I hate to be backing up dmaziuk, but I think what he meant was if the two sampling frequencies are different, not that one is actually 20s and the other is actually 40s.
Matching Dive Computer Algorithms
Please register or login
Welcome to ScubaBoard, the world's largest scuba diving community. Registration is not required to read the forums, but we encourage you to join. Joining has its benefits and enables you to participate in the discussions.
Benefits of registering include
- Ability to post and comment on topics and discussions.
- A Free photo gallery to share your dive photos with the world.
- You can make this box go away
The fourth dive of the last day. It is one of the few dives where the Petrel was more liberal than the Cressi...
The Cressi hit 99 at 28 minutes and 38 feet.
The Petrel at 27 minutes and 46 feet. Of course the Petrel had more NDL at the start of the ascent.
The Cressi hit 99 at 28 minutes and 38 feet.
The Petrel at 27 minutes and 46 feet. Of course the Petrel had more NDL at the start of the ascent.
Sampling frequency != recalculation frequency. Logging frequency != sampling frequency. Recalculation frequency != display update frequency. There really is a whole lot of different ways you can write this, and with multiple cores you can do those in parallel just to make it even more confusing.
What you really want to know is the recalculation frequency that produces meaningful results by the decompression model. Is "quarter-time" of the fastest tissue compartment good enough? "Eighth-time" is over a minute if your fastest compartment has 5 minutes half-time.
Yeah, right: measured every second with data acquisition interval of 20 seconds. One assumes the main loop reads the sensor every second and discards 19 readings out of 20.
I have no idea if it answers the question but you can easily see by the graphs that the Petrel shows more detail in depth changes where the Cressi smooths out the lines.
I assume this means the Petrel is recording more often?
I assume this means the Petrel is recording more often?
Yes, the terminology can be a bit confusing. Just to clarify, @dmaziuk, used the term; recalculation frequency, while @tursiops used the term; sampling frequency. To me, recalculation frequency refers to the time the dive computer recalculates the algorithm based on depth in real time. On the other hand, sampling frequency refers to time the dive computer stores the data in memory to be transferred to the digital log book at a later time. Based on the computer specification, depth is measured every second, so I assume the computer recalculates every second. The data acquisition interval (sample frequency) is every 20 seconds. So if the data acquisition interval (sample frequency) is different between two computers, then I agree the graphs would be different. In some cases the data acquisition interval can be changed to bring the graphs closer together. On my Tusa IQ750 the frequency can be 2, 15, 30, 60 (seconds). I'm not sure if you can change the data acquisition interval on the Petrel or Cressi. But with regards to recalculation frequency, I assume both computers are doing it at least every second. The difference then is the actual algorithm being recalculated and in this case the Pretel running ZH L16 GF and the Cressi RGBM. Does that make sense or have I just compounded the confusion?
Sort of, and...a little bit.
You can't just use a term (like I did!) and assume everybody else uses it the same way. I used "Sampling frequency" in the technical sense of how often the data from the sensor are extracted...not how often it is written to memory and stored/logged. Those pressure data may be combined with other stuff, or smoothed, or whatever, and written to memory at the same or (more likely) a slower pace. But I don't think it makes any sense to go down this rabbit hole; two different computers are not going to behave the same....and it may be due to the rate at which they sample, recalculate, store, display...whatever.
I have it on good authority that physicists throw away almost every data point coming from the sensors of instruments like particle colliders. Otherwise, the phrase "disk farms size of France" has been used to describe storage requirements.
Recalculation frequency may be limited to simply how long it takes to run through single iteration of the main program loop. That's the easiest, but one may want to slow it down to save the battery. For instance, recalculating every second when your sensor readings change every 20 seconds is just pointless. Recalculating every 20 seconds when your fastest tissue compartment has half-time of 5 minutes also seems quite pointless to me, but I haven't run the numbers. Those are finer details.
Comparing the graphs from 2 computers running at different frequencies is problematic, just like @CandiveOz says -- and that's without factoring in the switches in leading tissue compartment, of which one has 9 with fastest half-time of 2.5 minutes and the other: 16 with fastest half-time of 4 or 5 minutes, depending on which on-line source Shearwater programmers copy-pasted the "ZHL16" table from.
Similar threads
