Front/rear dashcam solution

mrx

New Member
Joined
Jun 15, 2015
Messages
5
Reaction score
1
Country
United Kingdom
Hi all,

As I'll be doing a fair bit of mileage soon, I figured it'd be a sensible idea to invest in a dashcam or two (front and rear).

My requirements are fairly basic, and I guess obvious for the most part:
  • 1080p/30fps recording both front and rear. I'm fairly interested in some of the models that are capable of recording 2560x1080 footage, wide angle would be handy (to get the front/rear corners of the car).
  • Excellent capture quality in very dark conditions, motorway/highway driving - England in the winter after all. Need to be able to capture plates as clearly as possible.
  • Don't really need GPS. The cameras will only be used for insurance/police purposes if anything were to happen (touch-wood nothing will); I doubt anybody will be disputing where an incident took place. Just time/date is required.
  • Compact, with/without a screen, no particular preference.
  • Parking mode not required, just using the 12V socket in the glovebox.
Additionally, the usual decent reliability/support are a must.

I have considered some of the two channel solutions, however there seems to be a big compromise somewhere or other - size and capture quality are the main two downfalls. So I'll likely go for two separate cameras.

I've done a fair bit of research into various models, some examples of what I've considered:
  • Marcus 4 - seemingly great picture quality in both the day and night, offers all the features really (2560x1080, 160°). Quite expensive, and image quality in extremely dark conditions is unknown, but probably alright. Number plates aren't necessarily completely clear at medium distances from what I can see.
  • Mini 0805 - reliable, cheap, however footage in the day appears to be a bit dim. At night seems OK, but I doubt it'll be as strong as the Marcus 4.
  • DOD LS460W - don't know much about this one, seems to be great for night footage too, doesn't offer the wide-angle that the Marcus 4 does (140° here).
Honestly, there are so many choices, including those from no-name brands that I'm not really any wiser than before. In terms of budget, £150-250 (GBP) ideally however I'm flexible either way. As above, picture quality is very much the priority here, particularly at night time.

Many thanks in advance :)
 
Thanks for the reply. Must say it looks great (and ridiculously good value), however unfortunately the night footage isn't quite as good as the other two. Cost isn't really a concern, so I'm willing to splash out more if need be.
 
...the night footage isn't quite as good as the other two...

Unless you've seen some simultaneously shot comparison footage, I don't think you can assert this as fact. The comprehensive review of the BlackSys by @niko said... "very good night video quality recording thanks to Sony Exmor CMOS sensor ( front camera ). There are no background noise at night, video is nice, clear and smooth".

niko has tested over 100 DashCams so, when he makes comments like the above, you can be sure that the CF-100 performs well at night.
 
Oops, I meant to say that the rear camera doesn't! Having a look at the front camera footage, I can't say there is too much to complain about. I'm going to have a surf through more of niko's posts to get some better ideas however, so thanks very much for that suggestion!
 
Oops, I meant to say that the rear camera doesn't! Having a look at the front camera footage, I can't say there is too much to complain about. I'm going to have a surf through more of niko's posts to get some better ideas however, so thanks very much for that suggestion!

Yeah, rear camera is only 720p and low bit-rate. We can not expect too much from this and actually for that resolution / bit-rate it's doing fair job. I see rear camera on CF-100 as just "free gift": - buy front camera, - get rear 720p free, just like a bonus to have something in rear. To have "something" is better than nothing, however of course we can not compare to 1080p high-end camera.

Within maximum level of your budget, you may also consider Thinkware F750. I did review on F550, which has same design as F750, but last one has 1080p + 1080p with extra bells and whistles.
@B20RRL stock them locally in UK within your budget and they have posted a review.
I have not seen yet any raw video footage, especially from extreme day / night time conditions to give some kind of evaluation. Hope others who bought it can share their opinion in Thinkware thread.
 
Last edited by a moderator:
Thanks for the reply niko. It's actually one I looked at before (saw the preview from CES), looks like a very well-polished solution. The price came out cheaper than expected so I'll definitely be taking a further look.

EDIT: @niko, having a further look, seems that the F750 may very well be the camera of choice, seeing as the form factor is good, the speed camera detection actually works unlike most other cameras, video quality appears to be good, and the overall package is excellent (build quality, app, etc). My only concern is detail at night time. You previously commented on the F550, stating that night footage was too dark, more so than competing cameras with the same sensor and processor. Is this something that Thinkware addressed in subsequent firmware updates?

Main thing is to be able to read plates from a decent distance. A higher resolution sensor (such as the unknown 4MP sensor in the Marcus 4) in theory should fair better, however so little detail can be clearly made out in such conditions that it probably won't make a huge difference. As long as the F750 can offer excellent image quality "virtually" matching that of the Marcus 4, it's probably what I'll go for.

Does anybody know of the bit rate the F750 actually records at? In the F550 review the last post mentions that it wasn't recording at the advertised bit rate (roughly 17Mbps), and that reliability suffered when the firmware was updated to offer it. The maths suggest that the F750 is recording at 17Mbps or so per camera, however I'd like to be assured that this actually is the case.

Final question, does the F750 use the same sensor front and rear? I'm presuming this is a yes, though I haven't read anything solid.

Thanks!
 
Last edited:
Thanks for the reply niko. It's actually one I looked at before (saw the preview from CES), looks like a very well-polished solution. The price came out cheaper than expected so I'll definitely be taking a further look.

EDIT: @niko, having a further look, seems that the F750 may very well be the camera of choice, seeing as the form factor is good, the speed camera detection actually works unlike most other cameras, video quality appears to be good, and the overall package is excellent (build quality, app, etc). My only concern is detail at night time. You previously commented on the F550, stating that night footage was too dark, more so than competing cameras with the same sensor and processor. Is this something that Thinkware addressed in subsequent firmware updates?

Main thing is to be able to read plates from a decent distance. A higher resolution sensor (such as the unknown 4MP sensor in the Marcus 4) in theory should fair better, however so little detail can be clearly made out in such conditions that it probably won't make a huge difference. As long as the F750 can offer excellent image quality "virtually" matching that of the Marcus 4, it's probably what I'll go for.

Does anybody know of the bit rate the F750 actually records at? In the F550 review the last post mentions that it wasn't recording at the advertised bit rate (roughly 17Mbps), and that reliability suffered when the firmware was updated to offer it. The maths suggest that the F750 is recording at 17Mbps or so per camera, however I'd like to be assured that this actually is the case.

Final question, does the F750 use the same sensor front and rear? I'm presuming this is a yes, though I haven't read anything solid.

Thanks!


1. there are no dashcams out there yet that could read numberplates at night from moving cars. Yes, despite F550 has also Sony Exmor, unfortunately their night sensitivity tune-up was too dark. I don't know if this was fixed with F750. Need to have it in hand and make pitch dark driving tests and later inspect raw / original footage.

2. More "Megapixels" of sensor does not give you a better video recording at 1080p. It is wrong understanding that 4Mpix will be better than 2.4Mpix at 1080p. Not at all. Actually 2.4Mpix is considered better choice for 1080p than 4Mpix, due to bigger area of each individual pixel. For 1080p you need only 2.06 Mpix effective area.

3. Bit-rate is important factor, but only till certain level. Over some certain value, adding more bit-rate deos not give much advantage. Per 1 channel golden-middle value is 15Mbps. I am not sure if F750 records 17Mbps per channel, I have not seen raw footage yet. Unfortunatelly Ambarella A7 maximum practical combined bit-rate is around 20Mbps, thats where it operates without bugs. , it means 10 + 10Mbps in 2ch systems. I dont know if this applies to F750, maybe they found a trick how to overcome those bugs which appear when using a too high bit-rate.

It looks like front and rear uses same IMX322 Sony Exmor CMOS sensor.
 
Thanks for all your help niko, it's really appreciated.

My hypotheses of the higher resolution sensor was that footage would essentially be downscaled, however I see you point (otherwise, the unit would actually be recording footage at that resolution then downscaling).

As for bit rate, I know what you mean. 10Mbps could be a bit low, that's all. Footage on YouTube is mostly useless, frustratingly (and it adds interpolation, so how well the unit handles motion is virtually impossible to discern).

After further consideration I think I'm going to go for the F750. I doubt footage at night will be quite as good as that of the absolute best units out there, however:
The overall fit and finish of both the hardware and the overall solution (as I said before, the app appears to be excellent) have also contributed. I was strongly considering picking up a couple of CF-100s but the convenience of the F750 just nips the CF-100's video quality to the post. Being able to use the app in the event of an emergency as opposed to having to carry a laptop around is always a bonus (I'm an iPhone user, so no microSD card slot) and the excellent speed camera alerts help too.

If I do purchase tomorrow, I'll try to share some raw footage in the next couple of weeks.
 
Thanks for all your help niko, it's really appreciated.

My hypotheses of the higher resolution sensor was that footage would essentially be downscaled, however I see you point (otherwise, the unit would actually be recording footage at that resolution then downscaling).

As for bit rate, I know what you mean. 10Mbps could be a bit low, that's all. Footage on YouTube is mostly useless, frustratingly (and it adds interpolation, so how well the unit handles motion is virtually impossible to discern).

After further consideration I think I'm going to go for the F750. I doubt footage at night will be quite as good as that of the absolute best units out there, however:
The overall fit and finish of both the hardware and the overall solution (as I said before, the app appears to be excellent) have also contributed. I was strongly considering picking up a couple of CF-100s but the convenience of the F750 just nips the CF-100's video quality to the post. Being able to use the app in the event of an emergency as opposed to having to carry a laptop around is always a bonus (I'm an iPhone user, so no microSD card slot) and the excellent speed camera alerts help too.

If I do purchase tomorrow, I'll try to share some raw footage in the next couple of weeks.

Raw footage from different scenes, especially from extreme conditions ( against sun, tunnel in / out, pitch black, near trees etc. ) - would be appreciated. This way can see how it copes in difficult conditions.
If your budget allows F750 over CF-100, then I think you made a good choice, however I must say for those who has a limited budget, CF-100 is very ahrd to beat, but they ar from "different planets" and can not be compared.
 
Back
Top