ESCAM QD410, a 4-megapixel camera: a comparative review

I've already done both a standalone review of the Hikvision 3.2 MP camera and a comparison of the Hikvision 3.2 MP and 4 MP models.
Unfortunately, the 4 MP camera has been installed at a site for a long time, so it couldn't take part in the test for "valid reasons." But I can certainly compare the Hikvision 3.2 MP, the Szsinocam 2 MP, and the 4 MP camera under review. Especially since I'll be doing it as part of a real-world task, which I'll describe in general terms toward the end of the review, where I'll also include the comparison tests.

Today's review will have lots of photos and little text. Since all the cameras have a lot in common, I'll just comment on the key points.

Specifications stated by the manufacturer

Model ESCAM Bolt QD410

Camera
Image Sensor 1/3" CMOS
Effective Pixels Main stream: 4MP 2592*1520@18fps；3MP 2048*1536@22fps
Sub-stream: 704*576 @25fps
Electronic Shutter Speed Auto/Manual, 1/3~1/10000s
Min. Illumination color 0.01Lux@F1.2, black/white 0.001Lux@F1.2
S/N Ratio >50dB
Camera Features
Max. IR LEDs Length Φ5 36pcs IR Led, IR Range 15m
Day/Night Auto(ICR) / Color / B/W
Wide dynamic Support digital
White Balance Auto
Gain Control Auto
Noise Reduction Support 2D/3D
Privacy Masking yes
Lens
Focal Length 3.6mm
Max Aperture F1.6
Angle of View 75.7°
Video
Compression H.265/H.264
Frame Rate Main Stream 4MP 2592*1520@18fps；3MP 2048*1536@22fps；2MP (1920*1080) 25fps
Sub Stream 1280*720, 704*576 @25fps
Bit Rate Supports CBR/VBR
Network
Ethernet RJ-45 (10/100Base-T)
Protocol TCP/IP , UDP , RTP , RTSP , RTCP , HTTP , DNS , DDNS , DHCP , FTP , NTP , PPPOE , SMTP , UPNP
ONVIF ONVIF Ver. 2.0
Max. User Access 10 users
Smart Phone iPhone, iPad, Android
General
Power Supply DC12V
Power Consumption <6W
Working Environment -10°C~+60°C, 10%~90%
Ingress Protection IP66
Dimension 185 mm* 65 mm*65mm
Weight 0.36kg
Package Dimension 190 mm*90 mm*95 mm
Package Weight 0.48 KG
Approvals CE FCC RoHS

So, let's start with the packaging.
The camera arrived in a fairly familiar-looking package, except that the power supply was taped on as an extra. Cameras—at least those from well-known brands—usually don't come with power supplies, but they generally support PoE. In modern video surveillance systems, not being able to power a camera over the network cable is considered bad form.
Unfortunately, the camera under review falls into the category of models without PoE. That's a drawback, and a fairly significant one.


Here's everything in the box:
1. ESCAM QD410 camera
2. Cable gland plug
3. Mounting hardware
4. Hex key
5. Power supply
6. Manual
7. Warranty card and "disc."


1. When I saw the "disc" in the box, I immediately thought of those postcard records where you could record a greeting and send it to someone. Younger people probably haven't seen one, but those over 40 have likely come across this way of sending messages.
Here, though, the disc is just a fancy way of presenting a link to the manufacturer's website. I'll have a few "kind, quiet words" for them later.
2. The mounting hardware and key are standard, but having a proper cable gland for the network cable is a plus. Even Hikvision only started using these fairly recently.
3. The power supply has a European plug and was most likely added by the store.
4. Output voltage is 12 V, with a maximum current of 1 A.


The manual is attractive and colorful, but it'll only be useful to beginners—and even then, only those who know English.


I took the power supply apart mostly out of habit, but I did find that it's a little different from the power supplies included with the other products I've reviewed.


The power supply is heavily simplified: the input filter is completely missing, even though there's room for one.


1. It has a fuse and an input capacitor with a tiny capacitance of just 4.7 µF, though it's made by Jamicon.
2. The interwinding capacitor looks a little sad, but it's the correct type.
3. At first I thought the power supply used a powerful PWM controller, but after turning the board over I realized it's just a self-oscillating circuit. That's why the transistor in the photo has no markings.
4. The output has a pair of capacitors, 470 and 220 µF, as well as a small inductor.


There are quite a few components on the underside of the board too, but as I've already said, I don't like self-oscillating circuits; I prefer a proper PWM controller.
But the power supply doesn't get hot during operation. The camera was in use for about a week, and no problems were found.


The camera has a "Bullet" form factor, one of the most common housing types.


As usual, the front has the lens, a cluster of IR LEDs, and a light sensor. To get ahead of myself a little, the LEDs are visible in the dark from about 10–15 m away if the camera is positioned on the same optical axis as the observer. The illumination can't be forcibly disabled, which is a downside, albeit a minor one.
The top of the sunshade has a plastic anti-glare insert.


There's a sticker on the bottom of the housing with the camera's address, login (there's no default password), and the access code for the "cloud" service.
The hinge and mounting foot are standard, but the connector for the local network is a little different from the others. Besides being designed as a cable gland, it also has an activity LED. That's convenient in one way, but on the other hand, it gives the camera away: if the LED is off, the camera is disconnected.
The power connector—or, more accurately, the power plug—fits into the socket very easily. At first I thought there was a problem with the socket, but later it turned out to be a feature of the plug.


This time I took the camera apart before testing it, rather than afterward as usual.


Inside is a "sandwich" of three boards: one has the IR illuminator, another the image sensor, and the third the processor (SoC).


First, I remove the illuminator. It's held in place by 4 screws, rather than 2 as in the previous cameras.
The decorative plastic ring is a little annoying; it sits between the two halves of the housing. I don't understand why they put it there.


The illuminator board is completely standard. Power comes directly from the shared connector; the second connector links it to the camera electronics and carries the signal that switches on the illuminator.


Since quite a few data lines are needed between the image sensor and the processor, ordinary wires won't do, so the manufacturer used a ribbon cable. The connector was secured with a small amount of glue or varnish.


The lens has no markings and is secured with a plastic locknut.


Nothing special on the underside: voltage regulators and the IC that controls the IR-cut filter shutter motor.


We remove the last board. There's a packet of silica gel inside. The main cable enters the camera; most of its wires go to the processor board connector (power and Ethernet), while a pair goes to the illuminator board.


The processor board has a heatsink. But the camera gets noticeably less hot during operation than the Szsinocam from the previous review, even though its resolution is twice as high.
There are lots of connectors on the back of the board, most of them unused, but I happened to find the pinout for these connectors.


I won't go into detail; here are just a few separate photos of the components.


The last photo shows the board marking. That's why I didn't remove the heatsink: the processor name is included in the marking.
BLK3516A means the camera uses the HI3516A processor (SoC). Judging by information online, it's most often paired with the 4 MP OV4689 image sensor.


As far as I can tell, the image sensor is quite good, although the cameras don't use all of its capabilities.



Overall view of the disassembled camera and the assembly/disassembly sequence.


Let's move on to the software.
First, I launched the monitoring software, which found the camera automatically. I changed the address in the software, which reported an error, even though the address had actually changed.


After that, I connected to the camera through a browser. ActiveX installation started, and everything went smoothly.


After confirming that I could see the camera through the browser, I decided to visit the manufacturer's website after all and download the software.
This gave me a bit of cognitive dissonance. As far as I know, Escam is quite well known, so I was doubly puzzled about why all the files were hosted somewhere else instead of on their own server.


But I was even more surprised when I tried to download everything provided for this camera.
I was shocked. A major video surveillance equipment manufacturer can't afford decent hosting?


All right, I'll figure out the software later. For now, let's look at using the camera through a browser.
The first time you launch it, a window appears asking you to choose the default operating mode.


I recommend unchecking the bottom box; otherwise, you'll see this window every time the motion sensor is triggered.


Next are screenshots of the camera settings menu. There are quite a lot of them, so I'd better hide them under a spoiler.


And now we can run some tests. A review is all well and good, but it's better to see it once than read about it a hundred times.
To answer the question you might be asking right away: no, it's spring here too. The review just sat in my drafts for a long time :)

The image is generally fine, but it does feel a little dark; I wish I could turn up the brightness.



The camera is installed at the level of the fourth floor, about 17 meters from the subject. At that distance, you won't be able to make out faces, even with a high resolution.


Dusk. The image gets worse—or, more precisely, hazier.


At night, the IR illuminator reaches quite far, but it's hard to make anything out. For example, in this photo, two people are walking or running.


A few photos from the other side of the building.
You can tell spring has arrived—unexpectedly, as always.


To swim or not to swim, that is the question :)


It's dark now. The streetlight is on, along with the IR illuminator, and the camera is in black-and-white mode.


Color mode is forced on, which means the IR filter is removed.


And now, as promised at the very beginning, it's time for some comparative tests.
But first, a bit about the task. An acquaintance asked me to help him set up video surveillance.
The requirements: shelving units with signs on them, like price tags. Each shelf is about 3.3–3.5 m wide, and the text needs to be readable. The letters and numbers are about 15–20 mm high, 25 mm at most.
For the test, I printed text and numbers ranging from 10 to 40 mm high, as well as several groups of lines with different spacing (5, 2.5, and 1.25 mm).

The photos were taken from the same distance, and you can see that the cameras have noticeably different fields of view.
First up is the camera under review, with a medium field of view. The scene is lit by incandescent bulbs.


IR illumination only.


800x600 crop


Since the cameras have different fields of view, for this test I adjusted the distance to the subject so that the horizontal field of view was roughly the same, covering about 3.4–3.5 m.


The image often gets noticeably worse at the edges, since sharpness is usually specified for the center of the lens. So I took photos of all four corners.
For this test, the distance to the subject was the same as to the center. In a real-world setup, the distance would naturally be greater, but here I was checking how much the image degrades at the edges.



Szsinocam SZ-IPC-7032CSW, with a 2 MP resolution. The conditions and order of the shots are the same.
This camera has the widest field of view of those tested.




Here, too, I positioned the camera to get the same horizontal field of view as the previous one.



Hikvision DS-2CD2032F-I.
The narrowest field of view.




For this test, I had to install the camera noticeably farther away than the Szsinocam, by about half a meter.



While preparing the photos for the review, I noticed something interesting. Although the camera under review makes text easier to read, everything else looks hazier, especially the wallpaper pattern. So I ran one more quick test, first trying to adjust the focus a little.
I ended up with four images:
1. Szsinocam SZ-IPC-7032CSW
2. Hikvision DS-2CD2032F-I
3. ESCAM QD410
4. ESCAM QD410 after adjusting the focus


The test was, of course, carried out under ideal “laboratory” conditions. In real life, there will be problems because you can’t always install the camera at a specific distance from the subject. This can result in unwanted information appearing in the frame, taking up part of the useful image area and leaving fewer pixels for what matters.
Possible solutions are to choose a fixed-focus lens or use a varifocal lens. The goal is to make the area covered by the frame as small as possible while still including all the necessary information.

The test showed that the camera under review is suitable for the task, the Hikvision is just about adequate, and the Szsinocam is not suitable. That doesn’t mean I’ll use this particular camera at the site, but the test helped me understand what the image will look like with a 4 MP camera. I’ll use different cameras, most likely Hikvision.

While I was testing, I also noticed a nice feature. This camera has a fairly powerful processor. While testing, I had to move the camera around, sometimes quite actively. At those moments, the camera’s processor receives a maximum amount of information and literally gets “swamped” by it. In normal operation, most of the image is static, so the data stream is much smaller.
The camera performed better than the Hikvision, which occasionally froze for a few seconds, and much better than the Szsinocam, which started lagging badly at the slightest movement (tested at a bitrate of around 10–12 Mbps). The camera under review performed respectably, only occasionally showing some sluggishness.


And now, a few videos.
The video clips were edited in VirtualDub without recompression; the only distortion was introduced by YouTube’s transcoding. I uploaded them to the site in 1440p.
The first clip was shot under an overcast sky. The lighting was better in the following clips, and I also enabled the extended dynamic range.


Here’s a short video at night. The two shadows you can see are people :)
It’s almost completely dark in the area being filmed.


The other side of the house, in the afternoon, in the shade.


The same spot, but in the evening. The first part is in black-and-white mode; color is forced on in the second part.


Conclusions.
Advantages
High resolution
Powerful processor
Power supply included
Waterproof cable gland

Disadvantages
Software is not very user-friendly
Poor and inconvenient technical support from the manufacturer
No PoE

My opinion. As far as image quality goes, I don’t think this camera is much worse than the Hikvision option (I mean the 4 MP version), but there are significant differences in functionality. First, the lack of PoE rules out a fairly large number of possible applications right away. You could use a power supply or adapters, of course, but in my opinion that’s fine “at home, for yourself,” not “on-site, for a client.”
If you’re buying it for personal use, it’s a perfectly decent camera with plenty of settings, though it’s not very convenient to use. Some aspects of both the interface and the functionality feel poorly thought out.
If you want 4 MP resolution and want to save a little, it should do the job. If you can afford to pay more, I’d probably go with Hikvision.
The camera is compatible with Tecsar software, which I liked: it’s quite functional and convenient. It’s included in the supplementary materials.

That’s about it. As always, I’m looking forward to your questions and comments.
Now, about the price. The link in the title goes to the European version of the camera with a power supply; the other cameras cost $98.58. Also, I’m not sure whether it’s in stock, but I can check with the manager.

As supplementary materials, I’m including several datasheets for the installed components, a description of the board connectors, and the Tecsar software.