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

[EMPIRETECANDY]

These pics taken from the 4th floor, and IPC-T5442TM-AS always with IR off with color pic, the IPC-HDW5231R-ZE always with IR on with black/white pic. My apartment near the street, so have good light.
The 4th pic is from iPhoneXs.

 

[Dramus]

Thanks for the photos, @EMPIRETECANDY . Shows what the new cam will do when there's "city light" available. Maybe you could add one with IR off and forced into colour mode for the 5231, to show the difference between it and the 5442 in the 3rd picture?

 

[ThomasPI]

Thanks Andy. We will have a fairly dark area as in no street lights but there will be exterior LED lights which we can leave on at night to help. You mentioned it may take considerable time to tweak these new cams and get them dialed in. Typically is that done with firmware updates or in the past with other cameras, are hardware changes made to achieve desired results ?

 

[looney2ns]

Great review as always @looney2ns .
Really helps us out with actual real world footage rather than manufactured to suit the cameras as a selling point.

Now I have a few 5232r-ze cams in use and as we all know they have been the best around for some time for night and low light.
I can't help but feel a little underwhelmed by the newer starlight+ 4mp variants.
On specs they should be the new king.
The samples you posted look great with the IR on, was that with smart IR enabled as it seems a lot of washout on your face when approaching in direct line.

Also the ones with the IR off and colour mode don't look great at all, a lot of noise and motion blur. I don't see near as much of that with the 2mp (maybe that just the higher pixels tradeoff).

I'd be interested to hear your first hand comparison between the 2 because you've had hands on with both at your location o will be a better comparison.

I have a 5231ze two feet to the left of the test cam, I will do a comparison tonight with the two.
This cam does not offer "smartir" it only has a choice of Auto, manual and off. I assumed, wrongly, that the auto would work the same as smartir. That washout was with the cam in auto mode, and as you can see the obviously the shutter/iris did not react as it does in a 5231ze.

@ThomasPI This could be changed with firmware updates

 

[reeves1985]

What would be interesting, would be to run the camera at H264 and see if you still see the same artefacts on screen.

I have a sneaky suspicion it's the compression playing a part here.

 

[reeves1985]

I have a 5231ze two feet to the left of the test cam, I will do a comparison tonight with the two.
This cam does not offer "smartir" it only has a choice of Auto, manual and off. I assumed, wrongly, that the auto would work the same as smartir. That washout was with the cam in auto mode, and as you can see the obviously the shutter/iris did not react as it does in a 5231ze.

@ThomasPI This could be changed with firmware updates

That's very interesting that it doesn't offer smart IR ,

You'd think that would be a standard issue on these next gen cameras.

As you mention that's a firmware related fix.

Will you try running the cam in h264 if you have time as that might be playing a part with the ghosting if it's a poorly implemented algorithm

 

[ThomasPI]

I have a 5231ze two feet to the left of the test cam, I will do a comparison tonight with the two.
This cam does not offer "smartir" it only has a choice of Auto, manual and off. I assumed, wrongly, that the auto would work the same as smartir. That washout was with the cam in auto mode, and as you can see the obviously the shutter/iris did not react as it does in a 5231ze.

@ThomasPI This could be changed with firmware updates

Thanks @looney2ns - since this to date does not have the vari-focal option, has the bullet version been reviewed yet? Naming convention or model ID ? This particular camera may work in 1 or 2 locations but more likely than not, we are going to really need a vari focal camera in most locations and the Turret for this model is not going to happen.

 

[looney2ns]

What would be interesting, would be to run the camera at H264 and see if you still see the same artefacts on screen.

I have a sneaky suspicion it's the compression playing a part here.

I will do that, by default it's using h264h. Frame rate at 30, 4096bitrate, which I thought was odd. As the other cams 4mp and up the defaults for the bit rates are higher.

 

[handinpalm]

I will do that, by default it's using h264h. Frame rate at 30, 4096bitrate, which I thought was odd. As the other cams 4mp and up the defaults for the bit rates are higher.

Thanks for your great reviews Looney, as always. I would like to see what the cam can really do at night and crank up the bit rate where resolution is not compromised when there are moving targets. Very interested in this one.

 

[bp2008]

The ghosting behind/around @looney2ns as he moved seems to me like the effect of the camera having to stack multiple frames together to produce the final output. If you were to turn off 3d noise reduction (and WDR if it is on) then this effect should be greatly reduced if not eliminated, but the amount of noise in the image would be insane at night so it isn't practical.

 

[EMPIRETECANDY]

Thanks for the photos, @EMPIRETECANDY . Shows what the new cam will do when there's "city light" available. Maybe you could add one with IR off and forced into colour mode for the 5231, to show the difference between it and the 5442 in the 3rd picture?

The firmware is not same as IPC-HDW5231R-ZE, i will check it out, right now still not available on dahua firmware FTP.

 

[EMPIRETECANDY]

Thanks for the photos, @EMPIRETECANDY . Shows what the new cam will do when there's "city light" available. Maybe you could add one with IR off and forced into colour mode for the 5231, to show the difference between it and the 5442 in the 3rd picture?

5442 is brighter than the 5231R-Z in the night when IR light turn off, i think we will have more pics from other guys later this week.

 

[wackystuff]

disappointed that there is no smart ir support...looney's image is really whited out by the ir when he gets close to the camera at night...

 

[looney2ns]

New assist arrived.
You be the judge, I know what I think. What say you?
5231ze

4mp

 

[reeves1985]

New assist arrived.
You be the judge, I know what I think. What say you?
5231ze

4mp

There's a definate improvement in the daytime image as can clearly be seen with those 2 shots.

I'm still tempted to get one purely to play with it to try and unlock it's potential in a few different scenarios.

Like @EMPIRETECANDY just posted those stills showing it being brighter than the 5231,
I've got a street lamp outside my house which will not me me force the 5232 into b&w unless I use schedule. (As a measure of how much light the street lamp gives off).

I think that would be a good barometer of how well it can do against the 5231 in a scene that does have some light available. It would also give another real world example nd use case.

As it looks like @looney2ns has a far darker ambient scene than myself as I have no drive and only a small front yard.

I'll post a still from my 5232 to highlight this

 

[looney2ns]

The IPC-T5442TM-AS set at 30fps, H264, 8192 bitrate. Lamp post on and off.



5231ZE.

 

[HappyBaker]

You be the judge, I know what I think. What say you?

Thank you for posting this comparison.

The 4MP has a higher resolution, but in my eye also suffers from far worse compression/sharpening artefacts; this cancels out most of the potential benefit. On the 2MP I can easily read up to line 7, on the 4MP up to line 8, that's only a small improvement. Maybe the settings of the 4MP can be tweaked to squeeze out a better image with less artefacts?

 

[VisionxOrb]

In this test the 4mp looks like a step up to me under both conditions.

 

[reeves1985]

I'm still torn with this.
I was hoping for more, you can clearly see it's an improvement of sorts but it kind of gets defeated by the huge amount of artifacts on screen whilst your moving.

A real strange one this is.

As a follow up to what I mentioned in my last post heres the ambient light I have in my use case,
It would be interesting to see how it failed in this situation.

 

[reeves1985]

In this test the 4mp looks like a step up to me under both conditions.

It does indeed and hopefully is, but it's being held back a little I feel by the artifacts and ghosting.

I'm sure this can be tweaked out at works by a firmware upgrade.

As you can clearly see it's a sharper crisper image overall