gasses as mentioned above, but regardless....

What makes sense? I don't think we have a clear answer there yet or data to define exactly where one should set the low GF to be optimal. Honestly I don't think it makes any difference for no deco dives, as you really are only concerned with the GF High value. Going into deco on single gas? I would probably opt for a higher GF low value myself, something in the middle range, planned multi-gas deco it just gets more complex, but in those cases I tend to dive team decided plans..