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

[redfive]

Well done, @jmcu !!
While the T54Pro-ZE has great, crystal clear static image, IMHO, the S3 still wins with moving targets (at least, with those settings).

 

[steve1225]

@jmcu thanks for video...

There are smaller/bigger problems with facial details recovery on all cameras..
Shutter time is too big - first we don't see big difference between cameras (hard to compare), second probably this is one of the reasons why we have flat faces (they are moved in long shutter time).

As I understands all cameras are in automatic modes (shutter time etc)?

I propose to setup them ALL to the same Manual mode & custom shutter with starting parameters:

• shutter set to 0..10ms (or 0..8,3ms is you in the US),
• gain 0..50,
• both 2D and 3D noise reduction set to universal 0..30

If final video will be too dark on all cameras (especially for human movement) then you can manipulate with those parameters. Or even please show us video footage at a few different configurations of those parameters (1/120, 1/60 etc).. but always the same config on all 4 cams...

WizColor cams show compression or image recovery problems around "moving human". This is especially visible on 54PRO-ZE, where all gras around moving human are very blurred.

Please check - if changing Encoding strategy to AI Coding (with enabled IVS), with other encoding parameters not changed (CBS, H264H, 10240kb) helps here anything?

ps. interesting is that cameras have different width / height ratio - it's very visible in first seconds of you entrance.. on 54PRO-ZE your are "thinest", on 34PRO you are "widest"

ps2. does all cameras have latest firmware? and both 54PRO have the same firmware - they look too different???

 

[jmcu]

Great footage @jmcu ... just to be clear, was the gain (ISO) "about" the same on all of them?

At a QUICK glance, the overall exposure level seems about the same on all of them ... but looks like some "boost" in the shadows especially for the IPC-54PRO-AS. I think (?) I saw some better dynamic range for the PRO models ... but that may have been because the color temperature seemed more natural to me. I think(!) the PRO's look sharper ... but hard to tell with different focal lengths ... and I think their focus points might be a little closer as I'm looking at foreground grass and the lines in the sidewalk. The S3's telephone pool looks a bit less fuzzy and car spokes (nice to have) look pretty decent.

So maybe I'm clueless, but didn't seem (?) like a clear winner ... I was hoping to see the new camera's clearly show markedly better low light performance.
Curious what others say as I'm sure I'm missing stuff ... plus what do you think?

I'm wondering if maybe either try in a darker area (although it's a good combination of lit/dark) and/or bump the shutter speed up ... so you are "pushing" the limits of the cameras.
I.e. push the exposure triangle so the image from the "laggard" falls apart even in the static scene ... so easier to see who "wins" ... and yes, no IR and no on-board illumination.

BTW, I really like how you provide the "quad" screen and then each individual cam. Plus I really appreciate how you walk through the scene ... so showing not just "static" IQ, but also dynamic IQ. White shirt is a good test ... but maybe some more color in the shorts/shoe - LOL maybe you have an colorful umbrella you can walk through with to exaggerate it. Ditto if you walk around with something that has some big letters on it (or your shirt) for sharpness testing! ;-)

Again, GREAT job ... lining everything up for similar FOV, match settings, etc. for these type of tests is a LOT of work.

All the cameras settings were set to auto, they are so close to each other in the auto mode, I'm sure if you spend the time manually adjusting you could get the most out of all of them, look at @bigredfish S3- he has that super nice looking like 4k !!
I don't spent the time fine tuning them unless I am keeping them for myself as most of my samples are just first look and I am always changing them out with new products. And your correct, it can be a lot of work.

I think you will see better quality samples out of the new cameras as some of the others who received samples @wittaj will have more time to fine tune them and give their full reviews.

I was leaning towards the S3 being a better picture ( as this has been my favorite for a long time) but I think the PRO will overtake it in the near furture as they fine tune the FW.
I like the IR a bit better and the 2 way talk is nice on the PRO, I think it will be a great front door cam. I found the LED white light usefull for close up but not helpful at a distance.

 

[jmcu]

@jmcu thanks for video...

There are smaller/bigger problems with facial details recovery on all cameras..
Shutter time is too big - first we don't see big difference between cameras (hard to compare), second probably this is one of the reasons why we have flat faces (they are moved in long shutter time).

As I understands all cameras are in automatic modes (shutter time etc)?

I propose to setup them ALL to the same Manual mode & custom shutter with starting parameters:

• shutter set to 0..10ms (or 0..8,3ms is you in the US),
• gain 0..50,
• both 2D and 3D noise reduction set to universal 0..30

If final video will be too dark on all cameras (especially for human movement) then you can manipulate with those parameters. Or even please show us video footage at a few different configurations of those parameters (1/120, 1/60 etc).. but always the same config on all 4 cams...

WizColor cams show compression or image recovery problems around "moving human". This is especially visible on 54PRO-ZE, where all gras around moving human are very blurred.

Please check - if changing Encoding strategy to AI Coding (with enabled IVS), with other encoding parameters not changed (CBS, H264H, 10240kb) helps here anything?

ps. interesting is that cameras have different width / height ratio - it's very visible in first seconds of you entrance.. on 54PRO-ZE your are "thinest", on 34PRO you are "widest"

ps2. does all cameras have latest firmware? and both 54PRO have the same firmware - they look too different???

Yes I agree it is hard to see the differences when they are just put in auto for everything. If I get some time I will try to tweek a little and change things up.

The 34PRO has a very cropped view compaired to the others. I do have the S3 and Pro ZE pointing more upward so I can sample the full zoom. I also agree the with the compression issues with moving objects.
I have played around with the different formats and bitrates and it is hard to pin down, at thimes I thing I have it figured out and then it comes back, the fixed ASE seems to be better for this.

Yes both 54PRO have the same FW at present.
I did get a new copy for the ZE but rolled it back when I was testing in the full dark area.

I also have up the 38 PRO in about the same location, I have just not spent any time with it, I will point it up and put this in the mix.

 

[duplo]

The new varifocal cams are harder to fine tune because they have manual iris control. Not only shutter/gain. The S3 has auto iris.

"p-iris"

 

[jmcu]

Here are some IR clips from the cameras in the full dark area.

A few notes:

It looks like there are spider webs on the IR for the 4M-T so not great IR - when I moved them, I seen it.

Look at the lens shadow on the top of the 54PRO, I thought it was because the cam was tilted up at a full 90° but even moving it down did not help.
I think this was an isolated issue with assembly on this cam because I have seen other samples and it isn't present.

Smart IR (S-3) or Illumination Overexposure Remover (PRO) is enabled

ps. Don't judge the crocks.

 

[steve1225]

Yes I agree it is hard to see the differences when they are just put in auto for everything. If I get some time I will try to tweek a little and change things up.

I would be very grateful for video comparisons with shutter times 1/120 and 1/60...

The 34PRO has a very cropped view compaired to the others. I do have the S3 and Pro ZE pointing more upward so I can sample the full zoom. I also agree the with the compression issues with moving objects.
I have played around with the different formats and bitrates and it is hard to pin down, at thimes I thing I have it figured out and then it comes back, the fixed ASE seems to be better for this.

there are 2 sources of those blurriness after movement:

• compression artifacts - increasing bandwidth, moving to AI.Coding or lowering Smooth stream to 20/30 should help...
• ghosting - effect of 3D noise reduction - decreasing 3D nosing reduction or increasing shutter time/gain/light output should help..

Yes both 54PRO have the same FW at present.
I did get a new copy for the ZE but rolled it back when I was testing in the full dark area.

I also have up the 38 PRO in about the same location, I have just not spent any time with it, I will point it up and put this in the mix.


View attachment 226704

Nice

 

[steve1225]

Here are some IR clips from the cameras in the full dark area.

A few notes:

It looks like there are spider webs on the IR for the 4M-T so not great IR - when I moved them, I seen it.

Look at the lens shadow on the top of the 54PRO, I thought it was because the cam was tilted up at a full 90° but even moving it down did not help.
I think this was an isolated issue with assembly on this cam because I have seen other samples and it isn't present.

Smart IR (S-3) or Illumination Overexposure Remover (PRO) is enabled

ps. Don't judge the crocks.

• T54PRO-ZE have much stronger IR than T54PRO-AS or T34PRO,
• T54IR-ZE have awful IR with big central hot spotting,
• T34PRO have awful default video processing for IR (you can try to decrease exposure compensation or play with Gamma settings to decrease background to black)..

 

[bigredfish]

Yeah T34PRO he said had spider web. It’s much better than that

 

[steve1225]

@jmcu thanks for video...
I propose to setup them ALL to the same Manual mode & custom shutter with starting parameters:

• shutter set to 0..10ms (or 0..8,3ms is you in the US),
• gain 0..50,
• both 2D and 3D noise reduction set to universal 0..30

One more thing after watching footage one more time - T54PRO-ZE have real P-Iris, so Iris settings should be at 100 to be comparable to fully open lens on fixed lens cameras..

 

[bigredfish]

The bullet version of the 5442-ZE does much better on the hotspot and the transition. Granted this is the 5442H-ZHE which I think is more like the 74XX series...but bullets always seem to do better with IR coverage, many because they have 2X more IR LED's

View attachment Home_ch2_20240923204827_20240923204909.mp4

 

[alekk]

Ditto what @steve1225 said that if you have time @jmcu, would be great to see some faster shutter speed footage to "push" the lens/cameras.

As he also pointed out, dial in some gain (which I don't know enough about the Dahua world)... but similar to camera testing, bump the (basically) ISO up some we can see if the image falls apart and/or excessive noise reduction is needed - might provide some insights into sensors/processing.

BTW, with the same shutter speed and gain (ISO), the cameras with larger apertures should be brighter (I saw @steve1225 talked about forcing it all the way open) ... but I didn't see that in earlier footage even though I recall there is about a full F-stop difference(?)

Again, really appreciate your excellent testing and sharing results.