A229 Plus Test & Review PP

[Nigel]

That Mini 2 looks stellar! How come the Plus seems quite far off?

We need some motion blur in the image so that we can see if the Mini 2 is using twice the exposure time, which is fine when there is no motion, but...

 

[Panzer Platform]

That Mini 2 looks stellar! How come the Plus seems quite far off?

Stay tuned.

 

[Panzer Platform]

HDR COMPARISON
This stationary HDR test is NOT the final judge of image quality.
Installing the dash cams in a moving car trying to resolve moving objects at speed is needed, and still requires independent peer review to confirm & verify.
I won’t be gathering “driving” test footage until we get an “acceptable” firmware fixing certain issues.

I’ve been testing / using the A139 Pro since November 2022, (11 months).
I’ve been testing / using the A119 Mini 2 since May 2023, (5 months).
The A139 Pro has the best overall DAYTIME image quality of any dash cam currently on the market.
The A119 Mini 2 has the best overall NIGHT TIME & LOW LIGHT image quality of any dash cam currently on the market.
If Viofo could somehow combine the A139 Pro, and A119 Mini 2 they would have the best image quality dash cam on the market DAYTIME & NIGHT TIME bar none.
The A229 Pro has the potential to be that camera, but it’s not quite there yet.
It has the hardware, but the software specifically the HDR tuning needs to be perfected to the same level of polish as the A119 Mini 2.

On with the games.
For this HDR comparison I place the camera test rig on top of my car 5ft high in the center of my 20ft x 20ft two car garage.
The test starts with a single light source illuminating the garage.
I have a Streamlight Stinger LED flashlight hanging from the ceiling with a diffuser 5ft behind the test rig in an attempt to evenly distribute the beam pattern.
The flashlight has 3 power modes HIGH/MED/LOW.
I use the LOW setting rated at 100 lumen for this test.

Stinger® LED | Handheld Flashlight | Streamlight®

I let the cameras adjust / stabilize to the very dim conditions for approximately 10 seconds.
I turn on a second flashlight with an extremely tight focused beam pattern to maximize the intense difference between the darkest, and brightest part of the image.
The second flashlight is a MAGLITE ML300L.
The flashlight has 3 power modes HIGH/MED/LOW.
I use the HIGH setting rated at 625 lumen for this test.
I achieved the tightly focused beam pattern by constructing a telescopic shroud from USPS Assembly Grade Cardboard, (see attached photo).

ML300L LED 3-Cell D Flashlight - Chipboard Box

Whether you're a professional using a flashlight for work or security, an outdoor enthusiast who takes it camping, or an occasional user wanting it for emergencies at home or on the road, our New ML300L D-cell LED is The Light For All Your Needs. With Eco mode, up to 298 Hours With the familiar...

maglite.com

Cameras on the rig;
A229 Pro 3-CH ..(IMX678) + (IMX675) + (IMX307)
A229 Plus 3-CH (IMX675) + (IMX675) + (IMX307)
A139 Pro 1-CH ..(IMX678)
A119 Mini 2 ……(IMX675)

Camera Settings;
Bitrate: …..Maximum
HDR: ……..On
IR LEDs: …Off
CPL Filter: Off

Camera Firmware;
A229 Pro Front: ..V1.0 0925
A229 Pro Rear: …V1.0 0901
A229 Plus Front: .V1.0 0922
A229 Plus Rear: ..V1.0 0901
A139 Pro: ………..V1.1 0629
A119 Mini 2: ……..V1.0 0912

After watching the recorded footage, and comparing screenshots side by side, 5 things stand out at me.
1.) The A229 Pro, and A139 Pro have similar HDR performance resolving the license plate.
Take note of the difference in size / shape / density of the beam pattern.
It’s almost like the cameras achieving the same goal but going about it in different ways.
Is this the difference in processors, (A229 Pro 96529 & A139 Pro 96687)?
Or a difference in firmware programming?
1a.) Everything not under direct illumination appears sharp, and focused on the A229 Pro, where as the A139 Pro looks blurry, an unfocused.
Take note of the 5 misspelled “VIOFO” paper signs.
Can you identify how each one is misspelled?
2.) The A229 Plus has a much brighter overall image than the A229 Pro.
However, the HDR performance is ineffective, and unacceptable trying to resolve the license plate.
This is a major malfunction that requires corrective action.
3.) The A229 Pro & A119 Mini 2 have similar brightness, and clarity of the overall image.
However, the A229 Pro is no match for the A119 Mini 2’s HDR performance at resolving the license plate.
Take note “California” in red lettering is missing from the A229 Pro, among the overexposure.
4.) The Rear cameras of the A229 Pro & A229 Plus have good overall brightness of the image.
However the HDR performance trying to resolve the license plate is absolutely non-functional.
This is a major malfunction that requires corrective action.
5.) The Interior cameras of the A229 Pro & A229 Plus have a dark overall appearance of the image.
However, the HDR performance resolving the license plate is far superior than the Front & Rear cameras of the A229 Pro & A229 Plus.
How does the “lowly” STARVIS 1 (IMX307) have better HDR performance than the STARVIS 2 (IMX678) & (IMX675) in the A229 Pro & A229 Plus?
Take note in the video footage when I turn on the flashlight.
It takes the Interior camera 3 seconds to adjust / react to the change in lighting conditions, and choose / select the appropriate exposure.
This is unacceptable, because all other Viofo HDR cameras accomplish this task in 1 second.
This requires corrective action.










Here is the test footage I took the screenshots from.

 

[Panzer Platform]

HDR COMPARISON - CHEVRON SIGN
This stationary HDR test is NOT the final judge of image quality.
Installing the dash cams in a moving car trying to resolve moving objects at speed is needed, and still requires independent peer review to confirm & verify.
I won’t be gathering “driving” test footage until we get an “acceptable” public firmware fixing certain issues.
I’ve found this sign to be helpful in gathering baseline HDR performance due to the individual digits are made up of intensely bright LED lights.
Also the background, and surrounding area is very dark.

Cameras on the rig;
A229 Pro 3-CH ..(IMX678) + (IMX675) + (IMX307)
A229 Plus 3-CH (IMX675) + (IMX675) + (IMX307)
A139 Pro 1-CH ..(IMX678)
A119 Mini 2 ……(IMX675)

Camera Settings;
Bitrate: …..Maximum
HDR: ……..On
IR LEDs: …Off
CPL Filter: Off

Camera Firmware;
A229 Pro Front: ..V1.0 0925
A229 Pro Rear: …V1.0 0901
A229 Plus Front: .V1.0 0922
A229 Plus Rear: ..V1.0 0901
A139 Pro: ………..V1.1 0629
A119 Mini 2: ……..V1.0 0912

After watching the recorded footage, and comparing screenshots side by side, 5 things stand out at me.
1.) The A229 Pro Front & A139 Pro 1-CH have similar HDR performance resolving the digits of the price sign.
However, take notice of the bush & tree.
This is the improvement in low light sensitivity of the A229 Pro I’ve been talking about.
Once the A229 Pro’s HDR has been fined tuned, and we get an “acceptable” public firmware it should beat the A139 Pro’s image quality in every aspect.
2.) The A229 Plus HDR performance trying to resolve the individual digits on the price sign is ineffective, and unacceptable.
This is a major malfunction that requires corrective action.
However, take notice of the bush & tree.
The A229 Plus has even better low light sensitivity than the A229 Pro.
If we can get the HDR fine tuned, and keep some of this “brightness” the A229 Pro will be t*ts.
Another interesting note: The narrower FOV of the A229 Plus makes the license plates in the image appear larger, and easier to read, (more on this later in the next post).
3.) The A229 Pro Front is no match for the mighty Mini 2.
Not in HDR performance resolving digits on the sign, or in low light sensitivity, and overall brightness of the image especially as shown in the fine detail of the bush & tree, and even the weeds behind the chain link fence.
4.) The A229 Pro & A229 Plus Rear camera’s HDR performance trying to resolve the digits on the sign is ineffective, and unacceptable.
This is a major malfunction that requires corrective action.
However, the overall brightness / exposure looks good.
5.) The A229 Pro & A229 Plus Interior camera’s have far superior HDR performance resolving the digits of the sign over the Front & Rear cameras.
It’s been explained to me by Alex (Viofo) I should not compare the Interior camera with the Front & Rear because the Interior has been designed / engineered with a focal length of 1 meter appropriate for passenger compartments.
I agree, but at the same time I’m amazed how well the HDR works.
I’ll be comparing the Interior camera with the A139 Pro’s Interior camera in a future post.
Q: Why do we need HDR on the Interior camera anyways?
A: I’ve performed preliminary testing comparing the Interior camera with the Interior camera of the A139 Pro, and the HDR does an excellent job when my dome light is illuminated.
I can see how this will be beneficial for professional ride share applications because they have regular dome light illumination from passenger ingress / egress.
Another interesting note: Take notice of the license plates in the image, the “low” 1080p resolution really shows compared to 4K & 2K.
I really wish Viofo would have used the (IMX335) for the Interior camera or better yet another (IMX675).










Here is the test footage I took the screenshots from.

A229 Pro Front;

A139 Pro;

A229 Plus Front;

A119 Mini 2;

A229 Pro Rear;

A229 Plus Rear;

A229 Pro Interior;

A229 Plus Interior;

 

[Panzer Platform]

FOV COMPARISON - CONFIRM & VERIFY
In this post I would like to call attention to the false & misleading FOV specifications of the A229 Pro & A229 Plus.
Q: What is FOV, and why is it important?
A: FOV or technically AOV, or “viewing angle” is a specification consumers use to make an informed decision.
Narrow FOV’s are usually around 120°-140°.
Wider FOV’s are usually 170° or greater.
Narrow FOV’s make the center of the image appear larger, and it’s easier to see fine details like license plates.
Wider FOV’s provide overall more information in a given scene, however objects appear smaller, and warpage / barrel distortion is greater on the edges of the image.
I prefer 120°-140° FOV because it’s easier to make out fine details like license plates, and anything missed outside of view can be captured by auxiliary left & right cameras.
Because, why not add more cameras to cover all blind spots? lol

Both the A229 Pro & A229 Plus 3-CH have listed specifications of;
140° Front + 160° Rear + 150° Interior Viewing Angle.
Based on these screenshots the true FOV is closer to;
A229 Pro: 160° Front + 160° Rear + 170° Interior
A229 Plus: 145° Front + 160° Rear + 170° Interior
It’s been explained to me the FOV specifications are provided by the lens manufacturer to Viofo.
If Viofo fails to confirm & verify actual FOV of their dash cams after final assembly this is a deficiency of minimum standards for quality control in my opinion.
The name on the dash cam is Viofo, and they are ultimately responsible for providing true & correct specifications to consumers.












Here’s the test footage I took the screenshots from;

A229 Pro Front;

A139 Pro 1-CH;

A229 Plus Front;

A119 Mini 2;

A229 Pro Rear;

A229 Plus Rear;

A229 Pro Interior;

A229 Plus Interior;

 

[JardaB]

HDR COMPARISON - CHEVRON SIGN
This stationary HDR test is NOT the final judge of image quality.
Installing the dash cams in a moving car trying to resolve moving objects at speed is needed, and still requires independent peer review to confirm & verify.
I won’t be gathering “driving” test footage until we get an “acceptable” public firmware fixing certain issues.
I’ve found this sign to be helpful in gathering baseline HDR performance due to the individual digits are made up of intensely bright LED lights.
Also the background, and surrounding area is very dark.

Cameras on the rig;
A229 Pro 3-CH ..(IMX678) + (IMX675) + (IMX307)
A229 Plus 3-CH (IMX675) + (IMX675) + (IMX307)
A139 Pro 1-CH ..(IMX678)
A119 Mini 2 ……(IMX675)

Camera Settings;
Bitrate: …..Maximum
HDR: ……..On
IR LEDs: …Off
CPL Filter: Off

Camera Firmware;
A229 Pro Front: ..V1.0 0925
A229 Pro Rear: …V1.0 0901
A229 Plus Front: .V1.0 0922
A229 Plus Rear: ..V1.0 0901
A139 Pro: ………..V1.1 0629
A119 Mini 2: ……..V1.0 0912

After watching the recorded footage, and comparing screenshots side by side, 5 things stand out at me.
1.) The A229 Pro Front & A139 Pro 1-CH have similar HDR performance resolving the digits of the price sign.
However, take notice of the bush & tree.
This is the improvement in low light sensitivity of the A229 Pro I’ve been talking about.
Once the A229 Pro’s HDR has been fined tuned, and we get an “acceptable” public firmware it should beat the A139 Pro’s image quality in every aspect.
2.) The A229 Plus HDR performance trying to resolve the individual digits on the price sign is ineffective, and unacceptable.
This is a major malfunction that requires corrective action.
However, take notice of the bush & tree.
The A229 Plus has even better low light sensitivity than the A229 Pro.
If we can get the HDR fine tuned, and keep some of this “brightness” the A229 Pro will be t*ts.
Another interesting note: The narrower FOV of the A229 Plus makes the license plates in the image appear larger, and easier to read, (more on this later in the next post).
3.) The A229 Pro Front is no match for the mighty Mini 2.
Not in HDR performance resolving digits on the sign, or in low light sensitivity, and overall brightness of the image especially as shown in the fine detail of the bush & tree, and even the weeds behind the chain link fence.
4.) The A229 Pro & A229 Plus Rear camera’s HDR performance trying to resolve the digits on the sign is ineffective, and unacceptable.
This is a major malfunction that requires corrective action.
However, the overall brightness / exposure looks good.
5.) The A229 Pro & A229 Plus Interior camera’s have far superior HDR performance resolving the digits of the sign over the Front & Rear cameras.
It’s been explained to me by Alex (Viofo) I should not compare the Interior camera with the Front & Rear because the Interior has been designed / engineered with a focal length of 1 meter appropriate for passenger compartments.
I agree, but at the same time I’m amazed how well the HDR works.
I’ll be comparing the Interior camera with the A139 Pro’s Interior camera in a future post.
Q: Why do we need HDR on the Interior camera anyways?
A: I’ve performed preliminary testing comparing the Interior camera with the Interior camera of the A139 Pro, and the HDR does an excellent job when my dome light is illuminated.
I can see how this will be beneficial for professional ride share applications because they have regular dome light illumination from passenger ingress / egress.
Another interesting note: Take notice of the license plates in the image, the “low” 1080p resolution really shows compared to 4K & 2K.
I really wish Viofo would have used the (IMX335) for the Interior camera or better yet another (IMX675).

View attachment 68109
View attachment 68110
View attachment 68111
View attachment 68112
View attachment 68113
View attachment 68114
View attachment 68115
View attachment 68116

Here is the test footage I took the screenshots from.

A229 Pro Front;

A139 Pro;

A229 Plus Front;

A119 Mini 2;

A229 Pro Rear;

A229 Plus Rear;

A229 Pro Interior;

A229 Plus Interior;

1 - Youtube images are not telling, the Youtube system degrades the image and video.
2 - If the compared images are not in the same resolution (2k, 4k) - again it is not comparable.
3 - the basis of comparison should be cutouts of parts of the original images in the default resolution size view 100%.

 

[Dashmellow]

Both the A229 Pro & A229 Plus 3-CH have listed specifications of;
140° Front + 160° Rear + 150° Interior Viewing Angle.
Based on these screenshots the true FOV is closer to;
A229 Pro: 160° Front + 160° Rear + 170° Interior
A229 Plus: 145° Front + 160° Rear + 170° Interior

PP, Please explain exactly how you are measuring the angles you claim based on your screen shots.

And please don't tell me you are "eyeballing" it.

 

[Panzer Platform]

Youtube system degrades the image and video

comparison should be cutouts of parts of the original images

Correct.
All test footage was uploaded to YouTube from the raw original files by inserting the SD Card into my computer.
You can confirm by adjusting the resolution to the highest quality, (2160p, 1440, 1080p, etc.) in your YouTube player.
So all test footage has been degraded “equally”.
As far as comparing these test clips to each other, it’s a “fair fight”.
Same for the screenshots, they were taken from the raw original files.
These test clips, and screenshots can only be compared with each other, not anyone else’s test footage.

 

[JardaB]

Correct.
All test footage was uploaded to YouTube from the raw original files by inserting the SD Card into my computer.
You can confirm by adjusting the resolution to the highest quality, (2160p, 1440, 1080p, etc.) in your YouTube player.
So all test footage has been degraded “equally”.
As far as comparing these test clips to each other, it’s a “fair fight”.
Same for the screenshots, they were taken from the raw original files.
These test clips, and screenshots can only be compared with each other, not anyone else’s test footage.

So you expect me to have a 4k monitor and play with it at home alone to get an exat comparison ? 4k video on a 1920/1080 monitor is not 100% recorded quality. If I were doing some tests to compare image quality and submitting them to others, I would take a little extra care to cast doubt on ad1 - about degradation on Youtube servers or ad2 - about comparing 2k and 4k footage on a full HD monitor.

Try playing your original video and the same one uploaded to Youtube ... for me, it's always an absolutely tragic surprise. For regular videos it's not that bad, but if you want to compare the detail and legibility of the registration letters and numbers it makes a big difference.

 

[Panzer Platform]

Try playing your original video and the same one uploaded to Youtube ... for me, it's always an absolutely tragic surprise. For regular videos it's not that bad, but if you want to compare the detail and legibility of the registration letters and numbers it makes a big difference.

Correct.
I agree with you 100%.
This comparison is only valid for these screenshots & clips.

 

[Nigel]

Correct.
I agree with you 100%.
This comparison is only valid for these screenshots & clips.

It would be good if you could post your 4K screenshots in 4K resolution; at the moment the 4K front doesn't look significantly better than the FHD interior!
It is a website problem with dashcamtalk, so unless you can get dashcamman to allow 4K image pastes, you would have to host them elsewhere and link them in here, which would take rather more effort and a little computing expertise, so if you don't do it, I will not be complaining even if others do.

 

[VVondervvall]

If Viofo fails to confirm & verify actual FOV of their dash cams after final assembly this is a deficiency of minimum standards for quality control in my opinion.
The name on the dash cam is Viofo, and they are ultimately responsible for providing true & correct specifications to consumers.

Sound like Viofo are foremost an electronics company and not into optics.

Optics which can be tricky, the same lens has a different FOV depending on sensor size. I wonder if they botch up the f-number too, as it depends/is linked to focal length and what kind of equivalence or not one considers. Even big phone manufacturers can get confused in their marketing on these things so not really surprised.

I would prefer the plus for recording less unnecessary stuff on the sides, likelyhood of being able to rear plates when on the move reduces much there anyway and I will have side coverage anyway where needed....

 

[JardaB]

Correct.
I agree with you 100%.
This comparison is only valid for these screenshots & clips.

Discussions on this forum discuss in detail the quality of different cameras in different conditions. Sometimes they try to be smarter than the manufacturers of the best quality dashcam .

Everyone knows there are limits with reading a license plate correctly in high speed conditions of extreme scene contrast, high brightness when illuminated by a car headlight, or lack of light.

Any such comparison by the user carries some weight and can point to the importance of the size of the area of a single sensor cell or the quality or lack of firmware for processing the image into jpg format. However, it only applies to the conditions the tester used - lighting conditions, static or moving shot, method of transmitting test results, where the youtube channel is only conditionally suitable. I just wanted to suggest that an exact comparison requires initial conditions especially so as not to be affected by YouTube degradation (comparing detail cutouts from video frames using static cutouts from a video frame in 100% view ),

Benchmarking costs a lot of effort and time and even money. I've also done some comparisons and hit the limit of HDR where the frame detail is doubled. Finding some optimal camcorder settings for different conditions would cost me a disproportionate amount of time and effort with the obvious uncertainty of arriving at an optimal result. That's why I choose max stream and auto HDR I know that I have a camera manufacturer on the cutting edge of software and HW development, I won't change it for a year (I had the previous one for almost 10 years :-( ), I will follow pro and amateur tests and firmware updates

We have to accept that every dascam model has some limits, there are proven models.

 

[JardaB]

It would be good if you could post your 4K screenshots in 4K resolution; at the moment the 4K front doesn't look significantly better than the FHD interior!
It is a website problem with dashcamtalk, so unless you can get dashcamman to allow 4K image pastes, you would have to host them elsewhere and link them in here, which would take rather more effort and a little computing expertise, so if you don't do it, I will not be complaining even if others do.

Cutouts of the same essential details from the video from the 100% view are enough. That is, a picture taken by VLC or another SW for video processing (darktable in my case) regardless of the resolution of my monitor and a detail cut from the picture in 4k or 2k resolution and a subsequent crop that is comparable to "any monitor".

Some videos on Youtube are edited so that the live traffic is running and it is frozen at that moment and a cutout of the license plate is inserted into it. This has a comparable value with minimal distortion. Personally, I would even prefer only clips from the original video as still images (comparative collages).

But that's just my "prince advice"...I won't do it Technically, yes, from the supplied material, but not from its acquisition from multiple types of dascams

 

[Nigel]

Sound like Viofo are foremost an electronics company and not into optics.

They don't make the lenses, the lenses are made for them, and like all other dashcam manufacturers they just quote the lens specifications from the lens manufacturer.

It is not even the correct diagonal angle. since the sensors are not square, some of them are 4:3 and some 16:9, so comparing a 4:3 sensor A119 with a 16:9 sensor A229 makes the lens angle measurements even less equivalent, except most people use the A119 in 16:9 mode anyway!

Optics which can be tricky, the same lens has a different FOV depending on sensor size. I wonder if they botch up the f-number too, as it depends/is linked to focal length and what kind of equivalence or not one considers. Even big phone manufacturers can get confused in their marketing on these things so not really surprised.

The FoV specified is that of the lens, so it is correct, but doesn't match what you see in the recorded image.
The f-number is also that of the fixed focal length, fixed aperture lens, and so will be correct, doesn't matter about the sensor used since it is the actual f-number, not an equivalence.

 

[Panzer Platform]

I wonder if they botch up the f-number too

Every specification must be confirmed & verified.
"Trust, but verify"
-Ronald Reagan

 

[Panzer Platform]

The FoV specified is that of the lens, so it is correct,

The person who buys the dash cam is not a customer of the Lens manufacturer.
They are a customer of Viofo.
Viofo is ultimately responsible.

If McDonald's sells me a Big Mac, and I get sick.
The cattle rancher is not responsible.
McDonald's is the one to take action on.

 

[Nigel]

Cutouts of the same essential details from the video from the 100% view are enough. That is, a picture taken by VLC or another SW for video processing (darktable in my case) regardless of the resolution of my monitor and a detail cut from the picture in 4k or 2k resolution and a subsequent crop that is comparable to "any monitor".

Some videos on Youtube are edited so that the live traffic is running and it is frozen at that moment and a cutout of the license plate is inserted into it. This has a comparable value with minimal distortion. Personally, I would even prefer only clips from the original video as still images (comparative collages).

But that's just my "prince advice"...I won't do it Technically, yes, from the supplied material, but not from its acquisition from multiple types of dascams

Wait for the Vortex Radar review if you want that!
I suspect PP does not have the ability or time to be posting 100% cutouts and edited video, he is doing a good and time consuming job already!
You could petition Dashcamman to allow PP to paste 4K framegrabs, would be a big improvement...

 

[Panzer Platform]

You could petition Dashcamman to allow PP to paste 4K framegrabs, would be a big improvement...

Please do not bother DashCamMan anymore. lol
He's already given me "extra" privileges by upping the file size for my screenshots.

 

[VVondervvall]

The FoV specified is that of the lens, so it is correct, but doesn't match what you see in the recorded image.

But FoV is a property of the combination of lens AND sensor size so that ship doesn't sail. Maybe the lens manufacturer spec it for a specific sensor size (but not clearly so?), which seems to be correct for 1/2.8" in this case?