Reviewing and Testing the new Vantrue E360 Ace Dashcam

[EricSan]

The E360 is the first of its kind.

Absolutely, this is what makes it such an interesting camera!

I wonder if video processing can be optimized (greater detail) for that horizontal middle 1/3 band where all of the critical action will occur, and de-emphasized for the top and bottom 1/3 bands where critical action is not occurring. I'm clearly venturing into things I don't understand about MPEG encoding now...

 

[Julian23]

Absolutely, this is what makes it such an interesting camera!

I wonder if video processing can be optimized (greater detail) for that horizontal middle 1/3 band where all of the critical action will occur, and de-emphasized for the top and bottom 1/3 bands where critical action is not occurring. I'm clearly venturing into things I don't understand about MPEG encoding now...

Vantrue saw our wishlist. We have to see what changes will be implemented in the future. I hope many of them.

 

[EmailPJC]

In his review thread, @Julian23 made a comment about the bitrate for the E360. I was curious about the bitrate/MPEG ........

I'm wondering if there is a way to play with the geometry/spacing of the E360 lens in relationship to the sensor chip that would maximize the usefulness of the frame that is captured (as the top and bottom bands don't contribute very much that would help following an accident), yet still retain the excellent peripheral detail (all of the way to the A-pillars!) that other cameras cannot capture (while still enabling the 360 function). I'm guessing Jeff and his team have been through this in great detail while designing the E360. It's a classic case of that "no free lunch" thing that exists everywhere.
.......

Interesting.....

I suppose that the file size and quality of the video is also closely related to the compression codec, but I totally agree with you that so much dashboard and sky contributes little to what we would like to obtain and all that useless information would be great to have in that central third... unless a plane falls on us...

-------------
Interesante.....

Supongo que el tamaño de archivo y la calidad del vídeo también está muy relacionado con el codec de compresión, pero estoy totalmente de acuerdo contigo en que tanto salpicadero y tanto cielo aporta poco a lo que nos gustaría obtener y toda esa información poco útil sería estupendo tenerla en ese tercio central.... a no ser que nos caiga una avioneta encima....

 

[TonyM]

I fully acknowledge that this may not be the most ideal way to break this down, but I wanted some sort of "normalized" comparison between the two cameras. What I came up with is a measure of the "number of pixels per kilobyte over time." I obtain this measure by multiplying the vertical and horizontal resolution of the video frame to arrive at the total number of pixels that are encoded and divide this by the resulting filesize. Since both cameras are recording at 30FPS and with 3min video loops, these two parameters are ignored in my analysis. I did this for both cameras:

Vantrue: Single camera file at maximum resolution of 2592 x 1944 (5,038,848 pixels) at 30FPS = 358,436KB file for 3 mins of recording. Give or take a little bit on filesize.

Viofo: Single camera at maximum resolution of 3840 x 1600 (6,144,000 pixels) at 30FPS = 1,019,708KB for 3 mins of recording. Again, give or take a little bit on filesize depending on exactly what driving scene has been recorded.

Here are the results of using my somewhat arbitrary formula:
Vantrue E360: 14.05 pixels per kilobyte of filesize over a 3min recording at 30FPS.
Viofo A139Pro: 6.02 pixels per kilobyte of filesize over a 3min recording at 30FPS.

If you want to use a standard parameter for comparing video compression and data rate, you can look at the output from Media Info - https://mediaarea.net/MediaInfoOnline
There is a calculated value for "Bits/(Pixel*Frame)" which is a measure of how much data is used for each frame at the given resolution

In this example from an E360 low bitrate parking video:

(4,915,000 bits per second) / (1920 pixels x 1080 pixels x 15 fps) = 0.158

Taking your example of the E360 recording 2592x1944P at 30fps, with a calculated bitrate of (358436kB x 8bits/byte) / (3min x 60s) = 15930kb/s:
(15,930,000) / (2592x1944x30) = 0.105

And the A139 Pro recording 3840 x 1600 at 30fps, bitrate (1019708 x 8) / (3 x 60) = 45320kb/s:
(45,320,000) / (3840x1600x30) = 0.246

The end result is the same as comparing your values of 14.05 and 6.02, the bits/pixel density is about 2.3 times higher on the A139 Pro than the E360.

I may have been sloppy in my bitrate conversions using 1000 rather than 1024, but for comparison purposes it's OK.

 

[EricSan]

HA - told you I don't know much about MPEG codecs 🙂. Guess I wasn't too far off by making up my own calculation, the overall ratios of compression rates seems to hold up pretty well. Gotta love math as a universal language...

 

[EmailPJC]

How much better it would be if, as mentioned, the recording capacity could be maintained but a large part of the excess sky and dashboard could be removed!
----------------
¡Cuanto mejoraría si como se comentó se pudiera mantener la capacidad de grabación pero quitando una gran parte de cielo y salpicadero sobrante!

 

[Nigel]

How much better it would be if, as mentioned, the recording capacity could be maintained but a large part of the excess sky and dashboard could be removed!
----------------
¡Cuanto mejoraría si como se comentó se pudiera mantener la capacidad de grabación pero quitando una gran parte de cielo y salpicadero sobrante!

Looking at the image from above, if you remove the top third, and bottom third (marked as not seen by Viofo), then remove the area covered by the A-pillars (also marked as not seen by Viofo), then you get an image that covers much the same area as the Viofo, so you then just compare the bitrates, and instead of:
"the bits/pixel density is about 2.3 times higher on the A139 Pro than the E360."
you get about:
the bits/pixel density is about 0.3 times higher on the A139 Pro than the E360.

But then you don't have a 360° dashcam anymore, instead you have a Viofo copy!

This camera is special because it is 360°, if you actually want a normal dashcam then this is not the camera you want.

(That assumes that having cropped the image, you then interpolate it back to a decent resolution, otherwise you will have a tiny image!, but of course, interpolation is not good for detailed images.)

 

[Karagandinez]

Some people want to remake Viofo into Vantrue,
and some, on the contrary, Vantrue into Viofo...

 

[TonyM]

I use the 16:9 aspect ratio (1440P) to crop some of the sky and dashboard from the taller 4:3 video (1944P)

 

[EmailPJC]

Maybe I'm not explaining myself well. I don't know what the Viofo camera is like, I don't have one, I can't compare, and the 360 degrees are very good, I like them a lot, and even more so if those images are joined together as well as Vantrue does, and I also like the windshield frames to come out, they seem like a great reference, but nevertheless, it doesn't help me much to see in the videos how my knees are or the roof of the car or so much sky and that's just what I wanted to say.

I imagine that there is no lens with a reasonable price capable of taking those cropped 360 degrees that I was referring to with all the possible quality or information condensed only in that "important" part. It was just a wish that I suppose we would all find fantastic (or so I think).

----------
Quizás no me explico bien. No sé como es la cámara Viofo, no tengo ninguna, no puedo comparar, y los 360 grados están muy bien, me gustan mucho, y más si esas imágenes se unen entre si tan bien como lo hace Vantrue y también me gusta que salgan los marcos del parabrisas, me parecen una magnifica referencia, pero sin embargo a mí no me sirve de mucho ver en los videos cómo tengo las rodillas o el techo del coche o tanto cielo y eso es solo lo que quería decir.

Imagino que no existe una lente con un precio razonable capaz de sacar esos 360 grados recortados a los que me refería con toda la calidad o información posible condensada solo en en esa parte "importante". Era solo un deseo que supongo nos parecería a todos fantástico (o eso creo yo).

 

[EricSan]

Some people want to remake Viofo into Vantrue,
and some, on the contrary, Vantrue into Viofo...

I see what you are saying and this is not my intention at all. The 360 degree field of view is awesome and what I like most about this camera. The important part of the 360 degree function is in the horizontal field of view.

My point is that due to the fisheye lens, there is also a tremendous amount of vertical information that is not necessarily valuable here. I'm wondering if there is a way to optimize the pixel utilization of the sensor through changes to the lens geometry and retain the 360 degree field of view without all of the vertical information. Kind of like adding an "anamorphic" lens to old home theater projectors to get the wider image display without sacrificing data resolution stored in the file.

 

[Nigel]

My point is that due to the fisheye lens, there is also a tremendous amount of vertical information that is not necessarily valuable here. I'm wondering if there is a way to optimize the pixel utilization of the sensor through changes to the lens geometry and retain the 360 degree field of view without all of the vertical information. Kind of like adding an "anamorphic" lens to old home theater projectors to get the wider image display without sacrificing data resolution stored in the file.

The lens is not the problem, it appears pretty good.
I think the question is only how much of the current lens image should be stored. Tony says he is using 16:9 aspect ratio which chops off the top and bottom, which helps, you could use 32:9 which would remove more of the top and bottom.

It would be reasonable to request a recording mode where the camera only stores what can be seen in the following image, but with the front view zoomed in to about match the Viofo image, these 4 views are the useful views, although we are missing the rear:

It could record a single video file showing the 4 views above, exactly as seen above, but with the front zoomed in a bit, which makes it very easy to view, and if it was a 4K file, it could store all the resolution available from both sensors, but with the top and bottom cropped. Or it could store it as 4 separate files, or as a very wide 360 file that would be nice to view with a good 360 viewer, but is less easy to see what happens because you cant fit it all on the screen at the same time, without it being very small.

 

[Augustus]

I did a quick and dirty install for the Vatrue E360 in my Volvo S60. The Volvo presented some placement challenges as it has a large "IntelliSafe" module for it's Pilot Assist and Collision Avoidance systems that is mounted directly behind the rear view mirror. This module is somewhat large and obstructs the 360 degree view of the camera.

View attachment 76222

Placing the E360 directly to either side of the IntelliSafe Module doesn't really work in terms of allowing the 360 degree function to work. Placing it at the top of the windshield or moving it down a few inches results in the same camera obstruction in addition to interfering with my sun visor. Because of the larger profile of this camera (greater distance between the windshield and the bottom of the camera), placement is a bit sensitive to range of movement of your sunvisor.

If you wish to access the E360 touch screen, you need to keep the camera on the driver's side of the mirror or centered under the mirror. If you place the camera on the passenger side of the car, it will likely be easier to operate it through the app.

View attachment 76223

The only viable install location in my S60 is centered below the IntelliSafe Module so the cabin facing camera can see below the rear view mirror as shown below. You can see my Viofo camera in the background.
View attachment 76224

EDIT: I didn't think to show this perspective earlier in the day, so here are some images to show the relative differences in "profile" of the Vantrue and Viofo cameras. The first shows the distance from the windshield to the Vantrue E360 camera in my low-profile sedan. In a larger vehicle, such as a truck or SUV, the windshield is likely to be more vertical, thus decreasing the distance between the windshield and the camera.

View attachment 76226

Here is the same view from the passenger side of the car. In the foreground is my Viofo A139 Pro. Centered under the rearview mirror, the Vanture is not obtrusive at all, though it does potentially introduce a blindspot for the driver, again depending on the specific configuration of your windshield and placement of your rearview mirror.
View attachment 76227

I just saw how big that Volvo module for lane assist and AEB is, that's huge!

 

[EricSan]

Yeah, it's pretty big. It hides completely behind the mirror, so I don't normally think anything of it. But once you try to tuck something else out of the way by putting it behind the mirror, you start to realize just how big it is...

 

[Augustus]

Yeah, it's pretty big. It hides completely behind the mirror, so I don't normally think anything of it. But once you try to tuck something else out of the way by putting it behind the mirror, you start to realize just how big it is...

Shame, mine isn't that big so I wonder why theirs is so big unless it's an older car etc where they can't minimise the profile of it ..

 

[Julian23]

Shame, mine isn't that big so I wonder why theirs is so big unless it's an older car etc where they can't minimise the profile of it ..

Cameras and sensors for the various safety systems. Because more is better haha
The new BMW models also have a huge camera unit in front of the mirror.

 

[Augustus]

Cameras and sensors for the various safety systems. Because more is better haha
The new BMW models also have a huge camera unit in front of the mirror.

True but by the sounds of it,his car has the exact same stuff that mine does. Which makes me wonder whether it really does have additional safety systems on top or they just using older tech. Also seeing there single or dual camera units. A lot of it can be optimised by updates too.

The Kia forte I drove that was a couple of years newer than my one definitely was better at all the lane assist and centering - makes me wonder if it was done in firmware or upgraded CPU/GPU cos was same size unit ..

 

[EricSan]

Dunno, mine is a 2019 Volvo S60. I don’t remember the last time I had the cover off and how much of that was actually used space vs just empty space. I’ll have to pop the cover off again and have a look.

 

[EricSan]

I picked up a fun little USB-C breakout board a little while ago so I could measure which pins control which function for the Vantrue E360 when connected to the Vantrue hardwire kit. I found the results to be somewhat interesting.

This is the board that I used: https://www.amazon.com/gp/B0BC3PYQHJ
What I like about this one is that it is a pass-through board with a male connector on one side and a female connector on the other side, so you just insert it into your USB chain and devices continue to work as usual, but now you can make voltage measurements on the wires. Pretty clever...

The differences in behavior that I found with using/not using the GPS mount are kind of interesting. I suppose that denotes the dual functionality of the camera: as a dashcam in your car with the GPS unit attached vs as a handheld 360degree camera that mounts to the battery stick without using the GPS unit.

Vantrue E360 camera powered by the Vantrue Hardwire Kit. Drive mode was enabled by providing +12v to both the red and yellow wires of the kit. Park mode was enabled by removing the +12v signal from the yellow (ACC) wire, while maintaining +12v on the red (B+) wire. I plugged one end of the breakout board into the USB-C cable end from the hardwire kit and the other end of the board was plugged into the GPS unit for the E360.

E360 with GPS unit attached and +12v on both Red and Yellow wires:
A1: GND
A2: 0
A3: 0
A4: 5.03v
A5: 0.02v
A6: 1.01v
A7: 1.01v
A8: 4.38v (0.04v with yellow wire disconnected from 12v supply)
A9: 5.03
A10: 0.0
A11: 0.0
A12: GND

A8/SBU1 (and also B8/SBU2 due to reversibility of the USB-C cable) seems to carry the drive/park mode trigger for the Vantrue hardwire kit.
The fun part here is that the A8/SBU1 pin is not present on the typical ready-made 5- or 6-wire USB-C cable that you can purchase, so making your own cable will necessitate using a bare USB-C connector that has solder tabs on it.
-------------------

Things became interesting when the E360 was used without the GPS unit attached. Like the first time, I started with +12v on both the red and yellow wires. I found no difference in voltages on the USB-C pins as a result of connecting/disconnecting the +12v yellow ACC wire, suggesting that drive/parking mode switching is controlled entirely within the GPS unit:

A1: GND
A2: 0
A3: 0
A4: 5.03v
A5: 5.4v
A6: 0.8v
A7: 0.8v
A8: 5.4v
A9: 5.03
A10: 0.0
A11: 0.0
A12: GND

 

[Karagandinez]

Things became interesting when the E360 was used without the GPS unit attached. Like the first time, I started with +12v on both the red and yellow wires. I found no difference in voltages on the USB-C pins as a result of connecting/disconnecting the +12v yellow ACC wire, suggesting that drive/parking mode switching is controlled entirely within the GPS unit:

https://vantrue-app.vantruecam.com/files/manuals/Vantrue-E360-User-Manual-English.pdf

9. Rear Camera Port
10. Power Port