Dahua 8mp NON-Starlight HDW5830R-Z review

happf

Getting the hang of it
Nov 21, 2016
84
51
This is a review of the 8mp NON-Starlight HDW5830R-Z camera. I will also do some side by side comparisons with a 2mp Starlight HDW5231R-Z.

The camera is all metal and appears to be basically the same size as the starlight model. You will have to take it apart to insert an SD card. One big difference is the small set screw (camera comes with a tool) is not the exact same size as the one on the starlight models, so don't loose that tool!

312311_thumb.png


Features:

1/2.5" 8mp progressive scan STARVIS CMOS (this is NOT a starlight chip!)
H.265 & H.264 dual stream encoding
15fps@4k(3849x2160), 25/30fps@3M(2304x1296)
Day/Night(ICR),3DNR,AWB,AGC,BLC
2.7mm-12mm motorized lens
Built In Mic
MAX IR LEDs Length 50m
IP67,IK10,Poe

Product Page
User Manual
Quick Start Guide
Spec Sheet

Bought through @EMPIRECANDY - Online store

I'm not going to clutter this thread with pictures of the web interface as they are 99% similar to the starlight model and you can see some in the original 2mp starlight review.



I currently have an all Dahua setup with an NVR5216 and 9 various starlight models. A common question that is asked over and over is about higher mp cameras. I really like the starlight models but I wanted to experiment with something with a bit more resolution. There were already comparisons between the 2mp starlights and other 3 and 4 mp models and the starlights overall won. I wanted to see what the 8mp could do.


I currently have a HDW5231R-Z covering my driveway zoomed about halfway in. Near the top of the driveway closest to the camera I feel I can get a good identity. Further down the driveway closer to the garage doors it really starts to drop off. I do have another starlight down there, but I am hoping that the 8mp can do the job of both, and that I could repurpose that other starlight for something else. I have a test pole I use for test placing cameras and I placed the 8mp right next to the 2mp underneath the soffit for testing and tried to get it zoomed as close as possible to the 2mp. The number zoom in the web interface wasn't the exact same, maybe due to slight changes in the lense and the different chips.

I printed out a license plate image I found online to the closest size as my real plates (and even printed a 50% reduced one) and setup this up on a stand. My driveway is somewhat lit and I currently have the IR turned off on all my external starlights (to avoid bugs) as there is sufficient external lighting. I have bought some external IRs but have not mounted them yet.

Below are various screenshots I've taken from SmartPSS comparing the 2mp to the 8mp. Without a doubt, in the daylight, as expected, the 8mp really really shines, IF you need to ZOOM in. I'm amazed at how much I can zoom in and still get good resolution. At the light drops, the 8mp has a harder time keeping it up. At it's worst, in a somewhat still lit environment, I'm basically seeing the same resolution as the starlight. At least it's not worse. As it gets darker, one can see the starlight start to excel showing more detail in poorer lit areas where the 8mp stays darker. In the night shots, I tested with both my garage lights on and off. Keep in mind, my neighbors have lights too that were always on. I did try turning on the IR but I only saw the ground near the top of the driveway getting brighter and didn't see any impact on my license plate test so didn't test any further. For the daylight test, since it was that much clearer, I went ahead and moved the test sign further down to see how much further I could still read them. The license plates are approx 60-70' (guessing) from the camera. For the daylight shot where I moved them further down that is maybe another 15-20' away?

I had my kids walk around in the daytime and I could definitely identify them at greater distances by zooming in. At night, that would only happen if the area was really well lit.

So it's a tough call. The 8mp camera does cost a bit more. You definitely get a huge day time improvement IF you are wanting to identity something further away. Looking at both models full screen, I can't see a difference if that's all you are after. At night, it basically degrades to a 2mp, that at least in some areas better lit might offer a tiny bit more detail in return for less detail in darker areas.

Right now, based on these results, my plans are to go ahead and replace that driveway starlight with the 8mp and maybe move my second starlight at the garage door further down to better cover that darker part of the driveway. Also going to update the driveway floodlights.

Daytime comparison (with the test license plate at the further distance)

Daytime.jpg

Daytime zoom of license plates at both distances - Left: 2mp Right: 8mp

Daytime_Zoom1.jpg Daytime_Zoom2.jpg

2mp Starlight with driveway lights on

2mp_LightsOn.jpg

8mp with driveway lights on

8mp_LightsOn.jpg

2mp driveway lights off

2mp_LightsOff.jpg

8mp driveway lights off

8mp_LightsOff.jpg

Side by Side driveway lights on - Left: 2mp Right: 8mp

LightsOn_Compared.jpg

Driveway lights on, zoomed in (closer distance) - Left: 2mp Right: 8mp

Night_LightsOn.jpg

Driveway lights off, zoomed in (closer distance) - Left: 2mp Right: 8mp

Night_LightsOff.jpg
 
This is a review of the 8mp NON-Starlight HDW5830R-Z camera. I will also do some side by side comparisons with a 2mp Starlight HDW5231R-Z.

The camera is all metal and appears to be basically the same size as the starlight model. You will have to take it apart to insert an SD card. One big difference is the small set screw (camera comes with a tool) is not the exact same size as the one on the starlight models, so don't loose that tool!

312311_thumb.png


Features:

1/2.5" 8mp progressive scan STARVIS CMOS (this is NOT a starlight chip!)
H.265 & H.264 dual stream encoding
15fps@4k(3849x2160), 25/30fps@3M(2304x1296)
Day/Night(ICR),3DNR,AWB,AGC,BLC
2.7mm-12mm motorized lens
Built In Mic
MAX IR LEDs Length 50m
IP67,IK10,Poe

Product Page
User Manual
Quick Start Guide
Spec Sheet

Bought through @EMPIRECANDY - Online store

I'm not going to clutter this thread with pictures of the web interface as they are 99% similar to the starlight model and you can see some in the original 2mp starlight review.



I currently have an all Dahua setup with an NVR5216 and 9 various starlight models. A common question that is asked over and over is about higher mp cameras. I really like the starlight models but I wanted to experiment with something with a bit more resolution. There were already comparisons between the 2mp starlights and other 3 and 4 mp models and the starlights overall won. I wanted to see what the 8mp could do.


I currently have a HDW5231R-Z covering my driveway zoomed about halfway in. Near the top of the driveway closest to the camera I feel I can get a good identity. Further down the driveway closer to the garage doors it really starts to drop off. I do have another starlight down there, but I am hoping that the 8mp can do the job of both, and that I could repurpose that other starlight for something else. I have a test pole I use for test placing cameras and I placed the 8mp right next to the 2mp underneath the soffit for testing and tried to get it zoomed as close as possible to the 2mp. The number zoom in the web interface wasn't the exact same, maybe due to slight changes in the lense and the different chips.

I printed out a license plate image I found online to the closest size as my real plates (and even printed a 50% reduced one) and setup this up on a stand. My driveway is somewhat lit and I currently have the IR turned off on all my external starlights (to avoid bugs) as there is sufficient external lighting. I have bought some external IRs but have not mounted them yet.

Below are various screenshots I've taken from SmartPSS comparing the 2mp to the 8mp. Without a doubt, in the daylight, as expected, the 8mp really really shines, IF you need to ZOOM in. I'm amazed at how much I can zoom in and still get good resolution. At the light drops, the 8mp has a harder time keeping it up. At it's worst, in a somewhat still lit environment, I'm basically seeing the same resolution as the starlight. At least it's not worse. As it gets darker, one can see the starlight start to excel showing more detail in poorer lit areas where the 8mp stays darker. In the night shots, I tested with both my garage lights on and off. Keep in mind, my neighbors have lights too that were always on. I did try turning on the IR but I only saw the ground near the top of the driveway getting brighter and didn't see any impact on my license plate test so didn't test any further. For the daylight test, since it was that much clearer, I went ahead and moved the test sign further down to see how much further I could still read them. The license plates are approx 60-70' (guessing) from the camera. For the daylight shot where I moved them further down that is maybe another 15-20' away?

I had my kids walk around in the daytime and I could definitely identify them at greater distances by zooming in. At night, that would only happen if the area was really well lit.

So it's a tough call. The 8mp camera does cost a bit more. You definitely get a huge day time improvement IF you are wanting to identity something further away. Looking at both models full screen, I can't see a difference if that's all you are after. At night, it basically degrades to a 2mp, that at least in some areas better lit might offer a tiny bit more detail in return for less detail in darker areas.

Right now, based on these results, my plans are to go ahead and replace that driveway starlight with the 8mp and maybe move my second starlight at the garage door further down to better cover that darker part of the driveway. Also going to update the driveway floodlights.

Daytime comparison (with the test license plate at the further distance)

View attachment 20783

Daytime zoom of license plates at both distances - Left: 2mp Right: 8mp

View attachment 20784 View attachment 20785

2mp Starlight with driveway lights on

View attachment 20787

8mp with driveway lights on

View attachment 20788

2mp driveway lights off

View attachment 20790

8mp driveway lights off

View attachment 20791

Side by Side driveway lights on - Left: 2mp Right: 8mp

View attachment 20792

Driveway lights on, zoomed in (closer distance) - Left: 2mp Right: 8mp

View attachment 20793

Driveway lights off, zoomed in (closer distance) - Left: 2mp Right: 8mp

View attachment 20794
Thanks for the report, so the camera in the night the 2mp starlight still be amazing. lol Daytime the 8MP much better.
 
Nice comparison. Thanks. You can really see the difference in resolution. Also the advantage of the Starlights in the dark.

One thing that I didn't really think about when I got them but I've ended up liking a lot about the Starlights is that I can turn off the IR and rely only on ambient light and the IR from other cams. No bugs, or glare from surroundings, washed out faces, etc. Hard to do that with anything else where mine are. With all of the rest it's like a spider and moth party all night this time of year.
 
  • Like
Reactions: onyidzin
Thanks! Have been very interested in seeing the 4K cameras vs the Starlights.

Curious if you have tried to force color in low light conditions.
 
@happf nice post. useful data. it would be nice to know the ~lumens where it drops to worse than the 2mp. Do you have a meter that you can use? Can you compare the data rate (H265) during day and night?
 
@happf Awesome comparison, thank you.
 
Finally a break in the rain so I did some experimenting in the back patio. I don't have a light meter so I took a bunch of shots just after the sunset to get an idea of the difference with/without starlight.

There are 3 columns for comparison. Right up front I want to warn you, the 1st LEFTMOST column is NOT an apples to apples with the right two columns. I have a Starlight HDW5231R-Z vari-focal and a fixed lens starlight HDW4231EM-AS in my patio. I place the 8mp HDW5830R-Z as close as possible to the corner where the fixed lens EM-AS is to compare what would happen with the 8mp covering the whole patio instead of the EM-AS. The 1st col R-Z, would stay no matter what, and that camera is about 15' to the left of the EM-AS and it's zoomed almost all the way in (guestimating 10mm from the setup math) while the EM-AS in the corner is a fixed 3.6mm. The 8mp R-Z next to it is zoomed in (approx 4.5mm of it's 2.7-12mm range) to more approximate the EM-AS.

Once again, the left most camera is much closer and way more zoomed in so don't use it as a direct comparison to the right two. I have it there so that I can have a better chance to identify anybody entering the patio from that side. Soon I will be mounting another 3rd camera zoomed in to the far right so I can capture anybody entering from the right side. The EM-AS is just a wide angle so that I can see a lot more of what's going on and it's the one that I leave up running on the monitor.

I added some pictures I found online to try and help determine if someone can be identified. The target is 32' away from the 1st col camera R-Z and 43' away from the right two cameras.

This time I played around with IR since I can better control all the lighting. I took pictures after the sunset till it got dark (with IR off on all cameras). I did leave my one main light on, until I finally turned it off, and there we can see that even the starlight can not really see enough in the dark. Then I turned on various smart IR to see what would happen.

I was curious to see if the 8mp could keep up from the further distance and zoomed out more. I will say in the daytime, with the deck stacked against it, it put up a good fight. At night once again, if there is no lighting at all, will go black, but then again, even the starlight with no lighting (and no ir) goes 'almost' black too. Turn on the IRs, and they all come back to life.

Feel free to let me know if I'm not doing these comparisons accurately or should change any settings to give any of the cameras a better chance.






Image00004.jpg


Now cameras switched to night mode
Image00005.jpg
Image00006.jpg
Image00007.jpg
Image00008.jpg
Image00009.jpg

Now I turned on the smart IR on just the 3rd col 8mp-R-Z

Image00010.jpg

Now I turned on the SmartIr for just the 1st col starlight 2mp R-Z
Image00011.jpg

Now I go back to my default settings, the SmartIR on for both the original cameras (inside a screened patio with less bugs)
Image00012.jpg

I think the EM-AS camera had an issue in that 2nd to last pic with the angle of the IR from the other camera maybe being just right to wash out the image.
 
Lots of pictures are good. These are very useful. My first take on normalizing the images is:

The middle image is about what the left camera would be at that same zoom level, with the Zoom difference appear to be ~3x ppf

The first image set shows that 3x the ppf is better than 2x of the 8Mp in both resolution and contrast. But here are almost 10x the number of usable pixels in the pre-cropped & zoomed frame to a usable image during the day. This is the trade off in favor of the 8Mp

The second and third set show that the 8MP is still providing value over the center 2Mp at that light level, even though it is less sensitive.

The fourth set is interesting. It shows a just usable image from the 5231 at that pixel density and an almost useful one at 1/3 of that density on the 4231, but a useless one on the 5830. As this drop in light occurred over only 10 min, and there is the entire night, more light is needed for all of them, with the lower pixel count cameras giving maybe 10 min more use per night before the light is needed.

Several images show how the fixed focus 4231 can be better than the varifocal in that its images would be out of focus sue to the two near field objects that are blurry in its frame, where as if they were more in the frame of the VF it would have used that for the focus and the sign would be blurry.

I wonder how a couple of hundred lumens of LEDs placed around the area compare to IR floods?

I would like to see some comparison of the bit traffic under H265 compression and similar conditions. Is the 8Mp only slightly higher data, or more like 4x?

Personally, where is see the 2Mp starlight shine is in full color mode with good moon light and some supplemental street lighting. There are obvious advantages to the 8Mp variants of the Sony Sarvis lineup.
 
  • Like
Reactions: bigredfish
Amazing comparison. Thanks a lot. I wonder if the 8mp with varifocal lens would perform better?

Inviato dal mio SM-G930F utilizzando Tapatalk
 
@Dave88LX , you still need enough pixels on target to read / identify the image. At some distance and enough light level the higher pixel count is going to win. that is the interpretation, as you can not make out the plate in daylight at 2MP (just barely). The 8MP effectively give you 2x the pixels per foot of the 2MP, or 2x the zoom to noise floor. This is not the case when light is limited from dusk to dawn without added light where the 8MP has insufficient light at a useful shutter speed, resulting in noise not resolution.
 
  • Like
Reactions: Arjun
Thanks! Have been very interested in seeing the 4K cameras vs the Starlights.

Curious if you have tried to force color in low light conditions.

What is the benefit of forcing color? I have a Dahua HFW5231-Z5, and based on the lighting conditions it will go B&W when the Day & Night sensitivity is set to low and color when sensitivity is set to high. I have no idea what the import of high or low sensitivity is, so I'm trying to figure out whether I'm likely to get better results in B&W or Color.

Thanks in advance.
 
What is the benefit of forcing color? I have a Dahua HFW5231-Z5, and based on the lighting conditions it will go B&W when the Day & Night sensitivity is set to low and color when sensitivity is set to high. I have no idea what the import of high or low sensitivity is, so I'm trying to figure out whether I'm likely to get better results in B&W or Color.

Thanks in advance.

Hi RJF,

Depends, on how much normal light you have. If you have for example decent street lighting - having the image in color is very useful to help ID cars and suspects.

If you do not have enough light, it is better to have it in IR / Black and White than very poor color images.
 
  • Like
Reactions: RJF