Vortex's Best of 2023 Compilation

[Vortex Radar]

Over the past couple weeks I've been spending time compiling a video covering the key features, pros, and cons of my favorite dashcams on the market. The list includes:

• Viofo A139 Pro
• Viofo A119 Mini & Mini 2
• Viofo A119 Plus Duo & A229 Duo
• Thinkware U1000
• Thinkware U3000
• Blackvue DR970X-2CH LTE
• Blackvue DR970X-2CH + CM100G LTE
• Blackvue DR770X Box
• Powercell 8 vs. Blackvue B-130X vs. Thinkware iVolt Xtra

There's a lot to cover of course and hopefully this does a good job going over the key things I think people should know.

 

[kamkar]

Ohter brands are also gearing up to launching dashcams with smart capabilities if you are that way inclined.
So cone 2024 i am sure we will see many more brands

 

[Panzer Platform]

My favorite part of the video was when the battery pack talked at 25:25. lol

 

[5.0ALM]

Looking forward to a Starvis 2 (only) 4k comparison using the IMX678 sensor:
Viofo A139 Pro
Thinkware U3000
Vantrue N4 Pro
70mai A810

So we're comparing Apples with Apples, it might be appropriate to review the IMX675 2k dash cams separately?
So far:
Viofo A119 Mini 2
Vantrue N5

 

[Miastar]

Looking forward to a Starvis 2 (only) 4k comparison using the IMX678 sensor:
Viofo A139 Pro
Thinkware U3000
Vantrue N4 Pro
70mai A810

So we're comparing Apples with Apples, it might be appropriate to review the IMX675 2k dash cams separately?
So far:
Viofo A119 Mini 2
Vantrue N5

Purely from the perspective of current image tuning results, the effects achieved through the current range of main control chips (most of which are based on the NTK series) paired with the 678 tuning still fall short of the desired outcome. Similarly, upon reviewing comparison videos of upstream supply chain offerings, the improvement of the 675 over the 335 is relatively minimal, considering their nearly $1.5 USD chip pricing. This might be due to the current main control chips not being able to fully leverage the capabilities of the 675. On the other hand, the parameters and tuning of the 335 have matured sufficiently. The 335 boasts a pixel size of 2μm, which is quite commendable for a 5-megapixel sensor.

This explains why we often witness the Sony 415 not performing as well as the 335 in low-light conditions. Pixel size plays a crucial role.

Looking at supply chain trends, the price of the 678 chip has started to gradually decrease (as more products such as security cameras and dash cams adopt this sensor). This is a promising signal for future mid to high-end 4K dash cams. We no longer have to endure the torment of the 415's nighttime image quality.

A 1/1.8-inch 4K sensor with OV's 08A10, boasting a pixel size of 2μm similar to the 678, yields equally impressive results. Some Japanese COMTEC models utilized this sensor before the 678 hit the market. In terms of night vision, Purecel technology optimization is available, and native support extends to specifications like 8MP@60fps. This implies that with enhanced main control, the prospect of 4K 30fps HDR quality is highly anticipated.

Considering only cost-effectiveness, I believe 70mai's A810 is quite appealing. Priced at only $100 USD with a 678 sensor for the front camera, it's an excellent option. Of course, one limitation is that it only supports 2.4GHz WiFi. The dual-camera version is equipped with the GC2053 sensor, resulting in relatively ordinary image quality. Sony's 678 is accurately positioned, compensating for the shortcomings of previous 415 night vision with its large pixel size. However, as for the 675, even though it carries the Starvis 2 tag, its cost-effectiveness is quite mediocre, and its product positioning is awkward. Apart from the Starvis 2 novelty, I believe it's debatable whether it's worth going over budget for this configuration.

 

[Panzer Platform]

Purely from the perspective of current image tuning results, the effects achieved through the current range of main control chips (most of which are based on the NTK series) paired with the 678 tuning still fall short of the desired outcome. Similarly, upon reviewing comparison videos of upstream supply chain offerings, the improvement of the 675 over the 335 is relatively minimal, considering their nearly $1.5 USD chip pricing. This might be due to the current main control chips not being able to fully leverage the capabilities of the 675. On the other hand, the parameters and tuning of the 335 have matured sufficiently. The 335 boasts a pixel size of 2μm, which is quite commendable for a 5-megapixel sensor.

This explains why we often witness the Sony 415 not performing as well as the 335 in low-light conditions. Pixel size plays a crucial role.

Looking at supply chain trends, the price of the 678 chip has started to gradually decrease (as more products such as security cameras and dash cams adopt this sensor). This is a promising signal for future mid to high-end 4K dash cams. We no longer have to endure the torment of the 415's nighttime image quality.

A 1/1.8-inch 4K sensor with OV's 08A10, boasting a pixel size of 2μm similar to the 678, yields equally impressive results. Some Japanese COMTEC models utilized this sensor before the 678 hit the market. In terms of night vision, Purecel technology optimization is available, and native support extends to specifications like 8MP@60fps. This implies that with enhanced main control, the prospect of 4K 30fps HDR quality is highly anticipated.

Considering only cost-effectiveness, I believe 70mai's A810 is quite appealing. Priced at only $100 USD with a 678 sensor for the front camera, it's an excellent option. Of course, one limitation is that it only supports 2.4GHz WiFi. The dual-camera version is equipped with the GC2053 sensor, resulting in relatively ordinary image quality. Sony's 678 is accurately positioned, compensating for the shortcomings of previous 415 night vision with its large pixel size. However, as for the 675, even though it carries the Starvis 2 tag, its cost-effectiveness is quite mediocre, and its product positioning is awkward. Apart from the Starvis 2 novelty, I believe it's debatable whether it's worth going over budget for this configuration.

I agree with most of what you're saying here.
Do you have any test footage to back this up?
I maintain Playlists of test footage of all the dash cams I tested since 2020 on my second YouTube channel.

https://www.youtube.com/@PanzerUnlisted/playlists

Here's a two clips from the IMX415 & IMX675 at night on the same stretch of highway, same car, same headlights;

 

[Miastar]

I agree with most of what you're saying here.
Do you have any test footage to back this up?
I maintain Playlists of test footage of all the dash cams I tested since 2020 on my second YouTube channel.

https://www.youtube.com/@PanzerUnlisted/playlists

Here's a two clips from the IMX415 & IMX675 at night on the same stretch of highway, same car, same headlights;

Sure, Panzer.
Let me install my DDPAI mini5 (415), 70mai A810 (678), and Vanture N2 (335). I believe these three devices provide a very intuitive way to observe the advantages brought by larger pixel sizes.

The mini5 I purchased is an early model, and as far as I recall, it was equipped with the Hisilicon 3559V200 processor. However, I believe this won't significantly affect the final assessment.

On a slightly different note, it's worth mentioning that the supply of the Hisilicon 3559V200 has gradually decreased due to sanctions, with production nearing a halt by 2020. Nevertheless, when considering the image quality, development flexibility, and cost-effectiveness, this MCU is likely one of the most powerful options for security cameras and dashcam in recent years.

An area worth paying attention to is that Hisilicon has recently developed a new main control chip, the 3519, which comes with built-in AI ISP and boasts impressive performance, making it a great choice for 4K-level applications. However, it's unfortunate that the chip is built on the 28nm process by SMIC, which has posed challenges in effectively addressing the heat dissipation issue.