Hi guys,
Today I present you a review of Dahua T5442T-ZE. This model is equipped with 2.7-12mm zoom lens and an excellent 4 megapixel 1/1.8 inch CMOS sensor. Compared to 5442TM-AS-LED and 5442TM-AS models, which I’ve already reviewed, this one has twice as much RAM available.
The housing is slightly bigger than TM-AS model. It has only 3 screws for mounting, and – in my opinion – better shape for final positioning and installation. It just snaps easier when you screw it for final position. There is also a sign on the housing, indicating the top part of the camera, so you don’t have to worry about mounting it upside down.
From Polish division of Dahua I know that the sensor used in 5442TM-AS camera is not Sony (they didn’t want to tell which one is used, but clearly stated it’s not Sony Starvis), so the only other option that fits with the specs is OmniVision’s OS04A10 CMOS. I think, it is safe to assume, it’s the same sensor in all 54XX/74XX series. Dahua uses image sensors from 3 different manufacturers at this moment – SONY, OmniVision and SmartSens. The last one also has 4MP 1/1.8 inch sensor, but it differs slightly in terms of resolution and pixel size. Sensors from SmartSens are presumably used in new 5MP (3K, 16/9 image ratio) HDCVI analog camera series.
The 5442T-ZE has slightly darker lens than TM-AS model. According to Dahua specs, it starts at f1.8 compared to f1.6 in TM-AS. This is 1/3EV difference. My tests, when comparing both cameras, showed a difference of almost 2/3EV. To balance the brightness level, I had to use 40ms shutter on T-ZE (zoomed out), while 27ms on TM-AS-LED. So either TM-AS-LED is f1.4 or T-ZE starts at f2.0 or the gain scale is shifted. Which one is true? I don’t know. This difference is however not that big and shouldn’t be treated as a serious drawback.
As in case of my other camera with zoom lens, the 5831R-ZE the lens is prone to flare and chromatic aberration. In fixed lens models this is less probable, and the reason for that is because zoom lenses have more optical elements, and light passing through layers of glass is more likely to produce flare and chromatic aberration.
There is no wobbling when the lens is zooming. Autofocus is slightly slow but precise. I haven’t observed any issues with wrong focusing or soft edges as in case of 5831R-ZE. When zoomed all the way in, the lens looses about 1EV of light. So lens brightness starts at f1.8 at wide end, and ends at f2.5 at tele.
Talking about the zoom… once I’ve wrote this post:
And while it is true for all my photography lenses and my 25-400mm zoom camera, it somehow doesn’t work with Dahua zoom cameras. I found it to be false when dealing with 5831R-ZE. I couldn’t figure out, what was going on, and left this problem for a while. But now, playing with 5442T-ZE, I thought it is good time to recalculate this issue again. I’ve spent quite a bit of time measuring, searching and calculating again and again. Finally came to one conclusion… Dahua is messing up with numbers.
In most cases their cameras are available in following variants:
Fixed lens: 2.8mm/3.6mm/6mm
Varifocal lens: 2.7mm-12mm
Looking at these values, it seems the zoom version has clearly better range, than all 3 fixed options together. In reality it is not as it seems. If you look at the FOV values for 5442T-ZE and TM-AS on Dahua’s page, you’ll see following values (horizontal):
Fixed lens: 113o/89o/56o
Varifocal lens: 114o – 47o
To achieve 56o FOV with 1/1.8 inch sensor, the 6mm fixed lens is actually a 7mm, and varifocal lens ranges between 2.6mm and 9mm.
I’ve compared T-ZE with EM-AS-LED, and measured their real FOV. In reality the wide end of varifocal lens starts at 105o while the fixed 2.8mm TM-AS is 110o. The long end of varifocal is indeed 47 o. This means the varifocal T-ZE is actually 3mm to 9mm.
There is a phenomenon called lens breathing. Essentially it is a shift in FOV, when changing focus from close distance to far distance. Good quality video lenses are almost free of this phenomenon. Photography lenses exhibit this issue at different levels. In case of Dahua zoom lens the breathing is quite severe. And so, if I shift focus all the way from close to far end, you can see the FOV clearly changes.
When focus is at its closest distance value, the lens indeed achieves 114o FOV. I suppose, Dahua is testing their lenses with a test board at a very close distance, hence the 114o value. In reality you would mostly focus on distant objects or infinity. In such case the FOV is only 105o.
In summary, when dealing with zoom version, we are not getting declared 114 o, only 105 o, and we definitely are not reaching 12mm (33o FOV) but only 9mm (47o FOV). Now, just compare the FOV values between fixed 6mm and varifocal 12mm. 56o vs. 47o - 9 degrees of difference. Not much. Here is, how it looks in practice:
Why am I writing about this? Because those, who want to use a tele view, will most likely choose varifocal version of the camera over fixed one, looking at focal length values, while in reality the difference is small. You are not gaining much magnification choosing varifocal, but loose 1.3EV - over 50% of light, because at long end the varifocal lens is f2.5 while the 6mm fixed lens variant is f1.6 (at least on paper).
I’m not saying the varifocal camera is worse than fixed one. Both cameras are good and each one has its porpoise. It is just good to know such things, before choosing the right gear for yourself. And don’t worry – the HFW7842H-Z has identical zoom specs as T5442T-ZE and it suffers similar issues (108o-47o). The HDW5831R-ZE is also inconsistent with focal length values, but at least it complies with official FOV values (110o-39 o) and so it has grater “range” than 5442 and 7842.
The image sensor on T5442T-ZE reaches ISO40000 at 100% gain. It’s the same as in TM-AS and TM-AS-LED models and is higher than all other camera models I’ve already reviewed (apart from 4239T-ASE). When compared with TM-AS-LED, both cameras have same noise level when looking at image with no noise reduction. After equalizing lens brightness difference (27ms vs 40ms shutter) they both look identical from gain 0% to 100%. The only difference is when the noise reduction is turned on. It seems NR on TM-AS-LED is slightly more aggressive. Even when I apply additional 2D noise reduction (available after upgrading to SMD 3.0 firmware) the noise is still more pronounced in T-ZE.
The only real competitor to this 4-megapixel sensor would be Sony’s IMX327 from Dahua 4239T-ASE. There is also an 8MP 1/1.2inch IMX485, already present in Hikvision’s camera, which could probably compete here. But unfortunately Dahua still has no camera model with this one.
At daytime, the T5442T-ZE wins hands down. It has of course better resolution, but also much better WDR function. The WDR on all 5442 models is actually the best of all my reviewed cameras. It has no image artifacts (apart from motion of course), turns on instantly, and keeps good image contrast and color, even at high values.
What happens during night? Here is a video of temporary blackout on the street during night. This is between T5442TM-A-LED and 4239T-ASE, but it applies also to T5442T-ZE. Notice the reaction of both cameras in terms of gain boost, when it gets dark. The 5442 reacts instantly, while 4239 is much slower.
I compared both cameras with basic image setup – factory one. I’ve set zoom position on T5442T-ZE to match FOV of 4239T-ASE. No exposure compensations due to lens differences, no enhancements. When looking at static image, the 5442 looks better. Especially darker areas are handled better and of course the resolution does its job too. But when motion appears, it turns out the 4239T-ASE is doing similar, or even tiny bit better than 5442T-ZE. If I turn on IR night mode, I get additional ~1EV of light. I makes the night image on 5442T-ZE slightly more detailed. The 4239T-ASE has no IR mode. The downside of IR, is that it sometimes affects drastically the appearance of objects on the image. Looking at the example below, it looks like I was wearing white clothes. It makes me more visible, but when someone asks you, what was the person wearing, what would you say? In reality I was all in black.
For me, the conclusion is that if both cameras are doing similar at night, and 5442 clearly wins during the day (in all terms), the final winner is 5442.
Summarizing, the T5442T-ZE is a nice piece of equipment. Being equipped with very efficient 4MP COMS sensor, I consider it as highly recommended for those, who need the best quality image and zoom function. If zoom is not necessary, go for fixed lens version. There is of course a 7442H-Z with f1.2 zoom lens, but it’s a completely different price tag. You can also consider a HFW5442T-ASE-NI with f1.0 fixed lens. For the FOV issue… this is most likely about all zoom cameras from Dahua, not only T5442T-ZE.
Today I present you a review of Dahua T5442T-ZE. This model is equipped with 2.7-12mm zoom lens and an excellent 4 megapixel 1/1.8 inch CMOS sensor. Compared to 5442TM-AS-LED and 5442TM-AS models, which I’ve already reviewed, this one has twice as much RAM available.
The housing is slightly bigger than TM-AS model. It has only 3 screws for mounting, and – in my opinion – better shape for final positioning and installation. It just snaps easier when you screw it for final position. There is also a sign on the housing, indicating the top part of the camera, so you don’t have to worry about mounting it upside down.
From Polish division of Dahua I know that the sensor used in 5442TM-AS camera is not Sony (they didn’t want to tell which one is used, but clearly stated it’s not Sony Starvis), so the only other option that fits with the specs is OmniVision’s OS04A10 CMOS. I think, it is safe to assume, it’s the same sensor in all 54XX/74XX series. Dahua uses image sensors from 3 different manufacturers at this moment – SONY, OmniVision and SmartSens. The last one also has 4MP 1/1.8 inch sensor, but it differs slightly in terms of resolution and pixel size. Sensors from SmartSens are presumably used in new 5MP (3K, 16/9 image ratio) HDCVI analog camera series.
The 5442T-ZE has slightly darker lens than TM-AS model. According to Dahua specs, it starts at f1.8 compared to f1.6 in TM-AS. This is 1/3EV difference. My tests, when comparing both cameras, showed a difference of almost 2/3EV. To balance the brightness level, I had to use 40ms shutter on T-ZE (zoomed out), while 27ms on TM-AS-LED. So either TM-AS-LED is f1.4 or T-ZE starts at f2.0 or the gain scale is shifted. Which one is true? I don’t know. This difference is however not that big and shouldn’t be treated as a serious drawback.
As in case of my other camera with zoom lens, the 5831R-ZE the lens is prone to flare and chromatic aberration. In fixed lens models this is less probable, and the reason for that is because zoom lenses have more optical elements, and light passing through layers of glass is more likely to produce flare and chromatic aberration.
There is no wobbling when the lens is zooming. Autofocus is slightly slow but precise. I haven’t observed any issues with wrong focusing or soft edges as in case of 5831R-ZE. When zoomed all the way in, the lens looses about 1EV of light. So lens brightness starts at f1.8 at wide end, and ends at f2.5 at tele.
Talking about the zoom… once I’ve wrote this post:
Using vari-focal camera to determine focal length of fixed-focus camera
I've read several times on this (very informative by the way) forum using this technique. I'm planning to do this too: New setup: 2 camera's and a NVR But here's the question; how exactly do you do that? Personally, I have the IPC-HDW5231R-ZE which goes from 2.7mm to 13.5mm. For another setup...ipcamtalk.com
And while it is true for all my photography lenses and my 25-400mm zoom camera, it somehow doesn’t work with Dahua zoom cameras. I found it to be false when dealing with 5831R-ZE. I couldn’t figure out, what was going on, and left this problem for a while. But now, playing with 5442T-ZE, I thought it is good time to recalculate this issue again. I’ve spent quite a bit of time measuring, searching and calculating again and again. Finally came to one conclusion… Dahua is messing up with numbers.
In most cases their cameras are available in following variants:
Fixed lens: 2.8mm/3.6mm/6mm
Varifocal lens: 2.7mm-12mm
Looking at these values, it seems the zoom version has clearly better range, than all 3 fixed options together. In reality it is not as it seems. If you look at the FOV values for 5442T-ZE and TM-AS on Dahua’s page, you’ll see following values (horizontal):
Fixed lens: 113o/89o/56o
Varifocal lens: 114o – 47o
To achieve 56o FOV with 1/1.8 inch sensor, the 6mm fixed lens is actually a 7mm, and varifocal lens ranges between 2.6mm and 9mm.
I’ve compared T-ZE with EM-AS-LED, and measured their real FOV. In reality the wide end of varifocal lens starts at 105o while the fixed 2.8mm TM-AS is 110o. The long end of varifocal is indeed 47 o. This means the varifocal T-ZE is actually 3mm to 9mm.
There is a phenomenon called lens breathing. Essentially it is a shift in FOV, when changing focus from close distance to far distance. Good quality video lenses are almost free of this phenomenon. Photography lenses exhibit this issue at different levels. In case of Dahua zoom lens the breathing is quite severe. And so, if I shift focus all the way from close to far end, you can see the FOV clearly changes.
When focus is at its closest distance value, the lens indeed achieves 114o FOV. I suppose, Dahua is testing their lenses with a test board at a very close distance, hence the 114o value. In reality you would mostly focus on distant objects or infinity. In such case the FOV is only 105o.
In summary, when dealing with zoom version, we are not getting declared 114 o, only 105 o, and we definitely are not reaching 12mm (33o FOV) but only 9mm (47o FOV). Now, just compare the FOV values between fixed 6mm and varifocal 12mm. 56o vs. 47o - 9 degrees of difference. Not much. Here is, how it looks in practice:
Why am I writing about this? Because those, who want to use a tele view, will most likely choose varifocal version of the camera over fixed one, looking at focal length values, while in reality the difference is small. You are not gaining much magnification choosing varifocal, but loose 1.3EV - over 50% of light, because at long end the varifocal lens is f2.5 while the 6mm fixed lens variant is f1.6 (at least on paper).
I’m not saying the varifocal camera is worse than fixed one. Both cameras are good and each one has its porpoise. It is just good to know such things, before choosing the right gear for yourself. And don’t worry – the HFW7842H-Z has identical zoom specs as T5442T-ZE and it suffers similar issues (108o-47o). The HDW5831R-ZE is also inconsistent with focal length values, but at least it complies with official FOV values (110o-39 o) and so it has grater “range” than 5442 and 7842.
The image sensor on T5442T-ZE reaches ISO40000 at 100% gain. It’s the same as in TM-AS and TM-AS-LED models and is higher than all other camera models I’ve already reviewed (apart from 4239T-ASE). When compared with TM-AS-LED, both cameras have same noise level when looking at image with no noise reduction. After equalizing lens brightness difference (27ms vs 40ms shutter) they both look identical from gain 0% to 100%. The only difference is when the noise reduction is turned on. It seems NR on TM-AS-LED is slightly more aggressive. Even when I apply additional 2D noise reduction (available after upgrading to SMD 3.0 firmware) the noise is still more pronounced in T-ZE.
The only real competitor to this 4-megapixel sensor would be Sony’s IMX327 from Dahua 4239T-ASE. There is also an 8MP 1/1.2inch IMX485, already present in Hikvision’s camera, which could probably compete here. But unfortunately Dahua still has no camera model with this one.
At daytime, the T5442T-ZE wins hands down. It has of course better resolution, but also much better WDR function. The WDR on all 5442 models is actually the best of all my reviewed cameras. It has no image artifacts (apart from motion of course), turns on instantly, and keeps good image contrast and color, even at high values.
What happens during night? Here is a video of temporary blackout on the street during night. This is between T5442TM-A-LED and 4239T-ASE, but it applies also to T5442T-ZE. Notice the reaction of both cameras in terms of gain boost, when it gets dark. The 5442 reacts instantly, while 4239 is much slower.
I compared both cameras with basic image setup – factory one. I’ve set zoom position on T5442T-ZE to match FOV of 4239T-ASE. No exposure compensations due to lens differences, no enhancements. When looking at static image, the 5442 looks better. Especially darker areas are handled better and of course the resolution does its job too. But when motion appears, it turns out the 4239T-ASE is doing similar, or even tiny bit better than 5442T-ZE. If I turn on IR night mode, I get additional ~1EV of light. I makes the night image on 5442T-ZE slightly more detailed. The 4239T-ASE has no IR mode. The downside of IR, is that it sometimes affects drastically the appearance of objects on the image. Looking at the example below, it looks like I was wearing white clothes. It makes me more visible, but when someone asks you, what was the person wearing, what would you say? In reality I was all in black.
For me, the conclusion is that if both cameras are doing similar at night, and 5442 clearly wins during the day (in all terms), the final winner is 5442.
Summarizing, the T5442T-ZE is a nice piece of equipment. Being equipped with very efficient 4MP COMS sensor, I consider it as highly recommended for those, who need the best quality image and zoom function. If zoom is not necessary, go for fixed lens version. There is of course a 7442H-Z with f1.2 zoom lens, but it’s a completely different price tag. You can also consider a HFW5442T-ASE-NI with f1.0 fixed lens. For the FOV issue… this is most likely about all zoom cameras from Dahua, not only T5442T-ZE.
