Is a 4K rear dash camera worth it?

In 2029 I expect dashcams to have AI enhancement of license plates with a searchable list of plates that were in your videos, and 12K.
 
In 2029 I expect dashcams to have AI enhancement of license plates with a searchable list of plates that were in your videos, and 12K.
12K?? I think it just makes more sense to be able to have license plate readers like the cops have.
 
Here in my home state of Pennsylvania, front plates are not required and not provided.
License plate capture is not a deal-breaker for me. Even more so given that our USA
plates are much smaller than most others across the globe.

I still want to capture as much as I can, within limits. The power draw and heat created
by greater processing and added features will probably always be part of the consideration.

The need for license plate capture in the U.S. is overstated. There are a handful of technologists and enthusiasts at DCT who think license plate capture is the be-all and end-all of dashcam technology. A good bet is that most dashcam users in the U.S. give license plate capture little thought or concern. Users are generally more interested in capturing the bigger picture of what is going on around them.

The Vueroid S1-4K utilizes an AI system that involves importing the video into an editor, selecting the plate area, and then sending that selected area to an AI processing hub. It works reasonably well, but has its limitations, as can be expected. But it is a good start and a good direction for dashcam users who want an edge on license plate captures.

As for ALPR, Automatic License Plate Reader. The systems used by law enforcement agencies are not cheap. Depending on quality, cams start around $800 and can cost thousands of dollars. Then there is the back-office system, which is not cheap. The back office AI processing is often a subcontracted agreement with a provider. Those readers capture an image and import it into a system that AI processes, then searches a database. Even if the database search part is removed from the consumer market, the overall quality of the images and AI processing by LEO ALPR systems is out of the reach of the average consumer.

I like the idea of capturing license plates, if only for that rare occasion when a user will need that information. Mostly, users do not need that information, and license plate capture is the Holy Grail that dashcam manufacturers are chasing after. I prefer to have an excellent all-around image, both front and back, that captures quality details of events. Given my options, I would prefer a telephoto lens investment over license plate capture.
 
Last edited:
I just use plate capture as a test criteria, personally i have never needed a plate capture, hell i have not even heard of any family / friend / acquaintance of me being the victim of a hit and run,,,,,, vandalisen sure, but no in traffic stuff.
People should not be too focused on this.

Though to be honest i would personally like if i was able to carry a 9 mm under my armpit of otherwise concealed on me,,,,, just in case, Sadly all Danes are ordered to take the perpetual probable or actual victim role.

My flat rate car wash have plate reader cameras that to a degree look like those on the roof of some police cars here, so i just pull up and select NO i do not want a lesser wash than the premium i pay for, and then roll in.
 
Onboard AI chips that can read a license plate even if only 100 pixels captured it
 
For me, and maybe for many as well, license plate clarity is most useful as a part of
the decision-making process before purchasing. It is a standard image that is very easy to find
when trying to figure out how best to spend one's money, as license plates are everywhere.
When searching YT videos, Project Farm uses all sorts of images, street signs, gas station
signs, etc. Not things one needs for evidence but really handy when comparing/shopping.
It was his video that lead me to Viofo.

I think I'll be happy if I can capture anything useful when parked. My vehicle has visible evidence
of being hit more than 6 times while parked, but never once involved in a moving collision.

It is a race to the top, and right now there seem to be plenty of viable contenders.
 
I think I'll be happy if I can capture anything useful when parked. My vehicle has visible evidence
of being hit more than 6 times while parked, but never once involved in a moving collision.
Did you get captures that were not useful when parked? I am always concerned with the field of view when parked mostly for side impacts.
 
Did you get captures that were not useful when parked? I am always concerned with the field of view when parked mostly for side impacts.

I should have mentioned that. I bought/installed my first dash cam 8 weeks ago.
This is probably like buying an expensive snowblower then watching it not snow
for 5 years!

In the week between receiving the cam, and installing it, I had 3 close-calls. Since installing
it - no issues, no close-calls, no more dings. šŸ™„

Yet. šŸ˜Ž
 
from my own experience with the T800 and Vantrue S1Pro MAX all i can say is i have to "cheat" to capture plates at night, meaning run with dipped high beams, otherwise with regular running lights my car do not put enough light into the other lane far enough out ahead.
But otherwise these 2 cameras can indeed capture plates at night, but outside factors can ruin that as i have experienced.

It is pretty wild you can do stuff like that today, at least thinking back to the days when 1080 p became main stream.

Daytime i for sure prefer to run any system vanilla, so no HDR or other fancy stuff and so on for me at that time of the day,,,,, if i can do that, that is.

PS: Dunno if it is me, but i no longer feel for pressing the event button and share idiots in traffic, but i doubt Danes have become better more considerate drivers.

Me. in the last 2 days i have been told in 2 different stores that i am a very favorite customer,,,,, but i also do my very best to be nice even if i do have grievances with pretty much all of my fellow countrymen.

For instance if i come to the checkout and there is a huge stack of shopping baskets ( yes Danes often do not need a full blown trolley to shop )
Well i then take that stack with me thru checkout, CUZ i am heading for the door where these are placed anyway, so for me, something like that is only natural, anyone else for dams sure think " that’s their job they get paid for " or even worse let the state steal money for it and then assume the care giver / protectors role, which it fail again and again.
 
Last edited:
PS: Dunno if it is me, but i no longer feel for pressing the event button and share idiots in traffic, but i doubt Danes have become better more considerate drivers.

I would like it to be easier to save videos and have them downloaded to my phone in the background, so I can share to YouTube or socials. I sometimes lock funny things I see, but turning on wifi, navigating various apps, and staying in my car while they download, I just don't bother. I think any dashcam company that adds something like that, might get more exposure as they see people sharing from their dashcams.
 
I just use plate capture as a test criteria
license plate clarity is most useful as a part of
the decision-making process before purchasing
street signs, gas station
signs, etc
I agree with everyone capturing license plates is unnecessary.
However, what else are we suppose to use to verify HDR effectiveness.
They are the perfect mobile eye chart.

Better or Worse .webp


When we first got HDR with STARIVIS 2 (IMX678) A139 Pro (November 2022) I was using illuminated road signs to show / compare the difference with other cameras.
Also to verify if HDR firmware updates made things better, or worse.
Then I got ā€œa talking toā€ by Mtz, and Nigel to go out, and capture moving license plates.
I’ll never forget what Bill said in an email;
ā€œChuck, please go out, and drive your carā€
Of course Mtz, Nigel, and Bill are correct in that we need BOTH the camera, and subject moving.
But I still like doing stationary comparisons before I go drive my car.


No, and neither is a front 4K for the vast majority of use cases.
This is the most WEE-TODD-ED thing I’ve ever seen. lol
This is like telling;
1.) Galileo to stop making better telescopes.
2.) Thomas Crapper: ā€œNo Thomas, let’s keep using the outhouseā€
3.) Thomas Edison: ā€œNo Thomas, let’s keep using candlesā€
4.) Henry Ford: ā€œNo Henry, let’s just using horsesā€
5.) technology to stop because we have enough.

This kind of thinking will have us living back in the 80’s.
Actually that might not be so bad. lol
I want my MTV!

 
I agree with everyone capturing license plates is unnecessary.
However, what else are we suppose to use to verify HDR effectiveness.
That depends whether we want HDR to improve plate captures, or to enhance the overall scene.

If we want HDR to record a high dynamic range scene, then we should test that.

As Bill said, go drive the car. On a bright sunny day, through some dark shadows, or at dusk and dawn with a low sun, or in & out of a tunnel or underground car park.
 
I think the "Do we need to capture licence plates" discussion deserves its own thread because it influences the decisions manufacturers take regarding hardware and firmware optimisation, it may affect the process or outcome of law enforcement or insurance matters, and for some of us it's also an interesting discussion topic. But it's off-topic for the subject of this thread about 4K rear cameras despite there being some obvious overlap.
 
Back on-topic, I think it's important to recognise that the front and rear cameras of a 2CH dashcam setup have different challenges, so what works best for the front may not be ideal for the rear.

Front:
- Built-in to main unit so high bitrate is possible
- Scene ahead at night is illuminated by the car's headlights
- Need to record cars moving in the same direction, and cars oncoming at high closing speed

Rear:
- Remote from main unit, more challenging to maintain a high bitrate over a long cable
- Scene behind at night is poorly lit by car's brake lights, and often dazzled by headlights of following cars
- Mostly need to record cars travelling in the same direction, or with lower closing speed
- Rear screens are more often tinted than front screens

For the front camera I think 4K is the ideal solution - the IMX678 is favoured by most manufacturers at the moment.
Personally I would prioritise excellent HDR and low-light capability over high resolution on the rear camera.
 
Personally I would prioritise excellent HDR and low-light capability over high resolution on the rear camera.
Yes!
If only we could have all 3. lol
 
That depends whether we want HDR to improve plate captures, or to enhance the overall scene.

If we want HDR to record a high dynamic range scene, then we should test that.

As Bill said, go drive the car. On a bright sunny day, through some dark shadows, or at dusk and dawn with a low sun, or in & out of a tunnel or underground car park.

Better HDR (software), or better optics and processor (hardware)?

It appears that the limitations of the current Novatek processors used with the Starvis II and optical lenses have reached a plateau. Obtaining perfect clarity at the position where license plates are found often causes clarity and detail issues in other parts of the image. If we go for the cinematic experience, then that plate clarity is diminished.

I think this is where we find the advantage of using a dual lens arrangement in a single main chassis: the standard lens and a mildly telephoto lens, whose purpose is to capture plates. I don't believe there is a single lens solution to all the 'wants' that users have. However, to make this dual lens arrangement happen, the user would need to give up the interior and rear channel cameras due to bandwidth, bitrate, and processing.

At the price range we are currently willing to pay, there are trade-offs.
 
Last edited:
Back on-topic, I think it's important to recognise that the front and rear cameras of a 2CH dashcam setup have different challenges, so what works best for the front may not be ideal for the rear.
.
.
.
Rear:
- Remote from main unit, more challenging to maintain a high bitrate over a long cable
- Scene behind at night is poorly lit by car's brake lights, and often dazzled by headlights of following cars
- Mostly need to record cars travelling in the same direction, or with lower closing speed
- Rear screens are more often tinted than front screens
.
.

That is why I do not use a rear camera as a second or third channel, but put a front camera on the back glass. Using a front main camera on the back glass overcomes all the negative aspects of using a second or third channel on the rear glass. Of course, this comes at a higher cost, including some inconvenience when accessing the videos, and I have to extend the wiring harness to the fuse box.

There are always trade-offs, I am afraid.
 
Better HDR (software), or better optics and processor (hardware)?

It appears that the limitations of the current Novatek processors used with the Starvis II and optical lenses have reached a plateau. Obtaining perfect clarity at the position where license plates are found often causes clarity and detail issues in other parts of the image. If we go for the cinematic experience, then that plate clarity is diminished.

I think this is where we find the advantage of using a dual lens arrangement in a single main chassis: the standard lens and a mildly telephoto lens, whose purpose is to capture plates. I don't believe there is a single lens solution to all the 'wants' that users have. However, to make this dual lens arrangement happen, the user would need to give up the interior and rear channel cameras due to bandwidth, bitrate, and processing.

At the price range we are currently willing to pay, there are trade-offs.

Even though I don't own, the Viofo A329T appears to rememdy a complaint of many. It includes a telescopic lens and a normal view, bettering one's chances of capturing plates. I am not sure building into a single chasis is ideal in all circumstances. A separate camera affords the consumer the ability to place the Telescopic Lens in a different location, should it be necessary, reduces the size of the front unit, and generates less heat than an "all in one" design.
 
One thing i did learn playing with the zoom feature, on Mobius and Joovuu cameras was, at least when using extreme 12 mm lens like i did most often.
You need a high degree of throw to make damn sure it is aimed where you like it to be, CUZ 10 degrees off and you arte not at all where you like to be.

Also as this was on a suction cup mount with a ball joint, having had the camera off to look at footage, i needed to realign it every time.

Anyway the viofo zoom cameras appear to be pretty regular remote housings, and so i assume great care must be taken installing it, cuz after that if any all you have is up / down adjustment.

My zoom cameras was on the very R side of my windscreen, actually just out of wiper coverage, so needed to aim them good

Technology and progress is just fine, but i dont embrace it all, and for sure not right away. I do not own a tin foil hat even many seem to think i have a collection.
There is also non technology things i will have to bypass, sadly participating in the democratic process is one of them.

My current dilemma. my MI 9 is acting strange, and i am afraid i will have to buy a new phone, even if i hoped this would be my last phone.
But mom still alive, but the damn phone are doing all kinds of crazy when i get a call, so it is actually problematic to pick up the phone when someone call.
Sure am happy i dont use that damn thing for many seconds every day, and it dont even ring every day, just like i am glad it is now several months since there was anything in my mailbox.

On the + side, the last two days in two different stores i have been told by people working there that i am their favorite customer, so at least some things i am doing right.
 
Back
Top