A139 (Pro) Telephoto Remote Camera

TonyM

Well-Known Member
Joined
Jul 30, 2013
Messages
5,411
Reaction score
5,094
Country
United Kingdom
Dash Cam
A139, M1S
For a number of years I have experimented with the use of third-party fixed telephoto and varifocal M12 lenses on a dashcam, the intention being to complement a primary wide-angle dashcam by capturing information that the primary camera might not record. A telephoto lens can do this by recording a narrower angle of view (AOV), thereby increasing the pixels-per-degree across the middle of the scene ahead, and therefore the possibility of higher detail on features such as number plates or the actions of other road users. A telephoto lens can also benefit from metering only the narrower AOV, perhaps ignoring a bright sky or dark dashboard, and can give better results on reflective number plates for example.

For this project I decided to repurpose the interior remote camera of the A139 3CH dashcam and fit a fixed focus 6mm F1.2 starlight lens. Experience has shown that telephoto lenses in the order of 6mm to 8mm focal length have sufficient magnification to make a worthwhile difference over a primary camera (which are approx 2 or 3mm focal length), whilst not over-magnifying the central portion of the scene and suffering too much from vibration, motion blur and depth of field issues.

https://www.aliexpress.com/item/32663560773.html
1667220750911.png

The remote camera is fairly easy to open, the lens mount is easily accessible, and the lens focus can be achieved without first reassembling the housing.
1667221010893.png

The remote camera is also easy to position on the windscreen, and has a constant-friction stepless up/down angle adjustment.
1667221041332.png

Note: I started with the interior camera because I am still using the rear remote camera for its intended purpose. However I may in future try the same lens swap with the rear camera to see if it works just the same.
 
These frames show that the 6mm camera records approximately 50% width and 50% height compared to the angle of view of the A139 primary camera.

2022_1031_082202_F.MP4_20221031_130635.518.jpg 2022_1031_082203_R.MP4_20221031_130636.726.jpg
 
Last edited:
This was my first test drive an hour before sunset, which demonstrates the potential benefit for number plate capture in marginal conditions



The 'sweet-spot' for plate capture is not at the same moment in time for each camera/lens, so I have extracted 100% crops from each original video showing the best details that each camera captured. I was accelerating to 40mph whilst the oncoming traffic was slowing down, so the captures become progressively worse...

[Primary camera on the left / 6mm camera on the right]

1667223284560.png 1667223779710.png

1667223306460.png 1667223716463.png

1667223337730.png 1667223688801.png

1667223377764.png 1667223659526.png

1667223412527.png 1667223630059.png
 
Last edited:
What is the motion blur like at night? The A129 pro reverse camera is always 1/30 of a second exposure at night which makes the motion blur terrible.
 
What is the motion blur like at night? The A129 pro reverse camera is always 1/30 of a second exposure at night which makes the motion blur terrible.
It's going to be 1/30s exposure time at night, even on an IMX291 with an F1.2 lens.
 
It's going to be 1/30s exposure time at night, even on an IMX291 with an F1.2 lens.
It does not have to be 1/30s exposure time. They could crank up the ISO number and live with a dark noisy picture if it does a better job of capturing a readable license plate more often. As it is, license plates are over exposed to white, and smeared across half the frame.
 
If you have the lenses, would you be able to compare the different focal lengths?
Would be interesting to see how big of a different there is between each one.
 
If you have the lenses, would you be able to compare the different focal lengths?
Would be interesting to see how big of a different there is between each one.
I don't plan to try a range of lenses in this camera - I don't have the time for that. I've spent the past few days checking focus on the 6mm.
 
A139 6mm F1.2 - Night City Driving

EV-0, HDR OFF, Max bitrate


There is some curious ghosting on the highlights that I have not seen before with this lens. I'm not sure if it's some kind of internal reflection?

EDIT - I believe there was a chip in the windscreen near the camera
 
Last edited:
A139 6mm F1.2 - Night City Driving

EV-0, HDR OFF, Max bitrate


Another example at night. I'm not happy about having these electric scooters on the road. Whilst there are regulations governing their use - my city is one of a number of trial locations across the country - my experience is that scooter users don't follow the highway rules for much of the time, let alone the specific rules around scooters, and are a risk to themselves and others.
 
I'm not happy about having these electric scooters on the road. Whilst there are regulations governing their use - my city is one of a number of trial locations across the country - my experience is that scooter users don't follow the highway rules for much of the time, let alone the specific rules around scooters, and are a risk to themselves and others.
What rules do they operate under in your trial location? Same as liquid fueled scooters, or same as electric bicycles, or ?

He didn't seem to do anything unreasonable in that clip, other than the dress code, which should require a helmet and a reflective jacket rather than all matt black clothing and no protection! I do wonder if you broke the rules for minimum distance to a cyclist, if those rules apply? With those tiny wheels, there must be a risk of him hitting a pothole and falling under your wheels, you don't want to be too close when it happens!
 
I do wonder if you broke the rules for minimum distance to a cyclist, if those rules apply?

Keep in mind that this is a telephoto lens, Nigel.

Those of us who have modified cameras with telephoto lenses have often commented about how these lenses can create the illusion that one's vehicle is much closer to other cars, objects and people than is really the case. That is why for any use of this footage for legal purposes where distance is at question, it is important to show both a traditional dash camera wide angle image along with the concomitant telephoto image. (and perhaps an explanation of how the image was created)

By way of example, 5 years ago I made the following remark in regard to a family walking in front my vehicle in a parking lot captured with a telephoto lens (see last image in the post)

(Jeez, this telephoto lens sure does makes you feel like you're right on top of things and way closer that you really are!) :eek:
 
Last edited:
What rules do they operate under in your trial location? Same as liquid fueled scooters, or same as electric bicycles, or ?

He didn't seem to do anything unreasonable in that clip, other than the dress code, which should require a helmet and a reflective jacket rather than all matt black clothing and no protection!
It's the standard UK Gov rules for E-scooter trials / York E-Scooter trial

Agreed, in this example he didn't do much wrong. He ran the red light so was crossing the junction after our light turned green, and ideally he should be wearing a helmet and something other than black but I don't think that's mandatory.

In recent days I've seen a few examples of e-scooter riders hopping between the highway and pavement (pedestrian-only walkway) as it suits them.

My general impression throughout the trial has been that e-scooter riders are less compliant with highway regulations than cyclists, they don't seem to treat the activity with the seriousness that it deserves. Obviously that's not true of all riders.
 
Last edited:
Last week, I was driving down a street in town at night. Suddenly, I realized there was a person crossing the street in front of me. He was wearing black slacks and shoes, a black jacket and a black cap. He had long dark hair, so you couldn't really see his face. Like some kind of apparition, I could barely see him, even after I realized he was there about 10 or 12 yards in front of me. He was crossing the fairly busy street slowly, jaywalking in the middle of the block nowhere near a crosswalk, seemingly oblivious to his surroundings. I had to slow down to let him pass safely but he didn't notice. A Darwin Award candidate if ever there was one!
 
Keep in mind that this is a telephoto lens, Nigel.
Being a 6mm, that is not obvious, which is a big problem when using it for insurance/police!
He ran the red light so was crossing the junction after our light turned green,
I missed that!

In recent days I've seen a few examples of e-scooter riders hopping between the highway and pavement (pedestrian-only walkway) as it suits them.
These things go too fast for pedestrian areas, and the braking is I assume far slower than a proper bicycle, so hopping on the pavement can't be allowed.
And with a maximum speed of "15.5mph", they are too slow for safety on the road, so I reckon they should be confined to the cyclepaths and bridleways, and maybe pedestrian road areas (as used by busses) in the town centres. And if towns are going to encourage their use, then they need to implement a decent cycle network, which is a good thing to do anyway.
 
Being a 6mm, that is not obvious, which is a big problem when using it for insurance/police!

The fact that it is not "obvious" is the whole point here. Whether the particular telephoto focal length is shorter or longer makes little difference as the overall effect is that the viewer is given the impression that the vehicle from which the video was shot is much closer to vehicles and persons than it really is. Haven't you been paying attention? Maybe you haven't bothered to compare Tony's concomitant examples from both cameras? Of course, the fact that this is a thread specifically about a telephoto lens modification should certainly make it "obvious" to the reader, so there's that too!

And speaking of not paying attention, why do you ignore my remark about why it is important to put the footage in context when submitting to law enforcement/insurance and then go on to state that the footage is "a big problem when using it for insurance/police!" which already addresses your so called "big problem"?

for legal purposes where distance is at question, it is important to show both a traditional dash camera wide angle image along with the concomitant telephoto image. (and perhaps an explanation of how the image was created)
 
A139 Main Camera & 6mm F1.2

Composite video showing the 6mm F1.2 alongside the A139 main camera, proving that I did not get too close to the white car in front!

I would only use the telephoto video for evidence purposes if it provides useful information beyond that captured by the primary wide angle camera, and I would of course supply both recordings rather than just the telephoto video. In these bright sunny conditions the main camera is good enough.


And here's a few screenshots without YT compression:

Screenshot 2022-11-04 131114.jpg

Screenshot 2022-11-04 131314.jpg

Screenshot 2022-11-04 131741.jpg

Screenshot 2022-11-04 132243.jpg
 
I think the more sensitive IMX291 in the A139 internal camera is more suitable as a base camera for a telephoto lens.
The IMX291 is one of the best dashcam sensors, only issue is the low resolution. The IMX415 has excellent resolution, but for some reason it is always disappointing on motion blur. The IMX515 is much better.

... until Starvis 2 sensors are available!
 
The IMX291 is one of the best dashcam sensors, only issue is the low resolution.
That's one of the reasons I started looking into telephoto lenses years ago. I was never happy with the low detail on an IMX291 wide angle camera.
 
... until Starvis 2 sensors are available!
I might relocate this telephoto camera to the rear screen when my Starvis 2 camera arrives
 
Back
Top