Dahua 4431C-A vs Dahua 4231C-A vs CCTV-TOP IMX322

[Kawboy12R]

Original review has been edited because of a bitrate problem with the 4431C on the first evening. I've added the 10k bitrate video and labelled itand left the original 6k video as an example.

I did a little impromptu unscientific shootout this evening because I had a 6mm Dahua 4431C-A arrive today. I mounted it beside my current CritterCam ($40 3.6mm CCTV-TOP/BESDER Store on AliExpress IMX322/HI3516C eyeball dome) and a Dahua 4231C-A 3.6mm turret. The Dahuas are running stock auto settings and the no-name eyeball dome has had the slow shutter turned off with, IIRC, everything else pretty much stock auto. I did say this shootout was a bit unscientific. All three cameras are running and shining their IR for each other, plus there's a 940nm IR illuminator lighting up that suspicious looking fellow and the trees in the background. Other than that there's almost no white light in the scene. If you notice a difference in the crackling audio between the 4431 and the 4231 it's because the 4431 has the gain turned WAY up compared to the other. The no-name eyeball dome is getting its audio through BI cloning the 4231C-A audio. Both of the Dahuas are running h.264 decoding, NOT the default h.264h. Blue Iris doesn't seem to like the h.264h and it really causes the video to be blocky and amateurish. Changing the 4231C-A to h.264 turned it from a $25 low-light camera to how well it performed against the others.

I expected the 4431 to perform better than it did, especially with the added IR from the other cameras, the 6mm lens, and the IR flood. The $40 CCTV-TOP/BESDER eyeball dome is great for the price, but after discovering this evening how running the h.264h decoding for BI crippled the 4231 and h.264 improved it so much, the extra $30 for the 4231 and its audio and more mature firmware IMHO turns it into the night-time winner. I had liked the 4231's night-time performance before compared to the CCTV-TOP cam but it seemed to be suffering from too much compression and blockiness which put it in a distant 2nd. Vanilla h.264 changed that for the 4231 and is definitely the answer for BI. Two large thumbs up for the $70 4231 at night.

The pics and video labelled CritterCam are the CCTV-TOP/BESDER Store 3.6mm IMX322/HI3516C eyeball dome- https://www.aliexpress.com/item/IP-Camera-PoE-720P-960P-1080P-SONY-IMX322-Optional-Anti-Vandal-Dome-Camera-IP-Outdoor-Indoor/32580365934.html?spm=2114.13010608.0.0.F0zjaI

The other pics/videos have 4231 or 4431 in the video overlay.

I'm not sure yet where some pauses in the videos came from. These are the first videos I've recorded with the new BI 4.4.6.2 so it is possible that it introduced the stuttering. I've had no problems with earlier recordings so it is either that or something else new causing issues that don't normally occur.



CCTV-TOP- https://drive.google.com/file/d/0ByifJwXvTr39bHhJZ1dYZWF6RUU/view?usp=sharing

4431 crippled at 6k bitrate and full FPS - https://drive.google.com/file/d/0ByifJwXvTr39bzh6WVR6aFRtaDA/view?usp=sharing

4231- https://drive.google.com/file/d/0ByifJwXvTr39MXhVVDdoaHRnSVk/view?usp=sharing

4431 breathing freely at 10k bitrate and full FPS- https://drive.google.com/file/d/0ByifJwXvTr39VUI4ZEh1aVQ1cHc/view?usp=sharing

 

[Suburban28]

Really hope my 4431's perform better. The first still shot looks amazing.

 

[Kawboy12R]

I found an error in the 4431 this morning. I'm not sure if it came this way or if I accidentally changed the bitrate while checking the CBR settings but the bit rate was at around 6000 this morning. I knew something wasn't right when the daylight video was choked up and blocky as well. I thought it was just the low-light performance last night. I bumped up the bitrate to 10K and it made a world of difference in the daylight quality. I had left FPS at max and it looks like it really needs those extra bits in a complex scene with all the leaves and grass and stuff. Everything looked fine inside with flat walls and not much detail. It is either that or I unintentionally changed the setting after I did the indoor setup and peeked through the menus.

I will post a fairer pic and video of the 4431 this evening after it gets dark.

 

[Suburban28]

I found an error in the 4431 this morning. I'm not sure if it came this way or if I accidentally changed the bitrate while checking the CBR settings but the bit rate was at around 6000 this morning. I knew something wasn't right when the daylight video was choked up and blocky as well. I thought it was just the low-light performance last night. I bumped up the bitrate to 10K and it made a world of difference in the daylight quality. I had left FPS at max and it looks like it really needs those extra bits in a complex scene with all the leaves and grass and stuff. Everything looked fine inside with flat walls and not much detail. It is either that or I unintentionally changed the setting after I did the indoor setup and peeked through the menus.

I will post a fairer pic and video of the 4431 this evening after it gets dark.

Thanks. Look forward to seeing it.


Sent from my iPhone using Tapatalk

 

[Kawboy12R]

OK, the editing didn't allow me to add more pictures to the first post, so I'll add the standing snapshot at the 10k bitrate to this post, plus add three blown up pictures at roughly the same height to compare detail.





This comparison of detail probably gives the edge to the 4431C, but it easily should win because it's got a 6mm lens and has the advantage, albeit equally shared, of having a 940nm floodlight illuminating the far distance and 4mp vs 2mp of the other cams.

The below three pictures show the IR spread of the three cameras with each other source turned off. Basically, the two Dahuas have a wide even throw while the CCTV-TOP IMX322 has a spotlight in the middle instead of a more even floodlight. The 4431 shows more of an advantage over the 4231 in the distance on the trees but that's most likely because of the 6mm vs 3.6mm lens difference in addition to the 6mm being mounted and aimed a bit higher and altering the light pattern.

 

[nayr]

there we go, thats more inline with what Ive experienced.. thanks for the update.

yeah these 4MP cameras really want the bitrate cranked up.. if I hit the default button on the Video tab it fucking puts it at 4096Kb/s.. thats stupid.

 

[Kawboy12R]

Well, something "interesting" is going on with the Dahuas. Remember when I mentioned that h.264h on the 4231C apparently made the video blocky because changing it to h.264 improved it so much? And that changing the bitrate from 6k to 10k on the 4431C also improved the choked up blocky video? Well, it looks like the act of changing something improved the video, not what it was changed to. I left the 4231C plugged in overnight and the video stayed clear until after the day/night change and then went to blocky crap that was twitching about twice a second exactly after it went to day mode. It was set at 8k. After I got home tonight I changed the bitrate to 6k from 8k and the video cleared right up again. Sooo, it's not the act of improving something, it appears to be the SAVING of a change that fixes something that's causing the problem because I went backwards on the bitrate. Weird. I'm guessing it's got the same firmware glitch that the 4431C has that made the drastic improvement when I changed and saved the bitrate on it and it just a coincidence that feeding it more bitrate improved the picture. I guess that shows the danger of making assumptions just because it makes sense.



Anybody else running version 2.460.0000.4.R with build date 2016-06-21 and having video problems?

 

[nayr]

mine does not do that, but I am switching profiles at sunset/sunrise and not using the Normal profile at all.. mebe that might make a difference.

 

[Kawboy12R]

I plugged the 4431C back in and the video was a bit fuzzy. I went into the camera settings and changed the bitrate from 10k down to 8k and saw an immediate improvement. The change is at the 53 second point in the video.

edit- deleted the link because what I linked to was fuzzy and definitely not what I was seeing saved from BI even though it's a link to a crisp video from in my Google Drive. Anyway, going backwards in bitrate was definitely an improvement. I'll see what it does in the morning when it changes to daytime.

 

[nayr]

oh i forgot your using BluIris.. is it also fuzzy on the WebUI or an RTSP stream? mebe some buffers are filled up in BI

 

[Kawboy12R]

I thought of that and testing by viewing the web stream but both Dahua cams give me a webplugin.exe corrupt or invalid error in IE when I go to install them. Both cams are in the trusted zone as well. I haven't gone so far as to figure out how to load their RTSP stream into VLC yet. I didn't find much on invalid/corrupt webplugin issues on here but user @Billn seems to have had the same issue. I wonder if he found a way to fix it?

Also, both cams stayed crisp at dawn this morning. I flicked their bitrates to doublecheck and no change. Still good.

edit- and now that you mentioned filled buffers, I bumped up the receive buffer size in BI to 10MB from 3 for each cam just in case.

 

[Billn]

No, @Kawboy12R... I haven't found a way to fix the invalid/corrupt webplugin issue - but I haven't spent much effort. I just use BlueIris.

 

[nayr]

SmartPSS?

 

[Kawboy12R]

Haven't had much free time at home since I noticed the problem. Will definitely work something out to isolate the problem from BI if it reoccurs.

 

[GuiBou]

mine does not do that, but I am switching profiles at sunset/sunrise and not using the Normal profile at all.. mebe that might make a difference.

I have exactly the same thing on mine.

 

[Kawboy12R]

Both 4x31s are still behaving themselves so far. No reversion to crappy blocky video. And I think I figured out how the 4431 got changed to 6k at the beginning somehow. When set on CBR and changing the FPS from 20 to 10, the bitrate drops automatically to 6k from 10k. I think I did that but changed it back to 20fps to give a more fair comparison with the other cameras but didn't notice the automatic bitrate drop. It doesn't climb back up to 10k when changing from 10 to 20fps.

 

[nayr]

iirc @fenderman suggests everyone default to like 20MB buffers; mebe this is why.

 

[Kawboy12R]

It had a day's good run before I changed the buffer size so no proof it helped but it certainly didn't hurt. Ken asked me to bump up the buffer size to 10MB once when troubleshooting framerate and stuttering issues on my Axis cams so apparently it can help in some circumstances. I half wish the fuzzy blockiness would vome back so I could see if it exists outside of Blue Iris.

 

[mando209]

i have the 4431ca and i think night image looks excellent.it looks crystal clear.i will upload a pic later tonight.

 

[mando209]

here it is.i have some light to the left and right.probably wouldnt be good to compare to the one you took since you took it in pitch dark?.mine looks good though