Thinkware U3000 - Testing / Review - RCG

[rcg530]

I have asked Thinkware for a U3000 to test/review, but all of the marketing units had already been sent out to reviewers. I thought my U3000 review was not going to happen. Then Ben @safedrivesolutions reached out to me and said he was willing to loan me a U3000 Dual channel and an IROAD QX2. I quickly said YES before Ben changed his mind. Ben, thank you very much for the loan of the two dash cameras. They should be arriving tomorrow.

To prepare for the Thinkware U3000 review, I decided to install the Thinkware iVolt Xtra BAB-95 battery pack that I purchased for my recent review of 7500 mAh dash camera battery packs. The main reason for the installation of the BAB-95 was to test the battery communications cable connection to the U3000.

The wiring for the BAB-95 is different than most other battery packs in that it requires a constant battery power connection to the input charging port and there is a green wire in the output connector for "ACC In" power. As the name suggests, the "ACC In" is connected to a switched/accessory power feed from the car. It is used to determine when to charge the battery pack and it's also passed along to the dash camera via the "ACC Out" wire in the output port of the battery pack.

Yesterday, I spent several hours removing panels and the rear seat from my car so I could run a 10 foot 12 AWG custom made charging cable and an 18 AWG accessory power wire from the trunk of my car to the installation location for the BAB-95. My car has its starting battery and a fuse panel located in the trunk. I'm running out of places to put a battery pack in my car, so the only place left was the floor under the front driver side seat. I used an inline (12 AWG wire) fuse holder at the battery to protect the charging cable power to the BAB-95.

So, I'm as ready as I can be for the U3000 testing.

Thanks again to Ben for loaning me the dash cameras.

 

[safedrivesolutions]

I have asked Thinkware for a U3000 to test/review, but all of the marketing units had already been sent out to reviewers. I thought my U3000 review was not going to happen. Then Ben @safedrivesolutions reached out to me and said he was willing to loan me a U3000 Dual channel and an IROAD QX2. I quickly said YES before Ben changed his mind. Ben, thank you very much for the loan of the two dash cameras. They should be arriving tomorrow.

To prepare for the Thinkware U3000 review, I decided to install the Thinkware iVolt Xtra BAB-95 battery pack that I purchased for my recent review of 7500 mAh dash camera battery packs. The main reason for the installation of the BAB-95 was to test the battery communications cable connection to the U3000.

The wiring for the BAB-95 is different than most other battery packs in that it requires a constant battery power connection to the input charging port and there is a green wire in the output connector for "ACC In" power. As the name suggests, the "ACC In" is connected to a switched/accessory power feed from the car. It is used to determine when to charge the battery pack and it's also passed along to the dash camera via the "ACC Out" wire in the output port of the battery pack.

Yesterday, I spent several hours removing panels and the rear seat from my car so I could run a 10 foot 12 AWG custom made charging cable and an 18 AWG accessory power wire from the trunk of my car to the installation location for the BAB-95. My car has its starting battery and a fuse panel located in the trunk. I'm running out of places to put a battery pack in my car, so the only place left was the floor under the front driver side seat. I used an inline (12 AWG wire) fuse holder at the battery to protect the charging cable power to the BAB-95.

So, I'm as ready as I can be for the U3000 testing.

Thanks again to Ben for loaning me the dash cameras.

View attachment 65779View attachment 65780

No problem. Show me what I missed on your review. Honestly got respect for you. I got tired of testing current draw on dash cameras. So I assumed the A139 Pro was same current draw as every other dash camera 2 channel. I saw your review and was like WTF. Keep me on my toes.

 

[Panzer Platform]

@safedrivesolutions reached out to me and said he was willing to loan me a U3000

@safedrivesolutions What a guy!!!

 

[rcg530]

Today, I've been running my power consumption tests for the U3000 2-channel dash camera. The thing that surprised me was that the radar parking mode did not function as described in the Thinkware U3000 product videos and user manual. The documentation states the radar parking mode should be a buffered parking mode (10 seconds before / 10 seconds after) the triggering radar detected event.



That's just not the way it's behaving. It behaves just like "Energy Saving" parking mode in that when an event is detected, the dash camera powers up and starts recording within 2 to 3 seconds of the event. I've searched the settings and the documentation to see if there's some setting I missed to make the "Radar" parking mode work as documented, but I can't find anything other than setting the parking mode type to "Radar". I've sent my findings related to the radar parking behavior to Thinkware. I created a video documenting the current manner in which the radar parking mode is behaving so Thinkware will have something to review how I tested it.

If you have 6 minutes you want to spend watching my radar testing video, here it is:

Testing the radar parking mode is a bit of a challenge since the radar detection will pass through things like boxes (based on my testing). I found it a bit of a challenge to get a test environment that allowed the U3000 to enter into radar parking mode (verified by when the power consumption dropped to a very low level) and stay in radar parking mode until I wanted to trigger it with motion. The motion I used to trigger it in my video is not visible motion to the cameras, but I'm moving my arm just outside of the box on the left side of the video.

The only other thing I noticed (maybe it's a Thinkware "thing") is that when using the app to have a live view session active over a Wi-Fi connection, the front camera video resolution drops from 3840x2160p 30fps to 2560x1440 30fps. When the live view session is completed, the front camera video resolution is restored to the original setting of 3840x2160 30fps.

Here are my preliminary power consumption test results and estimated parking mode run times.

 

[safedrivesolutions]

Today, I've been running my power consumption tests for the U3000 2-channel dash camera. The thing that surprised me was that the radar parking mode did not function as described in the Thinkware U3000 product videos and user manual. The documentation states the radar parking mode should be a buffered parking mode (10 seconds before / 10 seconds after) the triggering radar detected event.

View attachment 65840

That's just not the way it's behaving. It behaves just like "Energy Saving" parking mode in that when an event is detected, the dash camera powers up and starts recording within 2 to 3 seconds of the event. I've searched the settings and the documentation to see if there's some setting I missed to make the "Radar" parking mode work as documented, but I can't find anything other than setting the parking mode type to "Radar". I've sent my findings related to the radar parking behavior to Thinkware. I created a video documenting the current manner in which the radar parking mode is behaving so Thinkware will have something to review how I tested it.

If you have 6 minutes you want to spend watching my radar testing video, here it is:

Testing the radar parking mode is a bit of a challenge since the radar detection will pass through things like boxes (based on my testing). I found it a bit of a challenge to get a test environment that allowed the U3000 to enter into radar parking mode (verified by when the power consumption dropped to a very low level) and stay in radar parking mode until I wanted to trigger it with motion. The motion I used to trigger it in my video is not visible motion to the cameras, but I'm moving my arm just outside of the box on the left side of the video.

The only other thing I noticed (maybe it's a Thinkware "thing") is that when using the app to have a live view session active over a Wi-Fi connection, the front camera video resolution drops from 3840x2160p 30fps to 2560x1440 30fps. When the live view session is completed, the front camera video resolution is restored to the original setting of 3840x2160 30fps.

Here are my preliminary power consumption test results and estimated parking mode run times.

View attachment 65841 View attachment 65842

I actually thought this same thing happened to me as well. I called Thinkware thinking there was a big issue. The Radar unit has a very narrow field of view. They told me 55 degree. However I would like to find a way to test that. When it seen radar it would record the motion and if there was an impact it also had the impact in energy saving folder. so it recorded both.

 

[Panzer Platform]

Awesome testing Robert.
I watched the whole 6 minute video.
If you have to re-perform this test again after a firmware update for confirmation it would be nice to see the power consumption on your power supply next to the tablet.
Also, maybe some background noise constantly playing to confirm the camera is recording audio during these triggered events.
If you can't relocate your power supply, maybe a Fluke 223 with the remote display is in order. lol
https://www.amazon.com/gp/B002X7ZX00
Oh yeah, I guess you got the DLC 16-pin pigtail sorted out.

 

[rcg530]

Awesome testing Robert.
I watched the whole 6 minute video.
If you have to re-perform this test again after a firmware update for confirmation it would be nice to see the power consumption on your power supply next to the tablet.
Also, maybe some background noise constantly playing to confirm the camera is recording audio during these triggered events.
If you can't relocate your power supply, maybe a Fluke 223 with the remote display is in order. lol
https://www.amazon.com/gp/B002X7ZX00
Oh yeah, I guess you got the DLC 16-pin pigtail sorted out.

I have not worked on the Thinkware OBD-II power cable testing with a DLC 16-pin pigtail (yet). Still waiting for info back from Thinkware on the operational details of when it turns on/off the accessory (red wire) power to the Thinkware dash camera.

The Fluke multimeter was located near me so I could confirm the entry into low power radar parking mode and when the power ramped up when the radar detected motion was noticed by the Thinkware radar feature. I wasn't planning on making a diag video today, but it gets the point across.

 

[Panzer Platform]

but it gets the point across.

100% agreed.
I just like to see the power fluctuations on screen. lol

 

[Vortex Radar]

So the way I understand it, radar parking recording is an energy saving buffered parking recording. Usually the dashcam is not recording which is how it saves power to get long term parking recording. Only when the dashcam’s radar module detects something does it wake up and begin recording which is when the “buffered” part starts. If then the event is deemed a big enough threat, such as impact detected as the oncoming car or person approaches, then it saves the clip. Otherwise it doesn’t. It’s not designed to record before the radar detects anything, despite being buffered.

When it comes to buffered motion detection though, that I’m not as familiar with. I’m not sure what would trigger a motion detection clip. Is there a parking mode motion detection sensitivity threshold setting in the options?

I’d hope there’s a way to not get tons of unwanted recordings in high traffic areas which would also blow through the battery and shorten parking record times.

 

[rcg530]

So the way I understand it, radar parking recording is an energy saving buffered parking recording. Usually the dashcam is not recording which is how it saves power to get long term parking recording. Only when the dashcam’s radar module detects something does it wake up and begin recording which is when the “buffered” part starts. If then the event is deemed a big enough threat, such as impact detected as the oncoming car or person approaches, then it saves the clip. Otherwise it doesn’t. It’s not designed to record before the radar detects anything, despite being buffered.

When it comes to buffered motion detection though, that I’m not as familiar with. I’m not sure what would trigger a motion detection clip. Is there a parking mode motion detection sensitivity threshold setting in the options?

I’d hope there’s a way to not get tons of unwanted recordings in high traffic areas which would also blow through the battery and shorten parking record times.

With the v1.01.00 US firmware installed, the radar parking mode always creates a video (in my testing).

• If the radar unit picks up motion (which I made sure was not visible motion by blocking the camera's line of sight by using cardboard boxes between myself and the dash camera)
  • It boots up the dash camera in 2 to 3 seconds
  • Records 20 second videos (front and rear)
  • The dash camera continues to consume power at that higher power level for another 40 seconds
  • Then it goes back to low power mode

The user manual's definition of radar parking mode uses the phrase "10 seconds before detection to 10 seconds after detection". It could be that if there's another radar detected motion event within the 40 seconds while the dash camera is still powered up, those may be buffered recordings. I did not test that situation because I'm testing how the dash camera handles the harder/first radar detected motion event. If they are designed to operate differently, then it should be documented that way.



The "Motion Detection" parking mode has a very similar description to the "Radar" description. The U3000 in motion detection parking mode keeps the cameras powered up and when they detect movement within the field of view, the 10 seconds prior to the movement event and the next 10 seconds are written to a video file.



There are sensitivity settings for parking mode impact detection, motion detection and front/rear radar sensitivity.

 

[rcg530]

I installed the U3000 2-channel dash camera into my car. I drove my car on my weekly grocery shopping trip route. Here are my findings.

U3000 announcing the Thinkware iVolt Xtra BAB-95 remaining charge level

I have the communications cable from the BAB-95 connected to the U3000 that allows the U3000 to announce the battery charge level. Turning off accessory power and waiting the 30 seconds (as configured in the firmware settings) for the U3000 to go into parking mode, it announces that it’s going into parking mode, and it announces the iVolt battery charge level. What was “interesting” to me was right after the battery charge level announcement, the U3000 thinks an impact has occurred and you hear a “beep beep” from the U3000. If you turn on the accessory power again to exit parking mode, the U3000 returns to normal (continuous) recording mode and it announces that one event occurred. It happens each time I was present in the car when it entered parking mode. That will be a fun one to report to Thinkware.

U3000 – going into parking mode

[Turn off accessory power and wait 30 seconds] Parking recording will now start. The remaining of iVolt battery is about 100 percent. “Beep” “Beep”

U3000 – exiting parking mode

During parking mode, event detection recording one occurred. Please have a safe drive today.

Both U3000 cameras are out of focus

Let's just start off by saying, the U3000 front camera lens and rear camera lens were clean. The windshield glass (inside and outside) was clean.

The left 2/3rds of the front camera lens is out of focus. The rear camera is out of focus across the entire lens, but its slightly less blurry in the center. When I compared the U3000 focus with the Vantrue N5 front/rear exterior focus, the N5 front is in focus and the N5 rear exterior is not as in focus as I would like it to be, but it's far better than the U3000 rear camera.

The areas outlined in the red box/boarder are the out of focus areas.

U3000 Front - No CPL:


U3000 Rear:


Vantrue N5 Front - CPL Installed:


Vantrue N5 Rear Exterior:

 

[Panzer Platform]

right after the battery charge level announcement, the U3000 thinks an impact has occurred

Is it possible the vibration from the speaker is triggering the G sensor?
How the heck would they fix that?
External speaker, or G sensor? lol

 

[rcg530]

Is it possible the vibration from the speaker is triggering the G sensor?
How the heck would they fix that?
External speaker, or G sensor? lol

The other announcements are of similar volume, so I don't know why that one would trigger an event based solely on the volume of the announcement. When I had the U3000 in my office for power consumption testing, it did not behave this way. I did not have the iVolt Xtra battery available in my office to test the announcements. It's either the installation location or the additional battery data cable allowing the U3000 to make the charge level announcements.

 

[Leonauto]

The other announcements are of similar volume, so I don't know why that one would trigger an event based solely on the volume of the announcement.

Since the dashcam is "glued" to the windshield, the glass can be an additional resonator that perceives the sounds of alerts. Maybe this explanation will clarify the question? A similar reaction of the G-sensor was once observed on some dashcams with speedcam alerts.

 

[safedrivesolutions]

Is it possible the vibration from the speaker is triggering the G sensor?
How the heck would they fix that?
External speaker, or G sensor? lol

Installed the same setup today in a clients car. It did it the first time where speaker set it off. Second time it did not. Very strange. But I imagine it is something to do with speaker which means they need it to do the speaking before going into parking mode,. Should be an easy firmware update

 

[rcg530]

This video contains sample daytime video and audio from a Thinkware U3000 front camera, VIOFO A119 Mini 2 and a VIOFO A139 Pro front camera.

The U3000 that I currently have is out of focus, so I haven't collected any nighttime footage or attempted any plate captures tests. When I get a replacement U3000 dash camera, I'll collect more footage (day/night) and attempt a plate capture test.

 

[Panzer Platform]

This video contains sample daytime video and audio from a Thinkware U3000 front camera, VIOFO A119 Mini 2 and a VIOFO A139 Pro front camera.

One of these days you should do a “Pepsi Challenge” and not label which camera is which until the very end of the video.
That would be so much fun trying to figure out which camera is which just from the footage. Lol

FOV
It’s obvious to my eye the A119 Mini 2 has the narrowest FOV.
When you’re sitting behind the blue Audi at 6:08 I can clearly read the green & white street sign that says East Bidwell Rd.
The A139 Pro is next wider, and it almost looks like the U3000 is even wider, but I can tell because of the focus issues.

Colors
The Mini 2 has the “truest color representation”.
When you’re coming down the hill at 2:17 that is an amazing shot.
The blue sky with white clouds, the house rooftops on the valley floor look great.

Sharpness details
It looks like the Mini 2 has better detail sharpness than the A139 Pro.
The A139 Pro makes bushes & trees look hazy on both sides of the screen.

Audio
The Mini 2 is the loudest but it also sounds a bit tinny.
The A139 Pro sounds 2nd best, and also very balanced.
The U3000 sounds like it has too much bass, and it needs the treble turned up it was my least favorite.

At 3:25 @Vortex Radar has a work truck???

7:02 The Honey Pot truck. Lol

I still want to see blurry night time footage.
That will still be nice to compare how well the U3000 copes in low light situations.
-Chuck

 

[rcg530]

I went on an early morning drive to visit @Panzer Platform. It was great to meet him in person. We talked for hours.

I started my trip at 4:30 AM. Due to the time/distance of this trip, the max capacity of the included Thinkware microSD card (64 GB) was insufficient to capture and retain all of the continuous recording (cont_rec directory) video files. The first 30 minutes of the continuous recording video files were purged from the microSD card by the time I arrived back at my home. I also had a VIOFO A119 Mini 2 and a VIOFO A139 Pro recording as well.

I left the house 70 minutes before sunrise. I drove down a very dark section of the main road to get to the freeway and I then drove by some businesses with all of their business signs and parking lot lights on. I have screenshots from the A119 Mini 2 and the A139 Pro from the early part of the trip.

On the dark section of the road going towards the freeway, the A119 Mini 2 (2560x1440 30 fps / No CPL / HDR On) captured a better quality image than the A139 Pro (3840x2160 30 fps / No CPL / HDR On). Full resolution screenshots have been uploaded, but this website auto resizes them to a lower resolution.

A119 Mini 2 / A139 Pro


In Town / Business District / Well Lit Nighttime Area

The A119 Mini 2 does a better job than the A139 Pro does with these lighting conditions.

A119 Mini 2 / A139 Pro


The oldest continuous recording video file from the U3000 was captured about 20 minutes before sunrise. Please keep in mind, the U3000 that I currently have is out of focus on the front and rear cameras. The U3000 front had no CPL filter installed, Super Night Vision 4.0 was disabled for continuous recording mode. The U3000 front camera does not capture the illuminated road sign contents and there's a noticeable amount of noise in the image. Once again, the A119 Mini 2 seems to do the best job of the three dash cameras capturing the cars, license plates and the illuminated road sign contents. The A139 Pro was able to capture the illuminated road sign contents, but not as clearly as the A119 Mini 2. There's a higher noise level in the A139 Pro video as well.

U3000 / A119 Mini 2 / A139 Pro

 

[Panzer Platform]

Thank you Robert.
Meeting @rcg530 in person was like meeting my mentor / hero.
I’ve been following him for 3 years ever since his YouTube review of the A129 Pro Duo 4K published March 2020.
It was a real treat picking his brain asking him all kinds of questions about the dash cam industry.
I felt honored sitting with him, and he was very patient listening to all my stupid ideas to improve overall consumer satisfaction with the dash cam industry.
I hope we can turn this into a monthly event.
-Chuck

 

[kamkar]

It is funny how much better the mini 2 are in the low light VS others, it is not like its sensor pixels are substantially larger.
I am inclined to think it is the smaller FOV lens that do it.

It is always nice to GTG with like-minded / people with same interests, thats often hard for me CUZ the things i find interesting are extremely borderline here.
For instance me liking to shoot air rifle benchrest, well a few others do that here too, but it is not easy to find a place to meet and do that as it is not really something any rifle club do, same go for field target which pretty much do not even exist here.
And we sure as hell do not have a range where you can walk in and shoot your own stuff.

I am amazed that there are several IPSC pistol clubs / ranges however, i would have thought that kind of shooting was wayyyyyy too wild for Danes, or at least the Danes other Danes vote on to set the standard.

Mind you recent events with a shooter in a shopping mall here, are once again prompting panic lawmaking, so for shooters things are probably going to get worse once again, CUZ it is rare this kind of law giving have any actual results on the problem at hand.