Shearwater Perdix AI

[Lorenzoid]

You can make it to be, if you want to. Just play with GF.

it's possible that there is no Buhlmann GF setting that will closely mimic Suunto RGBM's behavior over consecutive days of repetitive diving. Suunto RGBM seems to give extra weight to the mere fact of multiple consecutive days of diving. Of course, the overnight surface interval probably matters. I know from experience that if I do a night dive after a full day of diving, the next morning the Suunto behaves more conservatively on the first dive (and probably the next dive, etc.--I haven't really paid attention). I haven't compared it with my Shearwater.

 

[scubadada]

Hi @MomZo

Based on 2016 ScubaLab testing results, it is obvious that Suunto RGBM allows lower NDL time than Buhlmann GF algorithm while doing repetitive dives (but not for the first dive)
Thank you for pointing me in that direction as it was exactly the kind of data I was interested in.
I am about to buy a Perdix soon and I will set it to OC REC medium conservatism in a first time and compare NDL with my Zoop that I keep in backup.

it's possible that there is no Buhlmann GF setting that will closely mimic Suunto RGBM's behavior over consecutive days of repetitive diving. Suunto RGBM seems to give extra weight to the mere fact of multiple consecutive days of diving. Of course, the overnight surface interval probably matters. I know from experience that if I do a night dive after a full day of diving, the next morning the Suunto behaves more conservatively on the first dive (and probably the next dive, etc.--I haven't really paid attention). I haven't compared it with my Shearwater.

In fact, you will not be able to find a single Buhlmann GF hi that will mimick Suunto RGBM. Remember, it is just the GF hi that determines surfacing for a no deco dive. The GF lo does not kick in until you exceed NDL, it then determines the depth of the 1st stop.

As you have already seen from the 2016 ScubaLab data, Suunto has pretty long 1st dive NDLs, just shorter than DSAT and longer than 45/95 until the last stop. You will also see that Suunto is more affected by the 1h SIs after dives 1 and 3 than DSAT or Buhlmann. However, Suunto recovers significantly after the 2h SI following dive 2 and has relatively longer NDLs for dive 3. The ScubaLab profiles are very straightforward and include normal ascent rates and SIs. They do not bring out any of the peculiarities that have been previously discussed for Suunto RGBM. Real life use may be more variable.

I have been diving a Dive Rite Nitek Q, running ZH-L16C, for more than 1 1/2 years/nearly 300 dives, as a backup to my Oceanic VT3, running DSAT. It has become very clear to me that no single GF hi is able to closely mimick DSAT. Some of the differences in these two algorithms are illustrated in the 2017 ScubaLab test results. I have ended up diving my Nitek Q at a GF hi of 90-100, depending on dive number (1st or subsequent), depth, and expected profile. To be honest, I'm not sure what GF I would choose if switching over permanently from DSAT to Buhlmann. I have very considerable experience with DSAT over the last 16 years. Each deco algorithm appears to have its own unique nuances that makes it different.

Let us know your experience after you've had the opportunity to dive Buhlmann and Suunto side by side for a while.

Good diving, Craig

 

[MomZo]

With Suunto RGBM, it definitely makes it more and more conservative after few daily dives, and even more after few consecutive diving days.
I used to go on a diving week 3 times per year, diving 3 or 4 times a day, for 6 or 7 days. After 1 or 2 days, my NDL is always lower than the one of my buddies, and I often have to ask them to go much shallower or I would go in Deco. The NDL difference is sometimes so huge...

 

[MomZo]

Hi @MomZo





In fact, you will not be able to find a single Buhlmann GF hi that will mimick Suunto RGBM. Remember, it is just the GF hi that determines surfacing for a no deco dive. The GF lo does not kick in until you exceed NDL, it then determines the depth of the 1st stop.

As you have already seen from the 2016 ScubaLab data, Suunto has pretty long 1st dive NDLs, just shorter than DSAT and longer than 45/95 until the last stop. You will also see that Suunto is more affected by the 1h SIs after dives 1 and 3 than DSAT or Buhlmann. However, Suunto recovers significantly after the 2h SI following dive 2 and has relatively longer NDLs for dive 3. The ScubaLab profiles are very straightforward and include normal ascent rates and SIs. They do not bring out any of the peculiarities that have been previously discussed for Suunto RGBM. Real life use may be more variable.

I have been diving a Dive Rite Nitek Q, running ZH-L16C, for more than 1 1/2 years/nearly 300 dives, as a backup to my Oceanic VT3, running DSAT. It has become very clear to me that no single GF hi is able to closely mimick DSAT. Some of the differences in these two algorithms are illustrated in the 2017 ScubaLab test results. I have ended up diving my Nitek Q at a GF hi of 90-100, depending on dive number (1st or subsequent), depth, and expected profile. To be honest, I'm not sure what GF I would choose if switching over permanently from DSAT to Buhlmann. I have very considerable experience with DSAT over the last 16 years. Each deco algorithm appears to have its own unique nuances that makes it unique.

Let us know your experience after you've had the opportunity to dive Buhlmann and Suunto side by side for a while.

Good diving, Craig

Thanks Craig for this detailed answer.
I am diving Suunto for 5 years now, and for my recreational diving, the Zoop is really a good computer. RGBM algorithm is quite good, I believe, to keep you safe, but I have to admit that it is bothering during a diving week when you always have to tell your buddies that you have to go shallower or it would become a Deco dive, making the computer even more conservative after (vicious circle)
That is the main reason I want to switch to another computer which is more liberal. I agree that the Perdix might be a bit "too much" for my type of diving at the moment, but it is nice sometimes to offer yourself some great gift, and I think it is a great computer that I could use for many years, allowing me to continue my diving learning to more tek dives if I want to.
I will continue to dive with my Zoop as a backup and will observe the difference in term of NDL compared to the Perdix.

 

[Lorenzoid]

Not based on any algorithm, but just for my own conservative peace of mind, on a 7-day trip I might take an afternoon off from diving in the middle. On a 12-day trip, I might take a full day off in the middle. Something along those lines. This conservative mindset is why I chose a Suunto as my first computer. I have since warmed to the idea that I can add conservatism to my diving in this manner regardless of what computer I'm using. My Shearwater suits me fine.

 

[MomZo]

That's also for their reputation and reliability that I chose a Suunto as my first computer, and the Zoop never let me down so far.
As you just described, it is good to take a day off (or even half a day off) in the middle of a diving cruise/week to desaturate. I often do it too as it is also a good way to visit the surrounding areas.
Good suggestion for those who don't do it yet!

 

[muzikbiz22]

Hi @MomZo
Remember, it is just the GF hi that determines surfacing for a no deco dive. The GF lo does not kick in until you exceed NDL, it then determines the depth of the 1st stop.

These 2 sentences just explained more fundamental sense about GF's, than the dozen or so articles out there on the internet. Just sayin'

 

[stuartv]

These 2 sentences just explained more fundamental sense about GF's, than the dozen or so articles out there on the internet. Just sayin'

Be aware that not all implementations of Buhlmann w/GF are the same. Some do work as @scubadada said. Some do not.

Some will not do anything with GFLo until it determines that you would exceed GFHi by executing a direct ascent to the surface. Then it will use GFLo to calculate the depth of your first stop. As scubadada said.

But, some will impose deco stops, based on GFLo, even though a direct ascent would not cause you to exceed GFHi. They do so because your ascent would cause you to exceed the GFLo. Personally, I think this seems like a poor way to implement the algorithm, but I have played with some different planners and found that some have in fact implemented in this way (and I don't remember which are which right now).

This is a grey area in the definition of the Gradient Factors algorithm. The algorithm was developed to calculate ascents for deco dives. It was not developed to use for calculating NDLs. So, this particular detail seems to be not clearly defined. This has been a subject of a few somewhat contentious debates here on SB.

 

[scubadada]

Be aware that not all implementations of Buhlmann w/GF are the same. Some do work as @scubadada said. Some do not.

Some will not do anything with GFLo until it determines that you would exceed GFHi by executing a direct ascent to the surface. Then it will use GFLo to calculate the depth of your first stop. As scubadada said.

But, some will impose deco stops, based on GFLo, even though a direct ascent would not cause you to exceed GFHi. They do so because your ascent would cause you to exceed the GFLo. Personally, I think this seems like a poor way to implement the algorithm, but I have played with some different planners and found that some have in fact implemented in this way (and I don't remember which are which right now).

This is a grey area in the definition of the Gradient Factors algorithm. The algorithm was developed to calculate ascents for deco dives. It was not developed to use for calculating NDLs. So, this particular detail seems to be not clearly defined. This has been a subject of a few somewhat contentious debates here on SB.

@stuartv is absolutely correct. Shearwater and my Dive Rite Nitek Q work like I said, I'm not sure about Divesoft, Garmin, and Ratio. I run MultiDeco planning software, it does not supply NDLs, it simply adds stops when needed, based on both the GF lo and the GF hi.

It's difficult keeping things simple, but remaining fully accurate

 

[Lorenzoid]

. . .
This is a grey area in the definition of the Gradient Factors algorithm. The algorithm was developed to calculate ascents for deco dives. It was not developed to use for calculating NDLs. So, this particular detail seems to be not clearly defined. This has been a subject of a few somewhat contentious debates here on SB.

As I recall, in one of those debates someone asked Shearwater how they implement it, and Shearwater posted a reply. I'm not going to search for the thread, but it should be findable with a little Google-fu.