Subsurface & Shearwater Perdix AI

[inquis]

Using the bluetooth or bluetooth4.0LE?

I'm not sure. It takes a while, so likely LE.

 

[_sgm_]

I'm not sure. It takes a while, so likely LE.

I would be interested in the average number of dives it is uploading .... LE took 2.5 minutes for 7 dives to a cellphone/-cloud. Serial devices were clocking on desktop 1 dive profile per second to ingest into subsurface.
Note after battery disconnect reconnect the Bluetooth bind was cleared and the droid application connected.. still the rate was slower than expected for 30K of data followed by 30K verify
more work is needed with linux and a pull down / align import into subsurface form shearwater cloud... ( alot of moving parts ) plus alot of bluetooth traces to compare ...
Will compare against Garmin data sets for speed and logging over bluetooth ... going to call the day in a hour or so... 18 hours of review does not make for a happy soul... [ more progress in the past 10 hours than I had in the past 180 days... ]

 

[rsingler]

@_sgm_ I'll DM you a pair of dives. One from Perdix AI, the other an import from the XML file from Shearwater Cloud and saved in the same format from Subsurface. The Subsurface file is rich in data, and the graphs are identical, but the dive time field makes no sense.

 

[_sgm_]

.

 

[_sgm_]

your dive should have logged as 54 minutes 12 seconds
actual dive was 52 minutes 13 seconds..
samples taken every 3.x seconds
Subsurfaces issue is doing a Div 60 vs a div 60000

 

[_sgm_]

option 1 on the perdix log would be to round out 1000 then the div 60 should align when the dive gets imported

 

[_sgm_]

there was a 182000 hop between samples /60000 it hits 3.0x seconds per sample
and samples registered 1652 from 0 start til 0+59 0 termination.. accounting for the computer calling the dive and saving the log
55 bytes accounted for ... as padding

 

[rsingler]

your dive should have logged as 54 minutes 12 seconds
actual dive was 52 minutes 13 seconds..

Extra time added on Perdix AI as a deliberate instruction to not end dive on surfacing. See this thread: Rising GF99 after surfacing???

Now ask I have to do is figure out how to

option 1 on the perdix log would be to round out 1000 then the div 60 should align when the dive gets imported

Thanks, @_sgm_ !

 

[_sgm_]

Extra time added on Perdix AI as a deliberate instruction to not end dive on surfacing. See this thread: Rising GF99 after surfacing???

Now ask I have to do is figure out how to


Thanks, @_sgm_ !

I will look at the spacing then run a awk script to separate the sequence and a sed to do a pull of 000 out from the number adding back in a new file the correct times
because I do not think you want to do 1600 cut and paste actions

 

[rsingler]

It worked! I opened the XML file in Notepad and deleted every instance of '000'. That changed current time = 64000, for example, to 64.

When I imported the file, dive time was correct.
Now I just have to find what else I screwed up.
Thanks @_sgm_ !

Do I really want to do this for 300 dives?...