Odd calculated deco "ripples" (was Re: RFC: color change for calculated deco)

[Linus Torvalds]

On Sat, Jan 12, 2013 at 3:40 PM, Robert C. Helling <helling at atdotde.de> wrote:
>
> On Jan 12, 2013, at 10:30 PM, Linus Torvalds wrote:
>
> Linus,
>
>> Any ideas? I don't think the ripples *matter* (they are not a big
>> effect), but I'd like to understand why they exist. The last time we
>> had odd effects in the deco graph, it was because the deco
>> calculations were wrong.
>
> Hmm. I see the ripples and my first ideas about their origin were wrong (the guiding compartment does nothing fancy and what what goes into add_segment() does look completely normal as well, I had thought that the gas mix might jump but it does not). The ripples are still there with GFhigh=GFlow=100% i.e. plain vanilla Buehlmann, so the gradient factor stuff is not the origin either.
>
> Understanding this requires more time than I have right now. But I will come back to this as soon as possible.

It's interesting, but I am starting to think that it has something to
do with divecomputer choice. I seem to *never* see ripples with my
Suunto logs. With the Uemis, the ripples ar at 20-second intervals,
and I just looked at Dirk's dives, and with the Atomic Aquatics, the
ripples are at 15-second intervals.

And in his dives, the OSTC doesn't shows ripples, nor do I find them
in his old Suunto dives.

We don't have a huge number of dives with lots of deco in them
(although changing the gradient factors can be used to turn just about
any dive into a deco dive ;), and not all dives show the rippling very
much. But there *does* seem to be some tie-in to dive computer.

The fact that two different dive computers seem to have *different*
beats to them (15s vs 20s - and it does seem to be consistent from
dive to dive, although I only looked at a few dives) makes me wonder.
I'm starting to think it might be some "beat" from the sample
frequency vs some rounding effect of the depth by the dive computer.
It may be there in the depth samples too, it's just not visible, and
hidden by the noise in the actual depth movement.  Although I'd assume
that that noise would also be big enough to then drown any ripples.

Our own depth rounding does *not* seem to be an issue. Yes, we do
depth to "only" a precision of mm, and we do this in integers, but the
fact is, the sample precision itself is much less than that (the
Suunto depth samples, for example, seems to actually be in whole feet,
so the dive computer itself will have rounded the depth to *much* more
than mm).

I dunno. I'm still trying to see the pattern. And I'm not seeing it.
And maybe the pattern I *am* seeing (ie the dive computer pattern) is
a false one. But it does seem to be consistent enough to be real -
even if not *all* Uemis dives show the rippling effect, and some show
only a very weak ripple, it does seem fairly consistent.

                   Linus