Interleaving comes with FEC overheads.
N - N Value (NFEC)
RS codeword size. Reports the actual number of Reed-Solomon redundancy per codeword used in the latency path. Used with the R value to express the rate of FEC applied.
R Value - RS check Bytes (RFEC)
Number of (redundant/parity) bytes occupied by the Error Correction process (overhead).
Often used to describe the level of error correction and can range from 0 to 20. The more check bytes that are used then the more errors can be corrected. The value 0 indicates no Reed Solomon coding. Use this value with N to express the rate of FEC applied as a percentage. eg R/N
If Chryralis posts his full stats you can see how much overhead is used by FEC on his line with Interleaving is enabled.
Post your own --stats and compare.
A line on FTTC that is fastpath or has G.INP will sync at the full rate and show a similar attainable.
Something like
sync: 41,111
attainable: 41,216
when interleaving is enabled there's an effect on both sync and attainable.
overheads reduce sync.
attainable is inaccurate * more info below.
for the same line interleaving would do something similar to
sync: 36,132
attainable: 43,324
sync reduced, attainable increased.
*
an older inaccurate "basic" method of working out the attainable used by most modems (and Broadcom chipsets) exaggerates the attainable with INP/Interleaving. newer Lantiq chipsets/firmware use a newer "improved" method of calculating the attainable and don't do this.
There's an example on the forum of a user with a Lantiq chipset who's interleaved posting their stats before and after a firmware upgrade.
Post 1 shows an Attainable way above sync and post 2 showed an accurate attainable.
1st time I've seen an accurate attainable with interleaving which ejs suggests shows Lantiq using a better method of calculating attainable on newer firmware which makes sense.
Can't find the post.
WWWombat detailed the different ways modems can calculate the attainable in
this post