SGGCX2PRO low light quality a little too good - suggestions for settings to mimic human night vision?

TheStoryGirl

New Member
Joined
Jul 22, 2019
Messages
5
Reaction score
2
Location
Seattle
Country
United States
Hi all!

Does anyone have any suggestions for how to tweak the various picture settings to best mimic human night vision? Currently, my new SGGCX2PRO is a little too good on my nighttime commute!

To explain:

I drive in an area that has a lot of severely addled pedestrians who frequently stumble or dart into traffic without warning. I drive VERY carefully, constantly scanning the sidewalks, but I felt I needed a dash cam for protection anyway, should someone ever walk into my car when it's absolutely impossible to brake in time.

Since the goal of my dash cam is to prove to a cop/lawyer/insurance adjuster that an accident was unavoidable, I want the video to show only what I was actually seeing at the time. Right now the video playback on my computer makes the streets seem more brightly lit and visible at night than they actually are in person. I don't want someone looking at the video and saying, "I clearly saw that guy a second before he walked in front of your car - why didn't you?"

I've been tweaking the exposure settings, but it still seems like the streets are pretty well-lit! So I'm not sure I'm doing it correctly, and I'm not totally clear on how bitrate and AE metering might impact the visual image.

Any suggestions would be hugely appreciated!
 
That is going to depend on how many carrots you have been eating!

And after the Uber incident where their self driving car mowed down a pedestrian at night and they then released some darkened footage to make it look like the pedestrian's fault when it wasn't, I'm not sure that a court is going to believe your darkened video anyway! Better to have full detail, and then explain to the court that it actually looked like this to you, and provide them with an edited version in addition to the original that has been darkened to match your memory. They then have no choice but to believe that what you say you saw is what you saw since nobody can prove any different, unless they check your grocery receipts and your dustbins to see how many carrots you have actually eaten!
 
I've expressed the same concerns you have in the past. That is one of the 'problems' with the newer dash cams that have improved low light capabilities - they see better than the human eye. To the best of my knowledge the only way to adjust this would be to dial back the exposure settings but that will likely affect daytime performance.
 
I've expressed the same concerns you have in the past. That is one of the 'problems' with the newer dash cams that have improved low light capabilities - they see better than the human eye. To the best of my knowledge the only way to adjust this would be to dial back the exposure settings but that will likely affect daytime performance.

Thanks DT MI! It's sort of funny to me that a lot of dash cams are unhelpfully "too good." For example, after initially trying a different brand, I quickly decided a 170 degree ultra-wide angle lens doesn't do a driver any favors by making objects look farther away on the video than they were in person, nor that the driver is moving faster than they were. That dash cam made me look like a Formula 1 driver on my commute when I in fact rarely get above 25 mph.

So after trying a different brand, I specifically picked Street Guardian for its sensible, human-ish 130 degree field of view, because I really didn't want to be questioned about why I didn't see someone coming from an extreme periphery.

I'd love to explain after the fact that my dash cam is "too good," but I worry about people's ability to absorb that kind of information. Usually people just go with their initial reactions and don't adapt to new information.

But thank you, it's good to know that there's only so much tweaking I can do here.
 
That is going to depend on how many carrots you have been eating!

And after the Uber incident where their self driving car mowed down a pedestrian at night and they then released some darkened footage to make it look like the pedestrian's fault when it wasn't, I'm not sure that a court is going to believe your darkened video anyway! Better to have full detail, and then explain to the court that it actually looked like this to you, and provide them with an edited version in addition to the original that has been darkened to match your memory. They then have no choice but to believe that what you say you saw is what you saw since nobody can prove any different, unless they check your grocery receipts and your dustbins to see how many carrots you have actually eaten!

Thanks, Nigel!

I'll keep your advice in mind, although I'm a little dubious about being able to convince people they shouldn't go with the regular video. People tend to naturally believe the first version of things they see or hear. It takes a lot of work to get them to really adapt.

It would be a lot easier to just show them a video that provokes exactly the same "OH SH**!!!* response that I might have, with exactly the same timing, rather than make them do forensic adjustments in their heads.
 
I'm with you on the field of view, too narrow or too wide and that will affect people's judgement a lot.

With the brightness, it depends on too many things, such as the brightness of the monitor it is viewed on, and is it daylight, room lighting, or lights off at night - your perception of the same video displayed on the same monitor will vary hugely.

I don't have this camera, but I wouldn't expect it to be far from normal human vision, so I wonder if it is actually a problem of the camera not matching the average human or your eyes not matching the average human? If it is the camera then maybe SG would consider adjusting the firmware, or providing a darker "human vision" option... a darker option might give an improvement in motion blur as a side effect.
 
You can always play the " i am only human" card.
And as long as we have not started to augment our physical performance that will be a good excuse, the dashcam are always there dont even blink its eye, you on the other hand as a human have a range of " faults" that do mean you are a potential danger in traffic.
Thats why we have traffic codes and driving licenses, so there are a minimum you know and skills you have, and the traffic code a set of rules that should mean the danger to life and material should be minimized if all abide by that set of common rules.
But it only take a fraction of a second of you doing something, like changing radio station or something, and BAM ! you are in a accident.

And if you are to blame, you can of course just "loose" your memory card, no one can prove it was there when the accident happened.
But i must admit people should take the blame if they have the blame, otherwise you will be the kind of person we try and protect us self against by having a dashcam.
 
I'm with you on the field of view, too narrow or too wide and that will affect people's judgement a lot.

With the brightness, it depends on too many things, such as the brightness of the monitor it is viewed on, and is it daylight, room lighting, or lights off at night - your perception of the same video displayed on the same monitor will vary hugely.

I don't have this camera, but I wouldn't expect it to be far from normal human vision, so I wonder if it is actually a problem of the camera not matching the average human or your eyes not matching the average human? If it is the camera then maybe SG would consider adjusting the firmware, or providing a darker "human vision" option... a darker option might give an improvement in motion blur as a side effect.

Oh, I totally agree with you that there are a bunch of subjective variables when you look at camera/computer monitor/individual eye-sight and so on. It's not possible to get all the way to "reality" for every person. I think my night vision is at least "normal," based on being able to read well in low-light and preferring dimmer settings on my device screens, but without formal testing, who knows? I'm very near-sighted, which is adequately corrected with glasses but not as "perfect" as a lot of people's eyesight or a camera's.

That said, the video of the route I drive every night really "feels" different when I watch it - blocks that I know are dark and shadowy if you're on foot look like Walmart parking lots on the video. The camera's contrast seems to be doing better than the human eye, maybe?

Maybe I'll just cross my fingers and hope that cops/lawyers/adjusters have enough experience to figure it out!
 
I would think going with either Center or spot metering would set the lighting more to being controlled by the illuminated area of your headlights. Then you can also dial in the EV value to a negative adjustment of two or three to try and darken the overall frame. Worth a try at least.
 
I didn't remember in detail but it is center-weighted. There is no spot. Not sure if low-cut would mean low pass or low reject? Probably low reject so that would be brighter.

Anyway full frame is certainly going to be trying to bring up the visibility in under lit situations at the edges so that's probably not the one you want.
 

Attachments

  • 20190905_202208.jpg
    20190905_202208.jpg
    131 KB · Views: 14
Last edited:
I didn't remember in detail but it is center-weighted. There is no spot. Not sure if low-cut would mean low pass or low reject? Probably low reject so that would be brighter.
the low cut option ignores the very bottom of the AE table, the center weighted option is generally fine but dark vehicles may benefit from using the low cut option
 
this may sound silly and is provably bad advice but what about using a darker filter in front of your camera lens to reduce the light input instead or in addition of the settings controls. Sunshades for your camera at night that you can remove during the daytime easy enough.
 
this may sound silly and is provably bad advice but what about using a darker filter in front of your camera lens to reduce the light input instead or in addition of the settings controls. Sunshades for your camera at night that you can remove during the daytime easy enough.

A dark filter over the lens, otherwise known as a ND filter (neutral density filter - available in different degrees of exposure modification) might achieve what the OP desires but it will cause an unwanted amount of motion blur in the video due to the slower shutter speeds it will evoke.
 
A dark filter over the lens, otherwise known as a ND filter (neutral density filter - available in different degrees of exposure modification) might achieve what the OP desires but it will cause an unwanted amount of motion blur in the video due to the slower shutter speeds it will evoke.
those for the drones like for DJI Phantom are not that expensive and they are near in size. ND3 or ND8 would be something to experiment with. the motion blur may likely be an unwanted effect like you say.
 
those for the drones like for DJI Phantom are not that expensive and they are near in size. ND3 or ND8 would be something to experiment with. the motion blur may likely be an unwanted effect like you say.

Maybe an ND filter would work (aside from the motion blur issue) but I question the notion of buying a camera that features outstanding low light performance only to intentionally compromise it. Perhaps a different camera with a different sensor might be a better approach?

Actually, if it were me and I were to encounter a situation like @TheStoryGirl is concerned about regarding a mishap with an impaired pedestrian, I would simply submit the bright, clear video to law enforcement, prosecutors, attorneys, or insurance adjusters along with a sworn affidavit describing the actual circumstances which would include a brief explanation about the differences between what the naked eye of a motorist can see vs what cameras with the latest technology can capture given the same low lighting conditions. Seeing the behaviors of an impaired pedestrian on video before an accident might actually be better evidence than just seeing someone stumble out of the darkness.
 
Maybe an ND filter would work (aside from the motion blur issue) but I question the notion of buying a camera that features outstanding low light performance only to intentionally compromise it. Perhaps a different camera with a different sensor might be a better approach?

Actually, if it were me and I were to encounter a situation like @TheStoryGirl is concerned about regarding a mishap with an impaired pedestrian, I would simply submit the bright, clear video to law enforcement, prosecutors, attorneys, or insurance adjusters along with a sworn affidavit describing the actual circumstances which would include a brief explanation about the differences between what the naked eye of a motorist can see vs what cameras with the latest technology can capture given the same low lighting conditions. Seeing the behaviors of an impaired pedestrian on video before an accident might actually be better evidence than just seeing someone stumble out of the darkness.
I totally agree with you. I think it would be easier for the law enforcement to understand how good of a camera that is against the idea that they are dealing with a doctored footage and not being able to explain it properly in your favor.
 
Maybe an ND filter would work (aside from the motion blur issue) but I question the notion of buying a camera that features outstanding low light performance only to intentionally compromise it. Perhaps a different camera with a different sensor might be a better approach?

Actually, if it were me and I were to encounter a situation like @TheStoryGirl is concerned about regarding a mishap with an impaired pedestrian, I would simply submit the bright, clear video to law enforcement, prosecutors, attorneys, or insurance adjusters along with a sworn affidavit describing the actual circumstances which would include a brief explanation about the differences between what the naked eye of a motorist can see vs what cameras with the latest technology can capture given the same low lighting conditions. Seeing the behaviors of an impaired pedestrian on video before an accident might actually be better evidence than just seeing someone stumble out of the darkness.
+1
 
...dark vehicles may benefit from using the low cut option
It also helps a lot with white vehicles as well in my experience.
 
...Actually, if it were me and I were to encounter a situation like @TheStoryGirl is concerned about regarding a mishap with an impaired pedestrian, I would simply submit the bright, clear video to law enforcement, prosecutors, attorneys, or insurance adjusters along with a sworn affidavit describing the actual circumstances which would include a brief explanation about the differences between what the naked eye of a motorist can see vs what cameras with the latest technology can capture given the same low lighting conditions. Seeing the behaviors of an impaired pedestrian on video before an accident might actually be better evidence than just seeing someone stumble out of the darkness.

Sorry for the late reply, for some reason I stopped getting notifications on this thread.

You make a really great point, @Dashmellow!

While I can personally swear to noticing a difference between my dash cam playback and real life, I'm sure there will be an out-of-pocket cost to have an expert swear that it's an issue with the tech. But oh well.

And I didn't think of this before, but I'm guessing it could be useful to have this thread itself as a time-stamped record of it being A Thing.

This is still way better than not having a camera at all. Yesterday morning I watched a cyclist run a stop sign while I was turning in a stop sign intersection (I was forced to brake for him). Then he overtook me on the right, zoomed straight at another stop sign intersection without any apparent intention of stopping, and almost crashed crashed into the side of the truck in front of me as it was legally turning right after a full stop.

Without a dash camera, that would have been a really bad day for the driver of the truck, as it would have been difficult to prove that he hadn't carelessly turned into the cyclist! I would have of course pulled over to provide a witness statement, but my camera footage would have been so much better.

If you're not a total jerk driver, dash cams are your best friend. :)
 
Back
Top