[Qgis-user] Hillshade Rough Blocks for 1m Rasters?

Nicolas Cadieux njacadieux.gitlab at gmail.com
Thu Jul 2 12:27:33 PDT 2020



> Le 2 juill. 2020 à 14:14, cgray at graytechsoftware.com a écrit :
> 
> The intention is for the pixel itself to be a certain shade, not the edges. It works correctly if your native raster layer does not exactly match project CRS.

Probably because there is reprojection going on, thus, some smoothing.

> There is something specific that does not work appropriately, but I cannot figure out what it is. I have also seen the pixel boarder issue for quite some time, but hadn't worked with enough other programs and variables to recognize it as a problem.
> 
> As shown in my first message, the desired result of having clear pixel colors does work in QGIS.
> 
> Where would bugs be reported? I can take some time to get it replicated and share the results for bug identification.
> 
> Thank you, Chris
> 
Nicolas
> 
> -----Original Message-----
> From: Nicolas Cadieux <njacadieux.gitlab at gmail.com> 
> Sent: Tuesday, June 30, 2020 2:17 PM
> To: cgray at graytechsoftware.com
> Cc: qgis-user at lists.osgeo.org
> Subject: Re: [Qgis-user] Hillshade Rough Blocks for 1m Rasters?
> 
> Hi,
> 
> To my knowledge, what you are seeing is just the result of zooming in too much using the nearest neighbour algorithm.  I don’t think it’s a bug. If you change the sun angle and the sun position, you may have better results. Then, play with the Contrast and layer blending.  I have seen this Pixel border effect on QGIS since version 1.8.  Someone correct me if I am wrong.
> 
> Nicolas Cadieux
> Ça va bien aller!
> 
>> Le 29 juin 2020 à 23:37, cgray at graytechsoftware.com a écrit :
>> 
>> Nicolas,
>> 
>> My x, y, and z are all in meters for my raster. The CRS for the project is also x, y, and z in meters. 
>> 
>> I tried your suggestion for changing the resampling. It definitely eliminated the problem, however it results in too much smoothing for my purpose. I do want to be able to see the actual pixel, but I want that pixel to be a single color representative of its value rather than a color with a boarder as shown in my screenshots. 
>> 
>> It seems this only occurs when the raster resolution matches the base unit of the CRS. I think this is a bug, but I have not exhausted the configuration options yet. 
>> 
>> Thank you, Chris
>> 
>> -----Original Message-----
>> From: Qgis-user <qgis-user-bounces at lists.osgeo.org> On Behalf Of 
>> Priv.-Doz. Dr. Maria Shinoto
>> Sent: Monday, June 29, 2020 7:15 PM
>> To: Nicolas Cadieux <njacadieux.gitlab at gmail.com>
>> Cc: Alister Hood <alister.hood at gmail.com>; qgis-user at lists.osgeo.org
>> Subject: Re: [Qgis-user] Hillshade Rough Blocks for 1m Rasters?
>> 
>> A month ago or so I had exactly the same problem and could solve it with these tips from Nicolas. Good luck!
>> 
>> Maria
>> 
>> 
>>>> Am 30.06.2020 um 07:32 schrieb Nicolas Cadieux <njacadieux.gitlab at gmail.com>:
>>> 
>>> Hi,
>>> 
>>> If your x and y distance unit is different from the z distance (like long lat degrees for x and y and meters for z), you will need a z Factor, or you need to save the raster in a crs that has x,y,z in the same unit (like meter).
>>> 
>>> https://www.esri.com/arcgis-blog/products/product/imagery/setting-the
>>> -
>>> z-factor-parameter-correctly/
>>> 
>>> You can also improve the look in the layer/Symbology/Resampling and select cubic for "zoomed in" and Average for "zoomed out".
>>> 
>>> Nicolas
>>> 
>>>> On 2020-06-29 6:23 p.m., Alister Hood wrote:
>>>> It does shade the pixel, but it also shades the edges unless the layer is being reprojected.
>>>> There is a ticket for this, but you can workaround it by enabling resampling for that layer.
>>>> 
>>>> Date: Mon, 29 Jun 2020 16:57:11 -0400
>>>> From: <cgray at graytechsoftware.com>
>>>> To: <qgis-user at lists.osgeo.org>
>>>> Subject: [Qgis-user] Hillshade Rough Blocks for 1m Rasters?
>>>> Message-ID: <024601d64e57$dcbddd00$96399700$@graytechsoftware.com>
>>>> Content-Type: text/plain; charset="utf-8"
>>>> 
>>>> If I use a 1m resolution raster on a CRS in meters and render it as 
>>>> a hillshade, it appears to shade the edges of each pixel instead of 
>>>> the pixel
>>>> itself:
>>>> 
>>>> If I use, a raster made in a CRS with ft for units displayed on a 
>>>> CRS in meters, it renders each pixel smoothly:
>>>> 
>>>> How do I get the hillshade to render smoothly always, especially in 
>>>> a CRS in meters when using metric units for the raster?
>>>> 
>>>> Thank you, Chris
>>>> 
>>>> 
>>>> _______________________________________________
>>>> Qgis-user mailing list
>>>> 
>>>> Qgis-user at lists.osgeo.org
>>>> 
>>>> List info: 
>>>> https://lists.osgeo.org/mailman/listinfo/qgis-user
>>>> 
>>>> Unsubscribe: 
>>>> https://lists.osgeo.org/mailman/listinfo/qgis-user
>>> _______________________________________________
>>> Qgis-user mailing list
>>> Qgis-user at lists.osgeo.org
>>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> 
>> _______________________________________________
>> Qgis-user mailing list
>> Qgis-user at lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> 
>> _______________________________________________
>> Qgis-user mailing list
>> Qgis-user at lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
> 
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user


More information about the Qgis-user mailing list