Note that PhyR is proprietary to Broadcom, while G.INP is ITU-standardised. ReTx seems to be the parameter name used by BT to control G.INP being turned on.

In reply to a post by Ixel:

Looking at the latest SIN 498, I believe this replaces the standard interleaved profiles, which makes sense. The other possibility is that retransmission either is a middle option between fastpath and interleaved, or that retransmission replaces fastpath and interleaved remains for those connections which perform considerably poorly.

My (desperately limited) understanding of G.INP is that there are some modes where it does affect latency for every packet (rather than just the ones that get re-transmitted), although less than interleaving does. I read this was related to the ability to reduce jitter, but haven't been able to confirm that this feature exists, or to see how it works.

If that is true, then I could see why DLM would want to maintain a "clear" fastpath as the default setting.

G.INP may still not be enough for a troublesome line, so the traditional FEC/interleaving solution will still be available, but now only as a second line of defence - with banding dropping down to a third line.

More info here, though it is in the form of a sales pitch for Assia's own DLM tool:
http://www.assia-inc.com/products-services/pdf/DSL_E...

So what does this mean for a person with perfect line stats?

I guess not much.

Yeap not much, and Openreach are still looking to see if they have launched this change that people are saying they have launched

Not much, as G.INP will likely be disabled for perfect connections and most likely only used as a first choice by DLM in an effort to improving lines which don't run as efficiently without it. I'm imagining nobody will notice G.INP on their line until DLM makes a change to that profile at some point in the future - including those who are already interleaved and stuck.

Edited by Ixel (Tue 13-Jan-15 21:53:08)

Openreach don't know if they've launched it? Haha, that sounds funny .

None of the xDSL stats have really changed so far...

I'm considering getting another HH5, even if it's just to use it as an AP.

In reply to a post by MHC:

Every modem will be slightly different - the complex termination impedance can effect the level of received signal and thus the signal available to the demodulator.

I have both Huawei and ECI modems here and speeds vary by several Mbps. You could get another HH5 and find it syncs faster and another HG612 which syncs slower.

Would that explain why my SP08 firmware unlocked HG612, I have noticed my upload sync is @ 15.35, yet with the BT HH5 it's @ 17.89. So technically I'm losing 2mb of my upload on the HG612.

I took off QoS as well, but should I consider trying the older SP10 and kill BT agent via Telenet as the SP08 is disabled by default in the WAN options?

You are not actually losing 2Mbps as it is not there to use. Every modem will sync at a different speed - you could get anoth pair of HH5 and HG612 and they could be te other way round.

QoS and BT Agent do NOT affect sync speed.

In reply to a post by MHC:

BT Agent do(es) NOT affect sync speed.

Oh really? What does it do then?

In reply to a post by BatBoy:

In reply to a post by MHC:

BT Agent do(es) NOT affect sync speed.

Oh really? What does it do then?

It uses some of the available bandwidth. There could be hundreds of small ISP or manufacturers monitoring applications running - they would still NOT alter the sync speed.