SAC Graph?

[550Maranelloman]

Hi guys, newbie here so please be gentle?

I've just discovered Subsurface and I know I have a LOT to learn, but tonite I'm searching for one thing in particular and that's how to display SAC graphically?

After reading through a number of threads, it appears it's not possible, mainly because the graph would look too "busy", however perhaps a logarithmic scale could solve this?

My fiancee is a beginning OW diver and I simply want to show her her breathing rate graphically and coach her how to improve it.

Thanks in advance for any thoughts/suggestions and a special thanks to Linus for authoring another piece of awesome software. 

John

[Dirk Hohndel]

The coloring of the pressure line for your tank gives you an indication where your SAC was low (darker greens) and where it was high (red-ish)

That gives you a good indication where on the dive you were using a lot of air, and were you weren't.

you can tell here that there were a few specific points during that dive where my consumption went up... that first time about 15 minutes into the dive is when I was working on taking a picture of a ghost shrimp in a bit of current :)

I hope this helps

/D

[Linus Torvalds]

On Wed, Jan 5, 2022 at 3:55 PM 550Maranelloman <john....@gmail.com> wrote:
>
> I've just discovered Subsurface and I know I have a LOT to learn, but tonite I'm searching for one thing in particular and that's how to display SAC graphically?

If you mean per dive, then that's actually already done - by color of
the cylinder pressure line. It obviously requires a pressure sensor on
your cylinder to work right..

It's all relative to the overall SAC rate for the dive, but red means
"higher momentary SAC than for the dive in general", while dark green
means "lower momentary SAC".

You can definitely see "I was swimming away in a panic from the very
aggressive puffer fish" in the graph that way, when your SAC rate goes
up a lot for a portion of the dive.

One note: it's not very precise - you'll often see the SAC rate graph
being red at the beginning of the dive because the pressure is going
down as the gas cylinder is cooling down (rather than you necessarily
breathing it down). So you can only really see big changes in it.

Linus

[rick.warren4]

Just a maybe not so obvious comment on the line color going red. 

As you descend you need more volume of gas in your lungs to fill them up (to take a normal breath at that depth). Since the line color seems to change as the momentary SAC increases from the running average (speculation on my part),  it would make sense that if you descend at a fast rate, you will be taking in more volume per minute as you take on additional gas volume to fill your lungs. Once you reach the desired depth, the rate will stabilize again and the running average will start approaching the new rate at that depth.

If you make a very slow descent the momentary SAC is not deviating enough from the running average to change the line color red.

This is all speculation on my part but, that seems to be what I am noticing in my profiles.

I am more aware of this since I dive a rebreather and I am constantly adding gas to my lungs (counter and physical) as I descend or else it becomes hard to breath as my lungs are squeezed.

Rick

--
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/5283D82B-0106-4729-91EE-837696FA1E49%40hohndel.org.

[550Maranelloman]

Thank you everyone for the insightful & helpful responses! They're very much appreciated and yes, I wondered if the change in color of the pressure line was from SAC changing, but wanted confirmation.

Just wish I could graph the actual SAC numbers to show my fiancee. Are they embedded somewhere in the XML log file or is it strictly a calculation for the screen?

[550Maranelloman]

I know it's lame replying to one's own post, but after thinking about it, I guess there's nothing to prevent me from importing the dive log to Excel and constructing my own SAC graph?

[Robert C. Helling]

Do you realise that in the info box, you can read off the numerical value of the SAC graph. But please be aware that due to the sample interval and precision of the tank pressure sensors, those values don't have much meaning on short time scales (taken to the extreme, between breaths the SAC is 0 while during breaths it is much higher. So you should at least average over several breaths which can be a handful of minutes). 

If you really wanted to compute the SAC rate, you need to take into account besides the depth and the time interval also the compressibility of the gas you are breathing as that can have a significant influence). But you don't have to do the calculation for yourself to get a graph: Export the dive using the "CSV using the profile data" and there will be a column containing the momentary SAC>

[Dirk Hohndel]

Please keep in mind what Linus mentioned... there are a lot of facts that go into your SAC numbers (e.g., temperature of the cylinder). And given the granularity of data that most dive computers provide, plus the margin of error of the pressure readings... I would caution you to read too much into the short term changes to SAC.

I don't think that a SAC graph really would give you a lot of useful data - but of course, you could calculate that info very easily, just the same way the info overlay does.

/D

[Jason Bramwell]

This is the info box rendition of SAC that Robert mentioned.

 

 

Sent from Mail for Windows

--

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/3DDF3A4C-7A0E-4BA1-89AE-055CDC0A25A3%40hohndel.org.

 

[John Cribb]

Thanks Dirk; I only want to give the fiancée a “macro” view of her gas consumption and overlay it with mine to show her how/why she’s sucking her tank down to 30 bar in 30 minutes and I’m still at 125 on the same dive, LOL.

 

From: subsurfac...@googlegroups.com <subsurfac...@googlegroups.com> On Behalf Of Dirk Hohndel
Sent: Thursday, January 6, 2022 12:15 PM
To: Subsurface Divelog <subsurfac...@googlegroups.com>
Subject: Re: SAC Graph?

 

Please keep in mind what Linus mentioned... there are a lot of facts that go into your SAC numbers (e.g., temperature of the cylinder). And given the granularity of data that most dive computers provide, plus the margin of error of the pressure readings... I would caution you to read too much into the short term changes to SAC.

--
You received this message because you are subscribed to a topic in the Google Groups "Subsurface Divelog" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/subsurface-divelog/OVp5dRDnceA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to subsurface-dive...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/subsurface-divelog/3DDF3A4C-7A0E-4BA1-89AE-055CDC0A25A3%40hohndel.org.

[Linus Torvalds]

On Thu, Jan 6, 2022 at 11:25 AM John Cribb <john....@gmail.com> wrote:
>
> Thanks Dirk; I only want to give the fiancée a “macro” view of her gas consumption and overlay it with mine to show her how/why she’s sucking her tank down to 30 bar in 30 minutes and I’m still at 125 on the same dive, LOL.

You can also use the "ruler" thing to see the SAC rate for a particular segment of the dive, something like this:

which gives you better information. Use the color of the pressure graph to find the rough area you're interested in, and then the ruler to give information about that section of the dive (obviously not just SAC - time difference, depth difference, etc etc).

                               Linus

[may...@gmail.com]

I'd love to second the request for a graph of the SAC rate:

For me it's much more intuitive to grasp and draw conclusions out of an graph than looking at the color coding.

The Garmin Dive app of my computer shows such a graph:

In comparison the same dive in Subsurface:

I could easily see in the Garmin app how severe that stress was that I had at ~50 minutes when I wanted to ascent (caused by a stupid mishap by myself). 

In Subsurface I do see that the color got red and stayed so for about three minutes. But it's just red. 

But between 20 and 25 minutes the SAC rate also left the green shades and became nearly red, looking like an area needing further investigation. Looking at the graph of the Garmin app instead I can easily see that it was just a time of slightly higher consumption that easily fits to the time before and thus doesn't bear any relevance as it's most likely just an effect caused by different current strengths.

About the short time changes of the SAC rate and the noise it might show: that's just a question of a sensible selection of the low pass filter.

[rick.warren4]

Go to stats and you can create graphs of SAC over time, Temps or many other factors.

-------- Original message --------

From: "may...@gmail.com" <may...@gmail.com>

Date: 6/15/22 19:11 (GMT-05:00)

To: Subsurface Divelog <subsurfac...@googlegroups.com>

Subject: Re: SAC Graph?

--

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/488c5728-ef17-4b58-b129-eec5ab3ab49bn%40googlegroups.com.

[rick.warren4]

Go to stats and you can create graphs of SAC over time, Temps or many other factors.

-------- Original message --------

From: "may...@gmail.com" <may...@gmail.com>

Date: 6/15/22 19:11 (GMT-05:00)

To: Subsurface Divelog <subsurfac...@googlegroups.com>

Subject: Re: SAC Graph?

--

[Gmail im Auftrag von Martin Gröger]

hm..

good idea - but when I try this with the sac subsurface suddenly closes (repeatable)

 (win11 latest, subsurface 5.0.8, no error-log) 

To view this discussion on the web visit https://groups.google.com/d/msgid/subsurface-divelog/62aab71b.1c69fb81.d68c6.16b1%40mx.google.com.

[Robert C. Helling]

I wonder how the garmin app does this: How fine is their tank pressure measured? At 15m of depth with a SAC rate of say 15 bar l/min and a 12l cylinder, the pressure will drop by roughly 3bar per minute. The Garmin plot seems to give several values per minute (close to the granularity of individual breaths!) and to display different values for SAC it has to have a precision of fractions of a bar (on a sensor rated to probably more than 300bar). Would you mind exporting such a dive and sharing it with me (hel...@atdotde.de) so I can have a look at the type of data that comes out of the Garmin pressure sensor (of course it is impossible to plot data that is not there)?

best

Robert

[Mark Stiebel]

The Garmin reports tank pressure every five seconds to the 100th of a bar (is it accurate to 100th of a bar, who knows). Here's a snippet of a Subsurface export of one of my dives with two sensors (pressure0 is me, pressure1 is my buddy). Hmm.. looks like I went more than 5 seconds without taking a breath there, so plotting that exactly would show my SAC dropping down to 0.

To graph SAC, you'd probably want to do something like graph a moving average over a handful of samples - say 6 samples for 30 second granularity.

  <sample time='28:08 min' depth='21.002 m' ndl='2:50 min' heartbeat='65' />
  <sample time='28:09 min' depth='21.028 m' pressure0='101.14 bar' ndl='2:48 min' />
  <sample time='28:10 min' depth='21.072 m' pressure1='98.93 bar' ndl='2:46 min' tts='2:18 min' heartbeat='67' />
  <sample time='28:11 min' depth='21.139 m' ndl='2:43 min' heartbeat='69' />
  <sample time='28:12 min' depth='20.987 m' ndl='2:47 min' tts='2:17 min' heartbeat='70' />
  <sample time='28:13 min' depth='20.981 m' ndl='2:46 min' heartbeat='71' />
  <sample time='28:14 min' depth='20.984 m' pressure0='100.66 bar' ndl='2:45 min' />
  <sample time='28:15 min' depth='20.948 m' pressure1='98.66 bar' />
  <sample time='28:16 min' depth='20.931 m' ndl='2:44 min' />
  <sample time='28:17 min' depth='20.864 m' ndl='2:46 min' tts='2:16 min' />
  <sample time='28:18 min' depth='20.896 m' ndl='2:43 min' tts='2:17 min' />
  <sample time='28:19 min' depth='20.878 m' pressure0='100.66 bar' tts='2:16 min' />
  <sample time='28:20 min' depth='20.868 m' pressure1='98.38 bar' ndl='2:42 min' />

On Thursday, 16 June 2022 at 18:57:38 UTC+10 Robert C. Helling wrote:

I wonder how the garmin app does this: How fine is their tank pressure measured? At 15m of depth with a SAC rate of say 15 bar l/min and a 12l cylinder, the pressure will drop by roughly 3bar per minute. The Garmin plot seems to give several values per minute (close to the granularity of individual breaths!) and to display different values for SAC it has to have a precision of fractions of a bar (on a sensor rated to probably more than 300bar). Would you mind exporting such a dive and sharing it with me (hel...@atdotde.de) so I can have a look at the type of data that comes out of the Garmin pressure sensor (of course it is impossible to plot data that is not there)?

best

Robert

may...@gmail.com schrieb am Donnerstag, 16. Juni 2022 um 01:11:37 UTC+2:

I'd love to second the request for a graph of the SAC rate:

For me it's much more intuitive to grasp and draw conclusions out of an graph than looking at the color coding.

The Garmin Dive app of my computer shows such a graph:

In comparison the same dive in Subsurface:

[Robert C. Helling]

I have done a similar thing: Here is a plot of sample pressures over time: 

There is some variation of slope (which would correspond to SAC) but not much. You can also plot pressure differences between samples (those would directly measure SAC):

As you can see, there is quite some noise and negative values do occur. Also, there seem to be a preference for multiples of 0.07bar which might be an indication of the granularity of the sensor (that the values are reported with two decimal places does not indicate the precision obviously).

I also did a histogram of how often the different differences occur:

So, yes, this clearly needs some smoothing. I must say I am still not convinced this actually contains more information (as opposed to data) than what we show in the color coding.

Best

Robert

[may...@gmail.com]

Now I've also played a bit with the raw data.

Here you can see one blue dot for each available pressure measurement, the green line is current tank pressure in comparison to that 8 data points in the past (i.e. roughly one minute ago, as points are not always 5 seconds long) and the purple one doing that for 2 minutes:

Setting the axis scale to the range of the Garmin app, the purple line looks very similar to that of the app (it seems the app might be smoothing even a bit longer):

And to see whether this can work for other dive computers as well I took the measurements of exactly the same dive from the Suunto D4i that my buddy was wearing. The the raw data is evenly spaced with one record every 20 seconds, so it's computed by comparing the pressure to that 3 / 6 data points in the past:

My conclusion:

Such a graph works fine for multiple dive computers. It also needs smoothing by calculating the SAC rate not between the last two measurements but in comparison to that a bit in the past. Using 2 - 3 minutes for this averaging will give good results.

Whether such a graph is more useful than the color coding will most likely depend on the user (taste, but I can also imagine that color blind people find it much more easy to use) and the use case (color for a quick check or a graph for a detailed debrief of the dive)

Kind regards,

Chris