First Dive Computer

[dmaziuk]

The 126 page 1994 DSAT RDP document has it's goal clearly stated. Scubalab's article also has its goal clearly stated. (wrt NDLs and taking the computers together on a dive).

DSAT RDP document states very clearly the problem with existing algorithm, the issues driving the development of the new one, the rationale behind their testing schedule, and discussion and the interpretation of the result.

And you don't see any difference between all that and what ScubaLab has? -- ScubaLab: How We Test Dive Computers under Objective test protocol (sic)

To gauge the performance of the computers’ algorithms, they were subjected together to a series of four dive simulations in the USC Catalina Hyperbaric Chamber.

Meant to simulate a day of diving, the dive profiles (shown in the four charts) were:
...

(The best rationale I have for those particular profiles is Craig said they're representative of recreational dives.)

At least in 2016 they added "even with the caveat that the data applies only for profiles and conditions like those in our test."

Scubalab's data shows that for their Dive 1 in 2014, 2016 and 2017 your statement is wrong.

Since they weren't testing the same computers, I am not sure what it is that you actually mean. I didn't see a same reference computer showing the same numbers every year. Without a baseline reference you might as well be comparing random numbers.

Admittedly my views on experiment design is heavily influenced by a couple of decades of running computers for scientists. I'm sure I'm being too harsh on the ScubaLab guys, they're just regular hard-working journalists, salt of the earth...

 

[scrane]

Terrible analogy. But I chuckled anyway.

Now, imagine a car that, if you exceed 35 mph in a 40 mph zone, locks up on you for the next 48 hours.

An analogy that generates a corollary can't be all bad.

 

[Jay]

Since they weren't testing the same computers, I am not sure what it is that you actually mean. I didn't see a same reference computer showing the same numbers every year. Without a baseline reference you might as well be comparing random numbers.

My comment was in response to your earlier comment where you had said "all available evidence suggests that on the 1st dive with no additional conservatism, all algorithms should be within 2 minutes of each other at the most." (post #42). I said you're wrong as Scubalab's data (for every year as listed) shows differences greater than that (post #64). It is simply a piece of evidence that your statement is wrong.

 

[dmaziuk]

And e.g. 2016 spreadsheet has on top dive 1 "SI: 24hrs". As I recall Dr. Wienke's paper, RGBM considers 30 hours as "same day" so if that "SI" means the computers were in the water a day before, RGBM ones may or may not have considered it "the 1st dive".

The problem is no confidence. Confidence interval for ScubaLab numbers is between zero and nil.

 

[Jay]

And e.g. 2016 spreadsheet has on top dive 1 "SI: 24hrs". As I recall Dr. Wienke's paper, RGBM considers 30 hours as "same day" so if that "SI" means the computers were in the water a day before, RGBM ones may or may not have considered it "the 1st dive".

The problem is no confidence. Confidence interval for ScubaLab numbers is between zero and nil.

That is really a stretch, a desperate stretch, to say Scubalab's first dives were possibly not clean. Check the manual of each computer tested for what is regarded as a clean dive, AND then check with Scubalab if you want to make that outlandish claim.

But your outlandishness claims don't end there! (insert infomercial voice; 'but wait, there's more!'); you know full well that a single test to provide NDLs for a particular profile will never amount to anything statistically significant, and of course there's no statistical Confidence Interval.

At least I thought you would have responded with some of your "all available evidence" to demonstrate your point about relative algo (NDL) differences being minor on the first dive.

I come to this forum to learn. I don't mind explaining to you twice why your comment regarding NDL differences (quoted in posts 64 & 83) is wrong. But you seem so hell-bent to argue that I now feel you didn't bother studying the data from Scubalab or were blind to what that snap-shot said. As such, I will be reading your future posts with great scepticism and won't be replying to any comment you might make in this thread.

 

[Centrals]

AL i300 is under US$200.00. It runs on Pelagic Z+ which is a wee bit conservative than DSAT. Personally I found the DSAT algorithm used on Oceanic far too liberal. But that is me.

Happy hunting.

 

[stuartv]

AL i300 is under US$200.00. It runs on Pelagic Z+ which is a wee bit conservative than DSAT. Personally I found the DSAT algorithm used on Oceanic far too liberal. But that is me.

Happy hunting.

You've used DSAT? You got bent? If you didn't get bent, then what factors caused you to conclude that it is too liberal for you?

 

[scrane]

AL i300 is under US$200.00. It runs on Pelagic Z+ which is a wee bit conservative than DSAT. Personally I found the DSAT algorithm used on Oceanic far too liberal. But that is me.

Happy hunting.

The important thing to consider with Oceanics is if you're uncomfortable with DSAT (and I can understand that) you have three other options to dial the conservative factor to your taste, outside of how close you want to get to the indicated NDL.

 

[dmaziuk]

That is really a stretch, a desperate stretch, to say Scubalab's first dives were possibly not clean.

You mean you were being serious? If so, read this (courtesy of @Diver0001) http://dspace.rubicon-foundation.or...dle/123456789/10149/VDC_2012_7.pdf?sequence=1 and be enlightened. And if you still don't get it afterwards, please killfile me: I'm all out of water that I can possible lead you to.

 

[Centrals]

You've used DSAT? You got bent? If you didn't get bent, then what factors caused you to conclude that it is too liberal for you?

I have been using Uwatec Aladin Air and Pro Nitrox well over 20yrs without any issue. Unfortunately one of them, Air, died few yrs ago so I bought a Oceanic Veo 180 as a "back up". Both the Aladin Pro Nitrox and Veo 180 have very similar ndl at max depth but the Veo 180 will "generate" a lot more on ascend. And the difference can be as much as 2 -3 times!!!
I can easily bend my Uwatec if I follow DSAT all the way to the surface.