This article outlines the minimum and recommended hardware and software specifications necessary to achieve optimal performance with PIX4Dfields. It also provides a comparison of processing times between different machines.
IN THIS ARTICLE
Operating system
Windows:
- PIX4Dfields 1.12 and newer versions: Windows 10 (64 Bit) and Windows 11 (64 Bit) supported.
- PIX4Dfields 1.11 and previous versions: Windows 10 (64 Bit) supported.
macOS:
- PIX4Dfields 2.9: Sonoma (14) and Sequoia (15). Native Apple silicon support.
Intel processors are not supported. - PIX4Dfields 2.8: Ventura (13.0) or above. All chips.
- PIX4Dfields 2.9: Sonoma (14) and Sequoia (15). Native Apple silicon support.
Earlier versions of PIX4Dfields are compatible with older macOS versions; however, please note that Pix4D does not offer support for any issues that may arise.
-
- PIX4Dfields 2.7: Ventura (13.0) or above. All chips.
- PIX4Dfields 2.4: Monterey (12.0) or above.
- PIX4Dfields 2.3.1 and previous versions: Big Sur (11.0).
Apple devices with Intel processors should continue using PIX4Dfields version 2.8 or previous, and avoid updates.
Minimum specs
 Windows |
macOS |
|
| CPU |
|
|
| GPU |
|
|
| RAM |
|
|
| Disk |
|
|
| Display | 1024×768 display resolution (or higher). | |
Recommended specs
| Windows |
macOS |
|
| CPU |
|
|
| GPU |
|
|
| RAM |
|
|
| Disk |
|
|
| Display | 1024×768 display resolution (or higher). | |
The hierarchy of components that significantly influence the performance of PIX4Dfields is as follows:
1 - CPU Cores.
2 - Fast SSD (NVME).
3 - RAM.
4 - GPU (Nvidia preferred).
Sign up for a 15-day trial license to test PIX4Dfields using our Example use cases.
Processing times comparison in PIX4Dfields 2.9
The five machines underwent testing by processing five RGB and five Multispectral (MSP) datasets captured by various cameras and drones.
The processing times are derived from the average of all times recorded for each machine across different datasets and processing pipelines.
| Machine | Specifications |
| Mac M1 | Macbook M1 Pro - 32GB RAM - OS Sequoia 15.2 |
| Mac M4 | Mac mini M4 - 24GB RAM - OS Sequoia 15.2 |
| Windows i7 | i7 10750H 6 cores - 16GB RAM - RTX 1650Ti - SSD NVME |
| Windows i9 | i9 13900H 14 cores - 64GB RAM - RTX 4070 - SSD NVME |
| Windows AMD Ryzen | Ryzen 7 3700X 8 cores - 32 GB RAM - RTX 3080 |
The results were as follows:
Fast processing pipeline
MacBooks (M1 and M4) deliver similar total processing times for the fast processing pipeline, beating the Windows options.
- The data show that in the Fast processing pipeline, even a modest Mac M1 or a mid-range Windows CPU/GPU can produce results in just a few minutes.
- For example, macOS M1 averages ~5 minutes (MSP + RGB). That’s very fast compared to the high-end i9 system (~7 minutes) or AMD (~7minutes).
- This implies that an expensive, top-of-the-line machine is unnecessary to run the fast pipeline quickly. A cheaper i7 (or an M1 Mac) could suffice if fast processing pipeline is used.
Accurate processing pipeline
The Mac M4 was the fastest processor for the accurate processing pipeline, followed very closely by the Windows i9 machine and Mac M1.
- When Accurate processing is needed due to the need for a detailed Digital Surface Model, differences in machine capacity are more remarkable.
- In such scenarios, utilizing a more powerful machine can significantly reduce processing times, potentially saving you tens of minutes or even hours on larger projects. If you frequently undertake substantial accurate tasks, investing in a mid-range to high-end system could prove beneficial.
Datasets used for the comparison
| Dataset/camera | Sensor | Images | Hectare | GSD [cm/px] | Size GB | Average processing time (minutes) | |
| Fast | Accurate | ||||||
| 1- Altum | 3.2 MP per band | 1404 | 8.4 | 2.8 | 6.97 | 1.1 | 5.5 |
| 2- M3M - MSP | 2 MP per band | 3092 | 22 | 3 | 27.2 | 5.3 | 23.6 |
| 3- M3M - RGB | 20 MP | 775 | 24 | 1.8 | 4.9 | 2.7 | 22.1 |
| 4- M3M - RGB | 20 MP | 1686 | 38 | 1 | 16.6 | 6.9 | 51.8 |
| 5- M3M - MSP | 2 MP per band | 3212 | 42 | 3.6 | 30.2 | 5.3 | 21.1 |
| 6- Sequoia - MSP | 1.2 MP per band | 1724 | 88 | 10.4 | 3.95 | 1.1 | 8.4 |
| 7- S.O.D.A. - RGB | 20 MP | 374 | 92 | 2.7 | 3.15 | 1.3 | 11.3 |
| 8- Phantom 4 - RGB | 12 MP | 1655 | 250 | 3.4 | 11.9 | 6.7 | 49.5 |
What is the maximum number of images that PIX4Dfields can process?
PIX4Dfields does not limit the number of input images or the size of the generated orthomosaic. The maximum is determined by hardware limitations (disk space, memory size, operating system maximum file size) and time limitations (the time to compute an orthomosaic increases quickly beyond an input dataset size of several thousand images). Projects of more than 20.000 images were processed using PIX4Dfields.
Example datasets used for these tests can be found here: Example use cases - PIX4Dfields.
Important: The information presented in this article is based on Pix4D's testing experiences and should be regarded as a general reference rather than a definitive assurance.