There's no direct correlation between the number of ES a line receives and the number of FEC the same line may receive.
I had similar ES numbers to Jon21 but over 100 times more FEC.
Some types of interference cause no ES at all but cause 100,000+ FEC per min (Homeplugs do this for example).
You made other posts, that said otherwise?
There would naturally be a correlation, FEC are fixed CRC, the more CRC, the more ES, which means more FEC on an interleaved line.
Now of course you can have situations where you have batches of CRC on low numbers of ES, typical in noise bursts, that would give a high CRC to ES ratio and also high FEC to ES ratio, maybe thats what you mean, but regardless, there seems to be no counter for fixed errored seconds counter, so given it doesnt exist how would DLM determine a means of moving an interleaved line back to fast path? I dont think it just uses CRC/ES as I have seen instances where there is no ES but moderate levels of FEC, then a higher noise margin is applied, the levels of FEC go down, then DLM acts to switch it back to fast path.
On your homeplugs example, are you saying you seen them cause FEC on an interleaved line but when that "same" line is on fast path they generate no CRC? Or rather they cause no CRC on interleaved lines, there is a clear difference between the two. Because I can tell you I have seen homeplugs most definitely cause huge amounts of CRC, and then the line becomes interleaved and they then cause FEC instead.