Pix4Dmapper generates tiles for the orthomosaic and the DSM during step 3. DSM, Orthomosaic and Index.
Orthomosaic and DSM tiles are automatically generated so that each tile is 5000*5000 pixels. Tiles may be slightly smaller at the edges of the project area.
The size of the tiles cannot be defined according to a different tile file size or total number of tiles.
Article feedback (for troubleshooting, post here instead)
Is this a feature that is going to be supported? If it is on here, it is obviously being asked for. The current default size is way too small. For projects that are soo big that they actually need to be tiled, the number of tiles at 5000 pix^2 is just overwhelming to manage.
Since there are a lot of users that wish to export tiles using custom sizes and the idea is very interesting we will pass your suggestions to the product manager so they can implement the option in future releases.
At this moment, in order to split the output Orthomosaic into custom tiles, we advise you to have a look at different GIS software like ArcGis and QGIS. There is a plugin for QGIS called Grid Splitter. It allows the user to spit the input Orthomosaic in custom size tiles. You can find more information about QGIS and Grid Splitter following this links:
Hi. Would love to see an option to allow the user to either choose from some preset options or enter their own values for the tile size. Hope to see this option in the future.
We also would greatly benefit from custom tile sizes so we do not have to re-cut final outputs in other software.
Hello Nathan, Susan
Thank you for your input and feedback. We have added it to our suggestions list.
Good morning. I'm finally using a computer with enough RAM that allows me to choose point Density=High in step 2.Point Cloud and Mesh. The point cloud that i obtain represents much better the surface (average density=4.23 points/m3 compared to 0.42 points/m3) but the resulting DSM it's not much different from the one i obtained with a coarser point cloud, probably because the GSD remains the same. Is there any way to produce a raster DSM compatible with the new point cloud resolution? Thank you.
You are right. The GSD will remain the same for the DSM. The min value of GSD (cm/pixel) you can obtain for DSM is 5 times the GSD of the ortho. You can have a look at https://support.pix4d.com/hc/en-us/articles/115004864586-How-to-generate-the-point-cloud-classification (Point cloud classification gives better DSM since you can edit noises in the point cloud) and https://support.pix4d.com/hc/en-us/articles/115001671066-How-to-improve-the-Outputs-of-Pix4D-Desktop#label6