Planification

61 views
Skip to first unread message

Eric Tanguy

unread,
May 9, 2022, 12:07:15 PM5/9/22
to Subsurface Divelog
Hello, i'm trying to use subsurface planification. The problem is that my dive computer (Mares Quad) use RGBM algorithm and subsurface use Buhlmann or VPM model. Is there a possibility to use RGBM in subsurface or other mean to obtain a match between subsurface planification and the computer during the dive ?
Thanks
Eric

Stuart Vernon

unread,
May 9, 2022, 12:15:25 PM5/9/22
to subsurfac...@googlegroups.com

RGBM in your Mares computer is proprietary and undisclosed in its details to the general public.

 

So, no way to implement an exactly matching algorithm in Subsurface. No way to really even know whether an implementation that seems close will actually be close in all dive profiles.

--
You received this message because you are subscribed to the Google Groups "Subsurface Divelog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to subsurface-dive...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/subsurface-divelog/e32d21aa-594a-4abe-b06c-1eb964c21a1en%40googlegroups.com.

Linus Torvalds

unread,
May 9, 2022, 12:20:55 PM5/9/22
to Subsurface Divelog
On Mon, May 9, 2022 at 9:07 AM Eric Tanguy <eric....@gmail.com> wrote:
>
> Is there a possibility to use RGBM in subsurface or other mean to obtain a match between subsurface planification and the computer during the dive ?

No. RGBM isn't actually really documented anywhere unlike Buhlmann or
VPM. There are overviews of what the logic for RGBM is, but no actual
full docs.

(Also, "RGBM" is just a general term for a whole suite of different
variations of deco algorithms, so even getting two different dive
computers that both are ostensibly RGBM to match is going to be
interesting.

So the best you can do is probably to just compare the deco
information from your dive computer to Buhlmann, and learn what the
pattern of difference is so that you can at least anticipate roughly
what your dive computer will do for a dive you've planned with
subsurface.

But getting an exact match? No. And the pattern can be rather
non-intuitive (ie famously repetitive diving will likely have very
different effects - so even if you can match up deco for individual
dives by picking some particular gradient factors, the second dive
might be very different *cough*Suunto*cough*).

Linus

Eric Tanguy

unread,
May 12, 2022, 12:25:21 AM5/12/22
to Subsurface Divelog
Thank you very much for the explanations. 
Where can i find explanations about VPM-B and Buhlmann (mainly GF) algorithms ?
Eric

Gmail im Auftrag von Martin Gröger

unread,
May 12, 2022, 2:57:58 AM5/12/22
to subsurfac...@googlegroups.com

--
You received this message because you are subscribed to the Google Groups "Subsurface Divelog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to subsurface-dive...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/subsurface-divelog/010df4a8-e829-48ce-b409-9b18d6dc5b15n%40googlegroups.com.

JB2Cool

unread,
May 12, 2022, 3:28:09 AM5/12/22
to Subsurface Divelog
Results from a quick Google.

Buhlmann

Comparison

My usual go-to is the book Deco for Divers written by Mark Powell, that explains things quite well.

Jason

To view this discussion on the web visit https://groups.google.com/d/msgid/subsurface-divelog/CAAc4xU%2BV2-5YrmZog9%2BrV3Du2VQV3utdor5b-ZupfK3MEZ6QDA%40mail.gmail.com.

Robert C. Helling

unread,
May 12, 2022, 4:47:29 AM5/12/22
to Subsurface Divelog
For gradient factors, I second the recommendation for Matthias Heinrich's text https://www.heinrichsweikamp.net/downloads/OSTC_infoheft_GF_web.pdf
which is really well written.

For VPM-B, the situation is much worse: There are the original texts by Young and Baker but I must admit, I find these terribly hard to follow in their logic (and I have a PhD in theoretical physics, so I read technical texts about similar subjects on a daily basis). In the end, for me I decided that the only true reference documentation of VPM-B is the source code of the original fortran implementation. All the prose about bubble properties and membranes etc I only take seriously as long as it has a reflection in the source code.

I have spent a considerable amount of my time bending my head around this model and the result of this effort is two-fold: It is the Subsurface implementation (in particular what is in deco.c) which we rewrote from scratch based on the bits we understood and in the end made sure that we produced the same numbers as the fortran program. I should mention Rick Walsh who did a lot of the work for this project. Also I have written a series of articles in my blog about what I learned: 


But those might as well not be the most pedagogical introduction as I have written those with a target audience that is well aware of how Bühlmann + gradient factors work. I think my understanding can be summarised as follows: Despite the prose indicating otherwise, VPM-B is not really  trying to model bubbles, it is still a diffusion/solution like model. But it takes its criterion of what kinds of over-saturation are still ok from a bubble type motivation where it punishes over-saturation at depth more than closer to the surface (as it results in bubbles that will still grow when the ambient pressure is released). It assumes there is a total budget for over-saturation integrated over time with this Boyle correction taken into account. And more important: The deco schedules produced by it tend to be too deep with not enough time spent at shallow depths with absurdities like those described in https://thetheoreticaldiver.org/wordpress/index.php/2017/11/02/shortening-deco-by-diving-deeper/
So, I would advise against using this model for actual decompression. So any time spent on understanding this model will only lead to (hopefully) understanding why it is bad and thus only academic.

But YMMV.
Robert 
Reply all
Reply to author
Forward
0 new messages