This has to be worth a try, mine are arriving tomorrow potentially, I will do a sort of mini review of initial findings so as not to clog up this thread maybe?! I guess NR, Gain and Gamma have to be worth a shot.
Less sharpness in T54PRO-ZE (versus 5442-S3) possibly due to noise reduction being applied even if set to ZERO?
- Thread starter alekk
- Start date
This has to be worth a try, mine are arriving tomorrow potentially, I will do a sort of mini review of initial findings so as not to clog up this thread maybe?! I guess NR, Gain and Gamma have to be worth a shot.
Per a suggestion from Andy, I tried rolling forward to the 0928 Firmware that he had temporarily released (recall this "reset" the gain scale so that 0-100 was 0-62 on the 0731 firmware) and the NR is still happening (easily seen on my street/driveway) - again, gain=ZERO and 2D/3D NR is set to zero. Daytime with shutter speed set to 1/100s (and same exposure at 1/200s) so there is plenty of light.
I ALSO tried using the 0827 firmware direct from Dahua ... but this also made no difference. The "gain scale" is easiest to check at night, so I'll do that to see if it is "0-100" or "0-62"
WRT Gamma: I tried this on all three firmware's and it made no difference in the NR. Setting it to a lower value results in a darker image ... and "35-40'" on the T54PRO-ZE was semi-close to what "50" shows on the 5442-S3 ... during the day - TBD at night.
WRT Sharpening: This does make a difference ... but looks like it is sharpening an image that has also been NR'd ... so if something has been "smeared" out (even with NR=0), there's ain't much to sharpen there!
WRT Iris: I did try changing from 100 to 50 ... but that made no difference on the NR issue. FYI that I did confirm it's "working" as (in manual mode), I set the shutter to 0.01msec (1/10000s shutter) and at IRIS=50, I got a dimmed image ... and then brightened up when I set IRIS=100.
I ALSO tried using the 0827 firmware direct from Dahua ... but this also made no difference. The "gain scale" is easiest to check at night, so I'll do that to see if it is "0-100" or "0-62"
WRT Gamma: I tried this on all three firmware's and it made no difference in the NR. Setting it to a lower value results in a darker image ... and "35-40'" on the T54PRO-ZE was semi-close to what "50" shows on the 5442-S3 ... during the day - TBD at night.
WRT Sharpening: This does make a difference ... but looks like it is sharpening an image that has also been NR'd ... so if something has been "smeared" out (even with NR=0), there's ain't much to sharpen there!
WRT Iris: I did try changing from 100 to 50 ... but that made no difference on the NR issue. FYI that I did confirm it's "working" as (in manual mode), I set the shutter to 0.01msec (1/10000s shutter) and at IRIS=50, I got a dimmed image ... and then brightened up when I set IRIS=100.
Last edited:
Appreciate the feedback and update, sounds like waiting for another firmware update at some point. Should I stick with 0731 at this point when mine arrives?
bigredfish
Known around here
Thanks, that all makes sense unfortunately.
Saved me some time when mine arrive, appreciate your efforts
Time to play with codec options?
I've tried various codec combinations - doesn't seem to change it.
I have a 2nd T54PRO-ZE coming (hopefully before the wife gets back!) and yea, I'll probably stick with 0731 (Andy has removed the 0928 firmware) unless someone shows something significant worth upgrading too.
To be clear, I'm pretty happy with the improvement in low-light of the T54PRO-ZE over the 5442-S3 and have already replaced that as the "production" camera.
But being a "pixel-peeper", there's a couple of things that I think can be improved upon ... and/or maybe I'm just not smart enough yet on how to fix 'em myself! ;-)
I have a 2nd T54PRO-ZE coming (hopefully before the wife gets back!) and yea, I'll probably stick with 0731 (Andy has removed the 0928 firmware) unless someone shows something significant worth upgrading too.
To be clear, I'm pretty happy with the improvement in low-light of the T54PRO-ZE over the 5442-S3 and have already replaced that as the "production" camera.
But being a "pixel-peeper", there's a couple of things that I think can be improved upon ... and/or maybe I'm just not smart enough yet on how to fix 'em myself! ;-)
bigredfish
Known around here
Based on this video from @Chokolinho I may be able to live with the softness during the day.
But night using its own onboard LEDs will decide whether I swap out my 5442's for it
But night using its own onboard LEDs will decide whether I swap out my 5442's for it
Last edited:
@Chokolinho : I'm looking at those Large "vertical" concrete rectangles that border the road ... plus the inside of the circular manhole cover.
Does the "dimpling" as shown in the video match how it appears in real-life?
This is the type of stuff that I'm seeing get "smeared" (even with NR=0) ... albeit mine is further away ... and it does look like it might be happening on the "smaller" vertical rectangle plus the asphalt in the distance ... although realistically, it might not be reasonable to expect that to be resolved.
I've lost track - did you have a 5442-S3 here ... and if so, how that that image compare in those areas?
THANKS!
P.S. There's a LOT of foliage in that picture ... very difficult challenge for any NR/sharpening algorithm!
Does the "dimpling" as shown in the video match how it appears in real-life?
This is the type of stuff that I'm seeing get "smeared" (even with NR=0) ... albeit mine is further away ... and it does look like it might be happening on the "smaller" vertical rectangle plus the asphalt in the distance ... although realistically, it might not be reasonable to expect that to be resolved.
I've lost track - did you have a 5442-S3 here ... and if so, how that that image compare in those areas?
THANKS!
P.S. There's a LOT of foliage in that picture ... very difficult challenge for any NR/sharpening algorithm!
I was thinking this also in relation to your comments re the driveway and road in your samples @alekk and to my eyes I am sure there has to be a loss of surface detail/smearing on the paving, brickwork and vertical slabs. This would potentially translate elsewhere but maybe more noticeable in those areas. I sincerely hope if this is the case that it can be turned around by Dahua and is not a hardware limitation as such. On a side note I would say focus point on this image above is the drain pipe/guttering end.
Planning to try and get a 54Pro and S3 turret installed in the next 24 hours for feedback, 4K-T to follow probably the weekend or early next week.
Looking forward to seeing your observations. I assume you mean a 54PRO-ZE & S3 ... plus is that a 4K-T or a T54PRO-AS (challenge keeping up with product names!)
More Pixel Peeping that it's hard for me to determine where the exact focus point is, but yea, could be the guttering ... but even if the Iris is full open, with a wide-angle lens, it's not hard to be hyperfocal ("good" to infinity) from fairly close up, especially with (relatively) small sensor like 1/1.8". This wood slats in the top-right (by the timestamp) look decent to me.
Not sure what to make about the bricks on the house - note the top 3 levels of bricks appear a bit darker (I assume more shadowed by the roof) and seem to have a loss of detail, especially once you get past the door. I wonder if that is some NR being applied.
Last edited:
54Pro ZE, 5442 S3 and 4K-T 8MP which is coming later this week. The 54Pro and S3 are installed as of the last few hours, setting up now and will give some feedback later today (evening now) which will be based on night image quality. Got my head around the new UI now...just about, looking as expected initially, coming from 2MP starlight for these replacements so a jump regardless. But I will benchmark them against the existing 5442 original series I have.