Weird intermittent blackouts with dahua starlight 5231's

[Philip Gonzales]

Well if i can do 60 FPS, why not? It looks so much smoother. And it can mean the difference between catching the license plate of a speeding car and not. Just saying, if they're rated to do 60, I'd prefer to set them that high... at least for now. But it worries me that they seem to be struggling at anything above 25, and now you're saying yours aren't reaching the 30 they're set at? So perhaps this is normal for these starlights? They're popular enough, you'd think someone would have thrown this out there by now if that were the case, or maybe nobody ever runs into the issue cause nobody ever bothers trying to run above 30 FPS lol.

As far as PAL vs NTSC, I assumed everything was shipped with US standards (NTSC) per my discussions with Andy. But if not and if it's just a matter of flashing new firmware, then that's no big deal to just upgrade. I can tell you that the camera settings default to 25 FPS, and 25 FPS is the maximum for the substream. However, there are both PAL and NTSC resolutions available in the drop down menu for main and substream resolutions, so i can't say for sure what it is. I'm just going to have to check on this one.

Sent from my VS980 4G using Tapatalk

I think catching the license plate would be a lot more likely with better camera position and the proper exposure setting. Not really the frame rate. I don't think anyone runs above 30 fps because there is no point.

 

[Mike A.]

Well if i can do 60 FPS, why not? It looks so much smoother. And it can mean the difference between catching the license plate of a speeding car and not. Just saying, if they're rated to do 60, I'd prefer to set them that high... at least for now. But it worries me that they seem to be struggling at anything above 25, and now you're saying yours aren't reaching the 30 they're set at? So perhaps this is normal for these starlights? They're popular enough, you'd think someone would have thrown this out there by now if that were the case, or maybe nobody ever runs into the issue cause nobody ever bothers trying to run above 30 FPS lol.

You can do whatever you want. But you're not really going to benefit in most circumstances. It's surveillance video, you're not making hi-res stop-motion movies and don't have all of the rest of the lens quality and lighting and sensor and image processing, etc., that you'd really need to do so. You're still getting poorly lit, lower-res, kinda noisy images captured through a relatively low-end lens just at a faster frame rate. And it's costing you a lot in processing and storage on the back end of that. With just two cams you probably can get away with it if you really want to. With more you'll need to start balancing that out more or spend a bunch more money to handle it for relatively little benefit.

I can set mine above 25 without seeing the same effect. What I was saying is that they don't necessarily run at a steady, constant 25 or 30 fps because that's you've picked. They'll bounce around. At least as reported by BI which is the only thing that I have handy which shows the running frame rate. Right now I have it set to 50 fps, CBR, 8,192. BI shows a frame rate of anywhere between ~15 and 37 fps, mostly staying between ~25-30. No problems running that way over about the last hour. So doesn't appear to be a problem running set over 25 fps with the cams generally. How that works in combination with the NVR may be different.

I do see the same thing for a moment when I hit save after changing the frame rate and the camera switches to a different frame rate. Best I can tell testing it a few times, that seems to be independent of what the FPS setting is when changed (i.e., it does it at 15 to 30 and back to 15). During the day now the difference isn't as large but you can still see that the exposure changes momentarily as it flips.

 

[Big Ry]

You can do whatever you want. But you're not really going to benefit in most circumstances. It's surveillance video, you're not making hi-res stop-motion movies and don't have all of the rest of the lens quality and lighting and sensor and image processing, etc., that you'd really need to do so. You're still getting poorly lit, lower-res, kinda noisy images captured through a relatively low-end lens just at a faster frame rate. And it's costing you a lot in processing and storage on the back end of that. With just two cams you probably can get away with it if you really want to. With more you'll need to start balancing that out more or spend a bunch more money to handle it for relatively little benefit.

I can set mine above 25 without seeing the same effect. What I was saying is that they don't necessarily run at a steady, constant 25 or 30 fps because that's you've picked. They'll bounce around. At least as reported by BI which is the only thing that I have handy which shows the running frame rate. Right now I have it set to 50 fps, CBR, 8,192. BI shows a frame rate of anywhere between ~15 and 37 fps, mostly staying between ~25-30. No problems running that way over about the last hour. So doesn't appear to be a problem running set over 25 fps with the cams generally. How that works in combination with the NVR may be different.

I do see the same thing for a moment when I hit save after changing the frame rate and the camera switches to a different frame rate. Best I can tell testing it a few times, that seems to be independent of what the FPS setting is when changed (i.e., it does it at 15 to 30 and back to 15). During the day now the difference isn't as large but you can still see that the exposure changes momentarily as it flips.

It's actually somewhat intermittent. There are periods where it doesn't do it at all, then there are periods where it does it ever 2-3 seconds non-stop. I can't find any rhyme or reason for this (happens any time of day/night with no changes in settings). More often than not, it's blacking out at least every 10 seconds. Last night when i was playing with settings, anything over like 28 fps caused the issue to start up, although the timing between blackouts would be a little greater. Didn't get to mess with it too much because it was already so late. Is there any sort of tutorial or guide on ideal/optimized settings for these cams or ipcams in general? I'm still very much a noob with all this. I just figured if the specs say it can handle it, i am ok to set it.

Sent from my VS980 4G using Tapatalk

 

[looney2ns]

15 FPS is plenty for security cams.
Constant bit rate.
Max out the bit rate if you like. Typically 4096 is enough.

 

[Big Ry]

15 FPS is plenty for security cams.
Constant bit rate.
Max out the bit rate if you like. Typically 4096 is enough.

I believe I'm at 8192 at the moment on the window cam. That's the NVR Max. Again, i was just going for the best available option. The back door is on the default, which i think was 2048. Also tried VBR with "quality" maxed out at 6 (best). That made little difference. Just curious, why CBR over VBR?

It's funny because according to tinycam, these cams aren't even hitting any of these numbers anyway. Like if i set it to 50 FPS, it jumps between like 10-20, dipping as low as 1 FPS. I never get the bit rate i set, not even close. Even the stable default cam isn't quite right. The frame rate is pretty close to 25, hovering below, but the bit rate is well below what's set.

Sent from my VS980 4G using Tapatalk

 

[tangent]

I think it's most likely some settings others have touched on but for the sake of being thorough, tell us about your cabling and consider trying a different cable.

 

[Big Ry]

I think it's most likely some settings other have touched on but for the sake of being thorough, tell us about your cabling and consider trying a different cable.

I talk about the cables in post #9 of this thread. Is there something else you need to know about them that might be helpful? I haven't tried new cables, but it seems highly highly unlikely that I'd have identical problems in 2 separate cams, on 2 separate channels, on 2 separate runs of CAT6 of very different lengths from different manufacturers. I guess for the record, the backdoor cam is on Home Depot special CAT6...don't think I mentioned that before. Either way, CAT6 is a standard, which has a standard wire gage and requirements for twisting, so I can't imagine there would be too much variability in brands unless I'm pushing the limits on length or exposure conditions or something.

 

[tangent]

I talk about the cables in post #9 of this thread. Is there something else you need to know about them that might be helpful? I haven't tried new cables, but it seems highly highly unlikely that I'd have identical problems in 2 separate cams, on 2 separate channels, on 2 separate runs of CAT6 of very different lengths from different manufacturers. I guess for the record, the backdoor cam is on Home Depot special CAT6...don't think I mentioned that before. Either way, CAT6 is a standard, which has a standard wire gage and requirements for twisting, so I can't imagine there would be too much variability in brands unless I'm pushing the limits on length or exposure conditions or something.

It doesn't sound like the cables are likely to be the issue, though I don't really like couplers especially with PoE . Sorry I didn't catch your earlier post on the cables you were using.

It isn't as standardized as you would hope and there have been some news stories about lots of cables that when tested fail to meet the spec. Unlike HDMI where you have to pay royalties to HDMI to legally use the term HDMI or get sued (though that doesn't stop counterfeit goods) there really isn't an organization aggressively going after fake Cat-6 cable makers. There have been quite a few people who've posted about problems that were ultimately caused by using 'flat' ethernet cables which are often 32AWG. Generally damaged cables either from rough handling or bad termination can be surprisingly frustrating and time consuming to diagnose. The integrity of your cabling is not something to just dismiss.

 

[Big Ry]

It doesn't sound like the cables are likely to be the issue, though I don't really like couplers especially with PoE . Sorry I didn't catch your earlier post on the cables you were using.

It isn't as standardized as you would hope and there have been some news stories about lots of cables that when tested fail to meet the spec. Unlike HDMI where you have to pay royalties to HDMI to legally use the term HDMI or get sued (though that doesn't stop counterfeit goods) there really isn't an organization aggressively going after fake Cat-6 cable makers. There have been quite a few people who've posted about problems that were ultimately caused by using 'flat' ethernet cables which are often 32AWG. Generally damaged cables either from rough handling or bad termination can be surprisingly frustrating and time consuming to diagnose. The integrity of your cabling is not something to just dismiss.

Good to know. Thanks. I know HD products aren't the best, but I would hope they're common enough to not be considered complete garbage. As for monoprice, they may exclusively source their wires from china on the cheap, but they've always had great products. The cables I bought had a ton of stellar reviews too (btw, looks like they're 24awg).

Monoprice cables: FLEXboot Series Cat6 24AWG UTP Ethernet Network Patch Cable, 5ft Black - Monoprice.com
Coupler: 8P8C RJ45 Cat6 Inline Coupler, White - Monoprice.com

I agree with you about the coupler, but the window cam is by no means a permanent install. I do also have couplers in the other line though. Partially because I wanted the option of moving operations to the basement with ease should I choose to in the future...but mostly because I underestimated how much wire I needed lol

 

[Mike A.]

You can move one of them temporarily to a short cable near your NVR to test whether you get the same with other cabling.

BTW I've left 2 of mine up all day as above just to test and I've not seen any similar problem.

Guessing not but do you have any other way to power the cam away from the NVR? If so, then you could try that to eliminate the cam itself. Are there frame rate or other settings on the NVR that could be conflicting with what's set on the cam? Might check things are set the same way if so.

 

[tangent]

Good to know. Thanks. I know HD products aren't the best, but I would hope they're common enough to not be considered complete garbage. As for monoprice, they may exclusively source their wires from china on the cheap, but they've always had great products. The cables I bought had a ton of stellar reviews too (btw, looks like they're 24awg)

I don't have any issue with cables sold by monoprice or home depot. Cables generally become a bigger issue at longer distances anyway.

You could always try powering from a 12v adapter or PoE injector. could also try hooking to your pc / network instead of nvr using one of these power sources.

Interesting to note that the time stamp disappears briefly.

 

[Big Ry]

You can move one of them temporarily to a short cable near your NVR to test whether you get the same with other cabling.

BTW I've left 2 of mine up all day as above just to test and I've not seen any similar problem.

Guessing not but do you have any other way to power the cam away from the NVR? If so, then you could try that to eliminate the cam itself. Are there frame rate or other settings on the NVR that could be conflicting with what's set on the cam? Might check things are set the same way if so.

What you are saying about the short cable near the NVR... That's exactly what the current setup is for the window cam. It's my NVR sitting next to my TV stand with a cat6 cable running across a pile of dirty laundry to the camera propped up on the window sill. Btw, i managed to stretch the 5' cable to reach, so i no longer have the 3' cable & coupler in the line. It didn't affect anything.

I do not have any other way to power the cams. I'm brand new to all this, so i don't have extra equipment laying around or anything like that. I could probably pickup a power cord on the cheap for troubleshooting purposes, but i wouldn't want to go dropping big $$$ on switches, injectors, PCs, software, etc. just to find out why my current equipment isn't working.

In my troubleshooting last night, i basically tried every combination of frame rate and bit rate on both cams, and i could not get anything above 25 fps to work without blacking out - even after restoring both cams to default settings and only changing those two settings one at a time. Even with bit rate at the lowest setting it wouldn't allow me to up the frame rate without issue. I tried changing the audio sample rate, compression to 265, resolution to 720p... Nothing allowed me to up the frame rate without issue.

On another note, i tilted the back door cam down to cut the nearest street lamp out of frame, and that really cut down on those steaks you can see on the image, though they are still present. If i turn off 3D NR, the image is noisy as all hell, looks like the screen is crawling with insects, so I'm assuming those steaks are a result of whatever the 3D NR does. They also seem to emanate from the street lamp, and there are other very clear light artifacts in the image from the street lamp. Can other light sources cause interference to cams? Can they damage the cams? I feel like the cam looks much noisier than before, so I'm getting worried that this thing is deteriorating just like my reolink cams did. But what on Earth could be killing the cams if so, LED light from the street lamps!? And before anyone asks, I cleaned the glass covering the lens and the sensor while i was out there, but they were pretty much already spotless.

How could i tell if these issues were related to a power issue? Like for the cams, i know i can buy an external power source cable, but what if the issue is insufficient power at the NVR? I was pretty surprised at how thin the power cord to the NVR was, especially since the European one was still in the box as well and that looked thicker as expected. It could obviously just be skimping on insulation, but it had me worried at first that Andy gave me the wrong cord. The NVR seems to operate perfectly fine throughout the UI, and the fan runs constantly without issue. So i would guess there is no issue, but i wanted to check. Hell, it would have probably caught on fire at this point if there was an undersized wire lol

Oh and i tried to determine if the equipment was PAL or NTSC, but i couldn't find it in the settings. I swear i saw it there before, but i looked around for a solid 10 minutes and couldn't find anything. I got the firmware version for the cams (which is dated 7/20/2017) if that indicates which it is. I didn't see anything for the NVR that explicitly said "firmware" but there was a section with a list of versions of various things. I'm assuming one was firmware. It lists a build date of 4/18/2017. So I guess I'll check those against the latest firmware dates to see if there's something more recent.

Sent from my Pixel 2 XL using Tapatalk

 

[tangent]

PAL/NTSC doesn't really matter anymore.

You can get a get a 5 port PoE switch for $40. https://www.amazon.com/BV-Tech-Port-Switch-Ports-Uplink/dp/B00WKRBI0W/
A single injector runs around $14 https://www.amazon.com/BV-Tech-Single-Port-PoE-Injector/dp/B005BQUNEG/
You may have a 12V power supply lying around that would do the job if you wanted to power it that way.

If you were to use a 12v wall wart or single PoE injector you could connect an ethernet cable directly to a computer. You'd have to assign your computer a static IP in the same subnet as the camera you may lose internet access on that pc while you do this. A PoE switch could be hooked up a little differently. This type of setup would allow you to take the NVR and it's PoE out of the equation and directly access the camera's web interface.

 

[Big Ry]

PAL/NTSC doesn't really matter anymore.

You can get a get a 5 port PoE switch for $40. https://www.amazon.com/BV-Tech-Port-Switch-Ports-Uplink/dp/B00WKRBI0W/
A single injector runs around $14 https://www.amazon.com/BV-Tech-Single-Port-PoE-Injector/dp/B005BQUNEG/
You may have a 12V power supply lying around that would do the job if you wanted to power it that way.

If you were to use a 12v wall wart or single PoE injector you could connect an ethernet cable directly to a computer. You'd have to assign your computer a static IP in the same subnet as the camera you may lose internet access on that pc while you do this. A PoE switch could be hooked up a little differently. This type of setup would allow you to take the NVR and it's PoE out of the equation and directly access the camera's web interface.

You're going a little over my head here. I don't know a whole lot about networking yet. I was just trying to get a setup that's plug n play, so i can get monitoring setup ASAP. Then i planned to sort through the networking details, including including remote access and control later as i learn more. But I'm not at that point yet, so you're speaking in tongues right now lol

Sent from my Pixel 2 XL using Tapatalk

 

[looney2ns]

I have a solution for you, set the frame rates to 15fps.
Something in your setup, can't handle the frame rates you are throwing at it. The cam, the nvr, the monitor, something.
You admit you don't know what you are doing, you have asked for advice, it's been given, yet you continue to cause yourself heart burn.

Set everything back to defaults, cam and NVR.
Just because you can, doesn't mean you should...relating to frame rates.

 

[Philip Gonzales]

I have a solution for you, set the frame rates to 15fps.
Something in your setup, can't handle the frame rates you are throwing at it. The cam, the nvr, the monitor, something.
You admit you don't know what you are doing, you have asked for advice, it's been given, yet you continue to cause yourself heart burn.

Set everything back to defaults, cam and NVR.
Just because you can, doesn't mean you should...relating to frame rates.

Wait, what are you saying? I shouldn't hold the accelerator pedal all the way down to the floor when I drive? !?!

 

[Big Ry]

I have a solution for you, set the frame rates to 15fps.
Something in your setup, can't handle the frame rates you are throwing at it. The cam, the nvr, the monitor, something.
You admit you don't know what you are doing, you have asked for advice, it's been given, yet you continue to cause yourself heart burn.

Set everything back to defaults, cam and NVR.
Just because you can, doesn't mean you should...relating to frame rates.

To use the same analogy as the poster below you:

That's like buying a new car that can go 80mph, but then driving it at 40mph tops because the engine stalls above that. It still gets you from point A to point B at a reasonable speed, yes, but i paid for an 80mph capable vehicle. So while i may not need to run the cams at 60fps or even 30fps, i still want to know why i can't if they're supposed to be able to. I'm not dismissing your recommendations, I'm trying to troubleshoot the problem. I paid $1000 for this gear. I want it to work.

Sent from my Pixel 2 XL using Tapatalk

 

[tangent]

To use the same analogy as the poster below you:

That's like buying a new car that can go 80mph, but then driving it at 40mph tops because the engine stalls above that. It still gets you from point A to point B at a reasonable speed, yes, but i paid for an 80mph capable vehicle. So while i may not need to run the cams at 60fps or even 30fps, i still want to know why i can't if they're supposed to be able to. I'm not dismissing your recommendations, I'm trying to troubleshoot the problem. I paid $1000 for this gear. I want it to work.

Sent from my Pixel 2 XL using Tapatalk

There are settings or combinations of settings that CAN cause problems. Just because they didn't fully idiot proof it doesn't mean you should be an idiot.

 

[Big Ry]

There are settings or combinations of settings that CAN cause problems. Just because they didn't fully idiot proof it doesn't mean you should be an idiot.

How is that being an idiot? If you are going to imply that I'm being an idiot, then at least back it up with some justification as to why these cameras can't operate at stated specs instead of "just because". When i set my new Pixel 2XL that I'm typing on now to record at 60fps, it records at 60fps. I don't see why any other camera wouldn't if the specs are true.

Sent from my Pixel 2 XL using Tapatalk

 

[tangent]

How is that being an idiot? If you are going to imply that I'm being an idiot, then at least back it up with some justification as to why these cameras can't operate at stated specs instead of "just because". When i set my new Pixel 2XL that I'm typing on now to record at 60fps, it records at 60fps. I don't see why any other camera wouldn't if the specs are true.

Back to your car analogy, doing something that you know will make the car stall while on the highway wouldn't be very smart would it. If the car had a self destruct button would you push it?
In the case of the metaphorical car there are things your could do to increase it's performance, there's nothing you can do here.

I know first thing I did was max out recording resolution and FPS on the main and sub1 streams, and I set the "outdoor" or "street light" image setting for the cam out back with that glaring LED street light right in front of it. I already tried changing the outdoor/streetlight/auto/etc. settings to no avail, so I'm assuming that's not the culprit. I supposed I could try dropping the resolution and/or FPS, but that kinda sucks if I can't get the quality I'm expecting out of these cams.

Honestly the best thing you can do is going to be reset everything back to the default and then make small changes to the settings and try not to change too many things at once. Think of adjusting it more like a science experiment where you only change one thing at a time, keep some notes about what you did and whether it made the picture better or worse or introduced other problems. While you're at it do some searching to learn about what each setting does. Leave the frame rate and bit rate and sub stream resolution alone to identity if there's anything more than your settings causing your issue, otherwise you're just wasting everyone's time.

Life isn't a video game, high frame rates aren't very useful in most security applications they just result is a lot of wasted storage meaning you wipe out your recordings a lot sooner. 10-15 FPS is actually pretty common. You aren't likely to benefit from anything above 30fps. The substream's purpose is to provide a lower resolution image that can be used for things like displaying multiple cameras on the NVR's video output or when viewing all the cams at once on your computer without using up all the resources available.

The cameras and NVR each only have a certain amount of CPU and RAM. It's a balance between certain features as to what the camera or NVR can do. There are network and storage bottle necks too even if they're internal to the NVR. Only so many mbps of data are going to get written to the hard drive, this is arguably the most important spec on an NVR. When you enable motion detection for example it drives up the CPU usage. If you increase the frame rate or bit rate it increases the storage usage and CPU usage and network bandwidth. Some codec are more storage intensive others are more cpu intensive. Some settings on the camera will actually disable others that it doesn't have the resources for. The composition of your image can even effect resource use and lead to camera blackouts in some cases. Your Pixel 2 XL has way more ram and processing power than the camera or nvr. There's likely some combination of settings that would work at 60fps, but trying to find it is completely pointless and would mean disabling all sorts of far more useful functions.

If you try to record only on motion detection you'll miss events you wish you recorded. Alternatives to the traditional motion detection like line crossing or intrusion result in far fewer false alerts.