Search for an inexpensive rec computer running Buhlmann

[scubadada]

I think that the Deepblu Cosmiq+ and the Watoom Cyano have not met expectations. What else is there to offer?

Genesis Centauri: 3 proprietary conservation settings, said to be similar to Shearwater presets.

Serenity S1 Sport: 3 proprietary presets, unknown conservatism

Seac Action, Action HR: 6 conservative settings:

I'm not at all sure we are there, I would love to hear from divers who have used these computers.

 

[wetb4igetinthewater]

I'm curious if there are different performance/memory requirements for different algorithms. I can see some justification for differences in prices, but given how each 32-bit ARM processors and memory is, I don't see that as a factor. What is/are the big factor(s) in price difference?

Currently, I only teach 2 OW students at a time, so each gets a Shearwater. When I open up my own dive op, I don't think I can be providing all customers who want to rent a DC a Shearwater. I'd like to have a successful enough business to provide such a service, but that certainly won't be the case in the beginning.

 

[dmaziuk]

I'm curious if there are different performance/memory requirements for different algorithms. I can see some justification for differences in prices, but given how each 32-bit ARM processors and memory is, I don't see that as a factor. What is/are the big factor(s) in price difference?

Full "iterative" RGBM is apparently pretty compute-expensive. (And so the low-end computers run some "folded" versions that come with all kinds of limitations in the folds and creases.) ZH-L and its variants are simple, though a poor fit for no-stop profiles. DSAT should be about as simple. VPM is a bit more expensive but not by that much.

Power requirements of an ARM-level CPU are quite a bit more than the ASICs in the low-end DCs. And then you have the screen: once you go dot-matrix, you need a GPU with its own power requirements and so on and so forth. But if you want to develop a new ASIC for a dive computer, you're probably facing more R&D costs than your sales will ever pay for.

So I think at this point you're basically stuck between smartwatch hardware with an open algorithm, or an existing Seiko/PPS/Suunto puck with a fresh coat of paint like BT uplink. Smartwatches are $500+ even without pressure-proof bodies and custom software, so that's your starting price tag.

 

[tursiops]

Full "iterative" RGBM is apparently pretty compute-expensive. (And so the low-end computers run some "folded" versions that come with all kinds of limitations in the folds and creases.) ZH-L and its variants are simple, though a poor fit for no-stop profiles. DSAT should be about as simple. VPM is a bit more expensive but not by that much.

Power requirements of an ARM-level CPU are quite a bit more than the ASICs in the low-end DCs. And then you have the screen: once you go dot-matrix, you need a GPU with its own power requirements and so on and so forth. But if you want to develop a new ASIC for a dive computer, you're probably facing more R&D costs than your sales will ever pay for.

So I think at this point you're basically stuck between smartwatch hardware with an open algorithm, or an existing Seiko/PPS/Suunto puck with a fresh coat of paint like BT uplink. Smartwatches are $500+ even without pressure-proof bodies and custom software, so that's your starting price tag.

This is a remarkable and unusual post: it is clear, concise, informative, and void of snarky and cryptic comments. Good job!

 

[wetb4igetinthewater]

Full "iterative" RGBM is apparently pretty compute-expensive. (And so the low-end computers run some "folded" versions that come with all kinds of limitations in the folds and creases.) ZH-L and its variants are simple, though a poor fit for no-stop profiles. DSAT should be about as simple. VPM is a bit more expensive but not by that much.

Power requirements of an ARM-level CPU are quite a bit more than the ASICs in the low-end DCs. And then you have the screen: once you go dot-matrix, you need a GPU with its own power requirements and so on and so forth. But if you want to develop a new ASIC for a dive computer, you're probably facing more R&D costs than your sales will ever pay for.

So I think at this point you're basically stuck between smartwatch hardware with an open algorithm, or an existing Seiko/PPS/Suunto puck with a fresh coat of paint like BT uplink. Smartwatches are $500+ even without pressure-proof bodies and custom software, so that's your starting price tag.

Makes sense about power requirements, as DCs have relatively small batteries. Any idea on which ASICs are being used by different DC manufacturers? I'd like to look up the specs.

I am more of a fan of ZH-L16c with GF. Why do you feel it is a poor fit for recreational/no-stop diving?

 

[dmaziuk]

Makes sense about power requirements, as DCs have relatively small batteries. Any idea on which ASICs are being used by different DC manufacturers? I'd like to look up the specs.

No clue, I'm just ass-u-ming Seiko, PPS, and Suunto have purpose-made chips in their low-end offerings.

I am more of a fan of ZH-L16c with GF. Why do you feel it is a poor fit for recreational/no-stop diving?

The big selling point of Buhlmann's formula is it gives you the next stop's depth directly, with no intermediate hoops to jump through. Makes life very easy when you a priori know there's a stop. On a no-stop dive it's a bit of the opposite.

 

[wetb4igetinthewater]

The big selling point of Buhlmann's formula is it gives you the next stop's depth directly, with no intermediate hoops to jump through. Makes life very easy when you a priori know there's a stop. On a no-stop dive it's a bit of the opposite.

Okay, I've had my morning coffee, but it isn't kicking in. I'm not following. How does diving Buhlmann for recreation diving make life not easy?

 

[dmaziuk]

Okay, I've had my morning coffee, but it isn't kicking in. I'm not following. How does diving Buhlmann for recreation diving make life not easy?

Because if you don't have a stop, and only trying to figure out when you will have one, you have to run his formula in a loop adding a minute of exposure on each iteration, until you get one. Resulting number of minutes is your NDL. But every iteration drains the battery.

There is a version of Schreiner's equation that gives you NDL right away. But it needs Workman-style M-values, adjusted by back-converted Gradient Factor. So Buhlmann's numbers + GF are a pain to use with it. And then if your user overstays their NDL it all flips around and now Buhlmann's formula is convenient and Workman's M-values are not.

I wonder how Pelagic did PZ+. With a beefier CPU you'd just run it in a loop as per #1, but with something that runs for 2 years off a watch battery, it may be a bit more... interesting.

 

[wetb4igetinthewater]

Because if you don't have a stop, and only trying to figure out when you will have one, you have to run his formula in a loop adding a minute of exposure on each iteration, until you get one. Resulting number of minutes is your NDL. But every iteration drains the battery.

There is a version of Schreiner's equation that gives you NDL right away. But it needs Workman-style M-values, adjusted by back-converted Gradient Factor. So Buhlmann's numbers + GF are a pain to use with it. And then if your user overstays their NDL it all flips around and now Buhlmann's formula is convenient and Workman's M-values are not.

I wonder how Pelagic did PZ+. With a beefier CPU you'd just run it in a loop as per #1, but with something that runs for 2 years off a watch battery, it may be a bit more... interesting.

Ah, thanks for the clarification. I don't have a response as I want to think about it. I wish I had more insight to how the Perdix operates, as the battery consumption is simply amazing. I'm curious as to what Shearwater does in recreational mode. It is my intent to one day understand algorithm details of various algorithms (getting down to the mathematical equations). But that's not a small endeavor. Thanks again for the info.

 

[runsongas]

we'll see if this one does better than the deepblu/watoom

Atmos MissionOne Dive Computer Black-Black

Make the dive simple. 探索海洋就是這麼簡單