Hey everyone,
Tonight while I was trying to decode a remote I thought it would be a great idea to have a Flipper app that takes the infrared signals and provides a checksum of the value it perceives. This is helpful when trying to compare if two signals are exactly the same.
Now while you may just use IR Scope to see the signal, it would be easier if we could get a value to compare.
If the signal is known it will be shown as parsed. If it is not shown, how should the app compare the values?
4 captures of RAW does not necessarily look the same.
Even if the idea is great, I follow a similar thought in another project by doing it off board. I don’t think ‘comparing RAW data lines’ is the way to go, the better way onboard is to add parsers.
Just my opinion, maybe there are other views with better solutions.
I suppose an algorithm could be used and the samples could be scored by percent. It does sound like it would take some time.
It you ate able to identify a chunk, maybe. But the IR parser needs to be extended.
An IR raw signal can be:
4500 1500 300 600 300 300
So it can also be for a longer press:
4500 1500 300 600 300 300 60000 4500 1500 300 600 300 300
Or, if a repeat frame is defined:
4500 1500 300 600 300 300 60000 4500 300 300
(The values are fiction, no real protocol!)
As far as I know the parser on the flipper does not care about repeat frames, today.
My issue is, how to tell where to stop parsing. Is 60000 (mostly up to 20 times the base value) a good breakpoint? And from here on, look for duplicates or identify repeat frames?
In theory I know this is a problem for a state machine. In real life I have not enough time to follow this approach.
It does seems like energy would be best spent improving parsing as you suggested.