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

[EricSan]

Sorry to be so negative, and rain on your testing parade.

My goal is simply to be as objective and accurate as possible👍

One way is to request / obtain FCC ID# to review legal documents they’ve submitted for certification.
Documents may contain photographs of components.

I'll have to see what I can find with a cursory examination. I don't want to separate the lens assembly from the sensor board, I'll never be able to close it up again without introducing dust.

 

[Chuck McCoy]

My goal is simply to be as objective and accurate as possible👍

Well of course that goes without saying.
But why do you think no one else is testing these fake / counterfeit crap cameras?
Unless you’re purposely showing what a fraud they are.
Unscrupulous lying companies don’t deserve the exposure.

I don't want to separate the lens assembly from the sensor board, I'll never be able to close it up again without introducing dust.

That’s why you always request (2) units for testing.
If they refuse, no dice.

 

[EricSan]

Before installing the camera in my car, I made some power measurements:

My protocol for obtaining these measurements can be found here: https://dashcamtalk.com/forum/threads/reviewing-and-testing-the-vantrue-nexus-5s.54853/post-670930

Invoking parking mode is pretty straight forward. The camera powers up only when +12v is applied to BOTH the Red and Yellow wires. The Yellow wire carries the +12v power supply while the Red wire is the +12v ACC trigger that toggles parking mode. Remove the voltage from the Red wire, and if you've set Parking Monitoring Duration in the app to one of the pre-defined time periods (12h, 24h, or 48h) the camera will enter Time Lapse Recording. No other parking modes are available that I've been able to discern.

Overall, I like the power consumption figures: just over 4w for two channels of drive mode with GPS, ~3w for dual channel parking mode, and ~2w for single channel parking mode. Pretty nice.

Recommendation:
Given the relatively small amperage being drawn (I measured a maximum of 0.351A that ramps up slowly as the camera powers on), I'd recommend a 1A slow-blow fuse, or at maximum - and potentially to prevent the occasional random fuse blowing - a fast-acting 2A fuse instead of the 5A slow-blow fuse that is included in the power harness. 5A is simply too large to be effective and potentially invites other issues.

 

[EricSan]

Here is the set of options in SkyComm app menu - it's pretty basic in what it allows you to control:

Image Resolution can be set to 10M, 8M, 6M, 4M, or 2M. I'm wondering what kind of upscaling is going on to produce a 10 megapixel image out of what I believe to be a 5 megapixel sensor.
Video Resolution can be set to 4K + FHD, 2K + FHD, or FHD + FHD.
Loop Record Options can be set to 1M, 2M, or 3M.
Exposure Compensation can be set from +2.0 to -2.0 in the usual 0.3 increments.

If you scroll down in the app menu, here are the remaining settings - they span only two screens:

Collision Sensing can be set to High, Medium, or Low.
Parking Mode Collision Mode can similarly be set to High, Medium, or Low.
Parking Mode Duration can be set to Off, 12 Hours, 24 Hours, or 48 Hours. Parking Mode recording does not happen when this setting is "Off" - the camera just shuts down when voltage is removed from the Yellow wire.

And that's it, not too many options to fiddle with. This is a pretty basic camera without the usual set of bells and whistles..

Let's follow up on that parking mode setting:
The FAQ page on their website (https://mangoal.com/pages/faqs) indicates: "MANGOAL uses advanced power protection. When battery voltage drops below 11.6V, the dash cam enters sleep mode with minimal power consumption. Please ensure sufficient battery voltage before next engine start." (emphasis added)

This means that if you repeatedly run your camera with parking mode enabled, you are likely to run down your main start battery at some point in the near future and find that your car won't start. That will not be a happy day for someone! A quick search reveals that an 11.6v voltage cutoff all but guarantees your car won't start - 11.6v pretty much means that there is ZERO remaining battery charge. Depleting a car battery that far risks significant and permanent battery damage! A much better configuration would be to have a 12.2 voltage cutoff (which leaves 50% battery capacity which is sufficient to start your car) and then have the camera TURN OFF COMPLETELY, not continue to consume "minimal power."

Below is a screen capture from the CarMonitor website that highlights my concern:

Issues with the App:
All changes to settings need to be performed with the SkyCamm app. The user's manual seems to indicate this is a third party app, which raises some issues/questions. I'm wondering if this means that firmware/feature updates will be relatively slow and infrequent compared to other brands that directly oversee their own programming teams. The SkyCamm app for camera control was developed & written by Shenzhen Xiaozhen Technology Co., Ltd. This company also seems to be the author of at least 7 other dashcam apps: Nice DVR, AKEEYO, VeeCar, ZJTY, Onlook Cam, SARGO, NOBODY, and Viidure (also referenced in the manual).

There are very few options to select/set in the app, it's pretty minimal. The app also randomly kicked me back out to recording mode while I was tapping on options. I couldn't seem to make this happen with any kind of consistency. The app also failed to register many of my selections the first time I attempted to change a setting. I had to change many of them two or three times before the app registered my change. Thus, there appears to be some bugs in the app that need to be worked out. I'd also like to see a greater set of features and selections for controlling the camera than are presently available.

Finally, the 11.6v cutoff for the battery DEFINITELY needs to be changed to something closer to 12.2v to prevent killing an expensive car battery.

 

[EricSan]

Here is the camera installed in my car. I popped off the original panel from the underside of my ADAS system and snapped the Mangoal camera in its place. This is a quick and dirty install, so I didn't spend the extra time to tucked the wires neatly out of site (plus, it's very hot in the garage today). From my position in the driver's seat, the camera is barely visible, just a bit of the bottom pokes out. I removed one of my long-term cameras to provide a better view for this new camera. One camera remains in the upper field of view, it will just block a little bit of the sky and not interfere with recording cars or the roadway. I also installed the rear camera and will start collecting some footage tomorrow. I'm curious to see the footage.

 

[Karagandinez]

I’m starting to lose interest in this manufacturer...🤔
All dashcam manufacturers make their firmware publicly available.
I asked the manufacturer's representative for the firmware for my model.
Here is the reply:

I’m truly sorry that I’m not authorized to send you the original firmware file, as it counts as confidential proprietary product information.
If you need a firmware upgrade later on, we can send you the official upgrade file for installation.

 

[EricSan]

I asked the manufacturer's representative for the firmware for my model.

Hmmmm... I'm wondering how, when, and now "if" firmware updates happen. That seems like an odd reply, given that we are accustomed to dashcam brands that build our collective suggestions and feedback into the next iteration of their firmware and provide early test releases of new firmware directly here on the forums before the official public releases are posted to their web page.

Looking at the Mangoal support page and their FAQ (https://mangoal.com/pages/faqs), I cannot find any place where firmware or firmware updates are available for download.

Given the variety of apps that I found in the Apple App Store all authored by the same company, I'm wondering what kind of relationships (or separations) exist among physical manufacturing, firmware authoring, and retail sales and how much control over firmware and features actually exists at the retail level...

According the the screencap from the app I posted above, it looks like the firmware in this camera is already more than one year old: June 19, 2025.

 

[TonyM]

Before installing the camera in my car, I made some power measurements:

Can you add a heading or other information in your table to make it clear that timelapse is parking mode? Some dashcams have an option for recording timelapse whist driving.

 

[Augustus]

I'm watching this thread with interest, as pretty much everyone on this thread has had an offer of a review unit to be sent to them by Mangoal.

I have my doubts about the quality of this product, but if it is indeed a Sony IMX678 sensor used, then there may be a silver lining if it's been implemented to a good degree.

 

[EricSan]

Can you add a heading or other information in your table to make it clear that timelapse is parking mode? Some dashcams have an option for recording timelapse whist driving.

Interesting, I was unaware of a time lapse drive mode- thank for pointing that out! I've added a label to the table and replaced it. Is it more clear now?

pretty much everyone on this thread has had an offer of a review unit to be sent to them by Mangoal.

I suspected that might be the case... This is why I'm trying to document things as thoroughly and as accurately as possible. If my review isn't useful to others, then I've wasted everyone's time, including mine. As I dig further, I've gone back to edit & add some additional details to a few of my earlier posts.

if it is indeed a Sony IMX678 sensor used, then there may be a silver lining

I'm waiting on confirmation from Amber, but I'm thinking I have an example with a 2.5K Sony IMX335 sensor in the front camera. I made a bunch of videos today, but haven't yet had the time for a critical analysis and JPG comparisons. I should be able to post that within the next day or two. Have a big deadline at work this week.

 

[lufa6977]

I'm watching this thread with interest, as pretty much everyone on this thread has had an offer of a review unit to be sent to them by Mangoal.

I have my doubts about the quality of this product, but if it is indeed a Sony IMX678 sensor used, then there may be a silver lining if it's been implemented to a good degree.

Same here, watching to see how Mangoal responds. @EricSan has done a good job of highlighting the marketing issues and the indiscriminate use of the 2.5K/4 K resolution nomenclature.

 

[TonyM]

Interesting, I was unaware of a time lapse drive mode- thank for pointing that out! I've added a label to the table and replaced it. Is it more clear now?

Yes, that's clearer for anyone taking a quick look at the table.

 

[EricSan]

OK, here is the first of a series of frame captures comparing the Mangoal camera with two other cameras with known/verified sensors and resolutions. We'll start with the hard stuff: night time scenes. I tried hard to make sure my car and other cars are at the very same distance using other objects in the frame as markers of time alignment across the cameras. There is probably about a 40mph speed difference here, 20mph in each direction.

The black blob in the upper center of the Mangoal captures is another dashcam that is centered on my windshield. I was too lazy to move it out of the way because it was a pain in the butt to mount it centered and level the first time around 😉. The camera angle of the Mangoal camera is obviously not adjustable, so I was wondering about the mounting angle. It looks pretty close to having a 50% vertical split between sky and foreground. I tend to prefer a split that is closer to 60% foreground and 40% sky for better overall exposure levels (preventing dark foreground because of an overly bight sky).

Full frame from the Mangoal Camera with Wide Dynamic Range Enabled.

This capture is from a known 2.5K camera using the Sony Starvis2 IMX675 with Wide Dynamic Range Enabled:

And here is the very same scene from a known 4K camera using the Sony Starvis2 IMX678 with High Dynamic Range enabled:

Here are the corresponding 200% zoom/crops from those same images above, starting with the Mangoal camera. Lots of pixelation is visible, some interesting headlight glare is present, and the front plate is a complete washout. Moving back and forth a few frames reveals no readable plate capture at all.

Next up is a 200% zoom and crop from a known 2.5K Starvis2 IMX675. Much less pixelation, front plate still not readable. Much cleaner overall rendering this time. Moving back and forth a few frames reveals no readable plate capture at all for this camera.

And finally a 200% zoom and crop from a known 4K Sony Starvis2 678. There is some pixelation, but we ALSO have a readable plate!

Stay tuned for more night time comparisons...

 

[EricSan]

Here is the second set of night time captures. In this case, I'm moving at 15mph past the parked car. Same settings as posted above: HDR and WDR all enabled.

Mangoal full frame:

Known 2.5K Starvis2 sensor camera:

and the same known 4K Starvis2 camera as above:

And some corresponding 200% zoom and crop images from the series directly above:

Mangoal 200% zoom - very pixelated and plate is not readable.

Known 2.5K Starvis2 sensor camera zoom/crop. This is a much smoother image and I can see the rear plate, but nothing readable here...

And finally, the same known 4K Starvis2 sensor with a 200% zoom&crop. But this time, we have a semi-readable plate! The plate is actually "NGE 2262"

 

[EricSan]

Here is the easiest of tests for a night time scene: both cars moving in the same direction, so speed differences are close to 1-2mph.

Mangoal camera:

Known 2.5K Starvis2 camera, same scene:

and the known 4K Starvis2 camera:

Here are the 200% zoom & crop images in the same order:

Mangoal camera - we have a readable plate, but an overall image that contains significant digital noise / video compression artifacts.

Known 2.5K Starvis 2 camera, same 200% crop. Also a readable plate, but note the much cleaner presentation of the Starvis2 sensor over that from above:

And the last of the trio - a 200% zoom & crop from the 4K Starvis2, cleanest rendition of the three also with a readable plate.

 

[EricSan]

This time around, we have the very same image and distance/timing, but from the rear camera. This comparison is much more even, all three cameras produce a 1080p video feed, so resolutions are very nearly the same.

Mangoal Rear Camera, full frame, 1080p

Sony Starvis2 IMX675 sensor full frame, 1080p:

and from the rear camera of the 4K dashcam, also a Starvis2 IMX675 1080p full frame:

And now some 300 percent zoom & crops, in the same order as presented above:

Mangoal rear camera 300% zoom&crop from 1080P. Note the outline of the street lights and the red stop sign on the right. No lettering is visible on the sign at all. You can also see the back of a stop sign on the left side of the image - note the metal pole attached to the back of the sign is not visible at all. Four separate headlights and road reflections (from a dry road) all merge into a single bright blob in the center of the image. Those who are familiar with photography or other optics recognize the center of the field as the most linear/accurate area of lens geometry.

From the rear channel of the 2.5K camera, IMX675 1080P. Stop sign on the right is not readable, but it's clear that lettering is present. The rear of the stop sign on the left has a visible pole attached now. There is greater light bloom here, but we're starting to see some separation in the lights - but it is not yet clear and distinct.

Finally, a 300 percent zoom/crop from the best of the bunch - this is the 1080P Starvis2 IMX675 rear channel from the 4K camera. This time, the STOP sign is completely readable! The back side of the stop sign on the left very clearly has a metal pole attached.

All three cameras suffer from light bloom from the headlights of the car behind me - I expected this to be the case. The level of bloom is worst with the Mangoal camera and is best rendered by the rear channel of the known 4K camera - there are now four distinct light sources instead of two amorphous light blobs with the first two cameras. There is also clear separation between the headlights and their reflections on the road surface.

 

[EricSan]

A few conclusions at this point:

Given it's price point and (now) known specs, I was really hoping to see this camera perform more closely to a known 2.5K sensor/resolution camera. This didn't happen with the above night time images. I have some day time footage that I haven't yet had time to explore and will post that shortly.

- The Mangoal camera exposure level is a bit higher than the other two cameras, resulting is notably brighter images. While the images and video stream have a somewhat more pleasing appearance, it seems that the brightness comes from longer shutter times at night. This results in the greatest level of image blur among the three cameras.
- In each set of images, the Starvis2 sensors from the known 2.5K camera (both front and rear) outperforms the sensors in Mangoal camera. It seems the video resolution upscaling performed in the Mangoal camera that converts the 2.5K front sensor into an artificial 4k video stream introduces significant pixelation and digital artifacts that become visible in the video stream.
- In each set of images, the Starvis2 sensors from the known 4K camera (both front and rear) significantly outperforms both sensors in the Mangoal camera. There is a clear and obvious difference in video quality between the video feed that emerges from a 4K sensor vs a 2.5K video feed that is upscaled to 4K prior to video encoding.
- The video files from the Mangoal camera are in .TS format rather than the more traditional .MP4 format from other cameras. This makes the embedded video metadata a bit more difficult to inspect with my usual set of tools.
- The lenses on the Mangoal camera seem to exhibit the greatest amount of distortion at the edges of the frame, suggesting a lower grade lens is being used, despite having a very similar field of view as the 2.5K camera.
- All of this confirms (to me, at least) that the Mangoal camera uses a Starvis1 IMX335 front sensor as indicated in my user’s manual with native 2592x1944 resolution. This is clearly NOT a 4K camera as claimed. Though, through the use of video upscaling, the native 2.5K video sensor feed is mathematically converted to a video with 4K output resolution. This appears to be a highly deceptive marketing practice for this dashcam model.

I am unable to determine the video bitrate of the .TS files, but here are the resulting file sizes from different video resolution settings in the app:
Codec: MPEG-H Part2/HEVC (H.265) (hevc)
3-min video from the front camera at 4K menu setting is 408,987 KB
3-min video from the front camera at 2K menu setting is 326,241 KB
3-min rear camera at 1K is 151,488 KB

I was a little surprised to see the H.265 codec being used here - I was expecting to see H.264.

I made some captures during daylight hours that I'll post tomorrow. I'm curious to see if the daytime comparisons show the same magnitude of difference as the challenging night time scenes...

 

[Augustus]

A few conclusions at this point:

- The Mangoal camera exposure level is a bit higher than the other two cameras, resulting is notably brighter images. While the images and video stream have a somewhat more pleasing appearance, it seems that the brightness comes from longer shutter times at night. This results in the greatest level of image blur among the three cameras.
- In each set of images, the Starvis2 sensors from the known 2.5K camera (both front and rear) outperforms the sensors in Mangoal camera. It seems the video resolution upscaling performed in the Mangoal camera that converts the 2.5K front sensor into an artificial 4k video stream introduces significant pixelation and digital artifacts that become visible in the video stream.
- In each set of images, the Starvis2 sensors from the known 4K camera (both front and rear) significantly outperforms both sensors in the Mangoal camera. There is a clear and obvious difference in video quality between the video feed that emerges from a 4K sensor vs a 2.5K video feed that is upscaled to 4K prior to video encoding.
- The video files from the Mangoal camera are in .TS format rather than the more traditional .MP4 format from other cameras. This makes the embedded video metadata a bit more difficult to inspect with my usual set of tools.
- The lenses on the Mangoal camera seem to exhibit the greatest amount of distortion at the edges of the frame, suggesting a different lens is being used, despite having a very similar field of view as the 2.5K camera.
- All of this confirms (to me, at least) that the Mangoal camera uses a Starvis1 IMX335 front sensor as indicated in my user’s manual with native 2592x1944 resolution. This is clearly NOT a 4K camera as claimed, though through the use of video upscaling, the native 2.5K video sensor feed is mathematically converted to a video with 4K output resolution. This appears to be a highly deceptive marketing practice for this dashcam model.

I am unable to determine the video bitrate of the .TS files, but here are the resulting file sizes from different video resolution settings in the app:
Codec: MPEG-H Part2/HEVC (H.265) (hevc)
3-min video from the front camera at 4K menu setting is 408,987 KB
3-min video from the front camera at 2K menu setting is 326,241 KB
3-min rear camera at 1K is 151,488 KB

I was a little surprised to see the H.265 codec being used here - I was expecting to see H.264.

I made some captures during daylight hours that I'll post tomorrow. I'm curious to see if the daytime comparisons show the same magnitude of difference as the challenging night time scenes...

Thank you for the detailed, albeit, brutal, writeup of this camera. Suspected a lot of our worst suspicions from others here on the forum.

 

[Chuck McCoy]

I am unable to determine the video bitrate of the .TS files, but here are the resulting file sizes from different video resolution settings in the app:
Codec: MPEG-H Part2/HEVC (H.265) (hevc)
3-min video from the front camera at 4K menu setting is 408,987 KB

My abacus calculated approximately 21Mbps Bitrate.
This Fake 4K camera with only 21Mbps Bitrate should show its disadvantages at highway speeds with everything moving during daytime 💩

VIOFO 4K Bitrate is approximately 60Mbps.
VIOFO 2K Bitrate is approximately 30Mbps.

 

[TonyM]

The black blob in the upper center of the Mangoal captures is another dashcam that is centered on my windshield.

A few conclusions at this point:

- The Mangoal camera exposure level is a bit higher than the other two cameras, resulting is notably brighter images. While the images and video stream have a somewhat more pleasing appearance, it seems that the brightness comes from longer shutter times at night. This results in the greatest level of image blur among the three cameras.

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.