Review-OEM 4mp AI Cam IPC-T5442TM-AS Starlight+

(I've been lurking so far because I'm in the market for some more cams. Haven't been paying attention for about a year and was wondering if there's a good 5231 replacement)

With 8mbit h265 the compression artefacts are mostly gone and I'm actually pretty impressed with this cam. This, combined with the nighttime performance almost makes me get this over a 5231... I think there's another thread with a head-to-head shootout. What do you guys think (assuming 6mm is enough zoom)?

For most applications I don't think that it's worth the large resource hit for what's a relatively small (if any) improvement from what I've seen. Again, it's surveillance video. You're not making 4K documentary films (and the lenses, et. al. still kinda suck independent of the sensor so...). As long as it's good enough, it's good. Granted there are some specific cases where the higher resolution could be useful. For general use, the 5231 and other 2MP Starlights still seem like the best bang for the buck to me.
 
For most applications I don't think that it's worth the large resource hit for what's a relatively small (if any) improvement from what I've seen. Again, it's surveillance video. You're not making 4K documentary films.

#shotondahua

LOL


Sent from my iPhone using Tapatalk Pro
 
  • Like
Reactions: giel and watts
For most applications I don't think that it's worth the large resource hit for what's a relatively small (if any) improvement from what I've seen. Again, it's surveillance video. You're not making 4K documentary films (and the lenses, et. al. still kinda suck independent of the sensor so...). As long as it's good enough, it's good. Granted there are some specific cases where the higher resolution could be useful. For general use, the 5231 and other 2MP Starlights still seem like the best bang for the buck to me.

Whats funny for me is, given the option Id take 4k documentary quality any day of the week.
 
  • Like
Reactions: bigredfish
IF both cameras both produced results equal to the 5231, I’d be inclined to agree. This is not the case as of today. Nice start but it’s lacking considerably IMHO and YMMV.
 
  • Like
Reactions: looney2ns
For most applications I don't think that it's worth the large resource hit for what's a relatively small (if any) improvement from what I've seen. Again, it's surveillance video. You're not making 4K documentary films (and the lenses, et. al. still kinda suck independent of the sensor so...). As long as it's good enough, it's good. Granted there are some specific cases where the higher resolution could be useful. For general use, the 5231 and other 2MP Starlights still seem like the best bang for the buck to me.

Absolutely agreed.
 
I woud like to buy three cameras for my property within the next six weeks. Would you prefer the PC-T5442TM-AS or the IPC-HDW5231R-ZE?
 
The extra megapixels are worth it to me. I had 3MP 3.6mm amcrest bullets that struggled to capture license plates of cars parked on street and coming into our shared motor driveway. Upgrading to 4MP Amcrest fixes and 8MP Dahua 1831 subsequently fixed the problem. I don’t have to sacrifice wide viewing angle just to get plates. I can get both. Where I live, none of our cameras go into night mode because our community has lots of lighting.
 
I have one of these cameras inside and my tile looks weird with the lines. When I enlarge the video stream, everything looks normal. It looks the same on my mobile app. It also appears in my regular Dahua cam viewer through FireFox.

Tried change most of the video settings, including putting to 1080p.
s5Y17B1

5sZSwyC
 

Attachments

  • Capture.PNG
    Capture.PNG
    31.5 KB · Views: 232
  • Untitled-1.jpg
    Untitled-1.jpg
    47.8 KB · Views: 235
For most applications I don't think that it's worth the large resource hit for what's a relatively small (if any) improvement from what I've seen. Again, it's surveillance video. You're not making 4K documentary films (and the lenses, et. al. still kinda suck independent of the sensor so...). As long as it's good enough, it's good. Granted there are some specific cases where the higher resolution could be useful. For general use, the 5231 and other 2MP Starlights still seem like the best bang for the buck to me.

Agreed on the documentary although left Mark is more easily recognisable than right Mark. That to me is a significant gain and given that right Mark should be sharper and more recognisable that left Mark, with the increased sharpness and pixel count from 4mp, it's a sign that the current compression under h264 is too much to get a decent result.

From looking at some of those images full size I get the impression the new AI 4mp camera might be oversharpening things a bit in it's compression algo. The 5231 does not seem to do that. For now I think I prefer the 5231, even in broad daylight.

Yeah. I think they've turned up the sharpness a bit to compensate for the loss of detail due to compression. Unfortunately, as seen in the comparison, whilst it helps with the text, left Mark is more detailed than right Mark, something that shouldn't happen given the almost double pixel count on the right. The issue is, all the gains from the extra pixels are lost if simply throw them away again when you compress the image. At least that's my reading of what's going on.

I know some will scream at me for saying this. But I'd like to see the right picture at 12 and 15mbs for comparison. Not going to happen as neither the camera nor BI support that. However, I said it a long time ago, my approach to compression is simple, keep whatever you want as default for the large commercial guys to whom storage space is a premium, but give others the option of other settings in the firmware / software as for home users, small installations, installations where either quality is at a premium or storage is more freely available.

The only good news here is storage costs are falling fast. WD Purple drives are only a fraction of what they cost a year ago.
 
Last edited:
Agreed on the documentary although left Mark is more easily recognisable than right Mark. That to me is a significant gain and given that right Mark should be sharper and more recognisable that left Mark, with the increased sharpness and pixel count from 4mp, it's a sign that the current compression under h264 is too much to get a decent result.



Yeah. I think they've turned up the compression a bit to compensate for the loss of detail due to compression. Unfortunately, as seen in the comparison, whilst it helps with the text, left Mark is more detailed than right Mark, something that shouldn't happen given the almost double pixel count on the right. The issue is, all the gains from the extra pixels are lost if simply throw them away again when you compress the image. At least that's my reading of what's going on.

I know some will scream at me for saying this. But I'd like to see the right picture at 12 and 15mbs for comparison. Not going to happen as neither the camera nor BI support that. However, I said it a long time ago, my approach to compression is simple, keep whatever you want as default for the large commercial guys to whom storage space is a premium, but give others the option of other settings in the firmware / software as for home users, small installations, installations where either quality is at a premium or storage is more freely available.

The only good news here is storage costs are falling fast. WD Blue drives are only a fraction of what they cost a year ago.

I agree with your points on this and from the samples posted. But if you have a look at the 4mp bullet topic, which o think use the same sensor and tech, the samples posted there clearly show a difference in favour of the 4mp.

The night shots on that particular setup are slightly and I say slightly better than the 2mp.
But add the extra detail during daylight I'm swinging towards now favouring the 4mp.

It also seems that now they've sorted the auto IR it works slightly better than the smart IR of the 5231.

It's still definately a "it depends scenario"

But I can see the benefits
 
I haven't looked and so can't comment. There may be other reasons. However, one other point, from memory, h265 is twice as efficient, yet Dahua or BI (not sure which) from memory restricted h.265 to 5mbs max instead of 8mbs for h264. One way of helping a higher mp camera such as this without increasing the top bit rate would be to simply allow h265 to compress at 8mbs the same as h264. If it's twice as efficient, that should be the equivalent of 16mbs without any additional storage overhead.
 
I haven't looked and so can't comment. There may be other reasons. However, one other point, from memory, h265 is twice as efficient, yet Dahua or BI (not sure which) from memory restricted h.265 to 5mbs max instead of 8mbs for h264. One way of helping a higher mp camera such as this without increasing the top bit rate would be to simply allow h265 to compress at 8mbs the same as h264. If it's twice as efficient, that should be the equivalent of 16mbs without any additional storage overhead.
BI restricts nothing. H265 is not twice as efficient.
 
  • Like
Reactions: EMPIRETECANDY
It's almost twice as efficient in my video editor - with the same short clip. At 1080P 25fps, under h.264 it comes out at 14.8mbs and under hevc @ 8.3mbs. The respective data rates set by the encoder are 13,444kbs for h264 and 8,041kbs for hevc, although there's no really visible difference in final quality for the lower rate. Might be slightly worse in hevc, but it' so close as to be very difficult to tell. Maybe the encoder is set up differently in BI?
 
I will arrange it tomorrow, because right now we will do some more discountS for all of the products that we have stock in US.