SUNBA PoE+ PTZ Camera Outdoor, 36X Optical Zoom 5MP
- Thread starter Saltster
- Start date
just look at Uniview Global and could not find any that look like mine?
That is common. Many companies enter into a contract with another company to make a camera for them and they will spec out differences so that they do not look exactly alike.
As an example, Lorex and Amcrest and both Dahua OEM and they both sell cameras that do not look like anything Dahua sells.
Or the original company decides to make it different so that it doesn't look like something they sell at a higher price.
So the telltale sign is the firmware. If it has the look/feel of another manufacturer, it is likely OEM from that company.
The Automation Guy
Known around here
@Saltster
First, let me talk about the difference between a good performing camera and a poor performing camera. A good performing camera is going to provide footage with enough detail and clarity to clearly identify an unknown subject. The litmus test that should be used is this: Is the footage clear and detailed enough for the police to identify and ultimately arrest someone based on your footage alone? A poor performing camera is any camera that cannot provide this level of footage at all times of the day and night. It's not that the poor performing camera will be completely burry and unrecognizable, but even a tiny amount of blurring will cause all the important distinguishing features of a subject to be lost and therefore useless to identify an unknow subject. For example, with a poor performing camera you will probably still see a outline of a face with eyes, nose and mouth, but the image isn't sharp enough to provide all of the features needed to positively identify the subject. If all you can tell is that a subject is "a white male, 15-45 years old, wearing a hoodie" then that camera has failed at it's job which it to provide clear enough footage for the police to arrest someone. That is the goal we are talking about here.
During the day, almost any camera can provide detailed enough images to be a "good performing camera." However it's a night when you will see real differences between cameras.......
If the camera is set for auto exposure, at night the camera will slow the shutter speed down and crank the gain up until it produces a very nice bright image. These images look great - until there is something moving in the frame. Then the slow shutter speeds and high gain will result in blurry and ghosting subject (ghosting is where parts, or even all of the moving subject will disappear in some frames of the recorded footage). A PTZ with autotracking can have a slightly slower shutter speed than a fixed camera and still produce decent images because it will track the subject which counteracts the slow shutter speed blurring the motion, but there is still a limit before the subjects will be blurry.
We are judging these things on physics. Your camera has a known number of pixels (5,017,600) and a known image sensor size (1/2.8" which is roughly 1/3"). Experience has told us that the pixel size that these two things create is not going to give great low/no light images. Of course if you have your camera mounted in a bright area (perhaps there are some bright street lights nearby), then it may produce great results. But the vast majority of people are going to have a tough time getting the gain low enough to prevent ghosting, the shutter speed high enough to prevent blurring, and still be able to get an image that is bright enough to be useful.
On the other hand, a camera with only 2mp that uses the same size sensor (1/2.8") will have pixels that are more than twice as large as the pixels in your camera because there are only 2,073,600 pixels on the same size sensor vs 5,017,600 pixels that your camera has. The larger pixel is able to absorb more than twice as much light in the same amount of time as your camera's pixels can. This results in much brighter images and it is generally the highest resolution (2mp) that we suggest on a 1/2.8" sensor size to ensure that we can set the gain low enough to prevent ghosting, set the shutter speed fast enough to prevent blurring, and still have a bright enough image to be useful. Any higher pixel count and the image will be too dark after setting appropriate shutter and gain settings.
So what are you to do? Well as I already mention, this is highly dependent on the amount of available light in your setting. The more light you have, the greater the chance that you can break our "maximum pixel rule" and still get usable images. Outside of adding more available light, you will need to put your camera into manual exposure mode and play around with the settings (gain, aperture/iris, and shutter) to get the best compromise. You might have to live with some blurring and some ghosting in order to get an image bright enough to be useful.
Hopefully that helps. We aren't trying to crush your excitement about your new camera, but we do want you to look at the performance of it with open and honest eyes. You are going to have to play around with the shutter, gain, and aperture/iris settings with a moving subject at night to find the right combination of settings to give you the best result. If you leave the camera set on auto, it will look great with static images, but as soon as you have a subject moving in the frame, you are going to have really poor results (ie unusable footage to identify a subject). By finding the right combination of settings for your location, you will maximize the chance that you will get good identifiable footage of a moving subject at night.
I hope that helps explain it a little further. We are here to help, so don't be afraid to ask questions or even ask for help dialing in your settings.
First, let me talk about the difference between a good performing camera and a poor performing camera. A good performing camera is going to provide footage with enough detail and clarity to clearly identify an unknown subject. The litmus test that should be used is this: Is the footage clear and detailed enough for the police to identify and ultimately arrest someone based on your footage alone? A poor performing camera is any camera that cannot provide this level of footage at all times of the day and night. It's not that the poor performing camera will be completely burry and unrecognizable, but even a tiny amount of blurring will cause all the important distinguishing features of a subject to be lost and therefore useless to identify an unknow subject. For example, with a poor performing camera you will probably still see a outline of a face with eyes, nose and mouth, but the image isn't sharp enough to provide all of the features needed to positively identify the subject. If all you can tell is that a subject is "a white male, 15-45 years old, wearing a hoodie" then that camera has failed at it's job which it to provide clear enough footage for the police to arrest someone. That is the goal we are talking about here.
During the day, almost any camera can provide detailed enough images to be a "good performing camera." However it's a night when you will see real differences between cameras.......
If the camera is set for auto exposure, at night the camera will slow the shutter speed down and crank the gain up until it produces a very nice bright image. These images look great - until there is something moving in the frame. Then the slow shutter speeds and high gain will result in blurry and ghosting subject (ghosting is where parts, or even all of the moving subject will disappear in some frames of the recorded footage). A PTZ with autotracking can have a slightly slower shutter speed than a fixed camera and still produce decent images because it will track the subject which counteracts the slow shutter speed blurring the motion, but there is still a limit before the subjects will be blurry.
We are judging these things on physics. Your camera has a known number of pixels (5,017,600) and a known image sensor size (1/2.8" which is roughly 1/3"). Experience has told us that the pixel size that these two things create is not going to give great low/no light images. Of course if you have your camera mounted in a bright area (perhaps there are some bright street lights nearby), then it may produce great results. But the vast majority of people are going to have a tough time getting the gain low enough to prevent ghosting, the shutter speed high enough to prevent blurring, and still be able to get an image that is bright enough to be useful.
On the other hand, a camera with only 2mp that uses the same size sensor (1/2.8") will have pixels that are more than twice as large as the pixels in your camera because there are only 2,073,600 pixels on the same size sensor vs 5,017,600 pixels that your camera has. The larger pixel is able to absorb more than twice as much light in the same amount of time as your camera's pixels can. This results in much brighter images and it is generally the highest resolution (2mp) that we suggest on a 1/2.8" sensor size to ensure that we can set the gain low enough to prevent ghosting, set the shutter speed fast enough to prevent blurring, and still have a bright enough image to be useful. Any higher pixel count and the image will be too dark after setting appropriate shutter and gain settings.
So what are you to do? Well as I already mention, this is highly dependent on the amount of available light in your setting. The more light you have, the greater the chance that you can break our "maximum pixel rule" and still get usable images. Outside of adding more available light, you will need to put your camera into manual exposure mode and play around with the settings (gain, aperture/iris, and shutter) to get the best compromise. You might have to live with some blurring and some ghosting in order to get an image bright enough to be useful.
Hopefully that helps. We aren't trying to crush your excitement about your new camera, but we do want you to look at the performance of it with open and honest eyes. You are going to have to play around with the shutter, gain, and aperture/iris settings with a moving subject at night to find the right combination of settings to give you the best result. If you leave the camera set on auto, it will look great with static images, but as soon as you have a subject moving in the frame, you are going to have really poor results (ie unusable footage to identify a subject). By finding the right combination of settings for your location, you will maximize the chance that you will get good identifiable footage of a moving subject at night.
I hope that helps explain it a little further. We are here to help, so don't be afraid to ask questions or even ask for help dialing in your settings.
Last edited:
Cameraguy
Known around here
- Joined
- Feb 15, 2017
- Messages
- 1,486
- Reaction score
- 1,132
Your image looks to blur colors into each other. Maybe you can tweak some video settings This is my Dahua 59225 that's 4 years old and still love
Attachments
Last edited:
nice reply. I have 9 video cameras, this is my first PTZ
all the others were as I listed above, and none could give good face detail at night.
with that said, I am going to try a few things as you stated.
however, I purchased this 36X to do some daytime surveillance on some illegal activity and in that sense, the camera already paid for itself on the footage that I captured over the past few days. and its still captures what I need close up and in great detail.
all my other cameras are still house mounted.
I guess the bigger question is what were your settings on the other cameras - where you using default/auto settings or did you dial in the settings like shutter, brightness, contrast, etc. to your field of view?
Many here have those cameras and get quality facial ID.
What distance where you trying to get IDENTIFY pictures of people? The ZE maxes out around 30-40 feet daytime (20-30 night) and the Z12E maxes out at around 175 feet or so daytime (100 feet or so night) for IDENTIFY purposes, but both can see further in the dark if enough light is available.
So of course if you were trying to do too much with those cameras, then yes, you suffer from the biggest mistake most make of trying to get IDENTIFY well beyond the DORI numbers, whereas a PTZ with a lot more zoom capability can get those shots a little easier.
It does sound like Sunba has improved since I got mine, which is good. But also you will want to test out what we suggested (regardless of who makes the camera) to see if it can perform at night time with motion - that is where the cameras will struggle.
Yeah but usually the cameras will look similarr, I could not find any on their website that even looked close?
Again, that is common. As I said, there are many Lorex and Amcrest cameras as examples that look nothing like anything Dahua sells.
And some companies may have the internals be OEM but they make the external.
I know that Sunba did that with the Hikvision - Sunba provided the outer shell but Hikvision provided the sensor, motherboard, firmware, etc. Like I said, Sunba is transparent and they have told many people here that the internal of the model they had was Hikvision (although it was obvious by the GUI as well as looking at the motherboard).
The Sunba Illuminati model is a collaboration between Sunba and Hikvision. Sunba is responsible for the hardware and Hikvision is responsible for the software interface and lens module software / firmware. Here is their email to a member here confirming it:
This model firmware matches Uniview, so it is a guarantee that Uniview was involved in some way. A company the size of Sunba isn't going to copy/steal a GUI format. I suspect it is the same arrangement with Uniview that Sunba has with Hikvision.
So did you have a chance last night to capture any video with motion or try the 1/2,000 shutter test?
And some companies may have the internals be OEM but they make the external.
I know that Sunba did that with the Hikvision - Sunba provided the outer shell but Hikvision provided the sensor, motherboard, firmware, etc. Like I said, Sunba is transparent and they have told many people here that the internal of the model they had was Hikvision (although it was obvious by the GUI as well as looking at the motherboard).
The Sunba Illuminati model is a collaboration between Sunba and Hikvision. Sunba is responsible for the hardware and Hikvision is responsible for the software interface and lens module software / firmware. Here is their email to a member here confirming it:
This model firmware matches Uniview, so it is a guarantee that Uniview was involved in some way. A company the size of Sunba isn't going to copy/steal a GUI format. I suspect it is the same arrangement with Uniview that Sunba has with Hikvision.
So did you have a chance last night to capture any video with motion or try the 1/2,000 shutter test?
Last edited:
I read many posts on setting and saw the sample videos that any posted, and I have not seen any posts where you should snap a pic and zoom in without pixelation and blurred faces.
I have two Dahua cameras about 6 feet from my front door and capture people from the left and right sides, even then face recognition is difficult. So I had it with the 2 megapixels, So far with the camera, I can identify faces without any problem. So I going to add another SUNBA to the front doorway from a different angle and see if there is any difference.
I have most of my house angles covered now, but water a few PTZ cameras places so I can see 360 around my property from all angles, not to mention over 1500 feet up the road that I am now monitoring the activity to be used as enhance.
This photo is cropped from the video at over 1500 feet up the road and across a lake., with no enhancements at all
Again, it sounds like operator/install error. A 2MP Dahua camera can absolutely capture faces at 6 feet unless you had the camera installed up on the 2nd story, then all of the IDENTIFY distance according to DORI numbers is lost in the vertical, or you had the settings so out of whack.
I, as well as others here, get incredible captures with 2MP Dahua cameras. In fact, it is my 2MP cameras that get the clean face captures of perps at night that get the person arrested and my neighbors their stuff back!
A camera installed at the proper height and with the appropriate focal length for the distance to be covered will result in IDENTIFY captures without the need for digital zoom. Digital zoom only works in TV and the movies.
While your image above at 1500 feet is great and all, that is to be expected of a PTZ camera with the amount of zoom it has. Of course it will be better than a 2MP camera without as much zoom capability and has a focal length not designed to capture something that tight beyond 200 feet.
And again, a still image like that means nothing for most people. That truck is not moving and you can see a foot on the other side confirming it is not in motion.
These cameras need to perform with motion at its worse condition and for most that is a nighttime.
You won't convince or win many people over here with still images of day and night scenes. Video with motion and the ability to freeze frame and capture a clean image without blur is what people are after. Now if you use case doesn't require that, then great, but most here are not going to spend $500 on a PTZ that doesn't have that ability.
nope they are on the porch just over window around 7 feet up on botrh side of the porch.
Night
Day
been through all this before.
these are the best I can get, still shot and moving.
I do not think that is good.
What are your shutter speeds on that camera or is it all at default settings? The certainly look like default settings or settings out of whack. There should not be that much blur in B/W with infrared at that distance.
The day images are clearly night images with lights on - look out in the yard and it is dark and you can see light reflections off the tree.
The left most color picture has a clear foot and a blurry foot, leading me to believe this is auto/default shutter speeds and just a simple movement of your foot caused blur due to the lower shutter speeds.
And it could be a little bit of out of focus issues. These cameras are not infinity focus like many believe. I think this camera was the ZE varifocal and the focus was probably set with nobody in the frame so it picked some object to focus on a little further out. If you stood at that mat and did a refocus, I bet that would help it some.
But also keep in mind that these are surveillance cameras and not DSLR cameras, so we have to manage expectations. You will not make out individual hairs and pores on a face as an example. An iphone picture taken at 6 feet away will produce a better image, but that camera also isn't capable of 24/7 surveillance in all weather conditions either. With the exception of the blur pictures, a few of them have clear enough capture that the police could use and that is the goal of these cameras.
We look forward to your comparable captures from the Sunba.
The day images are clearly night images with lights on - look out in the yard and it is dark and you can see light reflections off the tree.
The left most color picture has a clear foot and a blurry foot, leading me to believe this is auto/default shutter speeds and just a simple movement of your foot caused blur due to the lower shutter speeds.
And it could be a little bit of out of focus issues. These cameras are not infinity focus like many believe. I think this camera was the ZE varifocal and the focus was probably set with nobody in the frame so it picked some object to focus on a little further out. If you stood at that mat and did a refocus, I bet that would help it some.
But also keep in mind that these are surveillance cameras and not DSLR cameras, so we have to manage expectations. You will not make out individual hairs and pores on a face as an example. An iphone picture taken at 6 feet away will produce a better image, but that camera also isn't capable of 24/7 surveillance in all weather conditions either. With the exception of the blur pictures, a few of them have clear enough capture that the police could use and that is the goal of these cameras.
We look forward to your comparable captures from the Sunba.
Last edited:
I would like to see posts that have clearer results because I have not seen any yet.
and the palms have been removed since these were taken.
The Automation Guy
Known around here
Your 2mp resolution is not the problem here. You proved that with your first "day" image. That image is clear enough and provides plenty of detail to identify you (even though the camera is actually focused on the bench behind you and not on you. Getting the focus right will improve the results even more).
The fact that your other images have blurring is a direct result of your settings, not the camera's resolution. As I mentioned in a previous post, you have to set the camera in manual mode and set the shutter speed, gain, and iris settings to get good consistent results. While there are some good "starting points" that we can give you, it really takes some trial and error to dial in the best settings for your location/light levels. There is no "one setting that works". If there was, then the manufacturers "auto" setting would provide perfect images in every situation/install.
We are happy to help you with the process of dialing in those settings.
Last edited:
Cameraguy
Known around here
- Joined
- Feb 15, 2017
- Messages
- 1,486
- Reaction score
- 1,132
Yep definitely a shutter setting issue. This is about 75 ft from camera . I don't zoom in too much cause I want to capture the entire cul de sac. Guaranteed if I was zoomed in you would clearly be able to Id that person.. my poor mail lady has to be my tester haha.
Other pic is a Dahua 1831C PIR camera..
Attachments
Last edited:
This Sunba is a 5MP on a 1/2.8" sensor with a focal length range is 4.5 to 162mm (focal length range is more important than saying 36x zoom as that is relative to the starting focal length). That is the same sensor as what 2MP cameras use, so 5MP will need over twice as much light at night to see the same field of view. As it relates to PTZs, that means a 2MP on the same sensor as the 5MP will see further in the dark.
For many people the 49225 is a great PTZ. It is a 2MP on the 1/2.8" sensor (so the correct MP/sensor combo) with a focal length of 4.8 mm–120 mm, so not quite as far as the one you have now, but would have better quality at least at night due to the more ideal MP/sensor combo.
The SD5A425XA-HNR is the next one that is 4MP on the 1/1.8" sensor (so the correct MP/sensor combo) with a focal length of 5.4 mm-135 mm.
While not on the ideal MP/sensor combo, the SD6CE445XA-HNR is a 4MP on the 1/2.8" sensor, but it has an incredible focal length of 3.95 to 178mm.
For many people the 49225 is a great PTZ. It is a 2MP on the 1/2.8" sensor (so the correct MP/sensor combo) with a focal length of 4.8 mm–120 mm, so not quite as far as the one you have now, but would have better quality at least at night due to the more ideal MP/sensor combo.
The SD5A425XA-HNR is the next one that is 4MP on the 1/1.8" sensor (so the correct MP/sensor combo) with a focal length of 5.4 mm-135 mm.
While not on the ideal MP/sensor combo, the SD6CE445XA-HNR is a 4MP on the 1/2.8" sensor, but it has an incredible focal length of 3.95 to 178mm.
As an Amazon Associate IPCamTalk earns from qualifying purchases.
Thanks for the response. If anything I would go up on zoom range but I cannot go down. as for I need to capture closeups of the activity.
I picked up two 10" X 20ft pipes today and will be painting and installing them in the ground as my camera poles. they will have a 4-5 inch pole bolted inside to mount the camera so that it is up higher them the 10" and will have better 360 views. The camera will be mounted facing the inside of the pipe for the center of gravity.
As an Amazon Associate IPCamTalk earns from qualifying purchases.
Justin Blackburn
Getting the hang of it
- Joined
- Jan 13, 2018
- Messages
- 63
- Reaction score
- 89
I would like a picture of your final mount when you’re done please. I got a 4” diameter steel pole 21’8” long and am trying to come up with what would be the best looking mount.