Mini 0803 vs A118 / Day & Night Footage comparison

[chris-s]

Hi all !

I've ordered first a 0803 for front and then an A118 for rear view.

As the GPS module for A118 will come soon, I want to compare both, and see if 0803 is as good as A118, day and night.

Will my 0803+GPS be replaced by an A118+GPS ?

All videos are uploaded to my Google Drive, you can download the real footage if you want to compare without video compression.

You will first notice that A118 has more dashboard reflections.
It is because the A118 is closer to the windscreen, compared to 0803.
At the top-center you can see some white A4 papers that I have placed on my dashbord.

Day test :

0803 : https://drive.google.com/file/d/0Bxdsi3_riDEkcFdOcnNUZkRGU2M/edit?usp=sharing
A118 : https://drive.google.com/file/d/0Bxdsi3_riDEkVnl3eEppU1lTRDA/edit?usp=sharing

0803 @ 60 km/h

A118 @ 60 km/h

0803 @ 110 km/h

A118 @ 110 km/h

Night test :

As I don't drive at night in this season, I've recorded a footage into a 2 tunnels.
First is short, the second is longer.

0803 : https://drive.google.com/file/d/0Bxdsi3_riDEkY2MwMW50VVJzcms/edit?usp=sharing
A118 : https://drive.google.com/file/d/0Bxdsi3_riDEkMlRyTnhjYW41WGM/edit?usp=sharing

As you can see, white blance is totally different, A118 shows the real yellow light in the tunnel but not the 0803.
0803 AWB setting is set to AUTO, EV setting set to +1.0 (my day setting)
A118 has standards settings.

The A118 provides a more sharp and detailed picture compare to 0803.
I don't know why @jokiin says that 0803 is better at night.
IMO, A118 is better.

Pictures extracted from this footage :
(click on picture to see original size)

0803 Tunnel 1

A118 Tunnel 1

0803 Tunnel 2

A118 Tunnel 2

Real night footage :

A118 & 0803 with standards settings (EV+0)

Bad conditions : was raining a bit...

0803 : https://drive.google.com/file/d/0Bxdsi3_riDEkblBhUlE2bjRkLU0/edit?usp=sharing
A118 : https://drive.google.com/file/d/0Bxdsi3_riDEkM0REMFJLQzlxbnc/edit?usp=sharing

First impression is that 0803 footage is brighter but does not seems to be real, like overexposed.
But second impression is that licence plates reading is much better on 0803 than on A118 by night !
I'm really surprised !
Was I lucky ? This was the only 2 cars I've recorded so close during my test footage.

On picture 1, we can read the licence plate of the left car with 0803 but not with A118.
On picture 2, we can read the licence plate of the right car with 0803 but not with A118.

0803 picture 1

A118 picture 1

0803 picture 2

A118 picture 2

 

[niko]

Nice test.
0803 is way better at night for sure, but this can not be seen from your tests. Need to drive in other more extreme night conditions with less street lights, also in pitch dark.

I have done over dozen extencive side by side tests for the past 2 month between B40, Mobius B lens, 0803, Panorama and can say that 0803 beats Mobius B lens and B40 / A118 at night. Also a 0803 is about 90% as good at night as Panorama 2.
I will share my tests soon with detailed comparison. Its all takes a lot of time to put together.
Other than that I am very happy with overall functionality and performance vs price. Great dashcam !

 

[chris-s]

OK I will try to make a test drive this evening at full night.

 

[niko]

OK I will try to make a test drive this evening at full night.

Try driving outside city ( coutry side ) or in places with no or minimum street lights.
For fair comparison need to set all dashcams settings to equal values ( factory default ).

 

[chris-s]

I'll go back to EV+0 on the 0803 for this night test.
I live in country

 

[sludgeguts]

Something that troubles me, it all seems so subjective. One says 'IMO, a is better than b' whilst another says 'b is better than a'
If we have two cams set to default and running alongside one another, is there any actual science that can pick out the differences?

 

[chris-s]

That's why I have uploaded original files.
So, everybody can have its own opinion

 

[niko]

Something that troubles me, it all seems so subjective. One says 'IMO, a is better than b' whilst another says 'b is better than a'
If we have two cams set to default and running alongside one another, is there any actual science that can pick out the differences?

only once to high-jack this thread in order to replay you , - as per sensitivity at night, then Mobius B-lens is better than A-lens and it can be seen only from side by side comparison tests.

v0.57 Mobius A lens vs B lens vs DOD LS300W Rainy Night

Albatross II vs Mobius A & B lens vs DOD LS300W Rainy Night

 

[maccardiolog]

I agree with chris-s - all you have presented above is in favor of A118 obviously ! It was a side by side comparison ! We'll see the results of the ,,dark,, night footage , but at the PRESENTED test the evidence is one and doubtless ! Sincerely , I am dissapointed about other opinion than that , particulary from the important members of the group forum !

 

[chris-s]

only once to high-jack this thread in order to replay you , - as per sensitivity at night, then Mobius B-lens is better than A-lens and it can be seen only from side by side comparison tests.

I don't think that he was talking about lens A vs lens B from Mobius.
I think he was talking generally about "cam a" and "cam b", or "cam x" and "cam y".

 

[niko]

I don't think that he was talking about lens A vs lens B from Mobius.
I think he was talking generally about "cam a" and "cam b", or "cam x" and "cam y".

That also make sense.

 

[niko]

I agree with chris-s - all you have presented above is in favor of A118 obviously ! It was a side by side comparison ! We'll see the results of the ,,dark,, night footage , but at the PRESENTED test the evidence is one and doubtless ! Sincerely , I am dissapointed about other opinion than that , particulary from the important members of the group forum !

Yes, it was side by side comparison, but due to limited scenario use and not equally fair settings, - this test is not fully valid and don't show full performance value of dashcameras in extreme environment, where true potential of each dash-camera can be put to the test.
I am here not to argue, but to give some tips on how to do it write based on my many years experience, 100-ds of dashcams and 100-ds of side by side tests back my shoulders.
Non of us is born as an expert, some knows more, some less ( I also dont know yet many things on how to hack FW, etc ). We all here to help each other to learn and improve knowledge and to share our experience

 

[sludgeguts]

Yes, I was talking about cam x Vs cam y.
TBH, my eyesight's so shot due to years of self-pleasure that the difference, to me, seems negligible.

 

[niko]

my eyesight's so shot due to years of self-pleasure that the difference, to me, seems negligible.

don't worry, give me 1 day, I will process one of my night tests and will "fix" your eyesight

 

[jokiin]

I didn't state that the 0803 was currently better at night, the picture is lighter and may look better during playback, paused though it loses detail, potentially though the hardware is capable of much better results, whether or not the potential is realised is another matter

 

[chris-s]

Added first results of real night test footage at the end of first post.

 

[FourEyes]

Where's the bottleneck in case of 0803? In the A7 SDK?

 

[jokiin]

Where's the bottleneck in case of 0803? In the A7 SDK?

There are definitely problems in the SDK, in the case of the 0803 some of what needs fixing is outside the capabilities of the engineers so they've had to turn to Ambarella for the fixes which is a slow process unfortunately

 

[FourEyes]

Thanks @jokiin

 

[niko]

@chris-s

I have analyzed "tunnel videos". There is a 1 sec jump-back on B40 / A118 video at 8:22:42.

A118 : https://drive.google.com/file/d/0Bxdsi3_riDEkMlRyTnhjYW41WGM/edit?usp=sharing

Is this video file you shared is original, from one continuous loop recording video clip file or its stitched together with Registrator Viewer or other programme ?