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

OK, I am quite certain this latest 1028 firmware is a step forward, both in daytime and night images. Feels like some of the automatic processing has been pulled back, I am seeing more natural looking (sometimes a little fine noise, which is OK) but colour at night and daytime is looking better in my two test locations. Also, my preference is still to have NR and sharpening at well below 10 or zero/disabled in both cases, I will take a grainy image with some 'sparkles' if it means retaining detail and enabling ID.
 
Last edited:
  • Like
Reactions: JDreaming
What are the sharpness settings on those?

I mean I can make the original FW do same moving sharpness from 15 to 75

Assume all other settings NR and Exp are same?

Hard to follow this thread,
some guys love posting text like book chapters without posting any videos/pictures. so cannot read anything.

You notice it if you put a test chart like usaf1951 (US letter size print) in a range around 10ft/3m at a mounting height of 6-8ft/2-2.5m. Day. Turn down exposure compensate to zero to be able to read the letters on the chart.

Digital zoom the chart.
On all previous firmware the test chart looks like a 2MP cam. On 1028 one could say now its 4MP.
The picture is now sharper, like 5442. Better overall image to me.

It looks like the firmware has some nasty software manipulation aka wizmagic. One day picture looks Ok, other looks not good. Somehow related to what colors are in the picture. So its some kind of alghorithm to me. The behaviour is unpredictable.

They removed/reduced it now somehow, but now have bit more noise. Can be settings related.
For me its now much better, less blurr etc. But noisier than before if you turn down NR too much.
 
Last edited:
Hard to follow this thread,
some guys love posting text like book chapters without posting any videos/pictures. so cannot read anything.

Sometimes text like book chapters are just as good as videos and pictures, especially if they are filled with wisdom and experience :thumb: - try the suggestions, if they work well, add that person to the 'they know what they are talking about list' :lmao:

It looks like the firmware has some nasty software manipulation aka wizmagic. One day picture looks Ok, other looks not good. Somehow related to what colors are in the picture. So its some kind of alghorithm to me. The behaviour is unpredictable.

I would agree with this 100%, as the scene changes, so does the video output in many cases, this needs controlling.
 
Hard to follow this thread,
some guys love posting text like book chapters without posting any videos/pictures. so cannot read anything.

You notice it if you put a test chart like usaf1951 (US letter size print) in a range around 10ft/3m at a mounting height of 6-8ft/2-2.5m. Day. Turn down exposure compensate to zero to be able to read the letters on the chart.

Digital zoom the chart.
On all previous firmware the test chart looks like a 2MP cam. On 1028 one could say now its 4MP.
The picture is now sharper, like 5442. Better overall image to me.

It looks like the firmware has some nasty software manipulation aka wizmagic. One day picture looks Ok, other looks not good. Somehow related to what colors are in the picture. So its some kind of alghorithm to me. The behaviour is unpredictable.

They removed/reduced it now somehow, but now have bit more noise.
Can be settings related.
For me its now much better, less blurr etc. But noisier than before if you turn down NR too much.

I concur
 
  • Like
Reactions: JDreaming
The T54PROs came with firmware 7/31, which matches the firmware on Andy's website. The latest firmware on the Dahua website is 8/27. Andy's latest firmware is 10/28 with several updates prior.

What's the correlation between what Dahua releases and what Andy releases? Do they overlap/intersect at some point? Or are they all the same firmware with just later builds?

When Dahua adds a new feature or enhances or fixes something, how does it roll into Andy's firmware, as an example. Or again is it all just later builds of the same code path.
 
The 1028 is a beta fw based on the 0827, this will be use for the baseline, so if the 1028 have good progress on pics day and night, we will use it for the later firmware update, there will have another update soon. I will collect more information about this 1028 using feedback. The AI functions will also have another update too after we get the best pics.
 
Last edited:
So this morning, we have an image taken from a T54PRO-ZE using some brand-spanking new 1028 Firmware that @EMPIRETECANDY just sent me - thanks!

The upgrade was seamless (had to do by file rather than by folder ... but just did one at a time and it reboots after each) and below are the version number changes (BOLDED) from Maintenance Center->System Info->Device Info.

I don't see any difference at first glance (download of .dav files is still about 1.3MBYTES/sec even though the Firmware loaded at about 11 MBYTES/sec) but I'll be futzing around with it. The only "issue" is that with the "Skull Brothers" now in the picture, the auto-focus tends to snap to them (which is to be expected!) but simple enough to use Area Focus. Settings are 5msec shutter, 100 iris, 0 gain, 0 NR, everything else "default" but Sharpness push from 50 to 75.

I did this on my "lower" camera - the "upper" T54 is still running 0827 ... but I'm thinking I might push that to 1022 so I can more easily see if 1028 changes anything.

@EMPIRETECANDY : I didn't see any change/release notes - anything you can share that we should look for with this latest release?

That looks truly aweful. Not only are the skull brothers out of focus but so is the scarecrow 1/2 way down the lawn and other items in dark areas. It looks as if the iris might be wide open causing a very shallow dof and the camera is strugling with the contrast. I see contrast issues with the 4kt in my drive location but on mine it appears to heavily compress & pixelates the bright areas.
 
I believe so. It does have an obvious benefit for my AS fixed lens. I have two cameras out back that shine added IR into the yard and dock and the AS responds well to the added IR. basically extends my IR range and allows me to run a 0-3 - 0-4 ms exposure without much noise or blur

View attachment 231202

I'm guessing you like soft pictures. That looks perfect to me and exactly how sharp I'd like my pictures to be. The detail and sharpness are sublime.
 
Yea, I think it looks a bit HDR'ish because there's a LOT of dynamic range ... plus as I'm sure you know from photography @CamCrazy, sharpness is really tricky on foliage because it can "overprocess it"

Image below is with the sharpness turned down from 75 to the (default) 50.

View attachment 231204

HDR reduces the dynamic range in the final picture on CCTV. It uses the ability to see a wider range of contrast to recover more detail and lighten dark areas to improve the picture. It's not the same as watching a HDR TV where HDR increases the amount of visible dynamic range to deepen dark areas.
 
  • Like
Reactions: JDreaming
What do you think, guys?

Shutter : 8,33 , Gain : 48 NR : 25 25

Sharpness : 10

Looks pretty good, have you tried a walk test on it with IR, the cat walking away will work in the 8.33 shutters favour, just make sure when you cross the scene left to right that motion blur isn't a factor. If this is the case then shutter will have to be lower and gain pushed up a touch.
 
  • Like
Reactions: Chokolinho
I'm guessing you like soft pictures. That looks perfect to me and exactly how sharp I'd like my pictures to be. The detail and sharpness are sublime.

Actually I aim for a balanced video. Not too much noise nor blur..I’ll accept more noise than blur if it makes enough difference in being able to positively ID the target.
But I know I'm at the limits of either the camera or available light when I can't find an acceptable balance. The answer is almost always, add light.

I find much of the tweaking of a static snapshot without a moving target to be mental masturbation. Sure on first initial dial in I strive to get a clean image, but without a (human, vehicle, dog, gator, giraffe ) moving target in the various lighting conditions, one doesn't really know.

I tend to spend wayyyy more time on nighttime scenes than day. Frankly any decent sub $100 camera can get reasonably good pics during the day, I want to be able to see that there are separate blades of grass at 25ft, but I don't need to be able to count them. But at night, and in the hardest close-in scenes where facial ID is critical, there are usually multiple light conditions even within 0-15ft of the camera and I find it much more difficult to get a great snapshot of the bad guy testing my door locks.

On the other hand some folks live in very beautiful remote settings like some of our our Alaska members for instance, where few if any bad guys come calling. There, still shots of the breathtaking scenery or aurora's are essential and many have cameras dedicated to doing just that. I envy them.

This car snap has a small amount of blur, its not perfect. But its more acceptable to adding noise from a faster shutter. I don't see any gain to be had that will make a difference to LE in ID of the vehicle.

Home_Color4K-X_main_20241014211732_@5.jpg
 
Actually I aim for a balanced video. Not too much noise nor blur..I’ll accept more noise than blur if it makes enough difference in being able to positively ID the target.
But I know I'm at the limits of either the camera or available light when I can't find an acceptable balance. The answer is almost always, add light.

I find much of the tweaking of a static snapshot without a moving target to be mental masturbation. Sure on first initial dial in I strive to get a clean image, but without a (human, vehicle, dog, gator, giraffe ) moving target in the various lighting conditions, one doesn't really know.

I tend to spend wayyyy more time on nighttime scenes than day. Frankly any decent sub $100 camera can get reasonably good pics during the day, I want to be able to see that there are separate blades of grass at 25ft, but I don't need to be able to count them. But at night, and in the hardest close-in scenes where facial ID is critical, there are usually multiple light conditions even within 0-15ft of the camera and I find it much more difficult to get a great snapshot of the bad guy testing my door locks.

On the other hand some folks live in very beautiful remote settings like some of our our Alaska members for instance, where few if any bad guys come calling. There, still shots of the breathtaking scenery or aurora's are essential and many have cameras dedicated to doing just that. I envy them.

This car snap has a small amount of blur, its not perfect. But its more acceptable to adding noise from a faster shutter. I don't see any gain to be had that will make a difference to LE in ID of the vehicle.

View attachment 231397
You make many great points. A person needs to ask himself, “Do I want a pretty picture or a useful picture?” I’ll opt for the useful picture every time. Sure, it would be nice to have everything perfect, but I’ll take the occasional blown out highlight if it means the subject is properly exposed in the image, rather than hidden in a shadow. Your meticulous setup really shows in your videos. A lot of thought and planning is evident.
 
Actually I aim for a balanced video. Not too much noise nor blur..I’ll accept more noise than blur if it makes enough difference in being able to positively ID the target.
But I know I'm at the limits of either the camera or available light when I can't find an acceptable balance. The answer is almost always, add light.

I find much of the tweaking of a static snapshot without a moving target to be mental masturbation. Sure on first initial dial in I strive to get a clean image, but without a (human, vehicle, dog, gator, giraffe ) moving target in the various lighting conditions, one doesn't really know.

I tend to spend wayyyy more time on nighttime scenes than day. Frankly any decent sub $100 camera can get reasonably good pics during the day, I want to be able to see that there are separate blades of grass at 25ft, but I don't need to be able to count them. But at night, and in the hardest close-in scenes where facial ID is critical, there are usually multiple light conditions even within 0-15ft of the camera and I find it much more difficult to get a great snapshot of the bad guy testing my door locks.

On the other hand some folks live in very beautiful remote settings like some of our our Alaska members for instance, where few if any bad guys come calling. There, still shots of the breathtaking scenery or aurora's are essential and many have cameras dedicated to doing just that. I envy them.

This car snap has a small amount of blur, its not perfect. But its more acceptable to adding noise from a faster shutter. I don't see any gain to be had that will make a difference to LE in ID of the vehicle.

View attachment 231397

Not much more needs to be said :goodpost:
 
Thanks. I've now used up all of my daily allotment of wisdom and will generally be my usual cranky old asshole self the remainder of the day :rofl:
 
Last edited:
  • Love
Reactions: JDreaming
Woeful lack of video from me, but here are some stills from various cameras looking at the same vehicle, identical time of day just simply different angles. Useful to compare with the 54ProZE (1028 firmware), overcast conditions, all 100%, no editing, vehicle moving around 32-35mph :


54ProZE (50% optical zoom - 1182) 16 feet...

54Pro-Fedex.jpg



4K-T 2.8mm 20 feet...

4K-T-Fedex.jpg



2MP Z12E (75% optical zoom - 980) 70 feet...

2MP-Z12-Fedex.jpg


2MP Starlight (30% optical zoom - 460) 14 feet...

2mp-Starlight-Fedex.jpg
 
Last edited:
That looks truly aweful. Not only are the skull brothers out of focus but so is the scarecrow 1/2 way down the lawn and other items in dark areas. It looks as if the iris might be wide open causing a very shallow dof and the camera is strugling with the contrast. I see contrast issues with the 4kt in my drive location but on mine it appears to heavily compress & pixelates the bright areas.

It's not surprising that the Skull Brothers are out of focus.

I wrote earlier that I'm running 100 iris to minimize shutter speed and gain ... and since this is a "scene" shot with the camera at fairly wide-angle with the middle of the driveway being ~50' away, most everything should be (reasonably) within the depth-of-field even wide-open. Although I just added something that is about 4' away which is a little fuzzy ... ;-)

The only "issue" is that with the "Skull Brothers" now in the picture, the auto-focus tends to snap to them (which is to be expected!) but simple enough to use Area Focus.


They'll come down after Halloween and then the picture will be cleaner.

Yea, strong sunlight and shadows can be tough - here's a snapshot from a week ago without those - reasonably decent?
BTW, if I view images "directly" on the forum, things look fuzzier because it's rescaling for whatever size I have the window.

192.168.1.172_ch_1_20251024_074506.png