Algorithm question (Perdix AI vs DSAT)

[Trailboss123]

I have used Oceanic VEO’s for the last 5 years (about 600 dives) and recently purchased a Perdix. Prior to purchase I was concerned about the differences between the DSAT and GF algorithms and @scubadada recommended setting a Perdix to the low conservative setting of 45/95 to approximate the DSAT model. When I fell into a good deal on a Perdix, I figured I would try it and if I didn’t like it I could always sell it.

I test dove it in July in Cozumel and did a pretty mellow 2 tank per day schedule for 5 days with a VEO 3.0 and the Perdix strapped to my arm. Normally I do a little more aggressive schedule of adding an afternoon two tank trip every other day but during this trip there were no other folks for the afternoons. Next time I hope to get in more dives per day but this particular trip represents the average vacation diver trip of a week with 10 tanks over 5 dive days. I am using lots of nitrox (I watched a good friend do 11 chamber rides in May, I am using more Nitrox than I used to).

Initially the DSAT algorithm seemed definitely more liberal but as the repetitive dives were added it leveled out. I do not have enough data but it seems like the GF 45/95 reacts more liberal to repetitive diving and hopefully on my trips coming up over the next few months I can compare them on a more aggressive dive schedule.

Attached are screen shots of the 10 dives from Oceanic’s ‘Ocean log’ with text boxes showing the comparative logged NDL values for the two computers taken from Ocean log for the VEO and Shearwater Desktop for the Perdix. This is probably a more real world comparison as opposed to a dive planner. Personally I feel it is important to know how a computer handles repetitive diving.

FYI, yes I go up and down a lot while hunting Lion fish and no way will I give up my Perdix

Thanks for posting- I found that quite helpful. I will be out diving for 2 days over the coming weekend with my boss. He just bought the Perdix and will be diving with it and my computer algorithm is DSAT. I was wondering how our NDL's were going to compare. We will be diving the same mixes, since # of dives and same profiles. So, if I understand and @scubadada correctly, if he will be diving his computer in Recreational mode, we should set the computer on LOW conservatism to approximate as closely as possible my NDL's?

 

[tbone1004]

Thanks for posting- I found that quite helpful. I will be out diving for 2 days over the coming weekend with my boss. He just bought the Perdix and will be diving with it and my computer algorithm is DSAT. I was wondering how our NDL's were going to compare. We will be diving the same mixes, since # of dives and same profiles. So, if I understand and @scubadada correctly, if he will be diving his computer in Recreational mode, we should set the computer on LOW conservatism to approximate as closely as possible my NDL's?

he should set the computer to the conservatism level that he is comfortable with based on his understanding of the algorithm and how it applies to decompression theory. If that correlates to your DSAT NDL's, then so be it, but if it doesn't, then you dive to whoever hits their limit first.

 

[scubadada]

@hammet

Here is the expanded 1st dive NDL analysis I alluded to. Best GF hi match for DSAT is shown in yellow.You'll notice best match for 32 and 36% may be a GF hi of 90 or 95 for shallower depths, while it is 100 for 80 feet and deeper. Of course, this does not address repetitive dives. I have not paid enough attention to this, I intend on doing so in the future.

 

[gopbroek]

On post #14 I posted a series of screen shots showing the comparison of 5 days of 2 tank dives in Cozumel. I recently went back down for a weekend and dove a little more aggressive schedule of 4 tanks a day for 2 days. Attached is another series of screen shots showing the difference in calculations between the Oceanic DSAT and the Perdix @ GF 45/95. As I noticed on the 2 tank dive a day routine the DSAT schedule is initially more liberal but it becomes less liberal to the GF 45/95 as the number of dives increase. On a 4 tank a day routine this became more pronounced and by the fourth dive on the second day it became very pronounced. just something to consider if transitioning from DSAT to GF for recreational diving.

 

[scubadada]

I'm not sure this is the explanation but Buhlmann seems much more liberal at shallower depths. The algorithm also gives back much more time when ascending to shallower depths

 

[hammet]

I took my Geo and new Perdix on 5 dives this weekend and am going to compare the results.

 

[Jay]

I had the same question as the OP and did some homework.

Using Scubalab's 2017 DC data from their website (4 dives) 11 New Dive Computers Tested By ScubaLab In 2017, I averaged and ranked each time in a dive, summed the averages and ranks per dive, then created an overall set of two ranks.

Results were:
Oceanic PPX close with Perdrix AI (low consv)
Sherwood Sage
SeaJack
Perdrix AI (med consv)
then the others in a ~close group.

Some handy info I came across:

View media item 204278 from Scubadada (here)

View media item 204279 from Oceanic / SimplyScuba

For the diving I do (warm, tropical water - recreational) I was leaning towards an Oceanic or Shearwater ... until Garmin's latest creation came along.

 

[stuartv]

For the diving I do (warm, tropical water - recreational) I was leaning towards an Oceanic or Shearwater ... until Garmin's latest creation came along.

When they make a Descent with AI, it is going to be a serious contender!

 

[MomZo]

Hi.

I am thinking about changing my Suunto Zoop for a Shearwater Perdix. The main reason is that my Zoop is way too conservative during repetitive dives, and even more if I dive few days in a row.

After a first dive, I have noticed that my Zoop always gives me the lowest NDL compared to my buddy computers (other brand), no matter the depth. Thus, I am always the one asking to go shallower to not enter mandatory decompression stops. Moreover, the Zoop has a pivot value at 10 meters which makes that you can have a very low NDL when you are at 12 meters (for example), and it goes up to 45 minutes NDL when you go shallower than 10 meters (like 9.5 meters) It is all or nothing, not proportional.

I would like to know how is the Perdix behaviour in term of NDL, for repetitive dives, compared to Suunto computers (if set to OC REC with medium or low GF conservatism) ?
Would the Perdix give me more NDL than the Zoop during repetitive dives?

Thank you for your answer.

 

[tursiops]

Would the Perdix give me more NDL than the Zoop during repetitive dives?

Yes, depending on the Gradient Factors you set in the Perdix.