Responses to user support requests. Weekly overview. Part 1.
Q: After more than one year, I can finally changed the OS, the program seems to work well, but I have a problem… When I install the program on the RockPro64 (ARM), I have no access rights to the archive folder if I want to access it through krusader or pcmanfm – using Armbian latest build.
What can I do to make the archive folders (and possible contents) not as root accessed but user accessed?
Using the 21.11.18 version now.
A: This was the standard logic for older versions of Xeoma when working on Unix: Xeoma’s server runs as root, so the files it creates are also owned by root, which adds one extra layer of security. The current version of Xeoma follows a different logic: the user who initiated the installation of Xeoma becomes the default owner of its files. With this in mind, the complete solution would be to get a renewal and update.
Alternatively, you can change the owner of archive files manually en masse (e.g. using chown -R) and automate this process (e.g. using chrontab).
Q: Are there any disadvantages if I use the xeoma.app (binary)? Because it seems it works well with the folders, and I don’t use/prefer to use the program as a service – I open it from time-to-time, so don’t need 24 hour surveillance.
A: The only notable disadvantage is reliability: if Xeoma’s process gets interrupted or the server machine gets rebooted (e.g. due to a power outage), then Xeoma will NOT restart itself automatically.
If this isn’t relevant in your case – yes, you can run Xeoma as a process without using a service, thus bypassing the need for root privileges.
Q: the manager of the local tourist port, asks me for a solution that allows to control the presence of boats at the dock stalls; obviously all this should happen automatically, and I thought that we could adopt the xeoma solution with the parking module. Have you had the opportunity to experiment with a solution of this kind at some customer and, above all, do you think it is adoptable, considering that boats, unlike cars, are subject to small lateral drifts due to the wave motion of maneuvering and this could confuse the algorithm used, falsifying the presence/absence of the boat. Do you have any ideas on this subject that you can suggest?
A: Yes, you can put “Object Recognizer” (set up to react to boats) before “Parking Spots” in a chain of modules – this way the latter will track the positions of the objects detected by the former.
The boats moving slightly shouldn’t cause trouble, because both modules react to the presence of an object, not necessarily to its movement. As long as the “point” you create in “Parking Spots” settings still touches the boat – this place is considered occupied.
We can provide you temporary licenses to test this directly with the customer’s cameras.
Q: The motion detector detects events that are not genuine motion, but are instead triggered by varying angles of sunlight
Unfortunately, I am unable to solve this issue. Decreasing the sensitivity of the motion detector by moving the slider to the right results in real motion (such as people) not being detected at all anymore.
A: At present there is only one reliable may to circumvent this issue – use “Human recognition” option in the “Motion Detector” settings.
Please note that this requires a Xeoma Pro license and enough CPU capacity to handle object recognition.
Q: I have a problem with one specific camera type:
Reolink Duo 3 PoE
The preview/substream is displayed correct, when accessing the server with windows xeoma client, but when accessing the server with the iOS mobile client, it shows for this specific cam “Error while decoding .. activate decoding on server .. ”
When I tap on this message, the main camera stream opens just fine. Probably there are problems with the unusual size/ Resolution for the preview stream?
A: Most probably, you get this message because it’s hard for a mobile device (iOS Client) to decode videos from all cameras.
Please follow the advice that you get in Xeoma: you should enable decoding on the server side. To do that, please open the Layouts menu in Xeoma on your Server machine (most right button on the bottom panel) – Client decoding settings – and choose “Disabled” for “Video decoding on the client side for live preview” and/or for “Video decoding on the client side for archive view” options. This way decoding will be performed on the server side.
Request in German:
Q: ich betreibe 4 IP-Kameras in Verbindung mit einer nicht mehr ganz neuen Software und möchte deshalb Xeoma testen. Für diesen Zweck habe ich mir die Testversion heruntergeladen. Nach dem Start des Programms kann ich die erste Kamera einrichten und sehe auch das Bild. Allerdings ist damit mein Rechner komplett überlastet, er reagiert nur noch sehr langsam auf Mausklicks, die CPU-Auslastung ist permanent bei 100%. Ein Test des Programms ist damit auch nur ansatzweise unmöglich.
Mein System: Windows 10/64, 4-Kern CPU 3700 Mhz, 8 GByte RAM
Ich kann mir nicht vorstellen, dass die Software diesen Rechner grundsätzlich überfordert. Möglicherweise ist nur eine Einstellung fehlerhaft. Was mache ich falsch?
A:
DE:
Ja, Xeoma versucht grundsätzlich, die CPU-Auslastung so gering wie möglich zu halten, um Ihr System nicht zu überlasten. Wir empfehlen Ihnen, diese Anleitung zur Optimierung der CPU-Auslastung zu befolgen: https://felenasoft.com/xeoma/en/articles/reducing-cpu-load-full-guide/
Sollte das Problem danach weiterhin bestehen, teilen Sie uns bitte Folgendes mit:
1) Reagiert die Benutzeroberfläche schneller, wenn Sie die Kamera (über das Modul „Universalkamera“) ausschalten, oder ist sie weiterhin langsam?
2) Sind in der Modulkette für diese Kamera Detektoren enthalten?
3) Was passiert, wenn Sie Xeoma Version 24.2.14 (hier herunterladbar) verwenden? Funktioniert diese schneller?
Bei Fragen, Anregungen usw. melden Sie sich bitte bei uns. Wir helfen Ihnen gerne weiter.
EN:
Yes, Xeoma generally tries to keep the CPU load as low as possible, so it shouldn’t overload your system. We recommend following this guide to optimize the CPU load: https://felenasoft.com/xeoma/en/articles/reducing-cpu-load-full-guide/
If the issue persists after that, please tell us the following:
1) When you turn off the camera (via the “Universal Camera” module) does the interface become more responsive, or is it still slow?
2) Do you have any detectors in the chain of modules for this camera?
3) What happens if you run Xeoma version 24.2.14 (downloadable here)? Does is work faster?
May 23, 2025