<div dir="ltr"><div><div><div><div><div><div><div><div>Hi all<br><br></div>Thanks Richard for your answer.<br><br></div>I made some tests. Here are my conclusions :<br><br></div>* By default, QGIS will use the current project ellipsoid to compute distances.<br></div>* You can change this behaviour in the Project Properties dialog ( Menu Project > Project properties ), tab "General"<br></div>* The safest distance calculation is done when you use the same projection between the project and the layers. I have seen some weird lengths while using the map tool with OTF on, with a project in Pseudo Mercator (EPSG: 3857) and a projected layer ( EPSG:2154 ). Like 30% differences (which is huge)<br></div><div>* You can select "Planimetric / None" instead of an ellipsoid in the project properties option to avoid discrepencies between distance measurements (for example inside projects with different OTR parameters). But you won't then benefit from the precise calculation done by using the projection ellipsoid.<br></div>* It seems many other softwares like ArcGis or grass , they use "Planimetric / None" to compute the distance (and I read some deactivate distance tool when OTF is on)<br><br></div><div>So it seems to me QGIS is advanced here in the field of calculating distances (and areas, etc.), but the end user must know what he (she) is doing to fully understand what is going on. Should we not use "Planimetric / None" as a default parameter, even if OTF is on ? Advanced users could benefit of more precision by using the projection ellipsoid, but everyone else will have consistent measurements ?<br></div><div><br></div>Cheers<br></div>Michaƫl<br></div><div class="gmail_extra"><br><div class="gmail_quote">2015-10-26 9:09 GMT+01:00 Richard Duivenvoorde <span dir="ltr"><<a href="mailto:rdmailings@duif.net" target="_blank">rdmailings@duif.net</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On 26-10-15 08:48, kimaidou wrote:<br>
> Hi all<br>
><br>
> Someone reported to me an issue with $length distance calculation and<br>
> the "Distance" map tool. It seems to him QGIS was not precise enough<br>
> compared to field data collection (about 10% or worst discrepency)<br>
><br>
> I remember reading some discussions here about wrong distance<br>
> calculations in some contexts.<br>
> Has anyone a "cookbook" or good practices about distance calculations in<br>
> QGIS, regarding different data sources, contexts, SCR, etc. ( on the fly<br>
> reprojection or not, data provider, etc.)<br>
<br>
</span>I know Nyall has been busy with this (see hist posts in both dev and<br>
user list).<br>
<br>
Regards,<br>
<br>
Richard<br>
<br>
</blockquote></div><br></div>