PDA

View Full Version : Suunto query



Ian smith
25-05-2015, 04:56 AM
Hi

Until recently I was diving with my 8 year old Suunto Vyper and my wife on her Gekko.
When she had the Gekko our NDL was broadly the same. She is now using her new D4i and it seems to be giving her between 10 and 15 minutes more bottom time. Not sure why. They are both Suunto, It's the same RGBM model (I think), both computers set to 100% RGBM and personal settings both at P0 (least conservative). We are diving in the Philippines at the moment and our profiles are very similar but of course she is in a strop because I'm cutting her bottom time down.

Any ideas? Has the RGBM model changed over the years to become less conservative?

any advice appreciated

Ian


Sent from my iPad using Tapatalk

smileydiver
25-05-2015, 06:00 AM
I'm guessing you're both set to air, or the same nitrox mix?

Nitnab Nhoj
25-05-2015, 06:12 AM
Ask Suunto but I bet the computers algorithm has been adjusted. I notice the difference between a D9 and D6i.

Pete Bullen
25-05-2015, 06:12 AM
Silly woman, tell her you always dive to the most conservative profile! End of argument :D (or start of a different one ;)

frogfone
25-05-2015, 06:16 AM
I have a Gekko and Roddy has the D4i. He often has more bottom time than me. He reckons as the air integrated computer takes your breathing rate into account therefore is more accurate. When we did the Beqa Lagoon shark dive I racked up 12 minutes of deco on my Gekko and he only had 4 on the D4i. At least it was a fantastic safety stop!

Sharon

Iain Smith
25-05-2015, 06:26 AM
Air integration shouldn't affect the rate at which your deco obligation racks up. If your bottom time is being limited by the amount of gas which your computer thinks you need, I hope it's taking into account the gas needed to get your buddy out of the water as well... [curtails rant about air integration being nothing more than an expensive way of monitoring SAC]

PS. "Deco" =/= "safety stop"


Sent from my iPad using Tapatalk

gobfish1
25-05-2015, 06:58 AM
Hi

Until recently I was diving with my 8 year old Suunto Vyper and my wife on her Gekko.
When she had the Gekko our NDL was broadly the same. She is now using her new D4i and it seems to be giving her between 10 and 15 minutes more bottom time. Not sure why. They are both Suunto, It's the same RGBM model (I think), both computers set to 100% RGBM and personal settings both at P0 (least conservative). We are diving in the Philippines at the moment and our profiles are very similar but of course she is in a strop because I'm cutting her bottom time down.

Any ideas? Has the RGBM model changed over the years to become less conservative?

any advice appreciated

Ian


Sent from my iPad using Tapatalk

Drop your RGBM to 50 see if you get a better match , or up her setting to old giffer mode ,
Sorry not much help , but if you fudge the setting at least you can do the same run times , :sweat:

Depth (m) P0/A0 times on d4i


9m 205min
12m 124min
15m 71min
18m 51m
21m 37min
24m 29min
27m 22min
30m 17min
33m 13mins

Up the po or ao and your diving the same dive
I'd up her to P1 and see how it go,s if still no joy P2 or A1 ,
A1 knocks a good lump of the ndl your down to 10mins for 30m so prob way to much ,
Stick with the P1 or P2 ,
I'd try the RGBM on your computer first ,

turnerjd
25-05-2015, 07:06 AM
It's just witchcraft.

I have now done a few dives with my Aladin 2G (Buhlmann ZH-l8 ADT MB PMG) also with mini-me's Cressi Leonardo (Weinke RGBM). There can be some quite big differences in NDLs, although I haven't yet taken the Cressi into propper deco territory..... But the differences are not exclusively one-way. As I would have expected from some of the comparisons I've seen, the RGBM often gives shorter NDLs and longer stops, however, on a couple of occasions I have had the inverse.

Now, my suspicion is that over the years I have played with some of the adaptive settings on the Aladin, and that I am not really comparing two similar things. I will have to have a look, but, anyway as the diving season has started, I won't get to dive with both computers for a while (or if I do I will have a pre-teanage strop to deal with after).

As for the difference between the different suuntos, it's almost certainly Bruce Weinke who (as all good scientists do) is playing with his model, tweaking it here and there, and making it more adaptive. You'll probably find that, in a similar manner to the Cressi/Aladin differences, there is some adaptive setting that is new, or differently set between the two.

Jon

Rob Dobson
25-05-2015, 07:15 AM
They will be running quite different algorithms. The term 'RGBM' is the cause of much confusion in this context and should simply be ignored. The term is simply licensed by Suunto to help market the computers. The Gecko will be running some proprietary algorithm derived from Haldane and the more recent computer a more sophisticated 'bubble model simulator' but again Suunto proprietary code. They will not be the same and therefore won't match.

Ian smith
25-05-2015, 07:31 AM
Thanks all lot of good comments there.

Smiley- yes both set to air

Nitnab - yes I think it might be worth trying to get a response from Suunto

Pete - do you mean start a new woman ??

Frog - did he stay with you for the other 8 mins. What a man!!

Gobfish - we are now in the old giffer zone so that might be a way forward.

Turner I think witchcraft just about sums it up. The other day we were diving with 2 Italians who I think we're using mares computers. I know that they were down deeper and longer on the first 2 dives yet on the third I was on 5 mins deco the guide was on 12 mins and back on board when I asked the Italian chap he said that he was no where near bottom limits. Makes you wonder who is correct. If their is a correct.


Sent from my iPad using Tapatalk

turnerjd
25-05-2015, 07:32 AM
They will be running quite different algorithms. The term 'RGBM' is the cause of much confusion in this context and should simply be ignored. The term is simply licensed by Suunto to help market the computers. The Gecko will be running some proprietary algorithm derived from Haldane and the more recent computer a more sophisticated 'bubble model simulator' but again Suunto proprietary code. They will not be the same and therefore won't match.

I thought that the Wienke / suunto RGBM was a true RGBM with additional Haldanean 9 compartment aspects modeling reduced off-gassing due to bubbles. Didn't they do this as a way to get a depth and deco stop limit?

Anyway, I'm sure the Suunto RGBM is more than just marketing -they've been throwing money at Bruce Wienke for years to access his work on RGBM algorithms.

Jon

gobfish1
25-05-2015, 08:41 AM
Iv just had a look at the viper numbers on P0 and A0
And the times given are all most the same as the other computer ,

So you should be getting the same ish dive ,
Make sure all your setting are the same ,

Not sure how old the info is, so could be reading info for a newer viper ,

Let us know if you sort out what's going on


I'd look and make sure your both on 100 RGBM ,

Ian smith
25-05-2015, 10:01 AM
Checked and both on 100% RGBM. I think I'll fire a query off to Sunnto when I get back. Maybe it's time to upgrade but I'd still like to know why it so different. As some have said maybe the model has been tweeked over the years. I've seen comments in the past that Suunto are too conservative, maybe this is part of there response.

Thanks to Rob also. I just saw your post. I wasn't aware of that. If I understand what your saying RGBM is just an umbrella, underneath you could have computers using various algorithms and codes as developments occur. Maybe 8 years is too long to keep one computer even if it has been a good one.


Sent from my iPad using Tapatalk

gobfish1
25-05-2015, 10:11 AM
Checked and both on 100% RGBM. I think I'll fire a query off to Sunnto when I get back. Maybe it's time to upgrade but I'd still like to know why it so different. As some have said maybe the model has been tweeked over the years. I've seen comments in the past that Suunto are too conservative, maybe this is part of there response.


Sent from my iPad using Tapatalk

Had a look for a old viper user guide , and the ndl time are all most the same , your viper give,s 18mins at 30m
And the d4i gives 17mins so god knows what's going on ,

Old viper numbers
9m all day
12m 124min
15m 72min
18m 52min
21m 37min
24m 29min
27m 23min
30m 18min
33m 13min
Enjoy your diving , and I'd like to know what suunto have to say ,

Ian smith
25-05-2015, 10:17 AM
If I get a reply from Suunto I'll post it. I wish I had known at the beginning of the holiday. When we arrived at the first resort there was a chap there who worked for Suunto on the D4i, he saw Karen fiddling with it and came and gave her some info on navigating around it.


Sent from my iPad using Tapatalk

gobfish1
25-05-2015, 10:38 AM
I thought that the Wienke / suunto RGBM was a true RGBM with additional Haldanean 9 compartment aspects modeling reduced off-gassing due to bubbles. Didn't they do this as a way to get a depth and deco stop limit?

Anyway, I'm sure the Suunto RGBM is more than just marketing -they've been throwing money at Bruce Wienke for years to access his work on RGBM algorithms.

Jon
RGBM
From what I remember was more about the repeat dives , and also would slap your hand on a 2nd dive if you had been arsing about on the first dive , ie coming up to fast or a shitty profile , I all ways had mine on 50 as I was all ways getting them blue bad marks on my dive profiles lol .

Rob Dobson
25-05-2015, 11:02 AM
Thanks to Rob also. I just saw your post. I wasn't aware of that. If I understand what your saying RGBM is just an umbrella, underneath you could have computers using various algorithms and codes as developments occur. Maybe 8 years is too long to keep one computer even if it has been a good one.


Not quite. RGBM is a very sophisticated bubble model and massively resource intensive to compute. 'Suunto RGBM' and various other such phrases are effectively marketing terms which relate to propriety algorithms which Suunto have implemented in their products over the years.

True RGBM requires massive CPU resources and anyone who thinks that the humble Gecko dive computer is actually running RGBM under the hood is misguided to say the least.

I'm not saying that the algorithms are bad or that the numbers are in any way invalid or anything like that, purely that the use of the term RGBM in this context is quite meaningless and probably misleading.

I guess that is the point of marketing at the end of the day.

Ron MacRae
25-05-2015, 11:53 AM
Two divers, buddied together, don't do exactly the same dive. Perhaps one is habitually just a bit lower, or a bit more saw tooth, or perhaps one computer is on an arm that is moving up and down a lot.
I carry 2 Suunto computers on the same arm and they don't always agree on the depth, sometimes by 0.2m.

I wouldn't worry about it and first one to reach the planned deco or gas limit calls the dive.

Baron015
25-05-2015, 01:43 PM
Not quite. RGBM is a very sophisticated bubble model and massively resource intensive to compute. 'Suunto RGBM' and various other such phrases are effectively marketing terms which relate to propriety algorithms which Suunto have implemented in their products over the years.

True RGBM requires massive CPU resources and anyone who thinks that the humble Gecko dive computer is actually running RGBM under the hood is misguided to say the least.

I'm not saying that the algorithms are bad or that the numbers are in any way invalid or anything like that, purely that the use of the term RGBM in this context is quite meaningless and probably misleading.

I guess that is the point of marketing at the end of the day.

Explanation from Suunto here

http://ns.suunto.com/pdf/Suunto_Dive_Fused_RGBM_brochure_EN.pdf

Rob Dobson
25-05-2015, 02:45 PM
Explanation from Suunto here

http://ns.suunto.com/pdf/Suunto_Dive_Fused_RGBM_brochure_EN.pdf

Explanation...? I guess it states which generation of their model is in which product which is nice, but if they had just called them 1st, 2nd, 3rd and 4th etc then it would have made things less confused. The insistence of using RGBM after the different pre-fixes implies that RGBM is actually being computed under the hood instead of simply being emulated. I'm not saying that you need it or even that it would be desirable to crunch actual RGBM on a dive computer (inefficient use of energy for a start), but the use of the phrase makes it a muddled discussion. Unnecessarily so IMO.

Iain Smith
25-05-2015, 03:26 PM
Explanation from Suunto here

http://ns.suunto.com/pdf/Suunto_Dive_Fused_RGBM_brochure_EN.pdf


When deep stops (Pyle stops) became a proven beneficial method for decompression,
they were added to the Suunto D9 and Vytec DS (2004) as a voluntary notification

I wonder if by "proven beneficial", Suunto mean "widely used" or, perhaps more charitably, "proven beneficial (by Level 5 (aka "GOBSAT"*)) evidence"

*GOBSAT": Good Old Boys Sat Around Table (more politely referred to as "Expert Opinion")

Tris
25-05-2015, 05:47 PM
Does she have deep stops enabled?

If you dont get to the bottom of it easiest solution is treat yourself to a new Suunto d4i dive computer as well.

Ian smith
27-05-2015, 01:11 PM
Hi just got back so a little jet lagged. That's an interesting article. So it looks like it's because the new computers use the, so called Suunto Fused RGBM model. And it may be time for me to upgrade. However although the timeline in the article suggests computers after 2012 have this algorithm it doesn't say so on the spec for her computer. It's all a little vague. Do all suunto's now use Suunto Fused or just some of them?


Sent from my iPad using Tapatalk

Tris
27-05-2015, 04:09 PM
Just particular models I believe Ian, Anything coming out of the factory recently I would suspect may have it. But I think its only in the computers more likely to go Tech diving. (D9tx-DX) Have you not had a response from Suunto yet? Best to ask direct as there is alot of confusion online

Ian smith
27-05-2015, 07:01 PM
Just sent the query off tonight. I'll post there reply.


Sent from my iPad using Tapatalk

DaveBarber
01-06-2015, 08:09 AM
Do you really want a computer that brings you out quicker?

Its just allowing you to walk closer to the edge.

AxeMan
01-06-2015, 08:49 AM
Air integration shouldn't affect the rate at which your deco obligation racks up.
It used to on the old Aladin Air Xs. The justification was that if you're breathing more, you're exerting yourself more. The trouble is that the computer doesn't actually know your air consumption because it didn't know how big your tanks are.

AxeMan
01-06-2015, 08:52 AM
True RGBM requires massive CPU resources and anyone who thinks that the humble Gecko dive computer is actually running RGBM under the hood is misguided to say the least.

I used to know someone who had an HS Explorer, which implemented RGBM. It was one of the first trimix computers. He had to send it back more than once.

nickb
01-06-2015, 09:59 AM
Explanation from Suunto here

http://ns.suunto.com/pdf/Suunto_Dive_Fused_RGBM_brochure_EN.pdfHmmm....

"When deep stops (Pyle stops) became a proven beneficial method for de-compression, they were added to the Suunto D9 and Vytec DS (2004) as a voluntary notification." I should have stopped reading right there.

A lot of people I know have started substantially increasing their low GF, yet the profile on page 12 has what are essentially stops being started a lot deeper than even a 10 or 20 low GF.

Tris
02-06-2015, 03:28 PM
Do you really want a computer that brings you out quicker?

Its just allowing you to walk closer to the edge.

Or is it? Is the algorithm getting you out earlier by not exposing you to additional unnecessary stops that increase nitrogen onload, so by getting out earlier your infact safer? A computer that is over conservative may infact onload more nitrogen.

AxeMan
02-06-2015, 04:23 PM
Do you really want a computer that brings you out quicker?

Its just allowing you to walk closer to the edge.

Is walking 20m away from the edge any safer than walking 10m away?

Garf
03-06-2015, 06:49 AM
If nobody can actually tell you precisely where the edge is, and if the position of the edge changes for every walk, then yes it is, very much so.


Sent from my iPad using Tapatalk HD

rongoodman
04-06-2015, 01:08 PM
Not quite. RGBM is a very sophisticated bubble model and massively resource intensive to compute. 'Suunto RGBM' and various other such phrases are effectively marketing terms which relate to propriety algorithms which Suunto have implemented in their products over the years.

True RGBM requires massive CPU resources and anyone who thinks that the humble Gecko dive computer is actually running RGBM under the hood is misguided to say the least.

There is a third party version of RGBM available for the Liquivision Xeo which is claimed to be a full implementation. I don't think I've ever seen any mention of it from an actual user though.

Ian smith
07-06-2015, 07:16 PM
Finally got a response from Sunnto on this. Copied below. Can't say I'm overly impressed. It wasn't particularly in depth and they referred to a zoop when I told them I had a Vyper. I tend to agree with the idea that the algorithm has changed over time.

Over the years I've often been involved with assessing equipment used for analysis and measurement and it always amazes me when you perform a controlled side by side evaluation and both machines give different results for a control assay, the manufacturers will always tell you that there machine is correct my argument to them is simple, they both can't be correct. I realise that it is very difficult for a diver to perform this type of assessment as your buddy's profile will always be slightly different to yours and because results are based on theoretical algorithms which differ from computer to computer. However I wouldn't mind betting that if you strapped two different computers using the same model to your wrist you would see different results.

I'm tapped out now but I think I'll try and get a new one before the next trip. It's been a good one but maybe it's time to update.

Thank you for contacting Suunto Customer Support.

Please accept our apologies for the belated reply.

Our technicians believe that it could be about the recording sample rate of the devices. This is most likely the case. Zoop records 30 s whereas with the D4i it is possible to setup to use 1s-30s.

Also the algorithms are a bit different since Zoop doesn't utilize deepstops and D4i does, but this really shouldn't change the NDL.

Hope this answers your questions. Have a nice day!

JimmE
08-06-2015, 06:31 AM
Whilst they haven't helped themselves referring to a zoop, on the Vyper you can also adjust the sample rate, so it could be partly responsible for the difference. Presumably you have also checked the conservatism setting on each?


Sent from my iPhone using Tapatalk

Tris
08-06-2015, 11:59 AM
Check your settings and perhaps dive with them both on your wrist to see how the results differ then. If they remain well out I'd send a follow up letter to Suunto pointing out the settings and mis-match if it remains.

jturner
08-06-2015, 12:17 PM
Whilst they haven't helped themselves referring to a zoop, on the Vyper you can also adjust the sample rate, so it could be partly responsible for the difference. Presumably you have also checked the conservatism setting on each?

The sample rate that you can adjust only affects what is recorded in the log book of the device, not the real time calculations, which is why the logs can differ from what you actually saw on a dive. However, it could be that the calculation sample rates that you cannot adjust also differ between the two devices too.

OrangeHead
08-06-2015, 12:33 PM
It seems to me that unless the sampling intervals are very small, and even if the rates are the same, the samples will not be synchronised and so the two devices may record significantly different profiles. The example that leaps to mind is that of momentarily raising an arm over your head - triggering an ascent rate warning on one computer - but returning to trim before the next sample occurs on the other.

Tris
08-06-2015, 01:13 PM
^ Sounds most likely to me what Orangehead has posted.

AxeMan
08-06-2015, 01:39 PM
It seems to me that unless the sampling intervals are very small, and even if the rates are the same, the samples will not be synchronised and so the two devices may record significantly different profiles. The example that leaps to mind is that of momentarily raising an arm over your head - triggering an ascent rate warning on one computer - but returning to trim before the next sample occurs on the other.

You would still need to do a lot of see sawing up and down to affect the NDL by 10-15 mins as in the original post. I'm assuming that this was a significant difference and not one was giving 200 mins NDL and the other 190 at a shallow depth.

drysuitdiver
08-06-2015, 01:53 PM
If nobody can actually tell you precisely where the edge is, and if the position of the edge changes for every walk, then yes it is, very much so.


Sent from my iPad using Tapatalk HD


if we don't know where the edge is and it changes every time we make the walk then it makes sense to not walk anywhere the edge could possibly be.

but I am not going to stop diving.

jturner
09-06-2015, 07:19 AM
It seems to me that unless the sampling intervals are very small, and even if the rates are the same, the samples will not be synchronised and so the two devices may record significantly different profiles. The example that leaps to mind is that of momentarily raising an arm over your head - triggering an ascent rate warning on one computer - but returning to trim before the next sample occurs on the other.

IIRC, the sampling rates for real-time calculation are very small (<1s) but I can't recall where I found this information, so cannot check and thus I might be wrong. Either way, 10-15mins seems to be a great difference. FWIW, my brother and his Mrs have 2 x D6 (one each) and 1 D6i (spare, carried by one of them... odd setup but there you go!) and they also often get a difference between the two displayed NDLs carried by one diver, often more than they would expect. The D6i uses a more modern implementation so my guess is there's something in that configuration that is causing this difference... deep stops maybe providing "credit" or something. ATEOD, they just followed the more conservative and didn't worry about it further.

dmcann
12-06-2015, 03:59 AM
From my understanding, with different models of Suunto Computers, the algorithms are updated, most of the time this ends up with providing the diver with a longer NDL.

I have buddies with Gekkos and Zoops. I use a D4i Novo and some of my friends use D4is too. The D4i definitely provides longer NDLs. The Gekkos and Zoops are more conservative.

Hope this helps.

Cheers,
Denise

Ken
12-06-2015, 06:16 AM
I dive with a Zoop and a Helo2. For single gas Nitrox dives they always line up within a minute for stops. Most of my buddies dive Suuntos of some sort. The only time they don't match on stops is if we are on different mixes or did different first dives. Note that some display the stops differently, for example the Helo2 shows the actual stop time and then the safety stop time rather than the sum of the two as the Zoop does.

The Suunto PC software allows for planning for a particular model of computer. You can try out plans for dives which exhibit the issue there to see if it claims any difference. You would have to play the high low game with the times though,mint calculates stops not NDL.

Remember that NDL is very sensitive to depth and pressure sensors do vary.

drysuitdiver
12-06-2015, 10:45 AM
I dive with a Zoop and a Helo2. For single gas Nitrox dives they always line up within a minute for stops. Most of my buddies dive Suuntos of some sort. The only time they don't match on stops is if we are on different mixes or did different first dives. Note that some display the stops differently, for example the Helo2 shows the actual stop time and then the safety stop time rather than the sum of the two as the Zoop does.

The Suunto PC software allows for planning for a particular model of computer. You can try out plans for dives which exhibit the issue there to see if it claims any difference. You would have to play the high low game with the times though,mint calculates stops not NDL.

Remember that NDL is very sensitive to depth and pressure sensors do vary.


they do vary. I did a dry dive last night to 50m my stinger recorded 49.9 apparently the pot gauge is highly calibrated so my stinger must be out. 2 other suuntos in there also read different 50.6 and 50.7

good thing is that at 21 12 and 6 it read bang on.

Ron MacRae
12-06-2015, 11:46 AM
they do vary. I did a dry dive last night to 50m my stinger recorded 49.9 apparently the pot gauge is highly calibrated so my stinger must be out. 2 other suuntos in there also read different 50.6 and 50.7

good thing is that at 21 12 and 6 it read bang on.

How accurate do you think these things are? I'm surprised they were all within 1m of the correct pressure.

AndrewRawlingson
12-06-2015, 12:05 PM
My two HelO2s are always within 0.1 metre of each other.

Suunto algorithms are a bit of mystery and a point of interest, but I am as confident in them as anything else that they will get me out the water in reasonable shape. The HelO2s don't "demand" that you get to 3 metres ASAP and I don't find them overly conservative compared to other algorithms.

I dived HelO2 and an old Vytec/Gekko. First dive of a series and they matched up until the ascent on 40 metre dives, then it all changed. Depending on depth, the HelO2 started clearing deco, the other two just racked up more deco. I think this is in keeping with what Suunto say.

gobfish1
12-06-2015, 06:26 PM
they do vary. I did a dry dive last night to 50m my stinger recorded 49.9 apparently the pot gauge is highly calibrated so my stinger must be out. 2 other suuntos in there also read different 50.6 and 50.7

good thing is that at 21 12 and 6 it read bang on.

id think most computers have some fudge in on what depth its working your deco from ,
a few of my desk top planner s add .03 per m of depth to the real depth to put you on the safe side ,

I used to run a 85m table for a 90mfw dive at dotty , by the time my computer did its numbers plus the fudge id have a 87m table