A139 Pro Test & Review PP

[rcg530]

When I gathered my power consumption data, I also gathered the Mbps bitrate for the various recording modes and I placed that in the "Notes" section of my power consumption spreadsheets that I posted in my A139 Pro review thread.

I've consolidated the four camera configurations 3-Channels (Front+Interior+Rear), 2-Channels (Front+Interior), 2-Channels (Front+Rear) and 1-Channel into one spreadsheet showing the bitrate differences in each of those configurations. All of those tests were with the bitrate setting set to "Maximum". As the number of cameras (channels) decreases or if the video resolution is reduced, the higher the video bitrate becomes and thus the larger file sizes. The one exception being the reduction of video resolution from 3840x2160 to 3840x1600 does not result in a higher bitrate.

 

[Panzer Platform]

Right on Robert,
That is an awesome bitrate spreadsheet.
And I do remember reading your original post about the different bitrates.
But I was not sure if there had been any changes from the pre-production to production units, and the different firmware versions.
I think I remember Viofo changing the bitrates slightly with different firmware versions.

Wow! 58 Mbps in 1-Channel mode, that sounds like a lot.
I remember the A229 Duo being only 29 front 25 rear.
And the A119 Mini being 30 Mbps.
-Chuck

 

[rcg530]

Right on Robert,
That is an awesome bitrate spreadsheet.
And I do remember reading your original post about the different bitrates.
But I was not sure if there had been any changes from the pre-production to production units, and the different firmware versions.
I think I remember Viofo changing the bitrates slightly with different firmware versions.

Wow! 58 Mbps in 1-Channel mode, that sounds like a lot.
I remember the A229 Duo being only 29 front 25 rear.
And the A119 Mini being 30 Mbps.
-Chuck

Those bitrates are with the v1.0_1105 test firmware and using my preproduction unit. I’m still waiting for my two production units

 

[TonyM]

I'm getting 58-60Mbps in 1CH 4K mode on my retail unit. It seems to run at a similar temp to my regular A139 in 3CH mode, just going by a simple touch test.

 

[Panzer Platform]

I'm getting 58-60Mbps in 1CH 4K mode on my retail unit. It seems to run at a similar temp to my regular A139 in 3CH mode, just going by a simple touch test.

Right on Tony,
Thanks for the confirmation.
-Chuck

 

[Panzer Platform]

A139 Pro - IR Interior POV’s

I just installed the A139 Pro’s IR interior camera. The traditional mounting location of center & top of the windshield was presenting a challenge due to lack of space. So I decided to mount the IR interior camera at the top of the passenger side A-pillar, (US right side).

Ever since I saw @rcg530 review of the new Thinkware Multiplexer I really like the external side view camera’s POV. I wanted to see how close I could emulate that with the A139 Pro’s IR interior & rear cameras. So I started with the IR interior camera by mounting it on a standard ¼ ball mount so it can be easily swiveled outboard to capture some side view action.

So far I think I like it, and will do the same on the driver side with the rear camera.
I’ve attached some screenshots of the POV’s under the various lighting conditions in my garage.
1.) Garage Lights + Dome Light
2.) Dome Light Only
3.) Garage Lights + Dome Light + IR Enabled
4.) Dome Light + IR Enabled
5.) IR Enabled Only
6.) Rearward Side View
7.) RCG530 Thinkware Multiplexer
-Chuck

 

[kamkar]

58 mbit are not that wild, at least not compared with action cameras that will easy do 100 mbit

But when you have been used to see 18 - 20 mbit for 1080p cameras for a long time, it is something you notice.

 

[Lothar]

When I gathered my power consumption data, I also gathered the Mbps bitrate for the various recording modes and I placed that in the "Notes" section of my power consumption spreadsheets that I posted in my A139 Pro review thread.

I've consolidated the four camera configurations 3-Channels (Front+Interior+Rear), 2-Channels (Front+Interior), 2-Channels (Front+Rear) and 1-Channel into one spreadsheet showing the bitrate differences in each of those configurations. All of those tests were with the bitrate setting set to "Maximum". As the number of cameras (channels) decreases or if the video resolution is reduced, the higher the video bitrate becomes and thus the larger file sizes. The one exception being the reduction of video resolution from 3840x2560 to 3840x1600 does not result in a higher bitrate.

View attachment 62836

I still see that the bitrate discrepancy on 3840x2560 (44.86Mbps) vs 3840x1600 (28.06Mbps) still exists while running in 2CH mode...
The same thing happens whether one uses Front + Rear or Front + Interior combination.

If anything...One would think that it should be capable of running either the same (or higher bitrate) at 3840x1600 vs 3840x2560 resolution, but yet the opposite is happening.
It makes me question the utility and benefit of running in 1600p resolution on any other camera combination except 3CH mode.

 

[TonyM]

58 mbit are not that wild, at least not compared with action cameras that will easy do 100 mbit

But when you have been used to see 18 - 20 mbit for 1080p cameras for a long time, it is something you notice.

Of course the main difference is that a 100Mbps action camera probably is not attached to a windshield, recording for many hours at a time

 

[TonyM]

I still see that the bitrate discrepancy on 3840x2560 (44.86Mbps) vs 3840x1600 (28.06Mbps) still exists while running in 2CH mode...
The same thing happens whether one uses Front + Rear or Front + Interior combination.

If anything...One would think that it should be capable of running either the same (or higher bitrate) at 3840x1600 vs 3840x2560 resolution, but yet the opposite is happening.
It makes me question the utility and benefit of running in 1600p resolution on any other camera combination except 3CH mode.

I am running 3840x1600 in 1CH mode at 53Mbps and that seems plenty to me. It saves a little bit on memory card space and heat generation compared to 58Mbps.

Taking a quick look at the numbers, 1600/2560 = 0.625. So the 1600p video has 62.5% number of pixels compared to 3840x2560.

It looks like the Viofo engineers started with 44.8Mbps as the max bitrate they could run in combination with 16Mbps on the rear camera, then reduced the 1600p video bitrate in proportion to the change in pixel area: 0.625 x 44.8 = 28.0Mbps
Oops - I used the wrong pixel dimensions - 4K is 3840x2160, not 3840x2560.

I have tried the A139 Pro in 1CH, 2CH and 3CH configurations. Personally, I didn't like the effects of the reduction in bitrate, but then I am used to having 4K cameras at around 60Mbps.

 

[Nigel]

Taking a quick look at the numbers, 1600/2560 = 0.625. So the 1600p video has 62.5% number of pixels compared to 3840x2560.

It looks like the Viofo engineers started with 44.8Mbps as the max bitrate they could run in combination with 16Mbps on the rear camera, then reduced the 1600p video bitrate in proportion to the change in pixel area: 0.625 x 44.8 = 28.0Mbps

Have you made a mistake, or are you saying that the engineers made a mistake?

16:9 4K is 3840 x 2160,
21:9 4K is 3840 x 1600.

1600 / 2160 = 0.74

If anything...One would think that it should be capable of running either the same (or higher bitrate) at 3840x1600 vs 3840x2560 resolution, but yet the opposite is happening.
It makes me question the utility and benefit of running in 1600p resolution on any other camera combination except 3CH mode.

The 16:9 has 35% more pixels, so needs 35% more bitrate to achieve the same image quality.

The only reason to run the 1600 at the same or higher bitrate is to achieve significantly higher image quality, but the engineers were not necessarily choosing the bitrates for maximum quality. There is always a compromise between image quality and other things, such as temperature, reliability, hours of video per card, power consumption (especially if running on battery in parking mode), etc.

 

[CuriousGeorge]

In my quick (unprofessional) testing comparing the image quality between 2160 vs 1600, it seems that the 2160 image has more details and is easier to read plates that are off center (for example, when the car is in another lane to the left or right). I compared the exact same scene in daylight, no motion, full sun.

 

[TonyM]

Have you made a mistake, or are you saying that the engineers made a mistake?

16:9 4K is 3840 x 2160,
21:9 4K is 3840 x 1600.

1600 / 2160 = 0.74

I made a mistake by reading other people's posts above quoting 3840x2560 and using those numbers, rather than 3840x2160 which is the actual resolution!

 

[CuriousGeorge]

In my quick (unprofessional) testing comparing the image quality between 2160 vs 1600, it seems that the 2160 image has more details and is easier to read plates that are off center (for example, when the car is in another lane to the left or right). I compared the exact same scene in daylight, no motion, full sun.

2160 Plate Crop (Grey Corolla)








1600 Plate Crop (Grey Corolla)

 

[TonyM]

2160 Full Image
View attachment 62869

2160 Plate Crop (Grey Corolla)
View attachment 62870



1600 Full Image
View attachment 62867



1600 Plate Crop (Grey Corolla)
View attachment 62868

Those two crops look identical to me in terms of resolution. I'm more concerned about the difference in colour / exposure between the two.

What are the settings for each video? (Bitrate, HDR on/off?)

 

[Nigel]

Those two crops look identical to me in terms of resolution. I'm more concerned about the difference in colour / exposure between the two.

What are the settings for each video? (Bitrate, HDR on/off?)

It looks like the 2160 image has the chromatic aberration correction turned off, maybe the 35% more pixels resulted in too much work.

I think most people would prefer the 1600 version?

In my quick (unprofessional) testing comparing the image quality between 2160 vs 1600, it seems that the 2160 image has more details and is easier to read plates that are off center (for example, when the car is in another lane to the left or right). I compared the exact same scene in daylight, no motion, full sun.

Note that if you stop and start recording, you don't necessarily get back to the same exposure, and very often get a different white balance, so don't analyse them too accurately.

As always, I recommend you put the horizon at 50% height, that is what the engineers will have calibrated the auto-exposure to work best at. You really don't need to record all that dashboard, nothing of interest ever happens there!

 

[CuriousGeorge]

Those two crops look identical to me in terms of resolution. I'm more concerned about the difference in colour / exposure between the two.

What are the settings for each video? (Bitrate, HDR on/off?)

Both Images have the same settings. HDR is on, Bitrate is set to maximum. Full disclosure, I'm not a pro tester like many folks here, just wanted to share my anecdotal observation that there is a difference in the actual image (2160 vs 1600)

 

[CuriousGeorge]

It looks like the 2160 image has the chromatic aberration correction turned off, maybe the 35% more pixels resulted in too much work.

I think most people would prefer the 1600 version?



Note that if you stop and start recording, you don't necessarily get back to the same exposure, and very often get a different white balance, so don't analyse them too accurately.

As always, I recommend you put the horizon at 50% height, that is what the engineers will have calibrated the auto-exposure to work best at. You really don't need to record all that dashboard, nothing of interest ever happens there!

Thank you! I will aim it up a bit.

 

[TonyM]

Full disclosure, I'm not a pro tester like many folks here, just wanted to share my anecdotal observation that there is a difference in the actual image (2160 vs 1600)

Most of us are amateurs, and your observations are worth sharing

 

[rcg530]

I made a mistake by reading other people's posts above quoting 3840x2560 and using those numbers, rather than 3840x2160 which is the actual resolution!

Sorry TonyM. I typed in 3840x2560 when it should have been 3840x2160. I've corrected my typo in the text of my post. The resolutions in the spreadsheet were correct.