SMD 3.0 dahua

Is the detection for humans with SMD improved in this version?
Yes, just get the firmware , will send to some guys do some testing now. Can contact me directly if want to test the smart Ir one
 
All

I’ve been testing the latest release of the SM 3.0 FW (21-05-06). For clarity, this release by Dahua was primarily focused on trying to improve the SmartIR function from the 21-04-03 version. Unfortunately, in my test, (which is tested on cams with bare minimum config to ensure cleanest possible testing) SmartIR is still an issue. In this video below I walk through both fixed and ranged shutter speeds, compare against the 11/23 FW (known good that we tweaked for SMD 2.0 and SmartIR) and show ways that this current FW still needs improving in areas of IR strength, exposure compensation (actual adjustments needing to be implemented and still not showing in code), response time to target and some key differences I see between 11/23 and 05/06.

I’ve fed this back to Dahua and @EMPIRETECANDY and asked Dahua to focus on using 11/23 as the baseline for SmartIR in this release. Thanks Andy, for also following up with the dev team. My understanding is that Dahua is already looking at this and working to correct.

As always, I've included timestamps in the YouTube video for easy reference and for those that want to jump and compare configs I cover here.

 
other than the SmartIR issue... how is the new FW for HUMAN / ANIMAL AI?
 
Maybe it's just me, but if Dahua was trying to fix the smart IR I have to wonder if they ever actually tested it prior to releasing it.
 
I think only us on this forum care about video quality LOL

Look at all the consumer grade cameras that exhibit this and worse and people just accept that is a limitation...

And most of the great firmware updates that Andy gets from them they never push it out to the main customers!
 
This is a tough crowd, Andy :rofl:
 
All, I was asked by Dahua to provide a further walkthrough on SmartIR testing on the 21-05-06 FW to help them pinpoint the issue. Therefore I re-ran the tests, included the menu settings, OBS screen recordings, my recommendations on how to fix with a demonstration on how it improves the image and an updated comparison to 11/23.

While 11/23 wasn’t perfect, we did get it tweaked by Dahua to pretty good after the testing / feedback here and with the small additional tweaks I showed at the time worked very well. For clarity, in the How To Fix section, I’m not saying that Dahua should be dropping Exp. Comp all the way down to what I show in the video but I use it as a guide for them on where they could start focusing to fix Exp. Comp as well as IR strength and response of these to changing target / lighting in scene. This would be a great place to start as I showed in the last video where the differences between FW in these values is apparent. Including the new video link here so you have access too.



 
got a ton of these to test on > IPC-HFW5241E-Z12E's< Sign me up :)
 
  • Like
Reactions: EMPIRETECANDY