Dive computers... SO many choices!

[dmaziuk]

"conservatism" presets for

OSTC sport:
20/70
20/75
30/80
30/85
50/90

Hollis TX1:
20 • 75
25 • 80
30 • 85
35 • 90
90 • 90

(both have "custom" option)

Edit: but note that TX-1 is an AI computer and will do GTR. Presumably it'll ring an alarm when its calculated deco schedule hits your gas reserve.

 

[scubadada]

Short Bio about me first..
Open water diver
Nitrox certified
Live in Florida (which means I dive alot)
I plan on taking the advanced open water course. No cave, or tec diving, ever....

So my question is:
Exactly what type of *wrist* computer is appropriate for an advanced diver?
I want to buy something nice and suitable so that I don't feel the need to have to UPGRADE later on.....

After 221 posts, I thought it appropriate to go back to the OP, edited post, above. I'm a strong proponent of understanding one's decompression algorithm, as many of you know. You may agree or disagree me on this topic.

The OP desired information for a good wrist rec computer. Look where we are now, in a hot debate whether GF lo is applicable and used in NDL calculations. @Nikole has not been here since September 26. I hope she got herself a nice computer

 

[scubadada]

"conservatism" presets for

OSTC sport:
20/70
20/75
30/80
30/85
50/90

Hollis TX1:
20 • 75
25 • 80
30 • 85
35 • 90
90 • 90

(both have "custom" option)

Edit: but note that TX-1 is an AI computer and will do GTR. Presumably it'll ring an alarm when its calculated deco schedule hits your gas reserve.

I rejected the Hollis because it only went up to 90/90, I believe, including custom

 

[LandonL]

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?

No, because at that point of your dive, your gradient factor has shifted along the line and is no longer set at 30%. You should make it to the surface fine. You may exceed 30%, but you won't exceed the current GF at that point of the dive, or the GF High.

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...

Haha. True point, no model is perfect, thats why it is a model, and yes bubbles do always form.

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

Yeah, that is correct. It really doesn't.

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?

GF High sets the intercept, if at any point you can't surface with a saturation below GF High, then it starts to calculate stops. GF Low sets the first stop depth.

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...."

It isn't rubbish, the GF High sets your intercept, and the two values are connected, they form the slope of the GF line. Your stops, along that line, are determined by GF Low, and you progress down that line, towards the surface, and GF High.

 

[scubadada]

As an entry level diver, I would be glad to have the Deep6 computer with the Shearwater GF presets.

 

[KenGordon]

GF High sets the intercept, if at any point you can't surface with a saturation below GF High, then it starts to calculate stops. GF Low sets the first stop depth.

Will you explain that in the manual? It is not GF behaviour as per Erik Baker. What is your rational for choosing that behaviour rather than choosing presets such as 70/70 etc?

I think I will up the GF hi on my main computer for a few dive to that on my backup (where the gf hi and lo are equal) and see how they compare in the water. Last weekend the backup was showing no stops while the primary was well into deco.

 

[KenGordon]

No, because at that point of your dive, your gradient factor has shifted along the line and is no longer set at 30%. You should make it to the surface fine. You may exceed 30%, but you won't exceed the current GF at that point of the dive, or the GF High.

In Stuart's example the first stop would be at 20ft surely? You can't get to 10ft without exceeding 30% so need to stop at 20 until you can. Or some intermediate depth between 10 and 20 if you chose to allow following the ceiling.

Where do you start the GF limit line? What depth is the anchor for GF lo where you start the interpolation to GF high?

In other words, if

gf now = gf hi - depth now * (gf hi - gf lo) / gf anchor depth (for depth now <= gf anchor depth)

how do you chose gf anchor depth?

 

[stuartv]

Where do you start the GF limit line? What depth is the anchor for GF lo where you start the interpolation to GF high?

I think that is a legit question that I was wondering about myself.

GF Hi will always be at depth of 0. So, what do you use as the depth for setting the other end of the GF line (at GF 0) and then interpolating to find the depth for GF Lo?

I'm guessing it is Current Depth is taken to be GF 0, and Current Depth is the depth at the time the diver exceeds their NDL (which is calculated solely using GF Hi). So, (Current Depth,0) is one end of the line and (0 (the surface), GF Hi) is the other end of the line. And you use interpolation to figure out the depth for GF Lo.

Once into deco, the first stop depth is set, unless and until the diver goes deeper - in which case the first stop depth is recalculated. But, if the diver stays at the same depth or goes shallower, the first stop depth says the same. If the diver ascends slowly, the stop time at the first stop could drop to 0 before the diver gets there, so the stop would drop off the ascent plan. But, the stops remaining would be for actual GFs higher than GF Lo. They would be the interpolated GF values that result in GF Hi at Depth==0.

That's just my guess. It's clearly more complicated than Baker's original implementation, which was not intended to be used in real-time, and was only for calculating based on a predefined list of dive segments, with a very clear "end of dive, start of final ascent". The comments in Baker's code even state very clearly that nothing in his implementation would stop you from putting in a dive segment that would result in an ascent above a deco ceiling. It just calculates the deco to be done AFTER all the dive segments you tell it.

 

[scubadada]

After 221 posts, I thought it appropriate to go back to the OP, edited post, above. I'm a strong proponent of understanding one's decompression algorithm, as many of you know. You may agree or disagree me on this topic.

The OP desired information for a good wrist rec computer. Look where we are now, in a hot debate whether GF lo is applicable and used in NDL calculations. @Nikole has not been here since September 26. I hope she got herself a nice computer

It's also interesting that this discussion is in Basic Scuba Discussions, all along, I have been thinking it was in the Computers... forum.

Interestingly, it is also the first, or one of the first, threads to discuss different implementations of the Buhlmann algorithm. My Nitek Q behaves just like the Shearwater computers, the NDL is controlled entirely by the GF hi. MultiDeco behaves differently and utilizes the GF lo in dictating the "NDL" and the 1st stop. I wonder how other computers that use Buhlmann behave? I like the @stuartv simple explanation that the computer simply calculates the NDL as the time you need to leave the bottom and make a normal ascent to surface to stay below your GF hi. So this is the computer NDL mode. If you overstay your time on the bottom and can no longer surface below your GF hi, you are in deco mode and the 1st stop is dictated by your GF lo. Of course, this may be entirely wrong but it is a good explanation for a simpleton like me. Computers 1st priority is to calculate the NDL for you, if you exceed that, it calculates the ascent plan for you to take care of the deco obligation. MultiDeco does not calculate or give you a NDL. It is specifically designed to give you an ascent plan for deco diving, either you have no stops for the dive profile, or you do. It will give you stops as short as 1 second, depending on how you have the stop times configured, 1, 10, 30 seconds or 1 minute. It will also give slightly different stop times depending on this configuration. The relatively small differences between my Nikek Q and MultiDeco do not bother me, the computer and the planning software have different functions. I wonder how the computer ascent plan compares with MultiDeco once you click over into deco mode, I've not had the chance to compare and my opportunities are few with my kind of diving.

 

[stuartv]

I think that is a legit question that I was wondering about myself.

GF Hi will always be at depth of 0. So, what do you use as the depth for setting the other end of the GF line (at GF 0) and then interpolating to find the depth for GF Lo?

I'm guessing it is Current Depth is taken to be GF 0, and Current Depth is the depth at the time the diver exceeds their NDL (which is calculated solely using GF Hi). So, (Current Depth,0) is one end of the line and (0 (the surface), GF Hi) is the other end of the line. And you use interpolation to figure out the depth for GF Lo.

I just realized I'm a doofus and what I said above is dumb. There is no interpolation required. At the point you go into deco, the depth at which you exceed 30% (or whatever your GF Lo is) of the M-value (in your leading compartment) is whatever it is. It can be calculated directly, without regard to what GF Hi is. And, the only way your depth would affect the calculation is if you choose to factor ascent time into the calculation.

If you choose to calculate it the way Baker does, which is to say ignore ascent time, then current depth would not be a parameter in the calculation at all. At that point, you are doing a somewhat simple calculation to answer the question "at what depth will my current tissue loading have at least one compartment that is greater than or equal to 30% (or other GF Lo value) of the M-value for that compartment?"

Of course, my ideal computer would factor ascent time into that calculation, so that it doesn't show me a stop that I will never hit (assuming I ascent at 30 ft/min). I believe the Perdix does work this way and, for example, my SeaBear H3 does not. I believe my H3 would show me a stop, as I ascend, that will disappear before or just as I get there. But I digress...

Anyway, I think the answer to @KenGordon's question would be that the anchor is set at the time you go into deco, by calculating the ceiling at that time, based on your current tissue loading and your GF Lo setting. As your tissue loading increases, it would be constantly recalculated. If your tissue loading decreases, it would stay set at whatever value it has. At least, that's my guess.