M2 footage dash and action

[c4rc4m]

And here's the full-time night test results, showing 30fps and 60fps this time as the metering options made almost no difference. And for good measure, there's a comparison frame from the SG9665GC.

Interesting to see how 1080P30 has more visible grain than the 60P version. It's also interesting how the exposure seems poor on the 1080P30 version in so far as some areas are burnt out. You can clearly see the 30 fps frame is brighter. However, it appears to me the exposure control is poor as the picture looks to me to be quite markedly overexposed. To that end, the 60fps version looks much better metered.

Looking at the noise levels, I'd have to guess that the exposure software has some issues as it appears to me as though the longer exposed 30fps frame has more gain applied than the short exposed 60fps frame, either that or the overexposed picture just highlights it more. However, common sense would suggest it should need less gain as with longer exposure times, it will gather more light per frame and thus require less gain plus be brighter with less grain.

 

[TonyM]

Currently trying to find a setting that works well in fog! 1080p30 with WDR is looking a bit murky. The GC with WDR has better clarity and colour balance, and I feel is a closer representation of the actual scene. I've turned down the sharpness, exposure and contrast in the M2 using the new adjustment options (all at -32).

 

[TonyM]

Tested the M2 without WDR in the office car park this afternoon. Still looking murky by comparison to the GC, but it's closer.

 

[c4rc4m]

On the top pictures, I'd say the difference could be the adjustments you've made. The picture looks flat compared to the GC, but it's typical of lack of contrast and exposure. Hard to tell what effect the reduced sharpening has had as don't look too dissimilar and without an even back to back comparison, (settings for settings), it's hard to assess.

However, I don't think you can blame the M2 for being flat when exposure and contrast are -32! Be perhaps better to see them on the same settings then critique from there, or take a series of pictures with a range of adjustments to see which match the GC, and critique from, the closest match. It will also identify the perhaps what the standard settings should be.

One interesting thing my eyes do pick up on though, is the lower level of reflection from the windscreen from the chequered area on the dashboard on the GC. Does it have a polariser fitted / built in?

 

[TonyM]

Yes, I have a CPL fitted to the GC. I've not yet finished my modification to fit an identical CPL to the M2.

I was trying out -32 for exposure & contrast as in good light I generally found the M2 to be a bit too bright and contrasty on the default settings. There's no guidance as to what the exposure scale is for the M2 image adjustment, although the range is +/-128. I don't know if that equates to +/- 2 stops, in which case -32 would be approx -0.5EV. You're right that a more detailed test with a series of image adjustments would be better - that's the subject of another test I intend to do one day.

 

[c4rc4m]

...and some peeps say polarisers don't work. I'd call that a dramatic improvement.

 

[Dashmellow]

...and some peeps say polarisers don't work. I'd call that a dramatic improvement.

Polarizers don't work? Can you provide us with an actual example of someone stating that here on DCT?

 

[SawMaster]

That's quite a wide range of adjustability and we can clearly see it has an effect. I think it's more a question of finding the right combination of settings which then Mobius can (hopefully) use as the default settings. As is often said here Mobius is an enthusiasts cam not necessarily meant for the masses and having this wide range of settings can only make that more so.

Phil

 

[c4rc4m]

Polarizers don't work? Can you provide us with an actual example of someone stating that here on DCT?

Yes you here, effectively in post 91 : https://dashcamtalk.com/forum/threads/m2-footage-dash-and-action.20624/page-5

I've never found a CPL to be very effective against road glare. That's a different problem in terms of polarized light and it is wishful thinking to expect a CPL to fix that because a CPL cannot ameliorate light coming at you from very different and multiple angles at the same time. Either you use a CPL to polarize the light reflected off your dashboard against your windscreen or perhaps try to use it to fix the constantly shifting light reflecting off the road surface but it won't do both at the same time.

No offence is meant when I say this, but in the above examples, especially the top two, in addition to the reduction in windscreen glare, you can clearly see the polariser has at the same time saturated the colours outside of the car, namely the other cars and most obviously the sky which looks quite blue compared to the the non polarised version above it. There's no standing water to compare reflection with and maybe it wouldn't have cause a benefit from a reduction in standing water reflection. However, there is a clear gain here from the polariser both inside and outside of the car which disproves the theory it's totally one or the other.

Yep I do agree from with what you said in the other thread, a CPL is a ND filter in effect and you will lose 1-2 stops of light at night. However, you can see how the camera is affected, and if it gains up too much at night, use the CPL in the day and remove it at night. The hardest part in fitting a CPL is aligning it and to that end, once a good alignment is found, I'd suggest marking both the lens and CPL to make future alignment just a matter of matching up the marks. The hardets part I think of finding a CPL is finding a good one especially when you're dealing with a miniature size which tends towards the novelty end of the market.

 

[Dashmellow]

All the adjustability and settings don't matter if the camera can't provide adequate image quality which at this point in time seems to be the case with the M2. The M1

Yes you here, effectively in post 91 : https://dashcamtalk.com/forum/threads/m2-footage-dash-and-action.20624/page-5



No offence is meant when I say this, but in the above examples, especially the top two, in addition to the reduction in windscreen glare, you can clearly see the polariser has at the same time saturated the colours outside of the car, namely the other cars and most obviously the sky which looks quite blue compared to the the non polarised version above it. There's no standing water to compare reflection with and maybe it wouldn't have cause a benefit from a reduction in standing water reflection. However, there is a clear gain here from the polariser both inside and outside of the car which disproves the theory it's totally one or the other.

Yep I do agree from with what you said in the other thread, a CPL is a ND filter in effect and you will lose 1-2 stops of light at night. However, you can see how the camera is affected, and if it gains up too much at night, use the CPL in the day and remove it at night. The hardest part in fitting a CPL is aligning it and to that end, once a good alignment is found, I'd suggest marking both the lens and CPL to make future alignment just a matter of matching up the marks. The hardets part I think of finding a CPL is finding a good one especially when you're dealing with a miniature size which tends towards the novelty end of the market.

What a completely BS, Red Herring of a reply!

I said that polarizers use in regard to dash cams do not work for road glare and they basically don't. For dash cams, polarizers are intended to ameliorate the reflection of the dashboard on the interior of the windscreen and that's what they are optimized for. You might get some ancillary effects from a CPL but it must be tuned and adjusted for the windshield.

But whether they do or do not work for more than just the windscreen is irrelevant to your actual remark that I questioned,........ "Polarizers don't work", which is simply not the case. Nothing in the quotes from me you offer in any way suggest that I ever claimed that CPLs "don't work".

BTW, as I mentioned previously, I just don't know quite how you reach some of your conclusions I often see posted here. You keep making assumptions stated as fact that are demonstrably untrue. This time you state that finding a good CPL is allegedly difficult because, "you're dealing with a miniature size which tends towards the novelty end of the market". Novelty? There are many professional level high quality small polarizers available from high end manufacturers and they have been around for decades all the way back to 8mm and super8mm movie cameras and earlier for that matter. Some small polarizers are also used in microscopes and other scientific instruments. Tiffen, produces a wide range of camera polarizer types of all sizes including CPLs as small as 25mm. Schneider Kreuznach produces CPLs as small as 10mm in size. Novelty indeed.

 

[jokiin]

from your earlier comments @Dashmellow the way I interpreted it is that a CPL filter could possibly reduce the road glare but that's not how they are adjusted and used in the way we use them on dashcams where we are using them to counter the specific dashboard reflections which are a known quantity to deal with and we can use a fixed position to attain the desired result, I'm no photographer but I'd reasonably expect if using a CPL to counter road glare that would be something that would not just require a different adjustment but would possibly require constant adjustment as the glare issue changed as that's not a fixed value we're dealing with like the dashboard, if that were the case I guess there might be some instances where the adjustment we used could also show some benefit to glare outside but due to the changing nature of the external stuff that would be just luck of the draw???

 

[Dashmellow]

from your earlier comments @Dashmellow the way I interpreted it is that a CPL filter could possibly reduce the road glare but that's not how they are adjusted and used in the way we use them on dashcams where we are using them to counter the specific dashboard reflections which are a known quantity to deal with and we can use a fixed position to attain the desired result, I'm no photographer but I'd reasonably expect if using a CPL to counter road glare that would be something that would not just require a different adjustment but would possibly require constant adjustment as the glare issue changed as that's not a fixed value we're dealing with like the dashboard, if that were the case I guess there might be some instances where the adjustment we used could also show some benefit to glare outside but due to the changing nature of the external stuff that would be just luck of the draw???

Yeah, basically. I mean, polarizers are generally used to block reflections coming from a fixed plane at a particular angle and are adjusted accordingly. If you adjust it for the windscreen reflections from a fixed position it will not be optimized for the road surface and especially so if the plane of the road is constantly shifting as well as the angle of the sun as one is driving along. A CPL with ameliorate other reflections in the FOV but it would be more of a random ever-changing effect as you point out. For example, I've noticed that a CPL will sometimes help me see through the windshields of oncoming vehicles and sometimes not.

As for my above commentary, I originally was questioning where c4rc4m came up with the notion that "some" people claim that "CPLs "don't work", and now he seems to want to imply he somehow heard that from me.

 

[jokiin]

Yeah, basically. I mean, polarizers are generally used to block reflections coming from a fixed plane at a particular angle and are adjusted accordingly. If you adjust it for the windscreen reflections from a fixed position it will not be optimized for the road surface and especially so if the plane of the road is constantly shifting as well as the angle of the sun as one is driving along. A CPL with ameliorate other reflections in the FOV but it would be more of a random ever-changing effect as you point out. For example, I've noticed that a CPL will sometimes help me see through the windshields of oncoming vehicles and sometimes not.
.

ok thanks for confirming

As for my above commentary, I originally was questioning where c4rc4m came up with the notion that "some" people claim that "CPLs "don't work", and now he seems to want to imply he somehow heard that from me.

I think with the written word it's not always easy to know if someone has interpreted the meaning correctly compared to when you speak with someone directly, it's not really a straightforward subject

 

[Dashmellow]

ok thanks for confirming



I think with the written word it's not always easy to know if someone has interpreted the meaning correctly compared to when you speak with someone directly, it's not really a straightforward subject

Yeah, especially on the internet, it would seem.

 

[c4rc4m]

Yeah, basically. I mean, polarizers are generally used to block reflections coming from a fixed plane at a particular angle and are adjusted accordingly. If you adjust it for the windscreen reflections from a fixed position it will not be optimized for the road surface and especially so if the plane of the road is constantly shifting as well as the angle of the sun as one is driving along. A CPL with ameliorate other reflections in the FOV but it would be more of a random ever-changing effect as you point out. For example, I've noticed that a CPL will sometimes help me see through the windshields of oncoming vehicles and sometimes not.

As for my above commentary, I originally was questioning where c4rc4m came up with the notion that "some" people claim that "CPLs "don't work", and now he seems to want to imply he somehow heard that from me.

Well I would saw that's what your answer implied in effect. It's certainly true what you say that CPL's need to be adjusted to the light. However, to say it's either the dash or road is a little too simplistic in reality because as the angle of the reflection off the road changes, or the glare from the sun, the degree of polarisation will change also. So yes sometimes it will be poor but at other times when the angle is right or near right, the effect will be more marked. So to simply say they're not much use unless used for one or the other isn't really true. What's more accurate to say is that if it's adjusted to reduce dashboard reflection, then the degree of reduction in reflection from outside of the car may vary. However, that also means there is some benefit for both purposes to having one fitted. The only real time the loss of light may become an issue is in low light conditions such as night. But even then, there's nothing to stop you from removing the lens. So personally I wouldn't state that "it won't do both at the same time", implying that it's of no use for that purpose, because it will do both, just to very varying degrees according to the angle of incidence.

Also, something no-one has touched upon, I wonder what effect as an alternative for night, a night driving lens would have? For the eye, these reduce glare and increase contrast and brightness. However, of course a sensor chip isn't an eye, so it would be interesting to see what effect it had, whether it just added a yellow colour cast, dropped the available light ND style or actually had some benefit.

 

[jokiin]

dashboard and instrument reflection can be an issue at night, you do get a darker image but a CPL can still be beneficial, some examples here https://dashcamtalk.com/forum/threa...ns-filter-for-street-guardian-sg9665gc.14256/

 

[Dashmellow]

Well I would saw that's what your answer implied in effect. It's certainly true what you say that CPL's need to be adjusted to the light. However, to say it's either the dash or road is a little too simplistic in reality because as the angle of the reflection off the road changes, or the glare from the sun, the degree of polarisation will change also. So yes sometimes it will be poor but at other times when the angle is right or near right, the effect will be more marked. So to simply say they're not much use unless used for one or the other isn't really true. What's more accurate to say is that if it's adjusted to reduce dashboard reflection, then the degree of reduction in reflection from outside of the car may vary. However, that also means there is some benefit for both purposes to having one fitted. The only real time the loss of light may become an issue is in low light conditions such as night. But even then, there's nothing to stop you from removing the lens. So personally I wouldn't state that "it won't do both at the same time", implying that it's of no use for that purpose, because it will do both, just to very varying degrees according to the angle of incidence.

Also, something no-one has touched upon, I wonder what effect as an alternative for night, a night driving lens would have? For the eye, these reduce glare and increase contrast and brightness. However, of course a sensor chip isn't an eye, so it would be interesting to see what effect it had, whether it just added a yellow colour cast, dropped the available light ND style or actually had some benefit.

My original query was a request for you to provide examples to prove your assertion that “people say CPLs Don't Work” and again here you are attempting to claim that I myself made such a claim or somehow “implied” such a thing which is laughably untrue.

And now, once again, we get another one of your verbose, sort of convoluted posts theorizing and postulating about the subject, making assertions that are essentially speculative, or at best irrelevant to how and why dash cam owners actually use CPLs and what they expect of them. It seems increasingly obvious from your comments that you likely have no actual hands-on experience with using a polarizer on a dash cam but rather are reaching your conclusions based on random screen shots seen here on DCT along with conceptualization about how you believe CPLS function in the average motor vehicle; not a valid or useful approach. And attempting to twist what I actually said elsewhere into an effort to prove some tangential point while still avoiding the original question seems a bit disingenuous.

In any event, CPLs work quite well on dash cams for their designated purpose but they have some limits and are not intended to mitigate road glare or other extraneous reflections even if they might indeed do that on occasion.

 

[TonyM]

Back to M2 testing...
Here's the M2 with EV and contrast adjustment set to "0", showing how it tends towards overexposure in overcast situations. Both the M2 and GC are on centre-weighted metering, and no CPL on either for the purpose of a fair comparison .

The M2 exposure-priority is set to low light. Previous testing showed quite a marked difference between low and bright light priority. Perhaps it could do with a mid-light priority? FWIW I think the GC exposure is close to the actual conditions.

60fps:

Even with the M2 set at 60fps, the GC @ 30fps and better metering has selected a shorter exposure time leading to less motion blur.


30fps:

The over-exposure and consequential longer exposure time almost makes the difference between whether a number plate is legible or not

 

[TonyM]

60fps and 30fps night-time screenshot comparisons from yesterday, with the same settings (default EV, contrast, no CPL).

M2 1080p60:

GC 1080p30:


M2 1080p30:

GC 1080p30:


For me, the M2 night-time 30fps is unusable. At least the 60fps footage maintains a more reasonable exposure with less noise.

I'll try running some comparisons with the M2 on bright light priority to see what difference that makes.

 

[c4rc4m]

Back to M2 testing...
Here's the M2 with EV and contrast adjustment set to "0", showing how it tends towards overexposure in overcast situations. Both the M2 and GC are on centre-weighted metering, and no CPL on either for the purpose of a fair comparison .

The M2 exposure-priority is set to low light. Previous testing showed quite a marked difference between low and bright light priority. Perhaps it could do with a mid-light priority? FWIW I think the GC exposure is close to the actual conditions.

The GC pics look the better or the two in my opinion.

I must admit, Mobius are not alone in this, clearly GC are included as well as no doubt others, but I personally think adjustment should be automatic and dynamic according to the lighting conditions. After all it's going to be a PITA, especially more so with a camera with no screen, to have to stop and adjust the light setting every time it goes dull, or you transfer from day to dusk to night.

I guess a "night" button could be added to alleviate the day / night part of this. However, would it simply not be better to try to have accurate metering that didn't require user intervention? I don't remember ever having to adjust any DSLR or compact camera or video camera in automatic mode to compensate for a change of lighting state. Rather the conditions were metered (usually) accurately and automatically. The issue is usually with focus in low light and that's a non issue with fixed focus.