Viofo A129 Duo duplicating/jumping frames on rear camera/Firmware Bug Discovered

You seem quick to jump to conclusions and make everything sound easy.

The example I posted was from a pre-production camera running pre-production firmware. That same firmware might have duplicate frames on a retail A129. Alternatively, the current firmware might have no duplicate frames on that pre-production hardware. I might test both in time. Hopefully my observations will help @viofo engineers to address the issue.

I believe earlier in this thread, we proved the front and rear cameras are recording at slightly different frame rates. Which after so many frames, relates to the duplicate frame being created.

If the hardware was underpowered, we'd be experiencing dropped frames. Meaning losses in footage. This seems not to be the case.
 
I believe earlier in this thread, we proved the front and rear cameras are recording at slightly different frame rates.
No, that was just BCHobbyist's opinion
 
No, that was just BCHobbyist's opinion

Based on my experience with hardware, if the camera was unable to handle the resolution and bitrate, we'd be dropping frames. So this clearly isn't an issue of the hardware being incapable. More likely, a software bug introduced somewhere along the line. Unless of course, there's a design flaw within the cable connecting the front and rear camera, which doesn't allow the proper amount of data to be passed through the cable. But I would suspect this would cause more inconsistencies than a duplicate frame every 90 frames.
 
I have located an old beta firmware for the A129, V1.11 from May 2018. I ran this on my A129 camera this morning and found no duplicate frames on the rear camera.
Note that I use my A129 as a side-facing dashcam, so the rear camera is looking to the left.

Original file : A129 Rear Camera - FW V1.11

If anyone else wants to test this FW on their camera it will be available HERE for a short duration.
IMPORTANT: This is beta FW for testing only. Please do not use it for regular driving.
EDIT : link to V1.11 now removed

I do not know what changed between V1.11 and V1.2 that introduced the duplicate frames. I will leave this with Viofo to address in their own time. I'm restoring my A129 to FW V1.9.
 
Last edited:
Thank you so much,I have tested downgrading till v1.2 and lose the hope,but you found the one which not contains the bug..Hope @viofo will take a look to this version to see the difference,this will help them a lot.. (y)
LE: I can see with naked eyes without testing that there are no dupes,smooth as butter..:love:
 
I have located an old beta firmware for the A129, V1.11 from May 2018. I ran this on my A129 camera this morning and found no duplicate frames on the rear camera.
Note that I use my A129 as a side-facing dashcam, so the rear camera is looking to the left.

Original file : A129 Rear Camera - FW V1.11

If anyone else wants to test this FW on their camera it will be available HERE for a short duration.
IMPORTANT: This is beta FW for testing only. Please do not use it for regular driving.

I do not know what changed between V1.11 and V1.2 that introduced the duplicate frames. I will leave this with Viofo to address in their own time. I'm restoring my A129 to FW V1.9.

So we were right....This is a bug introduced in the firmware and NOT a hardware problem =0).....Naysayers go silent I'm sure....

So the trick now is to locate WHAT changed between firmware version 1.11 --> firmware where it first happened.
 
Last edited:
Do you know the first flawed Firmware Version @TonyM?
I don't have any other FW versions after V1.11, until Viofo posted V1.2 on July 26 2018. And most importantly, I have no idea what changed between FW versions that made the difference.

I think I've done enough to demonstrate that duplicate frames were not an issue on early FW. Only Viofo know what has changed since then, so I'm leaving them to look at it. I have more important things to do.
 
I don't have any other FW versions after V1.11, until Viofo posted V1.2 on July 26 2018. And most importantly, I have no idea what changed between FW versions that made the difference.

I think I've done enough to demonstrate that duplicate frames were not an issue on early FW. Only Viofo know what has changed since then, so I'm leaving them to look at it. I have more important things to do.

You're efforts are greatly appreciated. You've narrowed the problem down to an exact timeframe before this issue occurred. And for that, much thanks!
 
I have downgraded today to v1.11 and abused my mouse again testing frame by frame,I can confirm there is no dupe in this version,front camera as well don't suffer jumping or dupe frames anymore.I have chosen this area because of the palm trees,is a difficult area for the chipset to process,anyway it passed the test perfectly fine.I will stay with this version,till viofo will receive the updated SDK from the chipset vendor and launches a update to fix this bug.
Download to check the RAW files:


 

Attachments

  • 20190929_170002.jpg
    20190929_170002.jpg
    116.6 KB · Views: 13
Last edited:
I have downgraded today to v1.11 and abused my mouse again testing frame by frame,I can confirm there is no dupe in this version,front camera as well don't suffer jumping or dupe frames anymore.I have chosen this area because of the palm trees,is a difficult area for the chipset to process,anyway it passed the test perfectly fine.I will stay with this version,till viofo will receive the updated SDK from the chipset vendor and launches a update to fix this bug.
Download to check the RAW files:



Are there any drawbacks to 1.11? Does it support parking mode? Low Bitrate Recording? Hardware Kit?
 
I think is only the buffered parking mode missing,I will have to check if other options are missing.For the moment I just wanted to test the dupe issues on the rear camera,and dupe&jumping frames on the front one..My eyes are so relaxed now when I watch the footage on v1.11..:)

LE:The only thing I will miss with this version will be the @BCHobbyist's higher bitrate mods..
 
Last edited:
I will stay with this version,till viofo will receive the updated SDK from the chipset vendor and launches a update to fix this bug.

Are there any drawbacks to 1.11? Does it support parking mode? Low Bitrate Recording? Hardware Kit?
I'd like to remind you that I posted V1.11 only for the purpose of testing the duplicate frame bug.

It was a pre-release firmware. I think the parking mode options are basic.

Now that testing of this version has done its job I propose to remove the file from my earlier link.
 
Agree with you @TonyM In my particular case I will not use the parking mode on the A129Duo,as soon as my A129 Pro Duo arrives,the old one goes on my wife's car.Anyway I'm almost not using parking mode on my car neither,both cars are sleeping in garage..Viofo said on other post:for those who are bothered by the problem of duplicate frames,just use V1.1 for the moment,so is fine for me till the update will fix the problem.
 
I'd like to remind you that I posted V1.11 only for the purpose of testing the duplicate frame bug.

It was a pre-release firmware. I think the parking mode options are basic.

Now that testing of this version has done its job I propose to remove the file from my earlier link.

When you say parking mode, are you referring to buffered parking mode with the 30 secs before an accident? Or are you talking about parking mode with the hardwire kit when camera records in low bitrate when the vehicle is turned off?
 
@TonyM @Nigel

Will the 4K Pro work on the same 3 wire hardwiring as the A129 Duo? Or will it require a complete new hardwire kit? Can the Front to Rear USB cable be re-utilized or does it require an upgraded cable?
 
@TonyM @Nigel

Will the 4K Pro work on the same 3 wire hardwiring as the A129 Duo? Or will it require a complete new hardwire kit? Can the Front to Rear USB cable be re-utilized or does it require an upgraded cable?
Same HK3 hardwire kit (EDIT : latest version) . Same front-rear cable.
 
Last edited:
When you say parking mode, are you referring to buffered parking mode with the 30 secs before an accident? Or are you talking about parking mode with the hardwire kit when camera records in low bitrate when the vehicle is turned off?
I did not have a hardwire kit when I tested V1.11.

I think the parking mode was simply a 30s recording triggered by motion detection, with no buffering. I recall the transition from normal to parking mode and back was not perfect.
 
We have updated the hardwire kit cable to compatible with A129 PRO some month ago.
You can test it with A129 PRO first, if that one can not work, can upgrade to the latest version.
 
We have updated the hardwire kit cable to compatible with A129 PRO some month ago.
You can test it with A129 PRO first, if that one can not work, can upgrade to the latest version.
Yes, I should have been clear that the latest version of the HK3 is compatible with both A129 and A129 Pro.
 
Latest version as in having no high pitched hum in the recordings?

How do you tell if it's the latest version, is there a version number on the label somewhere?

Sent from my SM-G930V using Tapatalk
 
Back
Top