Dahua WizColor 5x59-PRO and SmartLight 5x59-IL new series

[wittaj]

@bigredfish - If you have some specific settings you want me to do on the T54PRO and 5442-S3 (and tell me day and/or night), I'm happy to do it for 'ya.

If you have a fairly wide-angle camera with most stuff at 50' (or more away) with everything being at least 20' away, why not run the iris wide-open (unless the lens really need to be closed down) if you are in a low-light situation?

You call this low light

Not sure about the pro version, but folks that get under the hood of the firmware say it isn't good to run any parameters on the extreme as the camera will compensate by overriding another setting somewhere, and I would assume the pro version would be the same.

Remember, this is a digital iris, so maxing it on an extreme means the camera is compensating in some other parameter.

As bigredfish said, you can adjust the 5442 to match the color.

Most here will go with the camera that performs better in low light. Daylight can be adjusted to make most cameras look similar.

 

[bigredfish]

@bigredfish - If you have some specific settings you want me to do on the T54PRO and 5442-S3 (and tell me day and/or night), I'm happy to do it for 'ya.

If you have a fairly wide-angle camera with most stuff at 50' (or more away) with everything being at least 20' away, why not run the iris wide-open (unless the lens really need to be closed down) if you are in a low-light situation?

I'll try it. I guess my typical target interest is 0-20ft or so. Beyond that I just consider it Overview and haven't seen a real benefit but I'll admit to not really focusing on it. I have a camera thats overview that is ideal for that.

See the Z4 testing thread for my latest on it..... gimme a few minutes to post

 

[alekk]

You call this low light

Of course middle of the day is not low-light.

Here's a snapshot I just took with 5msec shutter, 50 iris, 50 gain, 0 NR.
Yes, all you are seeing is the driveway lights and a few others ... THAT is low light! ;-)

BTW, it's a clear sky with an almost full moon that is above the horizon, but not casting strong shadows yet ... ;-)

So if I want to run full color (without any extra illumination), what do you suggest I change to make this useable?

 

[bigredfish]

IR - that scene isn't appropriate for color

 

[Looksee]

Multiple 18k fresnel lights on scafolding or lifts should do it.

 

[wittaj]

+1 above. All cameras need light, either visible or IR.

Or you slow down shutter and crank up gain and iris for the nice static image that is a blur with motion.

 

[bigredfish]

Where’s @CanCuba when you need him?
You want light? Airlines mistake his street for runway 7R

 

[alekk]

LOL @lookee with the Hollywood lighting ... totally the RIGHT answer ... but NOT an option.

Here's an IR picture for 'ya ... parameters are now 10 msec shutter (was 5), 100 iris (was 50), and 50 gain (same as before). 0 NR ... so anything not totally black is super noisy.

 

[alekk]

Same parameters, but turned on the IR lights (both near & far) to 100. Also turned on the NR (50 for both 3D and 2D) because without that, it's "dancing dark pixel" party! ;-)

BTW, I'm kinda impressed on the IR light since the middle of the driveway is 50' away ... so that other side of the street is 100' plus.

 

[alekk]

Now lets go back to Color (no illuminators) and in order to get "something", we are now at 20 msec shutter, 100 iris, and 65 gain. But without NR, we've got the "dark dancing pixel" party again, so we put 3D & 2D to 50.

BTW, be VERY careful going above 65 for gain (which is already waayyy more than I'd like to do) because the images really, really, falls apart.

So now I can "see" something ... but (DUH!) those parameters are only going to capture the slowest of motion.

 

[alekk]

One more picture - same params as above ... except shutter is now at 50msec ... so again, unless someone stands still, motion capture is going to SUCK at 1/20s ... plus it's a noisy image with a bunch of NR ... so motion is a blurry mess.

I'm just showing you guys the light levels I'm dealing with ... and did you notice the full moon shadow ... albeit there are now some light clouds that are slightly diffusing it.

Looks decent (and I can clean it up a bit more) ... but clearly a "REOLINK" image ... although impressive that the T54PRO can "semi!" do this at 1/20s ... which is much better than the 5442-S3 which I'll post from pics from.

 

[alekk]

Here's a comparison of the T54PRO-ZE versus the 5442-S3.

Obviously color (although there's a LOT more of it from the T54) with no illuminators, 50msec shutter, iris 100, 3D & 2D NR at 50.
The gain is at 65 on the T54 and it is 100 on the 5442.
Clearly ridiculous parameters to try to capture any motion, but our image is just too dark.

This is what I mean by the T54 can do a much better static image that then 5442 ... albeit I'm sure it's not all sensor/lens, but some AI processing contributing.

However, at this point, neither are going to do well with a moving object ... at THIS light level.

But the progress is encouraging - give me the next generation sensor with a 1/1.2" sensor (versus 1/1.8") and F/1.0 lens (versus F/1.2) and we might be able to get the shutter speed up to 1/100s and not have to process the sh*t out of it! ;-)


HOWEVER, for those of you who have MORE light, it seems like the T54 should be able to do better than the 5442 ... BUT ... from what people have experienced/posted about, there seems to be some funky processing by the T54 that "artificially" improves the static image ... but moving objects are ... "moving" ... so those algorithms show erratic results.

 

[flynreelow]

Here's a comparison of the T54PRO-ZE versus the 5442-S3. Obviously color, no illuminators, 50msec shutter, iris 100, 3D & 2D NR at 50.
The gain is at 65 on the T54PRO-ZE and it is 100 on the 5442-S3.

This is what I mean by the T54PRO-ZE can do a much better static image that then 5442-S3 ... albeit I'm sure it's not all sensor/lens, but some AI processing contributing.

However, at this point, neither are going to do well with a moving object ... at THIS light level.

But the progress is encouraging - give me the next generation sensor with a 1/1.2" sensor (versus 1/1.8") and F/1.0 lens (versus F/1.2) and we might be able to get the shutter speed up to 1/100s and not have to process the sh*t out of it! ;-)

View attachment 231846

So in reality u are saying for good non blurry motion at night, you are going to need way more light, use IR only, or use the on board illuminators?

 

[alekk]

So in reality u are saying for good non blurry motion at night, you are going to need way more light, use IR only, or use the on board illuminators?

I'm saying for THIS SCENE at THIS LIGHT LEVEL, neither camera is capable of capturing non-blurry motion without adding more (color or IR) light.

However, the T54PRO-ZE is able to capture a much better static color image than the 5442-S3 ... so I think(?) it has a better sensor/lens, but based on what people (who are much more experienced/knowledgeable than me) have showed with motion, some (lot?) of that is due to processing ... which @steve1225 has talked about several times on these new cameras.

BTW, I may pull out my external IR illuminator to compare the T54 versus the 5442 in IR with that added light.

 

[bigredfish]

I need some time today to pull together samples for you but here's my premise:

The 54PRO is clearly "brighter" (when dialed in to reduce gross amounts of blur and noise like any camera), but there is still a "digital fuzz" present far greater than standard noise seen on the 5442.
Now some of this may be preference, but I think I can show that the "brightness" of the 54PRO does not necessarily result in a better moving image or freeze frame.

I think we're seeing exactly what the marketing folks wanted us to see. Brighter=Better. (this trend has been around for a number of years by various manufacturers. I recall the famous HiK "starlight video" put out by Sean at Nelly's 4-5 years ago)
And maybe with enough time we can learn to adjust out the fuzz and future FW can help.
Time will tell.

 

[CamCrazy]

I need some time today to pull together samples for you but here's my premise:

The 54PRO is clearly "brighter" (when dialed in to reduce gross amounts of blur and noise like any camera), but there is still a "digital fuzz" present far greater than standard noise seen on the 5442.
Now some of this may be preference, but I think I can show that the "brightness" of the 54PRO does not necessarily result in a better moving image or freeze frame.

I think we're seeing exactly what the marketing folks wanted us to see. Brighter=Better. (this trend has been around for a number of years by various manufacturers. I recall the famous HiK "starlight video" put out by Sean at Nelly's 4-5 years ago)
And maybe with enough time we can learn to adjust out the fuzz and future FW can help.
Time will tell.

100% on all points, I feel like there is a good sensor locked away in there, so far with no way of escaping! Grain seems to be a thing with these 54Pro's, which I do not mind if the detail is there. All IP cams should be setup from the factory to capture at least human motion, both day and night. Remember going into CCTV outlets years ago with the bright images setup on demo, of course upon playback the story changes!! It should stop but I guess it will not.

Both my (unboxed) Pro's are now set in such a way that they didn't switch over to colour until late morning, due to it being overcast! - such is the requirement for settings to freeze motion even during daylight hours.

 

[bigredfish]

I swear every time I set aside time to do some camera adjusting/testing/comparison, we get an ugly dark overcast day....

 

[[email protected]]

I swear every time I set aside time to do some camera adjusting/testing/comparison, we get an ugly dark overcast day....

It's like that day in day out in the UK, these days.

 

[CamCrazy]

It's like that day in day out in the UK, these days.

This is the truth, so dark here lately, good for camera testing at the extremes though, so there is that!