My Mobius 2 / M2

[country_hick]

Some of the complaints will stop. Others should be lessened.
Work takes time. We now have a much needed improved firmware version to cheer...

https://www.rcgroups.com/forums/showpost.php?p=36351617&postcount=819

Today, 12:25 PM

Tom Frank
Dance the skies...


FW V2.03 Released
This is a bug fix release... no new or updated feature settings.

Bug fixes are:

* fixed issue with intermittent "flashing" caused by oscillating exposure in certain low light scenes with certain configuration settings.
* fixed issue with shutter button click being recorded at the beginning of some clips, causing some video players to attenuate the sound level.
* fixed progressive out of sync sound issue with successive video clips shot at 25 fps without rebooting the camera between clips.

 

[Harsh]

Firmware - 2.03

1080p, WDR - off, Metering - multi-spot, Bitrate - highest (CBR), EP - low-light, 50Hz.

Night - Mild fog - 1

@ 30 FPS

@ 60 FPS

Night - Mildly foggy - 2

@ 30 FPS

@ 60 FPS

SG9665GC

Night

@ 30 FPS

@ 60 FPS

Market

@ 30 FPS

@ 60 FPS

 

[Harsh]

Firmware - 2.03

1080p, WDR - off, Metering - multi-spot, Bitrate - highest (CBR), EP - low-light.

Daytime

@60 FPS

30 FPS vs 60 FPS comparison.

Daytime - 1

@ 30 FPS

@ 60 FPS

Daytime - 2

@ 30 FPS

@ 60 FPS

Daytime - 3

Goes blocky and mushy on a tree lined road.

@ 30 FPS

https://www.dropbox.com/s/rlvjgdx330dy6qx/3-30.mp4?dl=0

@ 60 FPS

https://www.dropbox.com/s/duu1caoqt5lbzpx/3-60.mp4?dl=0

Evening

@ 60 FPS

Once the light levels drop, things start going south.

I see a lot of hot pixels with 30 FPS at night if there's no street lighting , whereas 60FPS shows only 3 distinct hot pixels (videos in previous post).

Getting the same 'crickety' SD writing noise/interference that I get with my M1 which's installed in the rear. (Turn up the volume, it's in all the videos.)

 

[c4rc4m]

For the daytime one with macro blocking - it looks very unclear for the resolution. Your windscreen looks very reflective in those shots so the light may have had some effect. So far as the blocking is concerned, it seems to happen with overhanging trees but not always until you're under them which to my mind rules out light issues. If I had to point the finger, it's clearly back at encoding / bit rate issues due to the levels of detail in the picture. It would be interesting to know what the bit rate is on those setting because it looks low. If it's high, then it's got to be codec issues to my mind.

The night time videos look improved. The GC looks a bit better but suffers from sparkling pixels caused by gain. The Mobius doesn't but isn't as good.

As for the 3 pixels on the Mobius, you need to send the camera back. Those look like damaged pixels which means you have a damaged image sensor. Typically you see worse examples of this where a camera suffers laser damage. However, any pixel however damaged can be damaged in such a way as to record black or white.

Here's an example of laser damage to a camera sensor - entirely different cause, but you can clearly see the permanently stuck white pixels and the line of white pixels where an entire row has been taken out.

In my opinion, the "hot pixels" as you describe them indicate from their lack of movement and their appearance in daytime, that your Mobius has sensor damage, although the exact cause ie faulty manufacture or sun / glare damage is undetermined:

https://i.ytimg.com/vi/J0TgaGePhJA/maxresdefault.jpg

 

[jokiin]

As for the 3 pixels on the Mobius, you need to send the camera back. Those look like damaged pixels which means you have a damaged image sensor.

all sensors have this, there is a software process which masks this effect, something they'll need to fix still

 

[c4rc4m]

However, manufacture based Hot pixels should only show up at high ISO's and long exposures. In daylight, that's a fault because the manufacturer should have mapped those out at the factory.

The other alternative of course is the sensor is getting very hot during normal usage. Again I'd regard that as a fault as the heating sinking should be better. Looking at the above videos, yes it's India, but it's also overcast so not as if the camera is in direct sunlight on the dash.

 

[jokiin]

that's a fault because the manufacturer should have mapped those out at the factory.
.

there's a bit more to it than that

 

[Harsh]

For the daytime one with macro blocking - it looks very unclear for the resolution. Your windscreen looks very reflective in those shots so the light may have had some effect. So far as the blocking is concerned, it seems to happen with overhanging trees but not always until you're under them which to my mind rules out light issues. If I had to point the finger, it's clearly back at encoding / bit rate issues due to the levels of detail in the picture. It would be interesting to know what the bit rate is on those setting because it looks low. If it's high, then it's got to be codec issues to my mind.

The night time videos look improved. The GC looks a bit better but suffers from sparkling pixels caused by gain. The Mobius doesn't but isn't as good.

The bitrate is approximately 42 MBPS at the highest setting when using CBR, not low by any standard. My front windshield has a VTL70 clear film by 3M installed, don't know if that makes any difference in how the reflections are handled during the day.

After a thorough cleaning of the windshield from the inside yesterday, I pulled off the CPL on the GC and discovered it to be quite dirty (I smoke), had to clean the glass thrice on both sides using a lenspen to get rid of all the muck. Will redo the comparison within the next few days.

 

[c4rc4m]

The bitrate is approximately 42 MBPS at the highest setting when using CBR, not low by any standard. My front windshield has a VTL70 clear film by 3M installed, don't know if that makes any difference in how the reflections are handled during the day.

After a thorough cleaning of the windshield from the inside yesterday, I pulled off the CPL on the GC and discovered it to be quite dirty (I smoke), had to clean the glass thrice on both sides using a lenspen to get rid of all the muck. Will redo the comparison within the next few days.

That's quite interesting then when you consider the Mobius 1 was around 15mbs (?) or similar from memory and the GoPro Hero 2 around 25mbs (non pro tune). 42mbs certainly isn't low although not the highest of any camera out there.

Looks like it could be some encoding issue therefore.

 

[TonyM]

The M2 does appear to have hot pixel issues at night, but not so much that I'd say it affects

For the daytime one with macro blocking - it looks very unclear for the resolution. Your windscreen looks very reflective in those shots so the light may have had some effect. So far as the blocking is concerned, it seems to happen with overhanging trees but not always until you're under them which to my mind rules out light issues. If I had to point the finger, it's clearly back at encoding / bit rate issues due to the levels of detail in the picture. It would be interesting to know what the bit rate is on those setting because it looks low. If it's high, then it's got to be codec issues to my mind.

Are you looking at the youtube version, or the re-encoded dropbox download? I see a lot of noise but not so much blocking in the dropbox version.

 

[c4rc4m]

The M2 does appear to have hot pixel issues at night, but not so much that I'd say it affects

Are you looking at the youtube version, or the re-encoded dropbox download? I see a lot of noise but not so much blocking in the dropbox version.

I've only looked at the Youtube version on here at 1080P. Maybe Youtube's to blame then, although Harsh was also commentating he'd seen the blocking. I presumed he meant in the original.

 

[Harsh]

The original files are a whole lot better but there's certainly a little bit blockiness when tree cover is lost momentarily.

 

[TonyM]

I've only looked at the Youtube version on here at 1080P. Maybe Youtube's to blame then, although Harsh was also commentating he'd seen the blocking. I presumed he meant in the original.

The original files are a whole lot better but there's certainly a little bit blockiness when tree cover is lost momentarily.

Youtube (blocky) vs Dropbox (noisy):

 

[Harsh]

When the M2 adjusts to compensate for the momentary loss of tree cover, I do see a bit of blocking on my laptop screen. For most part, it's exactly as compared above.

 

[Harsh]

Post cleaning the front windshield from the inside and the GC's CPL using a Lenspen, the next day I noticed the VTL70 clear film was peeling off the glass from one of the top corners. The 3M SP70 had always appeared to have a very milky haze to it under white LED or white T5 street lighting, the slight peeling gave me a strong urge to go park under a streetlight and check if I could compare the difference through the inch of (now) exposed glass. Wasn't really surprised as I did expect a cleaner/clearer view but never expected the difference to be as drastic as it turned out to be. Guess what happened next?





Didn't look as dark as can be seen in the pictures above when I'd gotten it installed. The samples stuck to the dealer's office glass door (that sold me) also looked like a non-tinted clear film. Seriously confused about the shade of the film that's come off.

Not too bummed as this was their inexpensive version. would it have been the CR70, that would've hurt.

Below are a couple of comparisons with the GC with its CPL removed.

M2 - Firmware - 2.0.3 - 1080p, WDR - off, Metering - multi-spot, Bitrate - highest (CBR), EP - low-light.
GC -Firmware - Beta 27 - 1080p, WDR - on, Metering - centre, Contrast - medium.

Night

@30 FPS

@ 60 FPS

GC

Market

@ 30 FPS

@ 60 FPS

GC

I've come to understand and realise that it's not possible to accurately test or compare different options/adjustments against each other unless you have more than one unit of the same camera. If two different options/parameters/settings aren't being recorded at the same instance using two cameras, the test is flawed.

Will not be doing anymore comparisons with the GC till the M2's firmware matures enough for it to be a fair comparison. Do intend on comparing it extensively to the M1 once the firmware gives me enough confidence (with or without WDR) to mount the M2 in the rear; next to the M1.

 

[SawMaster]

Thanks for the extensive vids Harsh The SG still looks better at this point, but it's known to be good in these conditions so Mobius has a very high bar to clear to equal or best the SG at night. I don't know enough to say whether that is possible with the hardware they use, and that is always a solid limit of performance and possibilities.

Looking forward to seeing the M1 vs M2 since I have an M1 but not an SG (yet!) and I will be better able to relate to that comparison. I'm still planning on getting a M2 and an SG, and now that I've got the other 3 positions covered in my van I can save up for those. I don't drive much at night but I still want better vids for when I do; the M1 is just a bit short of what I want to see.

Phil

 

[Harsh]

Firmware - 2.03

1080p, WDR - off, Metering - multi-spot, Bitrate - highest (CBR), EP - low-light.

No street lighting

@60 FPS

https://www.dropbox.com/s/zbxdr9udp0132qp/Mgoing.mp4?dl=0

 

[Harsh]

An example of how well the M2 handles the shadow detail, same goes for the M1.





M2's definitely over exposed though.

Santa's welcoming people to come get drunk. Probably the worst Santa I've seen!

Happy holidays!

 

[Harsh]

Finally mounted the M2 in the rear using a modified version of the 15° FPV tilt mount that holds my M1.

A quick rear mounted night-time comparison with the M1.


Mobius - (F.W V. 2.41) - 1080p @ 30 FPS, WDR - On, Bitrate - 18 MB/s, Exposure- -10, Contrast - +10, CPL - No.

Mobius 2 - (F.W V. 2.03) - 1080p @ 60 FPS, WDR - Off, Metering - Centre weighted, Bitrate - 42 MB/s, EP - Bright-light, CPL - No.


M1

M2

Combined

M1

M2

Combined

As per a post from Tom Frank on RCG, the M2's next firmware update will include some low-light improvement. Hoping that things finally pick up and I can leave the M2 mounted in the rear while the M1's being experimented with for a side install. (I could just experiment with the M2 but the M1 is a lot more flexible to work with when it comes to mounting options.)

 

[kamkar]

The current beta i have on my M2 dont seen to have any changes for the low light side of things, but the daytime footage do seem to have gotten a lift as far as i can tell comparing to older footage.