What features do you want the most in a dash cam?

[batbug]

Hey Guys! I am thinking of making a dashcam and would really appreciate your opinions of what features you value the most when choosing what to buy.

I've made the results public so we can all see

 

[DT MI]

You forgot the most important feature - Reliability.

 

[kamkar]

Good footage, not just the generic parameters for the video ( which most seem to do sadly )
Reliability.
I would like to see smarts, but not like in online connectivity or anything to do with any form of social media, i would like to see some of the OBD stuff like Lukas have tried and failed with, most important indicator use - brake use. I would also like to see other forms of storage and a smart way of backing up your footage ( to a thumb drive or something ) SATA or M2 drives in the dashcam, hence also mean the main unit have to be remote.
And for the love of god make that daylights savings automated like computers and phones.
Support.
Customer care.
Technical braveness / capability, IE bring in some new sensors instead of just using what everyone else use.
HDR. ( proper HDR not 2 consecutive frames stitched together )

 

[Lothar]

Why must we only be limited to 4 of the options?

 

[SawMaster]

HiRes videos alone aren't enough; detail capture matters a lot daytime and dark, and a lot of that is in how the FW works. Cams using the same sensor/processor combo can give drastically different results, with both being "HiRes".

I like wide-angle views, but that can bring distortion and also a loss of detail. Still better than not capturing anything because of a limited FOV..

I've come to like 2-channel systems as long as the rear cam is decent (not all are) and as long as there's no issues with reliability.

Best of luck on getting a cam manufactured to your specs. Many have tried already from people with good knowledge about them to people only seeing a specific need and filling it. Almost none have been successful, but every company in the business started somewhere...

Phil

 

[HonestReview]

I would like to add another Choice: Cloud Storage. I wish more manufacturers would add this capability.

Other than that:

1. Reliability - Remember, Dashcams are a safety net. If they fail, you could lose critical footage.
2. Image quality - Goes without saying but Good Day and Night Image quality is important
3. Continuous recording in parking mode. I.E. a Low Bitrate Recording using a hardwire kit while car is parked. Not Event Detection or Buffered Parking Mode.

Suggestion:

A 3 Channel Setup (Front + Rear +Interior Cam) where the main unit isn't placed in the windshield. 3 Channel setups are subject to overheating. Like what Street Guardian did with this 2 Channel Setup:

 

[EGS]

Hey Guys! I am thinking of making my own dashcam and would really appreciate your opinions of what features you value the most when choosing what to buy.

I've made the results public so we can all see

Will you be offering testing samples to testers here?

 

[DAP]

#1 should always be reliability. Part of reliability is to inform the user when something goes wrong. Too many cameras fail silently when the memory card goes bad, and you go for months without knowing that the camera is not recording. Useful error messages are also important. It does not help if the camera just beeps, or just displays the word "Error" when something goes wrong, the software knows damned well what it detected that was wrong, it should display a description in clear language what went wrong so the user can do something about it.

A large sensor, and large aperture lens should be used to improve low light sensitivity. Perhaps even an RGBC (Red Green Blue Clear) sensor where the C does not have IR blocking could be used for better low light sensitivity. A high frame rate should always be used to minimize motion blur. Even a second camera optimized for license plate reading would be a good idea (this is the only option where I would agree with a multi-channel camera. Perhaps they could merge the license plates into the normal camera view.) Going a bit far into wishful thinking here, but an AI license plate reader that could read the license plates in the raw image before the video compression obliterates them.

I don't like the idea of multi-channel cameras as it introduces a single point of failure, and wears out the memory card faster. You are better off just buying two cameras, and mount one in front and the other in the rear for redundancy.

A parking mode with at least 10 seconds of RAM pre-buffer that detects visible motion to trigger recordings. More wishful thinking here: an AI to recognize the difference between vegetation blowing in the wind, and people, cars, & animals.

Remote storage for the video so the recording can be stored in a secure location in the car. Although I have never heard a case of a dashcam being stolen, it is something a lot of people worry about, and if it happens, it would be nice if they could not also steal the recording of them stealing the camera when they steal the camera.

 

[thomas-dash]

an AI to recognize the difference between vegetation blowing in the wind, and people, cars, & animals

I like this idea.

I was thinking about having the option to save and load different profiles from an app on the phone with just a few touches. This would make it easy to change settings in one go instead of changing each individually settings.

This would allow you to have a profile optimized for night and another one for day without having adjust multiple settings each time.

 

[HonestReview]

#1 should always be reliability. Part of reliability is to inform the user when something goes wrong. Too many cameras fail silently when the memory card goes bad, and you go for months without knowing that the camera is not recording. Useful error messages are also important. It does not help if the camera just beeps, or just displays the word "Error" when something goes wrong, the software knows damned well what it detected that was wrong, it should display a description in clear language what went wrong so the user can do something about it.

A large sensor, and large aperture lens should be used to improve low light sensitivity. Perhaps even an RGBC (Red Green Blue Clear) sensor where the C does not have IR blocking could be used for better low light sensitivity. A high frame rate should always be used to minimize motion blur. Even a second camera optimized for license plate reading would be a good idea (this is the only option where I would agree with a multi-channel camera. Perhaps they could merge the license plates into the normal camera view.) Going a bit far into wishful thinking here, but an AI license plate reader that could read the license plates in the raw image before the video compression obliterates them.

I don't like the idea of multi-channel cameras as it introduces a single point of failure, and wears out the memory card faster. You are better off just buying two cameras, and mount one in front and the other in the rear for redundancy.

A parking mode with at least 10 seconds of RAM pre-buffer that detects visible motion to trigger recordings. More wishful thinking here: an AI to recognize the difference between vegetation blowing in the wind, and people, cars, & animals.

Remote storage for the video so the recording can be stored in a secure location in the car. Although I have never heard a case of a dashcam being stolen, it is something a lot of people worry about, and if it happens, it would be nice if they could not also steal the recording of them stealing the camera when they steal the camera.

I disagree here. Far more people prefer a multichannel setup because of hardwiring. For instance, let's say you got 2 x VIofo A119 V3s. You'd now need 2 x HK3 hardwire kits. Meaning twice as many fuse slots taken up.

But wait, there's a bigger problem. On a 2 or 3 Channel setup, where you have a "Main Unit", you will know if one has failed. Unfortunately, if you have individual units, you aren't going to know if a unit in back has stopped recording. Most people aren't going to bother to check that unit EACH and Every time they start their car. So unless that back unit beeps or give some notification, you're going to just assume it's recording.

The above is the exact reason a Multi Channel setup is preferred. Sure, the SD card has more data being written to it, but MicroSD cards are pretty cheap. So if you get 3 or 4 years out of it, they aren't that expensive to buy a new one.

 

[batbug]

Why must we only be limited to 4 of the options?

I'd like to see what are people's top rated choices. I think if we could, we would get EVERY feature, but some are more important than others.

It looks like reliability is something that I missed out that a fair few people have reached out over. I suppose it isn't technically a feature but I have a really great idea on how I am going to do that.

I like this idea.

I was thinking about having the option to save and load different profiles from an app on the phone with just a few touches. This would make it easy to change settings in one go instead of changing each individually settings.

This would allow you to have a profile optimized for night and another one for day without having adjust multiple settings each time.

Simplicity is so important too!

Will you be offering testing samples to testers here?

I think that's a great idea! I don't exactly have the budget of a big company, I'm just one guy, but gaining insightful feedback from a crowd who know what they're talking about is going to be worth every penny!

Thanks for all the replies!

 

[Mtz]

All options are in vain if you don't have the right source for a good quality lens.
There are small chances to start with a dahscam with good lens, most Chinese dashcams have low quality lens and this will be not changed tomorrow.
So it will be just another dashcam.

 

[kamkar]

A good lens is a must, otherwise it is like a fully original Ferrari car ( body / tires / drive line ) But you have a VW beetle motor in it and not the banging 12 Cylinder .

Image quality must always be #1
Though even if you have the sweet sensor and sweet lens, it do not matter if your company can not maximize the hardware by using the firmware properly, and then you still have a generic camera just with good parts.
So #2 on the list almost have to be a good software engineer that can tweak the firmware to the limit, and if you find such a guy, for the love of god dont let him go, or restrain him, he is the guy that will really get your investment to shine.

If you dont, you will just be swimming with the thousands of other similar mediocre fish.

 

[Mtz]

You are dreaming. The goal is only money, almost nobody will search enough to offer a real better image.

 

[dash riposki]

With a mainstream market cam you compete for the less sophisticated, but more numerous, buyers.
I think there is a niche market for a rugged, weatherproof, reliable cam that can be mounted externally Single channel. No crap features, just good performance and reliability.

 

[batbug]

I've been doing a lot of thinking and I have decided to go ahead and make a Dash Cam company! The response to the poll above came back better than I thought it would which I'm really thrilled about. The most requested features matched with the idea that I had for a dash cam. I knew what I wanted to buy and was really surprised to see that it doesn't exist anywhere considering how simple it is, so I'll make it myself! There is one significant feature of my idea which makes it much better than anything I've seen on the market.

This will be dashcamtalk's very own homegrown dash cam!

What people want seems clear: a reliable front-and-rear dash cam which is discreet, has a high quality image and wide viewing angle. What is particularly important is to not just pick a camera with good specs on paper- this can often be misleading and can still lead to a poor image quality in the recorded footage. Legibility of written text from a distance (such as license plates) and clear footage even under poorly lit conditions are essential. Accessing the footage has to be quick and easy too.

If you want to be the first to hear about it, drop your email in this google form I made: https://docs.google.com/forms/d/e/1...V3k782SwpgRb_7KBZDwHSBbQ/viewform?usp=sf_link

 

[Mtz]

I wish you good luck with the project.
Always I am the scary guy which is always finding flaws on any idea or product in the world. I never say ”Great idea, let's do it”. I always say „Hmm... something is not OK... there are small chances to make it work... lets see if you can solve these weak points!.”
So if you want a scary guy, here I am:
- you need a very good source, connection in China. Even companies from Shenzen (the City-God of electronics), don't know all the sources. „a reliable”
- a chinese source can always find another manufacturer for some component and give it to you at the same price, but with lower quality „a reliable”
- you can be lucky to discover a good quality source, but at some moment they can use your idea and sell your product much cheaper under another brand. Make at least your camera cases in a secret place.
- will be very hard for you to find a good lens. This will cut your chances for „high quality image”.
- you will need firmware team. Without it you will be just another brand on the market with factory firmwares on your dashcam. No firmware team will cut your chances for „high quality image”.
- you said „wide viewing angle”. The lens paradox will haunt you forever. People are asking for wide lens. Cameras with wide lens are selling always better. But such lens will offer the poorest image details. The only chance to sell a dashcam with great image details is to have a small angle lens and lie to the people that it is a wide lens. For example 110 degrees in reality and 170 degrees advertised.
- „High storage capacity” if you will make a camera with EMMC with at least 128GB it will be very expensive and if by mistake the EMMC will die all your camera will die. In most of the countries you need to offer 2 years of warranty.
- „Up to £50” you just can't create such a quality camera.

So be careful if you have no intentions to move in Shenzen for the next 1-2 years at least.

 

[DAP]

The world needs people who start projects with no idea what they are getting into. If they had any idea what they were getting into, they would never start the projects, and we would never get the few that succeed.

 

[EGS]

The world needs people who start projects with no idea what they are getting into. If they had any idea what they were getting into, they would never start the projects, and we would never get the few that succeed.

I agree. Everything in this life takes a risk. We need to quit being too comfortable if wee want to succeed.

 

[petepete]

maybe you should just replicate a mobius and add a facility to drop it into 5fps for when parked up ....
or maybe mobius should lol