-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Could disguised Model X actually be a Model 3 test mule?
- Thread starter ToddRLockwood
- Start date
-
- Tags
- Model 3
No. For other brands it is quite common for the first spy shots to appear about 3 years prior to when the car is introduced. So model 3 in 2018 actually fits quite nicelly into such a timeline. I can't say if it's a model x or model 3, but I certaonly hope it's a model 3 since that would mean they are on the right track for it.
I'm not a native English speaker, so hopefully let's not get stuck on individual words. By "statistically relevant" I was making the point that at low resolution, even the difference of a single pixel in the assumed wheel size can alter the outcome when extrapolated to calculate wheelbase length from that. Hope the intent there is clear.
As for the low resolution, camera angle, lens geometry differences, lossy video encoding imprecision - these things tend to add up. As does the effects of scaling, which happens if digital zoom happens on the camera or in post-processing.
Two questions:
1) Where does your 75 pixels of wheel diameter come from?
2) Where does your assumed ~20% difference in wheel base come from?
And can you trust these?
1) In the Full HD frame, the width of the wheel is around 30 pixels at best. At no point during the 1:32 minutes does it become better. The author in the original post scaled up and you measured a post-processed image. He probably also made his measurements based on an image scaled up through some digital algorithm, from an already poor-quality origin.
2) Even without any scaling, there is already significant pixel bleed in the wheels, making the tires and wheelwells almost disappear next to the bright, bleeding wheels. The wheels seem to drive on hard ground without any tire underneath. When the source material is 30px wide at best, get that assumption even a couple of pixels wrong and the difference extrapolated over the length of the wheelbase becomes signficiant.
I would like to understand where your date on the 75px and the ~20% come from. Maybe I missed something? But if they come from the original poster's post only, I think that may be a misleading data source.
The point isn't that there can't be be difference in the wheelbase. It is quite possible there is. But enough to suggest something other than Model X? More on my thoughts on the analysis of the wheelbase of this mule in the other thread.
Your English is very good!
I should begin by saying that I am not following Model X development as closely as others on this thread. If Tesla or another reliable source has indicated that the Model X will utilize a different size battery (physical dimensions) than the Model S, then I would agree that this test vehicle is more likely a Model X beta vehicle in disguise.
That said, (and this is pure speculation) if I were Tesla and I wanted to do some real-world testing of a prototype Model 3 drivetrain, I would make a point of disguising it to the max. After all, the Model 3 is perhaps the single most anticipated automobile of the past decade. I'm reminded of some of the spectacularly ugly test vehicles Ferrari has created over past twenty years—just to keep the media and the public guessing.
I do believe that the vehicle shown in this video does have a shorter wheelbase than the Model X prototypes previously shown. It also has significantly more bumper overhang at each end, although this could simply be part of the disguise.
Horizontal image compression has been addressed by others in this thread. If there is some, it would be having a equal effect on the rest of the scene, and that does not appear to be the case judging from the logos on the ships, etc—certainly not on the order of 20%.
The 20% factor was arrived at by dividing the wheelbase by the wheel diameter of the Model X prototype (5.3), versus the vehicle in the video (4.3). I included the two versions of the Model S to show that wheel diameter (19 or 21-inch) does not significantly skew the result. Granted, on this test vehicle it is difficult to determine exactly where the wheel ends and the tire begins, but a few pixels will not change the outcome in a big way.
Again, this is all speculation on my part. The Tesla owners community is populated with people like me, who see a disguised test vehicle as a riddle waiting to be solved. Many of those owners, myself not included, have significant technical prowess to bring to these endeavors. I've enjoyed reading the comments to my post, and I've learned a few things too.
rogbmw
Member
Isn't Speculation Cool........
......sure makes for interesting reading though, but the real fact is...we just don't know.
......sure makes for interesting reading though, but the real fact is...we just don't know.
SW2Fiddler
We Are Cognitive Dissidents
It strikes me that tacking on larger (say 23") diameter rims would be an easy and effective camouflage at this distance and resolution. See how much wheel size factors in to the wheelbase speculation? You have to throw much of that out if you admit you can't trust the visual wheel diameter...
That said, this is the most fun analysis I've seen since they buried Paul.
That said, this is the most fun analysis I've seen since they buried Paul.
Thank you. Where the limitations of foregin langauge often strike are the specialist words in areas more rarely discussed, certainly this is getting down to the nitty gritty...
As to your clarifications, I agree with what you write there. There is certainly room for speculation - and it is possible the wheelbase in this test vehicle is shorter than Model X prototype. Some say Model X prototype wheelbase is longer than Model S and that production Model X would be the same as Model S (I haven't confirmed this anywhere if this is true), which could also be an altenative explanation - although if your measurements of this being shorter than Model S are correct, then of course this wouldn't be right either.
In this picture in the other thread, with one assumption of where the wheel ends and the tire starts, the difference in axel distance is less than 4%:
In any case, I think the cases for and against have been presented as far as the evidence at this point allows. Until someone produces even more compelling analysis, at least.
It will be interesting to see what turns out to be the case, if anything about this mule is even heard again...Let's keep the great discussion going! Waiting is no fun without some rampant speculation.
- - - Updated - - -
Good point. 23" rims would certainly help explain why the wheels seem to take over the entire wheelwell and hug the ground. Pixel bleeding explains some of that, but 23" would explain more.
At that distance, photon diffusion through heat differences in the air and water molecules, etc. could cause the rims to look larger, and ununiform mirrage-like lensing could distort the position of objects in a single snapshot.
In practice using Gimp to measure keys I photographed, I was amazed at how exact I could decipher all the key codes. It was easy to do the math for pictures taken at angles. It was pretty much >90% accurate first try, and if I listed top 4 probabilities it went to >99% (for a hit). I improved my accuracy and speed immensely with practice. I got better at doing it by eye, too! All this proves is that Gimp is a fine way to measure stuff. Distortions would still be a problem like in this distance photo. Using a known good facsimile and comparing it would remove doubt, such as taking photos of bright rimmed cars in that distance and weather type and area.
Some video compression algorithms may move objects across the canvas at inaccurate relative positions (speeds), if there are codec bugs. I've never seen this, but I'm new to some codecs, too, and rarely extrapolated from vids so I'm not a good data point.
In my experience with fuzzy key photos, pixel distortions that are proportional to pixel distortions of other nearby dimensions allow math to calculate some real proportional dimensions if you are festidiously careful about discerning every relationship. That's a poor facsimile for this car pic.
There could be some item of a Model 3 they want to experiment with, such as cheaper suspension parts or cheaper frame material(s), that don't require the full undercarriage shape of a Model 3. Same goes for the Model X, too. This level of obfuscation makes discerning its allegiance seem even more random. Personally I find development timing guesses more accurate but I think it's fun to try to speculate from anything anyway.
In practice using Gimp to measure keys I photographed, I was amazed at how exact I could decipher all the key codes. It was easy to do the math for pictures taken at angles. It was pretty much >90% accurate first try, and if I listed top 4 probabilities it went to >99% (for a hit). I improved my accuracy and speed immensely with practice. I got better at doing it by eye, too! All this proves is that Gimp is a fine way to measure stuff. Distortions would still be a problem like in this distance photo. Using a known good facsimile and comparing it would remove doubt, such as taking photos of bright rimmed cars in that distance and weather type and area.
Some video compression algorithms may move objects across the canvas at inaccurate relative positions (speeds), if there are codec bugs. I've never seen this, but I'm new to some codecs, too, and rarely extrapolated from vids so I'm not a good data point.
In my experience with fuzzy key photos, pixel distortions that are proportional to pixel distortions of other nearby dimensions allow math to calculate some real proportional dimensions if you are festidiously careful about discerning every relationship. That's a poor facsimile for this car pic.
There could be some item of a Model 3 they want to experiment with, such as cheaper suspension parts or cheaper frame material(s), that don't require the full undercarriage shape of a Model 3. Same goes for the Model X, too. This level of obfuscation makes discerning its allegiance seem even more random. Personally I find development timing guesses more accurate but I think it's fun to try to speculate from anything anyway.
Last edited:
Here is a closer photo of the wheel, doesn't seem to be 23", but it also shows how much smaller the wheel looks from this distance than in the video!
The same image also shows very tell-tale signs of the new falcon wing door shapes and the front door handles of the rear doors bulging underneath the camo.
The likelihood of this being a Model X is, in my opinion, very, very high:
- - - Updated - - -
Or if it isn't the Model X, then it is something disguised to look like it.
I still think that if Tesla were running powertrain mules for something, say Model 3, disguising them as current products - aka Model S - would make more sense, like when other manufacturers test early future platforms with cars that outwardly don't immediately look like mules at all, until you check out the lower parts... Model S look-alike Model 3 mule would also make more sense from the perspective that neither are (expected to be) SUVs, right?
The same image also shows very tell-tale signs of the new falcon wing door shapes and the front door handles of the rear doors bulging underneath the camo.
The likelihood of this being a Model X is, in my opinion, very, very high:
- - - Updated - - -
Or if it isn't the Model X, then it is something disguised to look like it.
I still think that if Tesla were running powertrain mules for something, say Model 3, disguising them as current products - aka Model S - would make more sense, like when other manufacturers test early future platforms with cars that outwardly don't immediately look like mules at all, until you check out the lower parts... Model S look-alike Model 3 mule would also make more sense from the perspective that neither are (expected to be) SUVs, right?
Trev Page
Member
Holy cow you guys are totally over analyzing this. It's a Model X. Pure and simple. Stop obsessing about it.
I agree it is a Model X. However, speculation is still better than having nothing to talk about a future model. One day the mule may be Model 3 and someone who took out their inner Sherlock Holmes will inform us about it and we'll all be better off for it.
This video is producing a ton of buzz before the official release, that is likely to happen in the coming weeks. Is it possible Tesla is simply waiting for the right time, when people and the media are paying attention, and not pre-occupied by other events, to reveal the Model X? Activity on TMC , and speculation about the features of the Model X has been relatively low for the past few weeks.
Last edited:
Good point. You would still have to be able to accurately discern the ratio wheel vs. wheelbase ratio from the image, though? Unfortunately the only close shot doesn't show full axel distance and the far shots have so much distortion around the wheels that it is hard to say where the rim ends, tire begins and what is wheelwell...
The shot of the vehicle being loaded into the trailer seems to show 19" wheels. If so, this would be a new 19" wheel, unlike the current 19" turbine wheel with its staggered spokes—somewhat like the wheels on the Model X showcar. It looks smaller than the 21" wheel on the Model S in the background.
In the photo from the video, it is possible to get a pretty good fix on the wheel diameter (in pixels) if it's measured from left to right. Once you have that, you can easily determine the wheels' center point.
My brain says it's the x, but my gut reacts as if something is off size-wise. Proportions seem ok, closer to what I expected for an x (more camback and third row angle) But there are plenty of image noise factors messing up the gut reaction added on purpose so I guess they worked.
