Gradient Factors

OSTC's running hwOS sport or tech
Post Reply
Yves
Posts: 20
Joined: Monday 24. August 2020, 13:07

Gradient Factors

Post by Yves »

Hi,
I assume that all computers using the ZH-L16+GF model from Erik Baker are the same.
Should my buddy be using a non-OSTC computer (eg Shearwater), and configure the GF parameters with the same configuration as my OSTC, can I take for granted than the deco will be the same (same deco stops and same durations) ?
Thanks for your clarifications.
Clownfish
Posts: 169
Joined: Sunday 1. July 2012, 15:22

Re: Gradient Factors

Post by Clownfish »

From experience with OSTC v Shearwater, if configured with same GF, time to deco and deco are within a minute of each other.
jb2cool
Posts: 142
Joined: Tuesday 21. January 2014, 21:19

Re: Gradient Factors

Post by jb2cool »

The implementations of the algorithm are slightly different on different computers but yes, they should line up to within a minute or so.
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Gradient Factors

Post by Ralph »

Unfortunately other brands than Shearwater do not match that good (to either OSTC or Shearwater), as these tend to add internal safety factors that result in more deco obligation than a calculation with plain & strict Bühlmann and E.Baker GFs would yield.
Last edited by Ralph on Thursday 3. September 2020, 16:03, edited 1 time in total.
heinrichsweikamp
Posts: 4370
Joined: Sunday 13. May 2007, 18:07

Re: Gradient Factors

Post by heinrichsweikamp »

Hi,

Just to add: We'll finally have the english GF manual online within the next weeks.
infoheft_GF_en_titel.PNG
This is a recommended addition to the manual, since we no longer have these pre-configured conservatism settings of the older versions.

regards,
Matthias
dadefay
Posts: 115
Joined: Thursday 2. February 2017, 15:39

Re: Gradient Factors

Post by dadefay »

Great news, thanks Matthias !
Didier A. Defay
OC trimix instructor / Trimix CCR diver
France

OSTC 2 #18835
- OSTC Plus #16077
-- OSTC 3+ #4806
--- OSTC 3 #3999
Yves
Posts: 20
Joined: Monday 24. August 2020, 13:07

Re: Gradient Factors

Post by Yves »

Thanks Matthias, this is great !
I am looking forward to reading this document that should be riveting.
heinrichsweikamp
Posts: 4370
Joined: Sunday 13. May 2007, 18:07

Re: Gradient Factors

Post by heinrichsweikamp »

Hi,

You can download the PDF here: https://www.heinrichsweikamp.net/downlo ... web_en.pdf

regards,
Matthias
Yves
Posts: 20
Joined: Monday 24. August 2020, 13:07

Re: Gradient Factors

Post by Yves »

Thanks Matthias for this riveting didactic document.
However, I have still a question. In the OSTC "tissue chart" display you show on page 9, Figure 12, I see 2 sets of horizontal bars, apparently each for 16 tissues. What is the meaning of these 2 tissue bars?
I noticed on my OSTC+ tech four vertical bars spaced out under the horizontal tissue bars, with different colours: turquoise, orange, green, and red. What is the meaning of those?
Thanks and regards
heinrichsweikamp
Posts: 4370
Joined: Sunday 13. May 2007, 18:07

Re: Gradient Factors

Post by heinrichsweikamp »

Yves wrote: Wednesday 23. September 2020, 16:30 However, I have still a question. In the OSTC "tissue chart" display you show on page 9, Figure 12, I see 2 sets of horizontal bars, apparently each for 16 tissues. What is the meaning of these 2 tissue bars?
When Settings -> Display Settings -> More -> Graphics is set to "N2+He": Upper set is for N2 (Can be lower then current ambient pressure (Think of high alitude/flying), so they a longer on the left side), lower set is for He.
Yves wrote: Wednesday 23. September 2020, 16:30 I noticed on my OSTC+ tech four vertical bars spaced out under the horizontal tissue bars, with different colours: turquoise, orange, green, and red. What is the meaning of those?
When Settings -> Display Settings -> More -> Graphics is set to "Pres+Sat":

turquoise: zero oversaturation
green: GF low (Any tissue reaching to this line is oversaturated to your current GF low)
orange: GF high
red: M-Value


Hope this helps,
Matthias
jb2cool
Posts: 142
Joined: Tuesday 21. January 2014, 21:19

Re: Gradient Factors

Post by jb2cool »

One for nitrogen and one for helium I think.
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Gradient Factors

Post by Ralph »

The supersaturation bars show the sum of N2+ He, as this is what goes into the deco calculation and what the GF low, GF high and m-line is rated against.

The turquoise bar actually just looks like a bar when there is no supersaturation at all, indeed it the 16 horizontal lines that show the relative supersaturation of the 16 tissues. When supersat = 0 they are drawn with a length of 1 pixel, what makes them look like a vertical bar when all of them are just 1 px in length.

Pres+Sat = total PRESsure (upper graphics) and relative super-SATuration (lower graphics). Both times sum of N2+He.

-R.
Yves
Posts: 20
Joined: Monday 24. August 2020, 13:07

Re: Gradient Factors

Post by Yves »

Hi,
I read in page 9 of your referred document:
...you never come shallower than the indicated ceiling depth and never ascend faster than 10 meters per minute (this is anchored in the model)
How about if you choose "Variable speed" instead of 10m/min in your deco parameters?

I tried my simulations with 10m/mn and Variable Speed for the ascent for various depths (from 12m to 51m), and I came up with the same results.
Does it mean that the "variable speed" selection was ineffective, or that it is normal that the results are the same?
In other words, whether 10m/mn or variable speed ascent has no effect on the decompression calculation!

Other remark: when NOT using the bar graph display and selecting Variable Speed, the actual rate of ascent with figures is not really clear:
it gives +10m/min in white, then yellow, then red. But we have no idea of how much +10m/min.
Why not displaying the actual rate, eg 20m/min, 18m/min,..., 7m/min,..., at every depth ?
and displaying a % of the ideal variable speed ascent, with colours: white if >60%-100%, yellow if >100%-130%, red if >130% ?

Could you please clarify.
Thanks and regards,
Yves
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Gradient Factors

Post by Ralph »

Hi Yves,

"Variable Speed" is no longer in the current hwOS firmware. As of now, the OSTC always calculates the ascent with 10m/min, as defined within the Bühlmann model. This parameter or better to say speed limitation is important, as can be seen when looking at the half-times of the fastest tissues, especially if Helium is in the game: these are in the area of the total ascent time or even shorter, which means the reduction of tissue pressure during the ascent towards especially the 1st stop is figured into the design of the Bühlmann model. If you do slower, it works towards the safe side.

Ascent speed visualisation depends on layout mode: bargraph and numerical speed are shown in normal layout, whereas in the large layout only the numerical speed is shown for reason of available space on the screen. The color-coding depends on the "Speed .. adaptive" setting, which controls vertical speed color-coding only, not deco calculations! When set to 'off' white is for all descent and for ascent < 10m/min. Yellow indicates the 10m/min borderline, above is red. So vertical speed is shown...

The variable speed in the deco settings with older firmware had effect as it was regarded in the internal calculations. However, the effect was only noticable if the ascent from bottom to 1st took a time in the range of the half-times of the fast tissues and if these fast tissues were the leading ones.

BR
Ralph
Yves
Posts: 20
Joined: Monday 24. August 2020, 13:07

Re: Gradient Factors

Post by Yves »

Thanks Ralph,

Although I understand your explanations, I have a couple of remarks:
"Variable Speed" is no longer in the current hwOS firmware.
I have the latest firmware v3.12 for the OSTC+ Tech that I downloaded recently, and I still have this Variable Speed parameter available for selection. Since that parameter has no effect on the deco calculation, I do not understand why it is still there in the menu. See also my next remark.
Since 10m/min is the recommended ascent speed even at the time of Hadane (1908), let's stick to it and forget the variable ascent speed even if it may make sense at deep depths from eg 50m to 20m depths.
bargraph and numerical speed are shown in normal layout, whereas in the large layout only the numerical speed is shown for reason of available space on the screen.

Well, I tried this large layout, and all I saw was the current depth in large characters, and the ascent speed in normal characters next to the water temperature. And indeed, no graph display (because of not enough room on the screen, ok). But, this is the current ascent speed that I expected to be in large characters, not the current depth that, in my view, has no interest.
Are you sure there is not a mistake there?

Thanks in advance for your reply.
With best regards,
Yves
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Gradient Factors

Post by Ralph »

Where do you have a Variable Speed Parameter?

Main Menu -> ... ?
Yves
Posts: 20
Joined: Monday 24. August 2020, 13:07

Re: Gradient Factors

Post by Yves »

Hi ralph,

Main menu->Settings, (then after Simulator)->Display settings->in following screen, second parameter between Speed graph and Display PpO2: Variable speed (default=no):
Here is my OSTC screen display (in french):
20201002_var_speed.jpg
French "Préfér. Réglages" = Display Settings
French "Graph.Vitesse" = Speed Graph
French "Vites.Variabl" = Variable Speed
French "Afficher PpO2" = Display PpO2

Regards,
Yves
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Gradient Factors

Post by Ralph »

Ah, that's DISPLAY settings, having nothing to do with DECO settings! This is the "Speed .. adaptive" setting i wrote about: if switched on, the color coding of the vertical speed bargraph depends on current depth, becoming more erarly red the shallower you are.

10m/min is - at least as of Bühlmann - not >recommended<, but the >maximum< ascent speed the model is valid for. Doing it slower, especially on final ascent to surface, is strongly recommended. Remember: the highest tissue supersaturation and the highest relative pressure change occure on reaching the surface!

As of the depth displays: sorry, i didn't get the point: Current depth is rated the most important display and hence always visible. The ascent/descent speed is only shown when a vertical motion is sensed, and the additional vertical speed bargraph comes only on display when in normal layout and the direction is ascent. Was that your question?

cheers,
Ralph
heinrichsweikamp
Posts: 4370
Joined: Sunday 13. May 2007, 18:07

Re: Gradient Factors

Post by heinrichsweikamp »

Hello,
infoheft_GF_fr_titel.PNG
Here is the French language version: https://www.heinrichsweikamp.net/downlo ... web_fr.pdf for download.

Thanks to OSTC-forum member "Bardass" for the translation work!

regards,
Matthias
Spacefish
Posts: 32
Joined: Wednesday 30. August 2017, 12:57

Re: Gradient Factors

Post by Spacefish »

How is the following case handled?

Imagine GF Low: 10% GF High 100% (or any other non equal number where gflow < gfhigh)

So my first deco stop should allow 10% super saturation of the leading tissue, the one in the middle 55% and on the surface 100%.
This works out pretty well if i do a normal deco and stick to the plan.

However how is the case handled where i decent during deco / not stick to the plan?
Like i have a stop at 20m (10% Super saturation allowed) i stop and i am allowed to accent to 10m (middle stop 55% super saturation allowed).
But now i don´t continue to the next stop, but decent! My ceiling drops again and a new stop at 18m is added.

Which super saturation is allowed at 18m now? 10% or ~20%? / How is the M-Line adjusted? Which stop is considered to be the "first"? I see three options here:

- Deepest stop ever calculated in the current dive (18m would be~20% supersat)
- Last stop added to the deco plan if deeper than any other stop in the current plan (18m would be 10% supersat)
- Deepest Ceiling Factor calculated when calculating with gflow seen in the dive (pretty much the same as the first one but non discrete)

The linked deco model document does not give an answer to this question as far as i can tell..
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Gradient Factors

Post by Ralph »

I really had to look into the code to give a prceise answer. Here it is:

1) The GF-low reference depth is set to the depth of the deepest deco stop found during the whole dive.

2) When two plan are calculated (fTTS or OC bailout), each plan has its own GF-low reference depth.

3) When the GF factors are toggled between GF and aGF (alternative GF), the GF low reference depth is reset to the depth of the 1st stop as of now in the dive.

Hmmm, maybe 3) isn't that clever - shall this case be deleted in future firmware?

Anyhow, to also directly answer your question: the GF-low reference depth does not go up (as of now: unless you are toggeling GF/aGF) whatever your dive profile is, but it may go further down when -- for whatever reason -- the deco obligation builds up further and requires a deeper first stop.
So in your example, the new stop at 18m will be calculated with 20%, as the GF-low reference depth has been latched with the depth of the stop at 20 m. Still, the 18m will appear as the 1st stop in the lower right of the main screen.

The Bühlmann model and also E.Baker's proposal were given at a time were dive computers heartly existed. All models were made for rectangular profiles, because that was the only profile that could reasonably be put into dive tables and followed by the diver. If you ever played with the PADI Wheel, you know... So creating directions for any other dive profile simply wasn't necessary --- tasking us nowadays dive computer programmers to make best decisions on what would make the most sense within the original author's model framework and the todays practical diver's needs.

If asking me, i would add someting to the GF concept, and that would be a 'minimum GF-low reference depth': The merrit of a full tech dive computer is, that it will calculate a deco schedule for whatever profile i dive as long as i stay within the limits of the model. But to make sense, the GF factors used during the dive should be set in relation to the expected deco schedule. So when i set them for e.g. 50/80 but then have to abort the dive with only a minimum of deco accumulated so far, resulting in a single shallow stop, why shall i do that stop with 50 if i would have done that stop with nearby 80 if i had done the planned dive? If such a minimum GF-low reference depth would exist, i would be able to push the 50 down to say 21m at least (and it would automatically go down deeper if the deco puts the 1st stop deeper than that) and can have may shallow stop at a reasonably GF without enganging aGF or doing the deco "by hand" following the supersat% in the custom view.

BR
Ralph
Post Reply