Ceiling on 2nd dive

OSTC's running hwOS sport or tech
Post Reply
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Ceiling on 2nd dive

Post by harv »

I did some dives with my new OSTC Sport (running hwos sport 10.73) with fairly conservative settings, GF 50/65. On the first dive it was fairly close to my Garmin watch with the same gradient factors. However, on the 2nd dive I had to get much shallower for the OSTC to start counting down the deco time.

On the first dive the OSTC is pretty close to Subsurface's ceiling and the Garmin is in the same ballpark.
ostc_first.png
On the second dive the Garmin and Subsurface are pretty close:
garmin_second.png
The OSTC is much slower though. It doesn't end the deco until Subsurface reports a surface GF of 60.
ostc_second.png
Is this the expected behavior or is something off for the 2nd dive? Did I just accidentally end up with a profile where the OSTC didn't think I was off-gassing the leading tissues, while the Garmin and Subsurface just happened to end up on the other side?
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Ceiling on 2nd dive

Post by Ralph »

Hi,
when you set the GF high (GF at the end of the dive, i.e. when reaching the surface) to 65, then

"The OSTC is much slower though. It doesn't end the deco until Subsurface reports a surface GF of 60."

is exactly what you wanted it to do: bring me to the surface such that i will be at 65% of max. supersaturation when surfacing.
The difference between 60% and 65% is most probably due to slight differences in the 1st dive and surface intervall data between the OSTC and subsurface, plus the fact that the OSTC calculates with the real surface pressure whereas offline programs typically run on a fixed 1000 or 1013 hPa if not specifically set to something different.

BR
Ralph
jb2cool
Posts: 142
Joined: Tuesday 21. January 2014, 21:19

Re: Ceiling on 2nd dive

Post by jb2cool »

I don’t think the user is upset that deco takes longer to clear with GF high set to 65 than it does set to something like 75 or 80, I think the user is complaining that it seems to take a lot longer to clear deco on the second dive than the built-in Subsurface deco calculations based on the same gradient factors. For the first dive they were largely in line but on the second dive there is a large difference
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Ceiling on 2nd dive

Post by Ralph »

Well then, does subsurface know it's a repetitive dive and account for the residual load from the 1st?
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Re: Ceiling on 2nd dive

Post by harv »

Yea, as jb2cool says. What I am concerned about is the big difference compared to the two other Buhlmann implementations for the 2nd dive. I had another dive a couple of weeks ago which showed similar behavior, but on that dive I had difference settings on the dive computers and there where other changes, so wanted to make a more controlled test.

The Garmin watch also uses the real surface pressure and I think Subsurface uses the pressure from the OSTC log when doing the calculation. The surface interval is taken into account and you can see the initial nitrogen load in the heat map under the depth graph. In this case the Garmin data is slightly deeper on average, even though they have the same salinity, so Subsurface ends the deco slightly later for that data compared to the OSTC data.

I have a Suunto computer that is much more strict on the 2nd dive, which makes it really hard to have conservative settings on the first dive and I want to be sure that nothing like that is going on here.
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Re: Ceiling on 2nd dive

Post by harv »

For reference. On the first dive Subsurface calculates 64 as surface GF when OSTC ends the deco, and on the 2nd dive the surface GF is 60 when deco ends.

On the dives I did a couple of weeks ago I used 50/70 as gradient factors. On the first dive Subsurface calculated the a surface GF of 68 when OSTC ended the deco, and when the deco ended on the 2nd dive Subsurface calculated a surface GF of 64 (very close to 63).

On the most recent dives the surface pressure was 1000 hPa according to the OSTC log and on the other dives it was 1052 and 1058 hPa, so even if some constant is assumed it's not going to make a big difference.
User avatar
gr3yw0lf
Posts: 106
Joined: Thursday 28. November 2019, 22:18

Re: Ceiling on 2nd dive

Post by gr3yw0lf »

hm...

in my eyes you seem to youse your ostc not as the primary dc (it is #3!) so subsurface does not base its calculation on the ostc. so for a try use it the next dives as #1 and see whats happening ;)
from the profile I can not see remarkable stops. it is more a "up&down" thing. so maybe you should work on that a little bit. also you can take a little look at themes like "minimum deco".
maybe think over your GF. Gflow=80%GFhigh is taked around, but not realy recommanded. for me personaly works GFlow=50%GFhigh very well (e.g. GF40/80 at shallower dives around 30m).
at last - if you think it is a "miscalculation" please post it also in the subsurface-forum.
life is better in fins

OSTC Plus #18168
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Ceiling on 2nd dive

Post by Ralph »

Well, for me i still don't get the problem. As harv wrote, when the GF high was set to 65%, the re-calculation of the dives delivered 60 to 64%, and when set to 70% re-calc yielded 64 to 68%. That's a pretty good match, as wrote before, and also as the tissue conditions before the 1st dives are tracked by the OSTC and go into the calculation (btw: > 1050 hPa is an unusual high surface pressure). Also to note, the OSTC is updating tissue pressures every 2 seconds whereas the log is on a 20 sec intervall, so Subsurface has only data with 1/10 resolution in time to track the depth changes, which make quite an impact in the shallows, e.g. on the last deco stops.
User avatar
gr3yw0lf
Posts: 106
Joined: Thursday 28. November 2019, 22:18

Re: Ceiling on 2nd dive

Post by gr3yw0lf »

hm...
if I got it right, the issue seems to be, that the OSTC shows a longer "red bar" for the deco .

btw what are the saturation/desaturation settings of the OSTC? if they are not 100/100 that could be the explanation.
life is better in fins

OSTC Plus #18168
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Re: Ceiling on 2nd dive

Post by harv »

gr3ywolf, it makes absolutely no difference if I set the dive computer as first or third. Subsurface calculates the ceiling for the data from the dive computer that is currently displayed. These are all "NDL" dives, I'm not holding a deco stop (except for the end of the 2nd dive where I stayed at 3-4 m just to get the OSTC to clear) I'm just doing a normal dive but have a subset of my dive computers set to conservative settings so I know that my decompression stress is reasonable when surfacing. In this case I set the GFHigh lower than usual since I wanted to investigate the oddity that I saw on my first dives with the OSTC. GFlow is irrelevant for these dives, 30 might give me a 6 m ceiling while I'm at 20 m. Currently I see no reason to bring this to the subsurface forum. I have access to 3 implementations of the Bühlmann algorithm, Subsurface, OSTC Sport and Garmin Descent MK2i. I've used the Garmin on around 70 dives and it is very close to Subsurface, the odd one out is the OSTC.

And Ralph, I still think the difference here is quite big. On second dive, when Garmin and Subsurface thinks the deco is cleared, the OSTC still has a 5 min stop at 3 m and the @+5 also shows no difference in TTS, so something is different in the off-gassing.

Subsurface takes the tissue conditions from the previous dives in the log into account, and the Garmin is always on me and tracks the same information. If I remove the first dive from the log then Subsurface calculates a surface GF of 52 when the deco is cleared instead of 60.

>1050 hPA might be unusual high, but it only has minimal impact on the deco and it makes it slightly shorter, not longer. In the dives above the surface pressure was 1000 hPA though, which is well within the norm.

The Garmin logs roughly every second and difference that Subsurface sees between the OSTC and the Garmin log seems to be a reasonable difference for the change in sampling interval, a few minuted difference in when it thinks the deco is cleared, not a >15 min difference.

It's still possible that I accidentally triggered some weird threshold and if I just moved up 0.5 m then the OSTC would off-gass the tissues that affects the ceiling, but since I saw this behavior on both dive days that I used the OSTC I thought it was odd.

Yes, gr3ywolf, the issue is that the OSTC shows a longer red bar. The saturation/desaturation settings are 100/100.
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Ceiling on 2nd dive

Post by Ralph »

I get to believe that there is some misunderstanding somewhere or with something, but currently i have no clue what it is. What do you mean by "the OSTC shows a longer red bar"? Where / what is that bar? What also puzzles me is "the OSTC still has a 5 min stop at 3 m and the @+5 also shows no difference in TTS". If 'no difference in TTS' means @+5 is 5 minutes too, this would imply there is no off-gassing so no deco stop - what does not really make sense... Just to ask a silly question: we are sure the 5 min at 3 meters is a deco stop, not a safety stop?

In general, the accuracy of the OSTC has been enquestioned a couple of times, but in the end all doubts could be removed by better understanding the settings and their effects, sometimes also by better understanding what is shown on the display. There's a very good track record with heavy dives where the outputs of the OSTCs and the Shearwaters are very good in line with each other. I'd consider those as a good match partner who can be trust in.
User avatar
gr3yw0lf
Posts: 106
Joined: Thursday 28. November 2019, 22:18

Re: Ceiling on 2nd dive

Post by gr3yw0lf »

the last part I can confirm absolutely.
my shearwater and my ostc are in line the whole time. maybe they differ ±20cm depth and ±1min deco no matter of deeper or shallower dives. but for me thats a difference I can live with.
and the difference to subsurface calculation (checked in the diveplaner) is round about the same ±1min deko or maybe 1min more/less ndl.
the thing I wouldn't trust is the garmin (except the navi leading me to the divesite) 'cause of some years experience in diving missing compared to shearwater or hw.
if a garmin and a ostc showing completly different deco data (e.g. a 40min@45m wreck dive) I would go with the ostc (maybe after compareing with the runtime table)...
life is better in fins

OSTC Plus #18168
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Re: Ceiling on 2nd dive

Post by harv »

Yea, I think there some real misunderstanding with the data I'm trying to present. Might be that you're not familiar with how Subsurface displays the information. Or something else is wrong (I do notice that the order of the images is wrong when watching the thread under the reply editor compared when just watching it normally).

There are 3 graphs in my first post. The first one is from the first dive. Since I only could attach 3 images i used the OSTC, but the Garmin looks very similar. The red bar is the deco ceiling according to the dive computer, it ends at around minute 45. The green bar is the deco ceiling according to subsurface, it ends just before, at around minute 43. These are all reasonable differences. If a do a plan like 30 min at 30 m on Air on the Garmin and OSTC they end up with very similar run time and stops. The calculations for the first dive is not in question, they are all within the same ballpark. It's the 2nd dive that is the issue.

In the 2nd graph you can see Garmin log for the 2nd dive. The dive computer ceiling ends at minute 35 and the one calculated by Subsurface ends at around minute 37.

In the 3rd graph you can see the OSTC log for the 2nd dive. There the dive computer ceiling ends at minute 52 and the one calculated by Subsurface ends at around minute 35. Here is the log according to the OSTC itself:
IMG_20220414_200456.jpg
IMG_20220414_200510.jpg
IMG_20220414_200521.jpg
During the dive,as the Garmin was just clearing its deco obligation at around the 35 min mark, I was at a depth of around 8-9 m and the OSTC was still showing the same 5 minute stop that it had counted up to earlier during the dive and the @+5 showed the same TTS. According to the log it seems to take a bit more than 5 minutes to decrease the deco time by 1 minute, so it is off-gassing, but not fast enough for @+5. Very much slower than the off-gassing that Garmin and Subsurface calculated though.

I don't have access to a Shearwater, but if people have done repetitive dives with it and the OSTC and compared to the Subsurface calculations, then it would be very interesting to see. With less conservative settings you could compared the saturation at the end of the dive, instead of cleared deco ceiling, and see if it matches up. I got distracted when surfacing so only have the saturation from when the log was saved 10 minutes after surfacing. If it would turn out that Shearwater and OSTC largely agree and are this different from Subsurface and Garmin on the 2nd dive, then I would really like to hear about it.
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Ceiling on 2nd dive

Post by Ralph »

Well, i'm probably at my limits in helping you. I'm not that familir with subsurface, but what puzzles me is 3 things i cannot align:

- above you wrote you are doing NDL dives with not stops,

- but the OSTC logbook screenshot showes a deco obligation, i.e. it was a deco dive (at least with a GF high at 65%, with 100% original Bühlmann maybe not)

- i don't know exactly what the bar and its colors at the top edge of the subsurface plot mean, but as it ends abruptly it seems to be the 3 m deco stop and not a ceiling curve, so i see no information on how the ceiling depth evolved along the dive.

Where does the bar at the top edge of the subsurface plot origin from: from the dive computer log, or from Subsurface calculation?

The OSTC will clear the last deco stop when the ascent calculation yields that on arriving at the surface the leading tissue will be at GF high. Sometimes you may read that the leading tissue reaches GF high at the last stop, but what is wrong because when doing a linear projection from GF low at the first deco stop and GF high at the last, there will be a leading tissue's pressure > GF high at the surface.

As your dive seems to be a deco dive, with one stop only at 3 meters, your GF low of 50% will have locked onto the 3 meters stop and your surfacing was controlled by the GF high of 65%. As you stayed at 8-9 meters if i remember right, the extra 5 to 6 meters do slow down the offgassing which explains why the @+5 mins do not give a smaller number. The TTS always assumes you are starting your ascent now (or in @+X minutes for that 2nd deco calculation), i.e. go up to the indicated 3 meters in this case and then follow the issued deco plan. As you stayed deeper, you pushed the deco clearing with you towards later times. I don't know how other Bühlman with Gradient Factors implementations handle the case when the 1st stop is equal to the last stop regarding GF low / GF high application, but above is how the OSTC does it and it yields a smooth transition from NDL to deco.

This probably does not directly answer your questions, but maybe it provides some more bits into the understanding.

BR
Ralph
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Re: Ceiling on 2nd dive

Post by harv »

I still think there is some communication barrier here.

I wrote to "NDL" dives in my reply to gr3yw0lf since I thought his comment on my choice of gradient factors was pushing the discussion in the wrong direction. No recreational dive computer would consider these dives deco dives on default settings. According to the PADI RDP there would be a mandatory safety stop, I don't know if that makes it a deco dive or not, it doesn't really matter in this case.

I picked 50/65 as gradient factors since I knew that I then easily would get a 3-5 minute stop at 3 m and be able to see when that cleared on the different computers. So, according to the OSTC (and Garmin in this case) the dive was a "deco" dive. With this data I can then more easily compare the different Bühlmann implementations in Subsurface.

I could set 85 as GF high and there would be no deco. According to Subsurface I briefly reached a surface GF of 80 during ascent from depth but it was down to 79 by the time I reached 20 m. However, then it would be harder to compare the dive computers, especially afterwords, since the tissue data isn't available in the log.

The bar at the top edge has two sources. The red part comes from the dive computer log. The green one is calculated by Subsurface on the data in the log + it takes into account previous downloaded dives. You can see the the ceiling for individual tissues as curves earlier in the bar as darker shades. They quickly go down to almost 3 m and then slowly go back to the surface, with different tissues having the lowest ceiling. The green 3 m bar continues until all ceilings have reached the surface, but if you hover over the graph in Subsurface the actual deco calculation in the information window shows that deco is cleared a few minutes earlier and the target GF high will be reached when surfacing.

The question is really about tissue loading and off-gassing. You write that OSTC will clear the deco when it thinks that the target GF high will be reached upon surfacing. In the second dive this would happen at around 32-35 minute mark according to both Subsurface and Garmin (once I get to above 10 m it takes about 10 minutes to clear the 3 minute deco) , while for the OSTC it takes until minute 52 to reach the same GF high. So what I'm wondering is why the off-gassing seems to be so much slower at 8-10 m, or the tissue loading so much higher. It could of course be that both Subsurface and Garmin are way too fast in off-gassing or too low in tissue load, but with the data I have it's OSTC that stands out.
Ralph
Posts: 708
Joined: Saturday 24. June 2017, 11:31

Re: Ceiling on 2nd dive

Post by Ralph »

I don't know how Garmin nor Subsurface handle in detail the end-of-deco detection. My only help i can provide is to suggest to run some dives in the OSTC's build-in simulator and watch the custom view which shows the saturation % (actually it is the supersaturation value of the leading tissue), the ceiling (depth at which the leading tissue will touch GF high) and the tissue graphics to familiarize with the OSTC behaviour. The vertical lines in the graphics are GF low (green), GF high (yellow) and 100% (red). Maybe your're able to run the Garmin in parallel.
atdotde
Posts: 1
Joined: Tuesday 16. September 2014, 20:26

Re: Ceiling on 2nd dive

Post by atdotde »

Hi,

I feel responsible for the deco calculations in Subsurface. I can have a look at your dives if you export the two dives (or better all dives in the last 48 hours of the dive in question) to an xml file (you can use the anonymise function to black out any text) and mail those to helling@atdotde.de and I will have a look.

And @Ralph, you can contact me (do you have my number?) and we can try to track down the origin of the differences. It seems that Subsurface (and Garmin) and OSTC have very different ideas about tissue loadings or how to convert those to a surface GF in the second dive: They seem do differ by almost 20min about when the leading tissue has off gassed to a level that corresponds to GF65 at the surface. You should look at OSCT_second. The green curves are the ceilings for the different tissues, the green bar are those rounded up to the next multiple of 3m and the red bar is the ceiling as computed (and logged) by the OSTC.

Best
Robert
harv
Posts: 10
Joined: Sunday 16. January 2022, 11:20

Re: Ceiling on 2nd dive

Post by harv »

Thank you Robert. I set the dives through mail.
Post Reply