IPC-Color4K-T and IPC-Color4K-T180 230227 Firmware
- Thread starter EMPIRETECANDY
- Start date
Anyone else experience an issue with this latest firmware on the Color-4k-T where after about a week of uptime at some point through the morning after sunrise it just goes to complete white and all you can see is the text from the overlay. If I login to the web interface it takes a really long time to display if it does at all. Changing the settings between day or night or making other adjustments doesn’t seem to make a difference.
After I reboot the camera it works fine. This happened on 3-24 and then again today. Last time it was around 6:30 that it started and gradually progressed to full white by 7:45. This time it went from normal to full white in 1 second. After rebooting it works fine again, but this could lead to a real problem if it misses something due to doing this. I also noticed that according to the log on the camera that the backlight, NR, and exposure save config every night at 23:59:59 and then again at 00:00:00. I’m hoping that’s just displaying that though since I have a schedule set to change from night settings to day settings at 6:00 am to 7:00 pm every day. I do find it odd though that these same save config log messages do not show up at the 6:00 am and 7:00 pm time frame set by the schedule though.
After I reboot the camera it works fine. This happened on 3-24 and then again today. Last time it was around 6:30 that it started and gradually progressed to full white by 7:45. This time it went from normal to full white in 1 second. After rebooting it works fine again, but this could lead to a real problem if it misses something due to doing this. I also noticed that according to the log on the camera that the backlight, NR, and exposure save config every night at 23:59:59 and then again at 00:00:00. I’m hoping that’s just displaying that though since I have a schedule set to change from night settings to day settings at 6:00 am to 7:00 pm every day. I do find it odd though that these same save config log messages do not show up at the 6:00 am and 7:00 pm time frame set by the schedule though.
Attachments
I have been running the latest firmware fine without problem.
That image looks like the shutter is too slow so it is blinding out the image at sunrise.
I would confirm your schedule isn't somehow messed up.
Also maybe try a custom shutter range of like 0-8.33ms so that if it gets more light the shutter can adjust.
That image looks like the shutter is too slow so it is blinding out the image at sunrise.
I would confirm your schedule isn't somehow messed up.
Also maybe try a custom shutter range of like 0-8.33ms so that if it gets more light the shutter can adjust.
Shutter/exposure is actually set to auto after last reset. The only settings I’m adjusting with the schedule are HLC to 100 at night to prevent bright light from headlights and saturation/contrast to 40 each currently. If you look at the ones from today it was actually very overcast, then it instantly went to solid white within 1 second and stayed that way until 3:40 this afternoon until I rebooted the camera.
I'm running firmware version V3.120.0000000.14.R, Build Date: 2023-02-27. It happen to my camera when I was testing it the first morning.
But it was probably like wittaj mentioned because of the shutter speed is too slow at the time. I switch to auto and no problem since. The color
still change from night time setting to day time setting in the morning. But not blinding white like your screen shot. And I don't use HLC at all.
But it was probably like wittaj mentioned because of the shutter speed is too slow at the time. I switch to auto and no problem since. The color
still change from night time setting to day time setting in the morning. But not blinding white like your screen shot. And I don't use HLC at all.
I have this exact same issue running the latest firmware. The White LED's are staying on 24/7 as well. A reboot clears it up, but it didn't fix it for to long last time I rebooted it.
I ended up doing the factory reset 3 times and am using adaptive scene rather than setting a day night schedule now and I haven’t had the issue anymore.
I was using an adaptive scene when the issue happened.
Also when it happened there was nothing I could do to correct it. I tried setting manual exposure speeds, changing the current profile running, etc, but nothing would get it to look proper until I rebooted.
That’s the exact same issue I had but after factory resetting 3 times it’s been fine for a couple weeks now.
WSCam
Getting the hang of it
Fortunately I haven't had any issues like this. I've got that same 2023-02-27 firmware. I'm just using the adaptive scene. On all my other cameras I have day/night settings and use a script to switch them at sunrise/sunset (my own home-rolled, but basically the same as the sunrise/sunset script people have worked on here). I was going to do that on this camera too but so far I've had good success just using the adaptive scene settings. I did set my shutter/exposure, and since I have illumination where the camera is I turned off the lights on the camera itself, and it has worked properly this way since unboxing.
During testing, I turned the lights on/off in my dark network closet, and I was surprised at how well it did just using the lighting from all the network blinky lights in the room, without the cam lights. That encouraged me to leave them off when I put them in their ultimate spot and it works well.
The lighting where I have the camera, it's a security light (LED) that has the lights on all the time, but when there's motion it turns them on brighter for a couple minutes. The end result on the camera is that normally there's still plenty of light, and when something is there and those lights get brighter, there's a little time where the camera sees the increased light and it's brighter for a bit, but the exposure ranges adjust a bit and I wind up with a correctly exposed image again within a few seconds. It's not like the image is being washed out or anything during that short time, it's really just the street gets a little brighter.
I do wonder how it would look if I turned those security lights off... I may give that a try sometime just to see, not that I'd ever want to do that. The lights and camera itself are on a 4x4 post near the street right over our row of mailboxes, looking both ways (thanks 180 degrees!) and the lights + camera are hopefully a deterrent to the mailbox thieves (and catalytic converter thieves; yes, that has happened too).
This just happened again last night/morning even with the Adaptive Scene setting enabled.
@EMPIRETECANDY
I have had it happen twice and both times rebooting fixed it. The two times it happened were almost back to back, but now its been fine for at least a week. Although it ran fine for several weeks before the issue happened in the first place. Seems like something funky is going on in the firmware but doesn't seem consistent so far.
I think I had SSA turned on as well, which I believe I have turned off now, so not sure if that made the difference, will see if it happens again.
CCTVCam
Known around here
- Joined
- Sep 25, 2017
- Messages
- 2,674
- Reaction score
- 3,505
Andy is it possible to put in a request to Dahua for an updated firmware with another higher quality option for encoding?
One thing I've noticed recently is that putting the encoder into H265 gives an improved picture (albeit at the cost of some artefact risk).
However, breaking this down into bit rates and compression etc. if H265 is 35% more efficient than h264, then a bit rate of 16,800 on h265 is equivalent to a bit rate of 22,680 on H264. Now noticing there's an improvement in picture, I believe there is more to come and not specifically from the bit rate but the actual compression level.
When I look at the background of the 4K picture or if you zoom in, it's obvious that the quality drops off quickly on zooming despite there being 8mp. It seems full of compression artefacts. Compare this to a normal 8mp phone for example, and the quality on the phone is much higher. I'm not trying to say here it's possible to achieve phone or dedicated video camera quality, as clearly the storage required would be massive. However, I do believe there's more to give for those who maybe don't care so much about compressing the files quite as small eg home users and small business users who only maybe run 3 or 4 cameras or so.
Looking in Blue Iris, the 16,864kbs bit rate I have set in camera, appears to be compressed in camera to around a 2,000kbs streaming bit rate which suggests a compression ration of approximately 8:1.
What I was wondering if maybe we could request from Dahua is an in camera OPTION of High Quality Version of the 264H and 265 codecs that have instead maybe a 4:1 compression ratio ie an output stream of maybe 4,000kbs. I wouldn't even be upset by a 3:1 ratio personally. 4:1 would potentially double the required storage size, but would in turn probably massively increase the visual quality based on lowering the amount of material compressed and guessed and the resulting artefacts / poor resolution seen when zooming the picture even modestly. It would also have the advantage of taking some load off the cameras CPU (which is a little stretched as standard) as less compression = less calculations to perform when compressing and thus less load on the in camera cpu.
There's no doubt these compression options wouldn't be for everyone as big organisations may still favour standard compression and small file size over a low compression rate and larger file size albeit at higher quality. However, for those for whom storage isn't critical or an issue - and lets face it here, storage is becoming much cheaper for everyone almost daily, having the Option of upping the quality with extra high quality h264H and H265 presets, would be a nice addition. However, this is something that can probably only be done in camera through the in camera compression as increasing the bit rate can only go so far versus what the camera has compressed away in the 1st place.
BTW, if I have any of the theory or presumptions about compression wrong, I'm pleased to be corrected. I don't claim to be an expert.
One thing I've noticed recently is that putting the encoder into H265 gives an improved picture (albeit at the cost of some artefact risk).
However, breaking this down into bit rates and compression etc. if H265 is 35% more efficient than h264, then a bit rate of 16,800 on h265 is equivalent to a bit rate of 22,680 on H264. Now noticing there's an improvement in picture, I believe there is more to come and not specifically from the bit rate but the actual compression level.
When I look at the background of the 4K picture or if you zoom in, it's obvious that the quality drops off quickly on zooming despite there being 8mp. It seems full of compression artefacts. Compare this to a normal 8mp phone for example, and the quality on the phone is much higher. I'm not trying to say here it's possible to achieve phone or dedicated video camera quality, as clearly the storage required would be massive. However, I do believe there's more to give for those who maybe don't care so much about compressing the files quite as small eg home users and small business users who only maybe run 3 or 4 cameras or so.
Looking in Blue Iris, the 16,864kbs bit rate I have set in camera, appears to be compressed in camera to around a 2,000kbs streaming bit rate which suggests a compression ration of approximately 8:1.
What I was wondering if maybe we could request from Dahua is an in camera OPTION of High Quality Version of the 264H and 265 codecs that have instead maybe a 4:1 compression ratio ie an output stream of maybe 4,000kbs. I wouldn't even be upset by a 3:1 ratio personally. 4:1 would potentially double the required storage size, but would in turn probably massively increase the visual quality based on lowering the amount of material compressed and guessed and the resulting artefacts / poor resolution seen when zooming the picture even modestly. It would also have the advantage of taking some load off the cameras CPU (which is a little stretched as standard) as less compression = less calculations to perform when compressing and thus less load on the in camera cpu.
There's no doubt these compression options wouldn't be for everyone as big organisations may still favour standard compression and small file size over a low compression rate and larger file size albeit at higher quality. However, for those for whom storage isn't critical or an issue - and lets face it here, storage is becoming much cheaper for everyone almost daily, having the Option of upping the quality with extra high quality h264H and H265 presets, would be a nice addition. However, this is something that can probably only be done in camera through the in camera compression as increasing the bit rate can only go so far versus what the camera has compressed away in the 1st place.
BTW, if I have any of the theory or presumptions about compression wrong, I'm pleased to be corrected. I don't claim to be an expert.
Keep in mind I bet most of their intended market leaves things on default/auto settings (especially from all the anecdotal posts of someone coming here after paying an authorized installer and all they did was hang them up and run the wires), so I doubt they would entertain this idea, but you never know...
Most here will probably say a phone camera can produce a better image than these types of cameras.
A single 8MP image from a "real" camera could be upwards of 5MB of storage. In surveillance cameras, if you record at 15FPS, every second of video could be 75MB or more, which could equate to 6.5TB per day per camera. Obviously most are not going to have that kind of storage, so lossy compression algorithms are used to reduce storage and network bandwidth requirement, and that can add noise.
A phone camera is used mainly under ideal conditions, whereas a surveillance camera is going 24/7 in every type of element, so the design and size impacts its capabilities. Hang an iphone outside and see how fast you blow thru storage and how long the phone lasts LOL.
Remember these are surveillance cameras, not DSLR cameras, so you have to check your expectations. For example, you can see individual hairs and skin pores with DSLR photography equipment and you won't with these kinds of cameras. These are for a different use and different expectations.
Digital zooming is never really good with these cameras because the sensors are too small and the compression and not something we recommend, but you stand a better chance of some digital zoom with H264 rather than a large macroblocked and higher compressed H265. I can digital zoom on my overview camera and kinda make out the address number of the house across the street with H264, but not a chance with H265 as it macroblocked his whole house.
tigerwillow1
Known around here
I think this is easily proven by observing the significant image quality improvement that can be had with ROI. It makes me wish the camera could just ROI the whole image. A possible reason preventing this would be not enough CPU horsepower in the camera.
CCTVCam
Known around here
- Joined
- Sep 25, 2017
- Messages
- 2,674
- Reaction score
- 3,505
I do agree they're apples and oranges and video storage considerations mean you're never going to be able to set the same quality level. However, for those of us who have less storage critical applications, I can't see why there couldn't be a much higher quality option that maybe trades storage space for image improvement. As said, it wouldn't be for everyone but that's why I'm suggesting an option rather than a change. That way for those for whom storage space isn't at a premium or mission critical, there is the option to choose one over the other.
WSCam
Getting the hang of it
I have a hunch that if anyone had oodles of storage available and wanted the best footage available, with the least amount of compression artifacts, they're probably using some type of bespoke setup, maybe even analog instead of IP cameras and doing the A/D conversion on something besides the camera itself, or simply using higher end cameras, not consumer grade equipment. I don't know who those people are, but I'm sure there are use cases where too many compression artifacts is simply not acceptable. The filmmaking industry may have some options along those lines (well, I'm almost positive they would... "RED" cameras with massive local storage, etc.)
CCTVCam
Known around here
- Joined
- Sep 25, 2017
- Messages
- 2,674
- Reaction score
- 3,505
Dahua cameras aren't consumer grade. They''re aimed squarely at professional installers.
As for using bespoke DSLR / Video cameras for cctv. Never going to happen unless you own a data centre. The bit rates are way too high. Even some prosumer cams can reach over 100mbits eg the GoPro Hero 10 records at 100mbits per second and it's not even a video camera but a sports cam.
100 mbits = 100,000,000 kbs vs the 16,000kbs even a 4kt records at raw or to put it another way, it's 6250 x as much data rate before compression and still a hell of a lot more after.
As for using bespoke DSLR / Video cameras for cctv. Never going to happen unless you own a data centre. The bit rates are way too high. Even some prosumer cams can reach over 100mbits eg the GoPro Hero 10 records at 100mbits per second and it's not even a video camera but a sports cam.
100 mbits = 100,000,000 kbs vs the 16,000kbs even a 4kt records at raw or to put it another way, it's 6250 x as much data rate before compression and still a hell of a lot more after.
WSCam
Getting the hang of it
All true. I was just throwing out suggestions if there was a need for higher quality.
I don't know what type of compression a phone uses for their videos. H.264 or H.265? Are they worried about optimizing it for network bandwidth like a surveillance camera would? I don't know. But I hear what you're saying. On my camera (the model in this topic) I have noticed that zooming in seems to have (maybe) more compression artifacts than others, but I also haven't tweaked it too much. This is my one camera out of all of them where I don't have Blue Iris getting the substream feed. It's at the street and records 24x7 so I want high quality. I've had issues in the past where motion detection failed me and even though I was recording 24x7, I could see the incidents in question but at the lower res. So I just said forget substream, just record the high quality and get larger disks... it's worth it if it helps catch the lowlife scum who pass through our area stealing stuff.
Maybe I should spend a little more time tweaking the bitrates and/or H.265 quality to see what helps (or drop to H.264 and see if that's useful). I do have to be a little mindful of bandwidth since I have a single 1 Gb/s connection that feeds a couple cameras and a wifi access point, but I'm pretty sure there's enough to spare if needed. I mean, right now it reports that camera's bitrate is 915 KB/s to Blue Iris. Total, with all cameras, I'm at about 8900 KB/s / 223 MP/s and the computer keeps up well enough (i9-10850K) until there's a lot of motion by my front door which uses AI, at which time the CPU spikes to 100%. I need a GPU in there, I think.
CCTVCam
Known around here
- Joined
- Sep 25, 2017
- Messages
- 2,674
- Reaction score
- 3,505
I have a 1gbs connection to the switch but each connection to the camera is 100mbs. It's more than enough. As I said above, 100 mbs is 100,000,000 kbs and the 4Kt is putting out just 2,000 kbs to stream. You're unlikely to have a bandwidth issue therefore even with many cameras.