GitUp F1 with 5MP f1.6 6-22mm varifocal lens

TonyM

Well-Known Member
Joined
Jul 30, 2013
Messages
4,222
Reaction score
3,904
Country
United Kingdom
Dash Cam
Mobius, SG, Viofo
After experimenting with this f1.6 6-22mm varifocal lens on a Mobius 1, I decided to give it a go on a GitUp F1.



This particular lens is from Treeye via AliExpress who have an excellent track record of customer service. They have supplied varifocal, wide angle and telephoto lenses to a number of DCT members who have experimented with third-party lenses, primarily on Mobius cameras but also the F1 and some others.



The 6-22mm lens is designed for a maximum 1/2.5" sensor, which matches the Sony IMX317 inside the GitUp F1. The back focal length is specified as 6.2mm, although there is some tolerance in where the lens is set relative to the camera body. To position the lens at about 6mm from the sensor I used an M12 extension ring between the lens holder and the lens.



I set the lens to its widest zoom setting at the 6mm end of the range, which should mean the lens is operating at its widest f1.6 aperture - the aperture reduces as the lens is zoomed towards its telephoto end. CORRECTION: The lens was not set at its widest zoom, in fact it was much closer to the telephoto end of the zoom range.

On a static scene there is good sharpness and depth of field, although the detail drops off a bit to the sides of the frame. I should point out that getting the focus point just right for optimal depth of field can be tricky with these high magnification varifocal lenses, so my initial results may not represent the best possible performance from this camera & lens combination.

The F1 is set to 4K 30fps with superfine quality (72Mbps) and medium sharpening. Click the image then double-click to see it at full resolution.



In this mostly static street scene there is good detail from front to back, although the potted plants to the left of centre have lost some definition. The red car on the right is approaching at about 10mph.



This comparative image from a Mobius 1 (1080p 30fps) with the same 5MP 6-22mm varifocal lens has considerably less texture detail beyond the car immediately in front. The Mobius 1 image is zoomed in slightly more than on the F1, partially due to the crop factor on the smaller sensor in the Mobius 1. The focus point on the M1 is also set a bit closer, which affects the depth of field and accounts for some of the loss of detail in the background.


Moving on to another street scene at low speed there is generally good detail in the F1 image, although there is some loss of detail again on the left side:


The Mobius 1 once again has less detail throughout the image.


Driving on the open road, the F1 starts to show much more motion blur despite the f1.6 aperture:


By comparison the M1 delivers a more useful image with less motion blur:


The same higher degree of motion blur can be seen in this example from the F1:


Whereas the Mobius 1 maintains its performance advantage:


Both the F1 and M1 were mounted side-by-side on a metal plate attached to the same sturdy screen mount, so they should experience a similar degree of vibration or movement within the car.

In conclusion, the 5MP f1.6 6-22mm lens performs fairly well on the 8MP GitUp F1 in static and slow-moving scenes. This might make it useful in certain situations. However it shows much more motion blur at medium and high speeds than is acceptable for use as a dashcam.
 
Last edited:

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
Tony,
How about an experiment with the 6-22mm lens ?
:)
Photo mode, pointed at an evenly illuminated scene (sky? ) take a pic at wide, medium, and tele zoom.
Report what the EXIF data reports as shutter speed and ISO.
Maybe even a couple other lenses for comparison.
 
OP
TonyM

TonyM

Well-Known Member
Joined
Jul 30, 2013
Messages
4,222
Reaction score
3,904
Country
United Kingdom
Dash Cam
Mobius, SG, Viofo
Tony,
How about an experiment with the 6-22mm lens ?
:)
Photo mode, pointed at an evenly illuminated scene (sky? ) take a pic at wide, medium, and tele zoom.
Report what the EXIF data reports as shutter speed and ISO.
Maybe even a couple other lenses for comparison.
Good idea!
I'll try that with the M1 and Maxi sometime soon
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
Good idea!
I'll try that with the M1 and Maxi sometime soon

I don't know about the Gitup F1 but EXIF data from the Mobius 1 "may" have limited value. The camera is programmed for default values which essentially revolve around the original A lens (such as focal length and default aperture) that was installed on the camera when it was first introduced. The camera is not sophisticated enough to know that you have installed a 6-22mm varifocal ƒ/1.6 lens and so may or may not have the ability to accurately calculate the exposure numbers. Of course, it will indeed report ISO and shutter speeds but I was never entirely sure if those numbers were accurate after you install a lens different from the default. I repeatedly queried Peter about clarifying how the EXIF values function on the Mobius particularly as it relates to having different lenses on the camera as I was trying to work with the A2 lens camera that was sent to me for testing and wanted to know the focal length and default aperture but never got a response. Up until that time Mobius did not want to reveal the focal length and aperture of their lenses. I guess try it and see what you get. Personally, I've not tried looking at EXIF data with the varifocal installed so it should be interesting to see what you get.
 
Last edited:

Matt_V939

Active Member
Joined
Apr 25, 2015
Messages
413
Reaction score
208
Country
Taiwan, Province Of China
Dash Cam
Mobius1&Mini,ViofoA118C+A119V2,(Fake)DODL500, YI,GP3,808#23
...
Both the F1 and M1 were mounted side-by-side on a metal plate attached to the same sturdy screen mount, so they should experience a similar degree of vibration or movement within the car.
...
ummm, I'd say not, because their mass (weight) is different so their proper frequency is not the same
Vibration.jpg
 

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
Dashmellow,
My guess is the lenses aperture and focal length reported are 'firmware' values.
It does not seem to change with lighting.
The ISO and shutter seem to be constant with constant lighting.
If would be easy to compare the EXIF with an 'Open' lens and again with a ND filter to see if the calculation is actually light based :)
Hear are the result between an open factory lens on a MAXI and with a 'good' CPL.
Same scene, Kitchen, light on.
Shutter_ISO_MAXI.jpgREPLACED IMAGE
Reduced lighting and shutter went to 1/30", then ISO gain started going up to 800.
Then image got dark. Max exposure seemed to be 1/30", ISO 800.
Four images.

Left Dark limit,
Dim, no CPL
Bright, CPL
Bright, No CPL (Bright is normal kitchen light
 
Last edited:

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
Dashmellow,
My guess is the lenses aperture reported is a 'firmware' value.
It does not seem to change with lighting.
The ISO and shutter seem to be constant with constant lighting.
If would be easy to compare the EXIF with an 'Open' lens and again with a ND filter to see if the calculation is actually light based :)
Hear are the result between an open factory lens on a MAXI and with a 'good' CPL.View attachment 40173

Thanks,@Rocketvapor. I think we are saying the same thing. I guess I haven't been clear whether any programmed aperture firmware values affect the accuracy of the ISO and shutter speed readings in any way as seen in the EXIF data. For example, some of us have been experimenting with "Starlight" ƒ/1.2 lenses. It would be interesting to check the EXIF values when using lenses that go outside the firmware based aperture parameters.

I guess for Tony's purposes it wouldn't really matter if it was perfectly accurate or not as it will definitely show the basic differences between what the two lenses are doing.

Here I shot two still photos with a vintage Mobius 1 with the original "A" lens. One is a normal image and the other simply had the lens cap on. I wish I had some more time this morning as I would do some testing with different lenses under the same lighting.

exif2.pngexif1.png
 
Last edited:

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
My GUESS,
Is with the f/1.2 lenses the camera will set the exposure for the exposure, not knowing what lens you had on it.
As long as the weighted EV was within range. Looks like the original Mobius has an upper ISO limit of 900, the MAXI 800.
1/30" is the maximum shutter which makes sense with a camera set up for video.
Bet the maximum video shutter is faster for 60fps.

Exposure (with enough light) seems to be accurate with Looong lenses and aperture settings up to f/16.
 
Last edited:

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
My GUESS,
Is with the f/1.2 lenses the camera will set the exposure for the exposure, not knowing what lens you had on it.
As long as the weighted EV was within range. Looks like the original Mobius has an upper ISO limit of 900, the MAXI 800.
1/30" is the maximum shutter which makes sense with a camera set up for video.
Bet the maximum video shutter is faster for 60fps.

Exposure (with enough light) seems to be accurate with Looong lenses and aperture settings up to f/16.

You're probably right. One or more of us should do some testing with the ƒ/1.2 lenses vs the stock lens and see what results we get. It won't be me for now as just yesterday for some mysterious reason, the Mobius cam I've been using for the last year or more with various experimental lenses including the varifocals suddenly failed for no apparent reason and so far extensive troubleshooting hasn't revealed the problem. :( It's starting to look like I may need to replace the PCB. Between that and the third replacement windshield that was installed on my truck yesterday I'm in a bit of a camera scramble anyway so it will take a little time to get it all sorted out. I was advised it would be best not to put any pressure on the glass for 48 hours. Hopefully third time is the charm for defective windshields!

Anyway, did you notice in the EXIF data that for both images there is one number (ƒ/1.8) designated as the "F-number" and another ((ƒ/1.7) as the "Lens Aperture"? I don't understand what that means. Do you?
 

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
Sorry about the windshield problems.
Do the windshields show the corner curvatures before installation, or only after pressed and glued into place?
 

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
Again, a guess.
Maximum lens aperture might be a limit for calculations,
With the selected lens option being f/1.8.
Selecting another lens Option 'Might' give you an f-number of f/2.0 but keep the Maximum of f/1.7

Try selecting other options. What does the reported aperture look like?

Mine says 2.8 and 3 ????
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
Sorry about the windshield problems.
Do the windshields show the corner curvatures before installation, or only after pressed and glued into place?

The warped corners only really show up after the installation. Before installation, we all examined the glass on the workbench and it looked normal. Even after install, it looked OK. It was only after I got back to my house and saw the windshield with the reflections from the tree canopy above that the warping was so obvious. Last week we had the first heavy rain for quite some time and I discovered the my windshield wipers lose contact with the glass when they pass over the warped corners. The amazing thing to me (and the guys at the glass shop) is that the manufacturer claims the glass to be "in spec" and denies the problem. This newest one is from a different manufacturer and so far looks much better. Interestingly, it is much clearer to look through as well. The other two had a slight haze which I also complained about but had a harder time demonstrating.
 
Last edited:

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
The haze was probably caused by high humidity and heat coming across the Pacific in a steel container :)

I have a 2003 Tundra so I will watch for this if (when) I ever have to replace the glass.
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
Again, a guess.
Maximum lens aperture might be a limit for calculations,
With the selected lens option being f/1.8.
Selecting another lens Option 'Might' give you an f-number of f/2.0 but keep the Maximum of f/1.7

Try selecting other options. What does the reported aperture look like?

Mine says 2.8 and 3 ????

I'll have to check the EXIF data on the Maxi and see what I get. Obviously, the different chipset will provide different results than the original Mobius which is Novatek.
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
The haze was probably caused by high humidity and heat coming across the Pacific in a steel container :)

You're probably right!! :smuggrin: I don't know where the newest one was manufactured but the glass shop consulted directly with Toyota about the problem and went with a supplier they recommended.
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
Again, a guess.
Maximum lens aperture might be a limit for calculations,
With the selected lens option being f/1.8.
Selecting another lens Option 'Might' give you an f-number of f/2.0 but keep the Maximum of f/1.7

Try selecting other options. What does the reported aperture look like?

Mine says 2.8 and 3 ????

Sounds plausible. Those calculation limits are why I'm wondering about the EXIF results it spits out. It will be interesting to explore this further.
 

Rocketvapor

Active Member
Joined
Apr 7, 2016
Messages
165
Reaction score
98
Country
United States
The exposure calculation seems to accurately handle putting a CPL, with about 2/3 to 3/4 stop light loss in front of a bare open lens.
Someplace?? I have some ND filters to try.
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
Again, a guess.
Maximum lens aperture might be a limit for calculations,
With the selected lens option being f/1.8.
Selecting another lens Option 'Might' give you an f-number of f/2.0 but keep the Maximum of f/1.7

Try selecting other options. What does the reported aperture look like?

Mine says 2.8 and 3 ????

I was just re-reading your post here and wanted to clarify. On the Mobius 1 GUI you have several lens options....A - B & C. These options are color balance LUTS (profiles) for each individual lens. These selections do not affect any aperture data, only color.
 

Dashmellow

Well-Known Member
Joined
Sep 22, 2013
Messages
14,187
Reaction score
15,670
Location
Uncanny Valley (●_●)
Country
United States
Dash Cam
Umpteen
The exposure calculation seems to accurately handle putting a CPL, with about 2/3 to 3/4 stop light loss in front of a bare open lens.
Someplace?? I have some ND filters to try.

I think you are probably right about all this. I simply went with the Lens cap option,'cause it's the ultimate ND filter! :smuggrin: (and it didn't take any work)
 
Thread starter Similar threads Forum Replies Date
Z F1 13
TonyM F1 15
TonyM F1 113
TonyM F1 0
J F1 1
Top