I wonder if this is a long standing issue that's not limited to the latest 2012 maps.....
Yes it's an issue that I have with all my 25k "Selections" maps supplied over a period of several years (some quite recently) but the magnitude depends on the region of the UK. It's not a major problem for separate areas over the country, which can be calibrated individually by changing the "Map Properties" on the PC. However, I'm surprised more aren't complaining about the DMS maps which can contain activated regions from the whole UK but only have a single calibration offset available (and of course may be download directly to mobile device software which AFAIK doesn't permit "local" calibration).
For my maps, I've needed to use shifts of about +0N, +8E metres for the London area but +10N, +18E metres for Stirling (Scotland).
I wonder if this is to do with the distortions caused by inaccuracies in the original trig surveying?
Up to about 10 years ago the OS had spreadsheets (etc.) for doing coordinate conversions but as GPS became more mainline for surveying the OS started publishing various transformation utilities on their website, I think mostly because also at this time (the time, or shortly after, they went digital?) they "tweaked" the maps to remove the (known to them) distortions rendering the older calculations obsolete. I recall the distortions vary across the country depending on the accuracy of the original surveying and I think are "worst" in the mountainous areas.
Up till then the coordinate calculations were "linear" and may examples can still be found on the web or embedded in progams with source available. I think these legacy (as the OS seems to call them) approaches can also still be tracked down in an old corner of the OS web site for those who need to "undo" old calculations before "redong" them with the new ones.
I wonder if MM is still coded with the legacy approach rather than the new alternatives recommended (and detailed) on the OS web site here:http://www.ordnancesurvey.co.uk/oswebsi ... loper.html
And much interesting info on this (and other GPS & Positioning topics) here:http://www.ordnancesurvey.co.uk/oswebsi ... index.html (edit/update) Having read a bit more about OSTN02, the 2002 Transformation, it doesn't look like the differences introduced / corrected by that are anything like 19m & 7m being discussed in an earlier post