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

[Sixto]

... For most situations auto mode should work best... day/night/infrared/white led ...

yeah. was just playing with manual and it's fine during the day but really auto looks excellent for me ... so just switched back to auto for now.

We had a delivery last night and the delivery guy was crystal clear at night ... will continue to play, but I may just stick with auto in the short term.

This new T54PRO is very close to the front door so it's excellent at covering the front stoop ... and so much better in low light then the other front camera (5442 non S3) but that's much further away from the front door where the lighting is. The 5442 will get replaced with one of the next two T54PRO's that are now on order ...

 

[flynreelow]

yeah. was just playing with manual and it's fine during the day but really auto looks excellent for me ... so just switched back to auto for now.

We had a delivery last night and the delivery guy was crystal clear at night ... will continue to play, but I may just stick with auto in the short term.

This new T54PRO is very close to the front door so it's excellent at covering the front stoop ... and so much better in low light then the other front camera (5442 non S3) but that's much further away from the front door where the lighting is. The 5442 will get replaced with one of the next two T54PRO's that are now on order ...

post that video of my UPS man.

 

[alekk]

Does anyone else change the "main stream" resolution (Camera->Encode->Resolution) from native 2688x1520 to another resolution such as 2560x1440 or 1920x1080 ... and seeing some blurriness?

I set this to 1920x1080 on my 5442-S3 ... so did the same on the T54PRO-ZE ... but the images from the later (pulled via ffmpeg ... which is what is often used under the covers for NVR's, etc.) are noticeably blurrier and there appears to be some goofy jpeg compression going on ... which may explain some of my struggles last night.

I also see this if I simply use the web browser and do Live View. Same for 2560x1440.

HOWEVER, if I do 2688x1520, there's a noticeable improvement ... and just to confirm this, I use Snapshot to grab a 2688x1520 image, resized it in Photoshop to 1920x1080 ... and compared that to a Snapshot at 1920x1080 ... which was much worse on the T54PRO-ZE.

I did the same with my 5442-S3 (basically same scene) and the difference (doing a resize via Photoshop versus letting the camera do it) is minor ... but not near as much as the T54PRO-ZE.


So my theory (for now!) is that the internal camera resizing algorithm of the T54PRO-ZE may be different (and noticeably worse) than then 5442-S3. Or I could be doing something stupid! ;-)
And yes, the encode parameters were the same in all tests.

 

[wittaj]

It is best to use the native resolution. As you have seen, it can wreak havoc on downrezing.

 

[alekk]

It is best to use the native resolution. As you have seen, it can wreak havoc on downrezing.

Roger that ... I'm just pointing out that it appears based on my initial impression/testing, that "more havoc" is being done by the T54PRO-ZE than was done on the 5442-S3.

 

[steve1225]

Does anyone else change the "main stream" resolution (Camera->Encode->Resolution) from native 2688x1520 to another resolution such as 2560x1440 or 1920x1080 ... and seeing some blurriness?

I set this to 1920x1080 on my 5442-S3 ... so did the same on the T54PRO-ZE ... but the images from the later (pulled via ffmpeg ... which is what is often used under the covers for NVR's, etc.) are noticeably blurrier and there appears to be some goofy jpeg compression going on ... which may explain some of my struggles last night.

I also see this if I simply use the web browser and do Live View. Same for 2560x1440.

HOWEVER, if I do 2688x1520, there's a noticeable improvement ... and just to confirm this, I use Snapshot to grab a 2688x1520 image, resized it in Photoshop to 1920x1080 ... and compared that to a Snapshot at 1920x1080 ... which was much worse on the T54PRO-ZE.

I did the same with my 5442-S3 (basically same scene) and the difference (doing a resize via Photoshop versus letting the camera do it) is minor ... but not near as much as the T54PRO-ZE.


So my theory (for now!) is that the internal camera resizing algorithm of the T54PRO-ZE may be different (and noticeably worse) than then 5442-S3. Or I could be doing something stupid! ;-)
And yes, the encode parameters were the same in all tests.

All WizColor / AI-ISP cams have TOTALLY different video workflow comparing to older Dahua cameras.. Many settings works very different... The same was on TIOC-PRO (34PRO)...

You can try to enable full WDR (with values >= 50) - it reverse some (not all) original Dahua video processing - to test different resolutions on main channel.

 

[alekk]

As always, I appreciate your expert commentary @steve1225 and I've seen you mention that, but I didn't realize that (assuming my observations are correct) this would effect simple resizing.

I'm planning to work around it by simply have my ffmpeg script pull the native resolution and then do the resize myself. Which arguably I should have been doing in the first place, but it was handy to have the camera do it, plus saved me some bandwidth.

P.S. I'm also seeing some difference in White Balance and 5442-S3 appears to do a bit more color saturation than the T54PRO-ZE.

 

[steve1225]

As always, I appreciate your expert commentary @steve1225 and I've seen you mention that, but I didn't realize that (assuming my observations are correct) this would effect simple resizing.

I'm planning to work around it by simply have my ffmpeg script pull the native resolution and then do the resize myself. Which arguably I should have been doing in the first place, but it was handy to have the camera do it, plus saved me some bandwidth.

P.S. I'm also seeing some difference in White Balance and 5442-S3 appears to do a bit more color saturation than the T54PRO-ZE.

why you want to resize? you should use native resolution...

you can try also check if third stream (which can be resized) have the same problem?

 

[alekk]

Sorry to be a bit "post-happy" with my "new toy", but thought people might like seeing the 5442-S3 and T54-PRO-ZE side-by-side deployed in the field.

As others have pointed out, it's bigger (ditto the new PFB2203 Wall Bracket) but unless you have a really tight spot, should be a drop-in.

The wiring harness is a bit bulky (I put electrical tape on the ends of everything) ... but it all "fits" in there ... although I'd recommend getting the camera pointed in "about" the right direction before you stuff it all in there. I didn't do this and wasn't able to tilt the camera enough (cable was probably preventing) so had to remount it ... albeit that's super-easy to do (and I'll have to do again when I paint things) and it's no problem doing it by yourself while up on a ladder ... nice work Dahua





BTW, there was a 7 year old Vivotek camera in that location before that ... so figured it was worth showing the "crud" that got in the back-side of the camera (there was also some inside I didn't get a picture of) from all that time. I think most of this was brought in by insects since the location is fairly sheltered. I'm sure others have seen worse in the bazillions of installations they have done.