In article ,
Java Jive writes:
I recently asked Ordnance Survey why there was such a difference and
which is the most accurate. This was their reply:

BEGIN QUOTE

"The difference is almost certainly the result of the natural
differences between the collection methods and their source
projection. Land-Form PANORAMA contours are unique in their origin.
They were manually surveyed at 1:50 000 scale and the DTM variant was
derived from the contours. The SRTM data is remotely sensed high
altitude data at 90m accuracy. In essence, the differentiator is the
scale of each source dataset; SRTM is global, PANORAMA for Great
Britain."

END QUOTE

Which doesn't really tell us much.
snip

I think we can infer that the OS believe that their own heights are the
more accurate.
--
John Hall
"The power of accurate observation is commonly called cynicism
by those who have not got it."
George Bernard Shaw

In article , Java Jive
scribeth thus
I check public usage of the Terrestrial Calculator most nights,
including choosing Find The Likeliest for any UK location given, to
check the prediction is sensible. Tonight, for some unfortunate user
in Spain, it wasn't:
http://www.macfh.co.uk/PrivTest/BigginHill.png

I've been aware for a month or so, ever since discovering in a similar
way that the SRTM tiles for The Shetlands gave out above 60N and
having to make my own replacement ones from OS data to cover them,
that SRTM heights and OS heights do not always agree, but have been
wondering what I may need to do about it. However this is the first
time that I've actually encountered an absurdity arising from it.

IIRC and I might be wrong I remember somewhere that as good as the SRTM
data is it sometimes does need some correction as remember this was a
radar image and would take the height as what the timed reflection from
the earths surface" told" it.

Which for radio plotting use is quite valid in a way, and we've used it
for that and compared the results with OS data and very little
difference overall . Much less than the algorithm in use for the
coverage calculation anyway!...


--
Tony Sayer

On Fri, 9 Dec 2011 14:19:47 +0000, tony sayer
wrote:

IIRC and I might be wrong I remember somewhere that as good as the SRTM
data is it sometimes does need some correction as remember this was a
radar image and would take the height as what the timed reflection from
the earths surface" told" it.

Yes. One particular problem with it that I discovered early is that
it contains data voids. Accordingly, my algorithm to read heights
from the tiles circles around the point of interest until one or more
valid data points are found, and averages the closest.

Next there was the Shetland problem I've already explained. Basically
there's no proper SRTM data at or above 60N, and as the Shetlands
straddle that boundary, I had to remake all the Shetlands tiles from
OS Panorama data.

Which for radio plotting use is quite valid in a way, and we've used it
for that and compared the results with OS data and very little
difference overall . Much less than the algorithm in use for the
coverage calculation anyway!...

Well, that's certainly a very valid point. Thanks for a timely sanity
check!

There's also the point that at least some of the antenna heights
change(d) at DSO, for example I believe the mast at Tacolneston was
extended, and I know that some transmitters such as Black Hill were
rebuilt, but AFAIAA coherent information on this has not been
published anywhere. I've only discovered those that I know of almost
accidentally. Tacolneston because Mark mentioned it, and Black Hill
because I spotted the discrepancy in NGRs and queried it with Ofcom.

My immediate problem is that clearly I need to get all the data into
the same projection before running the calculation. I'm thinking that
it would be much simpler to convert the OS and user heights of the
transmitter and receiver into WSG84 rather than, say, to convert the
page to use OS Panorama heights.

The signal profile functionality was designed to support users not
just from UK mainland nations but also from NI and Eire, of whom there
are particularly quite a few of the former but also one or two of the
latter, and from the CI, of whom I don't definitely recall more than
one or two. As height data for these other nations may not be freely
available except via SRTM, I think we have to keep the data in the
same format as SRTM.

Also that would mean much less additional work to correct this
problem, which is not an insignificant factor, as I'm not sure how
much continuing commitment I can make to maintaining the calculators
for a while after the first third or so of next year - I'm kind of
hoping that most of any changes required due to retunes, etc, will
have been published and incorporated into the data by then, so the
beast will more or less look after itself for a few months.

However, there is a disadvantage to converting the receiver and
transmitter heights displayed in the calculator, in that the scales on
the side of the profile will then not agree with the heights in the
calculator, and even if I put a note in, say, the key, explaining
this, I suspect that there will be some users who will notice this,
won't bother read any explanation, and just say that the calculator's
'no good', or words to that effect.

So I suspect that at least for the display, I'm going to have to
convert all the data points displayed back into the local (usually OS)
projection.

Also, similarly, those who've already got used to it being as it is,
and who haven't encountered a problem such as that at Biggin Hill, may
take some convincing that the thing was not working perfectly well
before I 'broke' it by 'fixing' it!

What do people think?
--
================================================== =======
Please always reply to ng as the email in this post's
header does not exist. Or use a contact address at:
http://www.macfh.co.uk/JavaJive/JavaJive.html
http://www.macfh.co.uk/Macfarlane/Macfarlane.html

On Friday, December 9th, 2011 at 10:35:51h +0000, John Hall wrote:

I think that HMG will tell you that we're not at war

But can you believe everything HMG tell you?

Are we also to pretend that hostilities against Iraq in the
First and Second Gulf Wars were not actual wars?

and I'd like to believe that they don't disrespect our soldiers

The acid test for that is how well do they treat soldiers who have
returned and left the military.

Homelessness and suicide rates are significantly higher for
ex-servicemen than those who were employed in other government jobs.

Apparently there's some disadvantage in international law in doing so.

Yes and it gets around certain national restrictions as well.

..

On 09/12/2011 00:08, J G Miller wrote:
On Thursday, December 8th, 2011, at 22:35:21h +0000, Brian Gaff wrote:

but I won't hod it against you.

Are you no longer infected and can you say that in Hungarian?

de nem fogom ó, hogy Ön ellen. (1)

And my hovercraft _is_ full of eels.

Andy

--
(1) don't believe Google will have got it right.

Java Jive wrote:

There's also the point that at least some of the antenna heights
change(d) at DSO, for example I believe the mast at Tacolneston was
extended, and I know that some transmitters such as Black Hill were
rebuilt, but AFAIAA coherent information on this has not been
published anywhere. I've only discovered those that I know of almost
accidentally. Tacolneston because Mark mentioned it, and Black Hill
because I spotted the discrepancy in NGRs and queried it with Ofcom.

Taccy has actually had a new mast, the new mast is listed on the South Norfolk
Council planning website at being 206.1 metres high (old mast was 165m)

Rowridge has a new mast for DSO too, I don't have the height, but I do have
the mean height of the main DTT aerial array, as 177m agl

--
Mark
Please replace invalid and invalid with gmx and net to reply.

www.paras.org.uk

Brilliant! Thanks as always, Mark.

On Fri, 09 Dec 2011 20:24:56 +0000, Mark Carver
wrote:

Taccy has actually had a new mast, the new mast is listed on the South Norfolk
Council planning website at being 206.1 metres high (old mast was 165m)

Rowridge has a new mast for DSO too, I don't have the height, but I do have
the mean height of the main DTT aerial array, as 177m agl
--
================================================== =======
Please always reply to ng as the email in this post's
header does not exist. Or use a contact address at:
http://www.macfh.co.uk/JavaJive/JavaJive.html
http://www.macfh.co.uk/Macfarlane/Macfarlane.html

Been playing around with this over the weekend, and have come to the
conclusion that the only way to do this that makes sense is to use the
height of the transmitting array above ground level (AGL) rather than
Above Ordnance Datum (AOD). This means that I'm going to have to
convert the data file that contains the height data.

Actual conversion is not the problem, it's finding the correct heights
to put in the file! I can work out the analogue information from
historical data, but, AFAIAA, there is no publicly published
information anywhere giving the DTT antenna heights either AOD or AGL.
I can guess that relays will nearly all be the same as for analogue,
but the 80 or so all muxes / main transmitters do seem to be being
modified significantly for DSO, and I'm going to struggle unless
anyone can suggest a pukka source.

Any ideas, anyone?

On Fri, 09 Dec 2011 15:41:26 +0000, Java Jive
wrote:

My immediate problem is that clearly I need to get all the data into
the same projection before running the calculation. I'm thinking that
it would be much simpler to convert the OS and user heights of the
transmitter and receiver into WSG84 rather than, say, to convert the
page to use OS Panorama heights.
--
================================================== =======
Please always reply to ng as the email in this post's
header does not exist. Or use a contact address at:
http://www.macfh.co.uk/JavaJive/JavaJive.html
http://www.macfh.co.uk/Macfarlane/Macfarlane.html