How to set up P1468-LE with Blue Iris

[smiticans]

Same, no option for 15 fps. I'm running the latest Axis OS (12.1.64).


This is why I'm trying to compare moving objects. I know it's easy for cams to generate beautiful static pictures. I guess the night video of the car is not terrible, but I suppose I expected a sharper picture and more detail because it's a 8mp cam with 1/1.2" sensor. I can't compare it with a Dahua 5442 because I don't have that cam at the same spot, but I feel like I could capture a very similar picture with the same amount of forensic detail even though the Dahua is 5x cheaper than the Axis.

I'm still wondering though what's causing the blurring or "compression" artifacts. Is the shutter speed too slow? Is it WDR? Is it the level of compression and/or Zipstream? Or am I simply expecting too much and is the night capture of the car actually a good capture?

Hmm, I'm running the same firmware version too.

What is the specific purpose of this camera? Is it to get vehicles coming and going through the gate, ID people exiting the vehicle to open the gate or to capture license plates?

The problem is that you're running the camera in IR mode in complete darkness. The camera is 4K with a 1 1/2 image sensor but the aperture is still F1.6 where the Color 4K-T is F1.0 (lets in more light) You're only going to be able to do so much with a camera that doesn't have any light. I would start with turning off the WDR while it's in IR mode and see what kind of a difference it makes. You can also change the shutter to 1/100.

If you feel the image is still compressed, change the quality to 80, 90 or 100 on the media profile. I believe the quality setting is the compression setting. So running quality of 70 would be compression of 30. I don't run mine this way so you would need to test it and monitor your bitrate.

I would change one setting at a time (WDR, Shutter etc.) and see what kind of a difference it makes.

Axis cameras are way more expensive than Dahua cameras. But they're more expensive for a number of reasons including build quality, long term firmware support, cybersecurity compliance and a ton of other features that a average homeowner won't use. If you're expecting the Axis camera to have a 10x better image than the Dahua camera because it is 10x the cost its just not going to happen.

 

[smiticans]

The overview image of the front porch is a Dahua Color 4K-T and the front door camera is an Axis Q3538-LVE. Obviously they have different mounting positions so it's not a direct comparison but its as close as I can get at the moment. Back to my observation of the Dahua camera having a constant bitrate and having a nicer looking image where the Axis camera has a variable bitrate and does not look as nice when it's a static image without a person in the FOV. The Axis camera runs at a reduced bitrate and increases it as an object enters the FOV. If you look at the 3rd image I attached, as the Amazon driver entered the scene, the camera ramped up the bitrate to over 12 Mb/s. Without an object in the scene, the bitrate will vary between 4-8 Mb/s during the day and can decease to 1 Mb/s at night.

The Axis camera has a lot of image processing going on which most of the time results in a very good image of a person. I've noticed if I walk up to my front door camera at night and I'm right on top of it where my body is taking up almost all of the FOV, it adjusts the image so it can still capture my face. However, when it makes this adjustment, the rest of the image is slightly distorted. I think something similar might be going on with sample video you posted. The truck is taking up a large portion of the image and it's also shining light into the lens of the camera and it's trying to compensate for it. If I had to take a guess, I'd say the WDR is what's causing the slight blurring on the truck in your scene. In addition to that, if you're trying to get a good image of vehicles moving at the speed of that truck, you're probably going to have to increase the max shutter to 1/100 or more.

 

[tuvow]

Thanks @smiticans! You're absolutely right that I shouldn't expect 5x improvement of picture quality if the price of the camera is 5x. But I hoped for 2x improvement because of the 2x increase of pixels at same pixel/sensor ratio. Maybe I've been judging the Axis camera a little too harshly. It's fair to say that the position of the Axis camera is the most difficult scene that I have in terms of lighting. I do feel the quality has improved since I did a factory default.

Next steps are to clean up the spider web (it's insane how fast a spider covered the cam within a few days of installing the cam), test with 1/100 shutter time and lower/no WDR.

The only thing I can't seem to figure out is why I'm unable to select 15 FPS in the onvif media profile. I suspect you maybe have this option because you're using ACS and it seems to create additional settings/options once you add the cam to ACS. I think for debugging purposes I'm going to run the cam at 25 FPS for now. This shouldn't hurt picture quality as long as I have VBR configured, right?

 

[smiticans]

Thanks @smiticans! You're absolutely right that I shouldn't expect 5x improvement of picture quality if the price of the camera is 5x. But I hoped for 2x improvement because of the 2x increase of pixels at same pixel/sensor ratio. Maybe I've been judging the Axis camera a little too harshly. It's fair to say that the position of the Axis camera is the most difficult scene that I have in terms of lighting. I do feel the quality has improved since I did a factory default.

Next steps are to clean up the spider web (it's insane how fast a spider covered the cam within a few days of installing the cam), test with 1/100 shutter time and lower/no WDR.

The only thing I can't seem to figure out is why I'm unable to select 15 FPS in the onvif media profile. I suspect you maybe have this option because you're using ACS and it seems to create additional settings/options once you add the cam to ACS. I think for debugging purposes I'm going to run the cam at 25 FPS for now. This shouldn't hurt picture quality as long as I have VBR configured, right?

You should be able to improve the image quality by making some of those adjustments. I just didn't want you to have an unreasonable expectation of this camera. In my opinion, based on my install (replaced a Dahua 4K-T) and other locations I installed them, the image quality is better than the Dahua's because of its advanced image processing, doesn't suffer from pixelation if I don't run it at a high bitrate (issue is strictly with my 4K-T's) and doesn't have a close focus distance limitation. Even even tho this cameras has an aperture of F1.6 it can keep up with the low light sensitivity of the Dahua 4K-T with an F1.0 aperture. Just keep in mind, the more the camera is zoomed in, the smaller the aperture will get and less light it will let in which will reduce the low light sensitivity. I'm very happy with how this camera runs in full color at night and also how it handles IDing faces in IR mode. I haven't really tested it in IR mode with vehicles moving at the speed they are in your scene so I suspect the WDR has some responsibility in the ghosting of the vehicle. You're going to have to test and see which settings work best. If WDR is an issue, I can post a screenshot of how to create an event rule that will turn the WDR off when the IR is on. You want WDR on when the camera is in full color mode because it makes a difference in image quality and being able to pull important details from an image in difficult lighting conditions. If the purpose of this camera is to strictly get details of vehicles entering/exiting the gate, I would install a white light and run this camera in full color at night. There's only so much detail you're going to obtain from a vehicle while the camera is in black and white running IR. You loose a lot of detail this way. The other option would be to use it as a LPR so at least you have license plates at night.

When I have more time, I'm going to dig into that issue and see if I can recreate it. I don't see how or why ACS would affect the ONVIF media profile as both ACS and the other VMS I run this camera on do not use ONVIF to connect to this camera. My ONVIF account is disabled in my camera.

I don't think it'll hurt the image quality but I prefer to run mine at 15FPS with the other settings I posted. If you want, you can download a free trial of ACS Pro from Axis's website and see if that affects any of those settings.

Let me know how it goes when you make those adjustments.

 

[smiticans]

@wittaj

I was able to get night captures with my P1468-LE and Q3538-LVE in IR without any white light.

Axis P1468-LE Night IR video;

Max exposure was set to 1/60 for this test video.



Axis Q3538-LVE Night IR video;

Max exposure was set to 1/100 (This is what I normally run it at with white light). Note, the slight IR reflection is from the railings right in front of it, not the dome. The dome has an IR shielding coating on it so it does not reflect IR.



Axis Q3538-LVE Night video with white light;

White light is an A21 13w LED bulb which produces 1600 Lumens. The fixture is mounted right above the camera so unfortunately it shines directly into the dome, but it doesn't effect it enough for me to fix it.