Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

verygreen

Curious member
Jan 16, 2017
2,997
11,620
TN
Thanks, this is amazing
i wonder if you can get some images of the left and right pillars and fenders when parked with cars on both side of you.
trying to see if my theory of blind spot is true.
Hm... I don't think there's a blind spot other than at the front of the car (addressed by the zoom camera, but not by the main, I believe).
What's your theory, anyway?

Getting a picture with cars nearby is going to be problematic, so the best I can offer you is this from the left side:
l1.jpg l2.jpg
 
  • Helpful
Reactions: Bladerskb

Bladerskb

Senior Software Engineer
Oct 24, 2016
2,772
4,690
Michigan
Hm... I don't think there's a blind spot other than at the front of the car (addressed by the zoom camera, but not by the main, I believe).
What's your theory, anyway?

Getting a picture with cars nearby is going to be problematic, so the best I can offer you is this from the left side:

thanks, do you have a model x or s?
my theory is that when parked side by side with other cars and attempting to back out both the fender and pillar cameras will be blocked from seeing cross traffic until you pull out half way for the pillars to be unblocked.

Ofcourse that depends what type of car is parked next to you and if you have a x or s.
 

verygreen

Curious member
Jan 16, 2017
2,997
11,620
TN
thanks, do you have a model x or s?
my theory is that when parked side by side with other cars and attempting to back out both the fender and pillar cameras will be blocked from seeing cross traffic until you pull out half way for the pillars to be unblocked.

Ofcourse that depends what type of car is parked next to you and if you have a x or s.
Well, you are correct that the side cameras would be blocked by a nearby car, but this is not a problem, because backup camera is pretty wide angle and so could be used to see cross traffic quite well (still need to pull a little bit out so that it is not blocked obviously).
 

lunitiks

Cool James & Black Teacher
Nov 19, 2016
2,698
5,990
Prawn Island, VC
BTW, forgot to mention, but the sensor is apparently aptina, according to the code, anyway.
Any references to camera model/product numbers?
All cameras, except rearview cam, uses Aptina / ON Semiconductor AR0132. Rearview cam uses OmniVision OV10635.
 

stopcrazypp

Well-Known Member
Dec 8, 2007
12,636
8,532

lunitiks

Cool James & Black Teacher
Nov 19, 2016
2,698
5,990
Prawn Island, VC
do you have access to information for the gps parts tesla use?
The Parts Manual vector/SVG-drawing of the GPS antenna sports a Hirschmann logo, so I'm pretty sure the antenna is a Hirschmann Car Communication unit. I really can't tell which model, though. I'd describe it as a "square, glued down version with a blue Fakra/SMB connector" :) Someone please crack open your rear window glass and cap some close-ups!!! :D

Anyway, here's where the GPS antenna is mounted (TMS):


Here's the GPS input port on the back of the MCU (like I said earlier, I'm unsure how this works on AP2, since the GPS antenna-cable now goes to the PX2 (not to the CID), and what happens in the PX2 I don't know):

The GPS module inside the CID is (or -- was -- at least in AP1) a UBlox LEA-6R-0-001:


Got a little off topic here, sorry :eek:
 
Last edited:

AnxietyRanger

Well-Known Member
Aug 22, 2014
9,408
7,978
EU
You guys are on fire! Very nice, thanks.

Any more comments on the fisheye. Indeed rather surprising to see how much of it is obstructed. Then again, perhaps the reality at that location means that part would be obstructed by the bonnet anyway.

Still, one could imagine a different kind of lens extracting more usable data/resolution...
 

Saghost

Well-Known Member
Oct 9, 2013
8,224
7,094
Delaware
@verygreen Wow indeed! I can't believe this is the actual FOV from fisheye. Check out that glare shield obstruction! (Made some quick paintshop color/level adjustments. @JeffK maybe you can do a better job at this?)

If you look carefully, I'm pretty sure you're seeing the rear edge of the hood reflecting just above the glareshield, and the A pillars up to about the middle of the image.

Given that, I don't think there's anything more to see that the glareshield is blocking. They could have tilted the camera up more, but they've already got more sky than they probably have a use for now.

There's no practical way to get the horizontal coverage they needed with excess vertical coverage on the lens, and the imaging sensor is probably cheaper to get with standard coverage than with an exotic high aspect ratio design.
 

lunitiks

Cool James & Black Teacher
Nov 19, 2016
2,698
5,990
Prawn Island, VC
If you look carefully, I'm pretty sure you're seeing the rear edge of the hood reflecting just above the glareshield, and the A pillars up to about the middle of the image.
No I don't think so. Maybe you're right about the hood, but those obstructions on the sides are definitely not the A-pillars --- instead they're the sides of the glare shield cove
 
Last edited:

JeffK

Well-Known Member
Apr 27, 2016
6,997
6,932
Indianapolis
@verygreen Wow indeed! I can't believe this is the actual FOV from fisheye. Check out that glare shield obstruction! (Made some quick paintshop color/level adjustments. @JeffK maybe you can do a better job at this?)
My point was, if they are using a red/clear filter for it's greater dynamic range adding any color besides red on the red pixels is just colorizing intensity values... it's fake color.

When you zoom in to the originals you can clearly tell which pixels are the red ones and what the RCCC pattern looks like. I de-mosaiced them, but if you were to take each non-red intensity pixel and convert it to three values for RGB (extrapolation, because each non-red pixel isn't R/G/B, it's all of them) then convert the red pixels to RGB as well (just using red data though) you'd be able to render a human viewable bitmap for what the camera is actually picking up.

If you simply de-mosaic the image and normalize the histogram to get as much dynamic range on one viewable image (with common image formats). You should get something with a large amount of dynamic range resembling an Ansel Adams photo (but not as sharp).

What, you mean there's no ready-made tools to do that? ;)
haha I don't know, I had to use a super simple filter application I made years ago as a naive college student.
 
Last edited:

Saghost

Well-Known Member
Oct 9, 2013
8,224
7,094
Delaware
No I don't think so. Maybe you're right about the hood, but those obstructions on the sides are definitely not the A-pillars --- instead they're the sides of the glare shield cove

Hmm. After looking at the image under magnification and then going out to my car, I'm not sure. Is this image from an S or an X?

On my X, the A Pillars don't overhang the windshield at all, except at the very bottom edge, and certainly don't have the curved transition this appears to.

On the image, there's a clear transition maybe a third of the way up the "side" from a sharp reflective edge to a broader flat/dark edge. If the A-Pillar contour in the region matched up, I'd say this was the transition from A-Pillar to glare shield, but on an X the A-Pillar isn't shaped like that.

From the images I can find online, it seems like maybe the S A-Pillars do have a curved overhang at the bottom?

I have trouble believing Tesla would have a sharp reflective edge at the bottom of the glare shield producing glare unless they had no choice for some reason.
 

lunitiks

Cool James & Black Teacher
Nov 19, 2016
2,698
5,990
Prawn Island, VC
I have trouble believing Tesla would have a sharp reflective edge at the bottom of the glare shield producing glare unless they had no choice for some reason.
There's a plastic rim at the bottom of the glare shield enclosure. Might be poor adjustment and the sun being reflected in the plastic.

Forget the A-pillars, you're not seeing them in the picture
 

AnxietyRanger

Well-Known Member
Aug 22, 2014
9,408
7,978
EU
Thank you for sharing. Great stuff.

I agree that the fisheye at the front may be use the glare shield to cover the bonnet/hood. It also clearly overlaps with the B pillar cameras on the front corners, though still it would seem the effective resolution in the overlap area of front corners - due to the fisheye effect and basically only half of the vertical used - may not be massive.

Photoshop also opens them with 1280x964, Channels: Count 1, 16 Bits, Header 0. Byte Order IBM PC generates dark grey images, Byter Order Mac generates sort of bright reversed ones.

None of these settings, I think, are ekeing out the full dynamic range, though. They are pretty crappy. More research needed.
 

Products we're discussing on TMC...

About Us

Formed in 2006, Tesla Motors Club (TMC) was the first independent online Tesla community. Today it remains the largest and most dynamic community of Tesla enthusiasts. Learn more.

Do you value your experience at TMC? Consider becoming a Supporting Member of Tesla Motors Club. As a thank you for your contribution, you'll get nearly no ads in the Community and Groups sections. Additional perks are available depending on the level of contribution. Please visit the Account Upgrades page for more details.


SUPPORT TMC
Top