Hello everyone! It would be very useful to have the option to define the resolution of the Orthomosaic and the Digital Surface Model independently. In my case, most of the time, I prefer the highest possible resolution for the Orthomosaic but a lower resolution for the DSM. This option would help save time and resources during the DSM generation process.
HI Alejandro, that’s possible! There is an option for the resulting GSD in both of the settings, for DSM and for Orthomosaic. Have you found it meanwhile?
Hello drones6,
Thank yo for your reply. Actually, there is not an option for this on Pix4D Matic, on the resolution options for DSM you can use custom and choose the resolution you want, but, in the orthomosaic, there is no option to use custom resolution, it only reads “Based on DSM”, see attached image. If there is another option I’m not aware of, appreciate if you can show me.
Hello Alejandro, you are totally right! Sorry for the confusion.
Yes, that’s another feature missing! As one of the main outputs for photogrammetry it should be possible to control the orthomosaic resolution independently from the DSM - that’s a fundamental function in all photogrammetry codes I know so far. Wasn’t aware this is missing in PIX4Dmatic… 
Dear developers, please implement this!
1 Like
Hello Alejandro, thank you for your message!
Indeed, in PIX4Dmatic it is not possible to set the resolution of the orthomosaic to be different from the DSM resolution. The reason is that the orthomosaic is produced based on DSM (Orthomosaic - PIX4Dmatic) and a lower DSM resolution might result in artefacts in the resulted othomosaic, in particular in the areas with a significant elevation gradient.
What is the main concern in your workflow if the DSM and orthomosaic resolutions are the same? In our tests the time to generate a DSM is significantly shorter comparing to other processing steps such as calibration or densification. Do you observe a big time difference in the DSM generation if you increase the GSD?
Kind regards,
Alexey
Hi Alexey - follow up question for you.
I see in the Orthomosaic documentation as well as your above message that the DSM and Ortho resolutions are the same. That being said, when inspecting my vertical accuracy (with my DSM) I get great results (SD=4.8 cm), though my horizontal accuracy (with my orthomosaic) is very poor comparably (25 cm).
@alexey.gribovskiy Does Pix4D does not georeference the images for the orthomosaic very tightly to the detailed DSM? Does it use a lower-res DSM for the orthorectification (which would lead to the horizontal error being high for the orthomosaic)? And if so, is there a way to make it be more tightly controlled?
Best,
Addy
Hello Alexey,
Thank you for your reply.
The issue is related to the time you mentioned. While it’s not a significant concern for small datasets, it becomes a substantial factor with larger datasets, such as 10,000 images from the Zenmuse P1, for example. You mentioned that the time required is shorter than the other process, and I agree; however, in scenarios like this, I would say it becomes comparable to the orthomosaic generation process, which is in my opinion far too long.
I’ve used Pix4D and similar software for many years and have not encountered the problems you described when working with different resolutions. In my experience, DSM generation time should not even come close to that of orthomosaic generation.
Pix4D Matic is exceptionally efficient in the point cloud generation process, but I believe a significant portion of that time efficiency is then lost again during DSM generation.
I hope this explanation makes sense, and thank you again for your reply!
Hello Alejandro,
Thank you very much for your feedback.
I see your point and, indeed, if a project area is relatively flat, without buildings or trees, then the resolution of the DSM has little impact on the orthomosaic. We are considering enabling the option to set the custom resolution for the othomosaic, I will let you know when I have an update on this.
Kind regards,
Alexey
Hello Addy,
Thank you for your message.
How do you measure the horizontal accuracy? Do you use the checkpoints?
There should be no difference in absolute accuracy between DSM and orthomosaic.
I suggest you reach out to our support team, they might need to request an example dataset from you. Alternatively you can open a separate thread in PIX4Dmatic Questions/Troubleshooting - Pix4D Community, so we can discuss your issue there.
Hope it helps.
Kind regards,
Alexey
Update: I see that you opened a thread, let’s discuss further there