Mangoal MG-0xxX OEM-Style Dashcam for Volvo S60

The other dashcam obstructing a large proportion of the bright sky in the frame is possibly affecting the Mangoal auto-exposure, causing it to raise the overall exposure.
I had thought about this after I posted images last night. The other camera leaves a big blackout area in the sky which is the brightest area of the frame, thus it makes sense it would lead to longer exposures to compensate for overall exposure levels. An easy solution would be to retest with complete darkness, so the sky isn't bright.

Edit: Maybe I'll remove that other camera. I just looked and it's mounted with static film so it should be easy to mark its location and replace it later.
 
Last edited:
My abacus calculated approximately 21Mbps Bitrate.
Given the video upscaling, I suspect the "effective" bitrate directly from the sensor (for example, if we had a native 2592x1944 resolution video file) is actually quite a bit lower.
 
My use case for wanting the Mangoal is very different from average users, which is that I have a car I will be renting out and I want to install a camera that:
  1. Renters cannot easily disable and are generally unaware of (no sounds or lights, but lights can be taped over)
  2. Records events
  3. Audio recording is disabled
I am wondering if you can confirm the following for me:

Can audio recording be permanently disabled from the app?

Does an event trigger the camera to make a sound? If so, can this sound be disabled from the app?

Does starting or stopping the car engine trigger the camera to make a sound? If so, can this also be disabled from the app?
 
OK, here we are for take two with the Mangoal camera comparisons. I've removed two other cameras so the Mangoal gets a clear view of the road ahead without giant black blobs in the way that can potentially muck with the exposure settings. The strips of masking tape you see in the Mangoal image are my markers so I can put my other cameras back. I timed this comparison so the rear wheels of the GMC pickup truck are on the painted white line on the road surface. I am at a full stop and the truck is passing by at about 35mph.

Mangoal Full Frame:
Mangoal gmc 1.webp


Here is the same from a known 2.5K camera:
25k gmc 1.webp


And the same scene from a known 4K camera:
4k gmc 1 marked up.webp


Comparison of Plate Capture from Arrow #1 Above:
Here is a zoom and crop of the rear plate from each camera made from the images directly above when the rear wheels of the truck we directly on top of the white painted line on the road in the image above. In this instance, the image quality of Mangoal zoom&crop looks closer the the known 4K than it does to the known 2.5K.
GMC 1.webp


Comparison of Plate Capture from Arrow #2 Above:
Here is the same comparison from a few frames later, when the rear tires are just past the black shadow line on the road about 15 feet beyond the first comparison point shown above. The best capture is clearly the 4K camera on the right. Both the Mangoal and the 2.5K capture are readable, but the "Y" appears to have transformed into a "T" in both images. Here is where a true 4K sensor begins to show it's value. The Mangoal image is also starting to become indistinct at the bottom of the numbers (especially on the right side) when compared to the known 2.5K camera.

Also noteworthy in this second set of plate comparisons is that the known 2.5K and 4K cameras are now producing a mostly rectangular image of the license plate, whereas the Mangoal camera is still producing a plate that most resembles a parallelogram. This larger distortion level indicates Mangoal's use of a lower quality lens than the other two. Curiously, the 4K camera is mounted on my windshield TO THE LEFT of the Mangoal camera (which is mounted exactly in the center of the window), which is a disadvantage for the 4K camera since I'm focusing on a truck that is passing on the right side of my car (greatest viewing angle). The known 4K camera is clearly using a superior lens.
GMC 2.webp


Comparison of Plate Capture from Arrow #3 Above:
And a few frames later after the truck has moved up a few more feet and the rear tires are just past the line indicated by arrow #3 above. At this point, both the Mangoal and the 2.5K captures are unreadable while the 4K can still be read. The Mangoal zoom now looks least like a license plate of these three images and still exhibits lens distortion resulting in a parallelogram rendering of the plate.
GMC 3.webp


What started off as a nice presentation by the Mangoal upscaling in comparison #1 began to deteriorate and only a few feet/frames later, the known Starvis2 sensors produced a superior image. I have several additional sets of these image comparisons, but they all follow this same pattern, so posting them seems like a lot of work that won't reveal any new information.
 
Last edited:
Can audio recording be permanently disabled from the app?
Looking at the screen cap from the app, it looks like it is possible to disable audio recording.

Does an event trigger the camera to make a sound? If so, can this sound be disabled from the app?
If the speaker volume is turned down to zero (off), I presume it would mute sounds.
Does starting or stopping the car engine trigger the camera to make a sound? If so, can this also be disabled from the app?
I think this is the same as above. I need to test these to confirm first.

You also need to make sure there is a model that fits your exact brand/model of car. Another option is the Vantrue E1 Pro or the Viofo A119, both are small enough to hide behind the rear view mirror (hidden, but not invisible), cost half as much, and produce an actual 4K video that is of very high quality.
 
It is well known that the center of the lens/frame often features the best image quality because the lens typically features the lowest level of optical distortion at that point. With that in mind, here are three 400% zooms of a road sign for "KOST Tire and Auto Center" that is more central to each full frame image above during bright daytime sun. This zoomed area is indicated by the yellow box in the image posted above. This is the most ideal situation for any camera.

400% zoom, center of frame from Mangoal's 2.5K to 4K upscaled camera (original image above):
Mangoal Kost.webp


400% zoom, center of frame from the known 2.5K camera (original image above)
25k Kost.webp


400% zoom, center of frame from the known 4K camera (original image above)
4k Kost.webp


At this point in time, these comparisons are becoming predictable. The digital upscaling from a 2.5K sensor to a 4K video stream in the Mangoal camera leads to the softest and most indistinct features of these three cameras. The Native 2.5K sensor/ 2.5K video lands it in the middle as expected, and the native 4k sensor/video combo comes out on top again.

Each of the three cameras is using some edge enhancement (see the perimeter of the rectangular sign) and each HDR/WDR routine leaves some color gradient ghosting between bright and dark areas in the image - this is especially visible where the tree leaves meet the sky.

It's a shame that the Mangoal quest for a "faux" 4K camera has resulted in notably inferior image quality when compared to just using a 2.5K sensor "as is" and without the video upscaling routine. A lot of effort has been expended that results in worse performance. File sizes would be smaller, leading to an SD card that holds more videos before being overwritten, and overall image sharpness/clarity would be improved without the upscaling.
 
Last edited:
MANGOAL’s Fake 4K components & counterfeit marketing make this is more of a summer squash 💩 🥔
VIOFO had a prototype RVM unit on display 18 Months ago at SEMA 2024.
Haven’t heard anything about it since then, (see attached screenshot).
 

Attachments

  • SEMA 2024 .webp
    SEMA 2024 .webp
    144.1 KB · Views: 7
OK, here we are for take two with the Mangoal camera comparisons. I've removed two other cameras so the Mangoal gets a clear view of the road ahead without giant black blobs in the way that can potentially muck with the exposure settings. The strips of masking tape you see in the Mangoal image are my markers so I can put my other cameras back. I timed this comparison so the rear wheels of the GMC pickup truck are on the painted white line on the road surface. I am at a full stop and the truck is passing by at about 35mph.

Mangoal Full Frame:
View attachment 92389
The Mangoal video appears much sharper on the right than on the left, which is a shame. Since this is mounted central to your car I think we can discount any effects of windshield curvature on the focus.
 
** IMPORTANT NOTE ABOUT SPECIFICATIONS AND PRODUCT LABELING **

The user's manual indicates two details that caught my eye:
- Sony IMX335 Video Sensor for the front camera (which has a maximum resolution of 2592x1944)
- 4K Ultra-HD 3840x2160 @30 fps

Both of these specifications cannot simultaneously be correct, there is a mismatch in resolutions. Perhaps I have a newer camera that was packaged with an older manual, I'm not sure right now. This prompted me to do some digging that I usually don't perform.

The MANGOAL webpage indicates four available dashcam models for my 2019 Volvo S60 (https://mangoal.com/collections/all/2019-2022-volvo-s60-v60-t5-t6-t8-b5-esr2394724). Each is labelled as a "Front 4K" camera, two of which are also labelled "HRD with Starvis 2." Two are listed as having a rear camera, one of the rear camera models is listed as "1080P" and the other is listed as "Rear 2K HDR and Dual Starvis." Below is a screen capture with red highlights added:

View attachment 92268
I searched for available products for my car and found what appear to be the same 4 variants: 1CH or 2CH, with either an IMX335 STARVIS 1 or IMX678 STARVIS 2 front camera.
1784186021509.webp


It is unfortunate that Mangoal did not send you a STARVIS 2 model to test. Having read your review so far, perhaps they will consider sending you the other model so you can make a comparison.
 
I searched for available products for my car and found what appear to be the same 4 variants: 1CH or 2CH, with either an IMX335 STARVIS 1 or IMX678 STARVIS 2 front camera.
View attachment 92405

It is unfortunate that Mangoal did not send you a STARVIS 2 model to test. Having read your review so far, perhaps they will consider sending you the other model so you can make a comparison.
Is the STARVIS 2 model new? I don't remember seeing that, but maybe I missed it. If it is new, dollars-to-doughnuts it has other improvements and "fixes", as smaller companies may leave old stock on the warehouse shelf even when they know they shouldn't ship it.
 
The Mangoal video appears much sharper on the right than on the left, which is a shame. Since this is mounted central to your car I think we can discount any effects of windshield curvature on the focus.
This is curious - I just went back to look at the overall images again and see the same thing. Funny, during my first pass of pixel peeping, I didn't look for lens uniformity left-to-right across the whole frame. That lens is definitely sharper on the right. Maybe the lens is meant for UK drivers or left-lane highway drivers? 😁

Or, it's just not uniform and the "good spot" randomly points left??

VIOFO had a prototype RVM unit on display 18 Months ago at SEMA 2024.
Haven’t heard anything about it since then, (see attached screenshot).

I wonder if it is getting harder to replace the entire RVM with cars becoming more complex (integrated electronics) and the introduction of ADAS that sit behind the RVM on more and more models? Form factor is clearly important and there is a market for the "integrated" look. It would be great to have both an integrated form factor AND excellent video performance & feature set.
 
Back
Top