Help: Panorama Maker and the Curvature of the Earth

[Sunsignature]

Hi everyone, this is my first time posting such a thread. Hope to get it right, in case I make any inaccuracies regarding the posting guidelines please let me know. With that said, I begin:

Many here at Metabunk, and even more so on the Flat Earth forum, will know that Panorama Maker (https://www.udeuschle.de/panoramas/makepanoramas_en.htm) is one of the best websites out there when it comes to analyzing long distance video and photography
If once you are on the web, you go to "FIRST APPROACH", and then to "Extensive Help" you will find the following:
"In the calculation of the panoramas the earth is taken as being an ideal sphere. Atmospheric refraction is accounted for with the Gaussian refraction coefficient or 0.13. The 3'' DEM data are interpolated down to 1'' using cubic spline methods"

It was expected, since it is necessary to take into account the curvature to offer the most faithful views. And it really does, and we can see it: we all remember even that famous video from JTolan Media1 where it shows the St. Jacinto mountain, which was later verified both on YouTube and on Metabunk and other sites that it perfectly matched the spherical model. Well, Panorama Maker also fits perfectly with what is expected in the spherical model, here is a sample recreating the same position of Tolan's infrared photography on the website:


So okay, the curvature is perfectly represented. However, I did have an exchange with a flat-earther about Tolan's videos where I mentioned that those observations fit perfectly with this software, which assumes that the Earth is an ideal sphere + a bit of refraction. His answer was directly to affirm that the web did not really take curvature into account (directly contradicting the author himself, although many flat Earthers are already like that) since no matter how much you raise the camera, the horizon always looks flat. And the truth is that it is not a bad point: I have tried to raise the camera, increase / decrease the field of view, and modify the variables in general; however, the curved horizon is not appreciated at any time, not even at an altitude of 400 km.
Here's an example: https://www.udeuschle.de/panoramas/...guage:en$$$screenwidth:360$$$screenheight:760

Anyway, I imagine a mistake is being made here with the interpretation of the projection used.
I suppose the matter comes because they are precisely panoramas, as it is said on this site:

(https://flatearth.ws/panorama)

However, in that panoramic altitude image, things underneath look quite distorted. In Panorama Maker you don't seem to notice any distortion, in addition to being able to work with narrow fields of view (60° for example) and you don't seem to notice any curvature.
The creator of the web leaves two email addresses, try to contact him to ask him, I have not received a response yet.
Anyway, I only have very basic knowledge of photography and / or 3D projections.

The reason why I publish this thread is because perhaps for someone who does have the right knowledge the answer can be very simple, or at least it could help me to know what is happening here.

So any input and ideas ablut whats is happening are welcome!
Thank you!

 

[Mick West]

His answer was directly to affirm that the web did not really take curvature into account (directly contradicting the author himself, although many flat Earthers are already like that) since no matter how much you raise the camera, the horizon always looks flat. And the truth is that it is not a bad point: I have tried to raise the camera, increase / decrease the field of view, and modify the variables in general; however, the curved horizon is not appreciated at any time, not even at an altitude of 400 km.

It is as the FlatEarth.ws image says. Panoramas use cylindrical projection which can go up to 360°, and in this projection, the horizon appears flat.

A regular camera image is essentially rectilinear, which shows the curve of the horizon.

You see distortion in images with lots of straight lines at right angles, but those don't occur in nature at that scale.

A way to demonstrate this would be to stand in the middle of a circular area, like a circus ring, or a portion of a sports field. Take a panorama photo of half the ring, and a wide-angle photo (not fish-eye). The edge will be flat in the panorama (assuming perfect positioning and a perfectly circular ring), but will be curved in the regular photo.

 

[Mendel]

Play around a little with Walter Bislin's app at http://walter.bislins.ch/bloge/index.asp?page=Finding+the+curvature+of+the+Earth

In particular, go to a height that has obvious curvature (40 km in this case), then slide the viewing angle (65.008° in this picture) down to 1° or less. As you can see, the curve appears straighter as you zoom in.

Someone has demonstrated this "curved to flat" magic on a basketball, so it's not just "programmed into the software":
https://imgur.com/account/favorites/8dvnDn6

Now imagine that the panorama software computes many such infinitesimally small segments and pastes them together to make the panorama, and it becomes clear why its horizon appears straight while the curvature (the "drop") away from the viewer is still apparent.

 

[Sunsignature]

It is as the FlatEarth.ws image says. Panoramas use cylindrical projection which can go up to 360°, and in this projection, the horizon appears flat.

A regular camera image is essentially rectilinear, which shows the curve of the horizon.

You see distortion in images with lots of straight lines at right angles, but those don't occur in nature at that scale.

A way to demonstrate this would be to stand in the middle of a circular area, like a circus ring, or a portion of a sports field. Take a panorama photo of half the ring, and a wide-angle photo (not fish-eye). The edge will be flat in the panorama (assuming perfect positioning and a perfectly circular ring), but will be curved in the regular photo.

Completely agreed, Mick. The more I thought about it, the more my suspicions pointed toward that possibility. So basically, due to the way a panoramic image works, we would never have to observe the curvature of the Earth because the angle is kept constant, which is why it will be viewed straight in the panorama.
This is just my speculation, but I think that the safest thing is that the Panorama Maker software works in this way: The user enters the location and altitude of the camera whose view he wants to recreate, consequently the servers process a 360 ° panoramic image of the views of the place using their databases. Then, when the user enters how many degrees the field of vision will need to cover and the direction of the latter, the server will trim the part of the panorama that meets the requirements requested by the user. When viewed at fairly low altitudes (well, this app is designed to represent mountainous landscapes after all), and with a reasonably narrow field of view, you won't notice any difference from a rectilinear image.
Sin embargo, dado que nunca deja de ser un panorama, la curvatura del horizonte no se notará incluso a gran altura.
Hago un ejemplo con estas imágenes (Perdón por la mala calidad, pero esto sucede porque tuve que recortarlo del panorama original):


(All credits for AirPano)

The first image was taken from AirPano's original interactive panoramic view of the Caucasus Mountains and shows the curvature. It seems cropped, but I imagine that if they did, it was for artistic purposes to give some touch to the background.
( https://www.airpano.com/vtour_mobile/stratosphere-caucasus/ )
And the last image (lower resolution) was obtained by cropping a part of the projected panorama, which can also be obtained on the same page) If you look closely, the images fit perfectly, the only difference is the curvature on the horizon due to projection panorama. Thanks for the reply, Mick.

 

[Mick West]

Or use software that actually is designed to create accurate images of the Earth, like Google Earth.
https://earth.google.com/web/@40.06....60961014d,35y,168.48254009h,77.71751672t,-0r

 

[Mick West]

This is just my speculation, but I think that the safest thing is that the Panorama Maker software works in this way: The user enters the location and altitude of the camera whose view he wants to recreate, consequently the servers process a 360 ° panoramic image of the views of the place using their databases. Then, when the user enters how many degrees the field of vision will need to cover and the direction of the latter, the server will trim the part of the panorama that meets the requirements requested by the user. When viewed at fairly low altitudes (well, this app is designed to represent mountainous landscapes after all), and with a reasonably narrow field of view, you won't notice any difference from a rectilinear image.

It probably does not create a full 360° and then trim, there's no need to. There no inherent downside to culling the view outside of the view angle, and if done properly it will speed up rendering. But the end result would be the same either way.

 

[Sunsignature]

Hi Mendel, yes, I know Walter Bislin's application and I use it often to disprove flat Earthers, and I also understand what you mean (if Google Translate does not fail me): if I take an image with high resolution of the ISS, which orbits 400 km high, and I adjust the frames to contemplate a small part of the horizon, the curvature will be so small that it is almost imperceptible. However I was not referring to that, I was simply saying that if in Panorama Maker you put a height of 40 km (as you did in Bislin's calculator, although it may be higher), you choose any visual field (40 °, 60 °, 180 °, 360 ° or whatever) and keep the zoom at 1.0x and still there will be no observable curvature in the application, both at 40 km and at 100 km or more. What I was suspecting that Mick later mentioned is that when you get a panoramic image, since the curvature is held at a constant angle, it will look like a straight line in the panorama. It just seemed strange to me that since it is based on a spherical model, the horizon does not curve at that height. My mistake was in part overlooking the "Panorama" Maker thing, since the way a panorama works, the horizon should not bend regardless of altitude.

Correct me if I wrong, but I imagine that with a good use of software it should be possible to take the panoramic projection at 40 km altitude of Panorama Maker and the geometric template offered by Walter Bislin, then cut a little (so that the frames and field of vision equalize) and eliminating the remaining distortions a 'straight lens' style view could be projected roughly from the original panoramic image.

To Mick: Yes, I understand that Google Earth is the best representation of what the Earth would look like to the human eye at certain heights; only now I understand that the difference between Google Earth and Panorama Maker is that Earth emulates a rectilinear camera, but does not display panoramic views. Panorama Maker, on the other hand, lives up to its name: it creates panoramas. I simply prefer to use PM instead of Earth to disprove videos like Tolan's since it eliminates atmospheric interference, and allows zooming without having to change camera position.
Anyway, both are equally valid since they both represent the spherical model, only we could say that they use a 'different' type of image.
PD: I understand what you mean by the way the software operates, but as you say: the result will end up being the same anyway.

Thanks for your answers boys!

 

[Mendel]

Hi Mendel, yes, I know Walter Bislin's application and I use it often to disprove flat Earthers, and I also understand what you mean (if Google Translate does not fail me): if I take an image with high resolution of the ISS, which orbits 400 km high, and I adjust the frames to contemplate a small part of the horizon, the curvature will be so small that it is almost imperceptible. However I was not referring to that, I was simply saying that if in Panorama Maker you put a height of 40 km (as you did in Bislin's calculator, although it may be higher), you choose any visual field (40 °, 60 °, 180 °, 360 ° or whatever) and keep the zoom at 1.0x and still there will be no observable curvature in the application, both at 40 km and at 100 km or more

The people you are talking to may be familiar with the way a smartphone creates a panorama when you rotate your view, by stitching separate pictures together. I thought it would help for the people you are talking to to think of the Panorama Maker output as that kind of picture, except made up out of even more small slices than a smartphone uses--and then they can make the experience that a small slice of any curve, if magnified enough, looks straight. You can do that with a circle drawn on paper if you can zoom in enough!

While my explanation is simplifying the geometry involved (mick is correct when he calls it a cyclindrical projection), the "slice" method results in a cylindrical projection if the slice width gets infinitesimally small -- this explanation is geometrically correct, too.

 

[Rory]

If once you are on the web, you go to "FIRST APPROACH", and then to "Extensive Help" you will find the following:
"In the calculation of the panoramas the earth is taken as being an ideal sphere. Atmospheric refraction is accounted for with the Gaussian refraction coefficient or 0.13. The 3'' DEM data are interpolated down to 1'' using cubic spline methods"

Cool to learn it factors for refraction: last I recall, Google Earth doesn't, so can noticeably differ sometimes from reality.

 

[Sunsignature]

The people you are talking to may be familiar with the way a smartphone creates a panorama when you rotate your view, by stitching separate pictures together. I thought it would help for the people you are talking to to think of the Panorama Maker output as that kind of picture, except made up out of even more small slices than a smartphone uses--and then they can make the experience that a small slice of any curve, if magnified enough, looks straight. You can do that with a circle drawn on paper if you can zoom in enough!

While my explanation is simplifying the geometry involved (mick is correct when he calls it a cyclindrical projection), the "slice" method results in a cylindrical projection if the slice width gets infinitesimally small -- this explanation is geometrically correct, too.

Oh sure, I just got what you were trying to say (I really have to be very careful when reading things translated by Google).
You meant that this would be a more intuitive way of explaining how Panorama Maker actually works to someone who is not very familiar with the web, instead of saying "because it is an equirectangular panoramic projection" and leaving them scratching their heads; the truth is that I agree. However, we know that not many of these people do not want to learn by maintaining their conspiracy stance; Of all, if someone really shows an interest in learning, perhaps your way of approaching it is the most appropriate.

Well, it seems to me that the matter is already being settled, since it was a very simple question and it seems to me that it has already been clarified enough. So let's recap by way of conclusion:

- Both Google Earth and Panorama Maker are faithful representations of a spherical Earth.

- However, PM creates panoramic projections, which is why the curvature cannot be seen at high altitude (since the angle of the curvature is constant, and it will be viewed straight in the panorama). Anyway, and as the creator rightly considers, the curvature is one of the main things that they take into account to calculate the panorama. Even so, if the variables are adjusted correctly (corresponding field of view, the corresponding zoom and so on) it is possible to perfectly match the views at an altitude where the curvature is very slight in PM (to avoid distortions) with a equivalent view obtained in Google Earth.

- Google Earth for its part offers a more faithful representation of what a rectilinear camera or a human eye would see being there, since the latter two do not work creating panoramas. -So, compared to real life, both are exceptionally accurate at low altitude, although since GE simulates atmosphere, terrain details are often blurred; For its part, Panorama Maker offers options (zoom in and / or eliminate atmospheric effects, etc). However, when curvature becomes more decisive, Google Earth ends up being a realistic representation of what a person would see at that altitude.

Up to here is correct? Or are there some details that I keep missing?
Thank you!

 

[Sunsignature]

Cool to learn it factors for refraction: last I recall, Google Earth doesn't, so can noticeably differ sometimes from reality.

Yes, and in fact you can see that it does it excellently with the image of St. Jacinto of Tolan! Although, as we know, refraction is very dynamic, and assuming a fixed value can also cause that many times they find slight deviations of the model with respect to the observations. Anyway, I think it is better to directly omit the data and make a representation without standard refraction, as if we had no atmosphere!
Cheers

 

[Mendel]

I believe what you wrote is correct.

Google Earth for its part offers a more faithful representation of what a rectilinear camera or a human eye would see being there, since the latter two do not work creating panoramas.

I would avoid bringing up the human eye, or you might have to deal with the claim that the Earth appears curved because the eye is spherical.

Here is another way to explain it:

If you stand inside a circle (for example, a round room; or in the center of a circle drawn on the ground), any picture you take depends on the direction you are looking in. The circle will appear to curve down to both sides. You wrote, "the angle of the curvature is constant". If you look towards the "down" curve in that picture, the direction of your look changes, and down is now at the top. (This makes more sense if you can try this out in reality; I'll try to take some pictures today.)

A true panorama is supposed to work independent of the direction you are looking in, and therefore it cannot have this dependency on the viewing direction. It must be useful regardless of the direction you are looking at in real life, and therefore it cannot show a downturn of the horizon.

 

[Trailspotter]

A true panorama is supposed to work independent of the direction you are looking in, and therefore it cannot have this dependency on the viewing direction. It must be useful regardless of the direction you are looking at in real life, and therefore it cannot show a downturn of the horizon.

I'd like to add that a true panorama is supposed to be displayed on the wall of a round room, curving around the observer in the centre.

 

[Mendel]

I'd like to add that a true panorama is supposed to be displayed on the wall of a round room, curving around the observer in the centre.

And then the straight line will be curved!

 

[Sunsignature]

I believe what you wrote is correct.

I would avoid bringing up the human eye, or you might have to deal with the claim that the Earth appears curved because the eye is spherical.

As incredible as it may seem, it's true, many flaties often use that excuse. However, they should also explain how our eyes can observe straight lines (like a door frame) at the same time that we can confirm it with our other senses, such as touch. Anyway, I wouldn't be surprised if they made something up.

Here is another way to explain it:

If you stand inside a circle (for example, a round room; or in the center of a circle drawn on the ground), any picture you take depends on the direction you are looking in. The circle will appear to curve down to both sides. You wrote, "the angle of the curvature is constant". If you look towards the "down" curve in that picture, the direction of your look changes, and down is now at the top. (This makes more sense if you can try this out in reality; I'll try to take some pictures today.)

If I understand; It is similar to what Mick suggested when he gave the example of taking a panoramic view of a circus ring.

A true panorama is supposed to work independent of the direction you are looking in, and therefore it cannot have this dependency on the viewing direction. It must be useful regardless of the direction you are looking at in real life, and therefore it cannot show a downturn of the horizon.

From that point of view, it is perhaps one of the ways in which I best understand it. If the panorama showed some kind of curvature, it would be assuming a specific viewing direction, which cannot happen in the panorama. However, as trailspotter says, once the panorama is correctly projected this problem disappears, since thanks to the projection (which would also eliminate distortions) now the observer can acquire a viewing direction (as in Google Earth); and yes, the curvature appear again

Although, just out of curiosity, I am wondering if there is any way to create an interactive 360 ° panorama from an image created by Panorama Maker ... If that were the case (and if I am right), a panorama calculated for 40 kilometers high could be projected in Panorama Maker to emulate a rectilinear view similar to Google Earth, where the curvature corresponding to 40 km appears.

 

[Sunsignature]

If my previous comment is correct, then the question is settled, there is nothing to discuss.

I was not planning to post this here as it deviates from my original question. But come to think of it, the thread title is "Panorama Maker and the Curvature of the Earth", so it still has a relationship.
Maybe someone is interested: specifically what I did was take certain fragments of the video "How to denoise the world" Source: https://youtu.be/FxRshjR04WY
where Mr. Tolan affirms that the Flat Earth is tested, try to take reference points to know its specific location, obtain the coordinates in Google Earth, and later recreate the view in Panorama Maker (spherical model). That was the results:

First Image:

A view over Kendrick Peak (orange) in the right, and Navajo Mountan (blue) at bottom. They mostly coincide, however, the movement of the plane during the shot causes the background to be slightly more displaced to the left than it should, but it is a trivial error in the recreation. Anyway, it fits excellent with the spherical model.
Minute of the image on the video: 6:46
Original panorama link: https://www.udeuschle.de/panoramas/panqueryfull.aspx?mode=newstandard&data=lon:-113.07233$$$lat:33.65476$$$alt:10500$$$altcam:1$$$hialt:false$$$resolution:400$$$azimut:28.1$$$sweep:11.2$$$leftbound:22.5$$$rightbound:33.7$$$split:3$$$splitnr:4$$$tilt:-2.89583333333333$$$tiltsplit:false$$$elexagg:1.2$$$range:750$$$colorcoding:true$$$colorcodinglimit:500$$$title:test 1$$$description:sd$$$email:$$$language:en$$$screenwidth:1920$$$screenheight:1040

Second image:

Looking at Grand Canyon North Rim. There is no need to comment on this image, both views are exactly the same.
Minute on the video: aprox 7:02
Original panorama link: https://www.udeuschle.de/panoramas/panqueryfull.aspx?mode=newstandard&data=lon:-113.07233$$$lat:33.65476$$$alt:10500$$$altcam:1$$$hialt:false$$$resolution:400$$$azimut:21.2$$$sweep:11.2$$$leftbound:15.6$$$rightbound:26.8$$$split:3$$$splitnr:4$$$tilt:-2.80208333333333$$$tiltsplit:false$$$elexagg:1.2$$$range:750$$$colorcoding:true$$$colorcodinglimit:500$$$title:test 1$$$description:sd$$$email:$$$language:en$$$screenwidth:1920$$$screenheight:1040

No minor detail is that JTolan claims to be flying at an altitude of 9114 meters in these last two shots. However, to make them fit perfectly in panoramas, I had to use an altitude of 10,500 meters. There are two options: either JTolan lied about his altitude (since he never shows his GPS), or the refraction was higher than the standard.

Third image:

Composite image (both mine and Tolan's) of Captain Mountains HP (yellow) in front, and Colorado Mountains in background (blue), which are more than 1 km obscured by the curvature, but are visible thanks to their great height.
Minute on the video: 14:41
Original panorama link: https://www.udeuschle.de/panoramas/panqueryfull.aspx?mode=newstandard&data=lon:-104.88364$$$lat:31.8593$$$alt:10500$$$altcam:1$$$hialt:false$$$resolution:400$$$azimut:353.2$$$sweep:11.2$$$leftbound:347.6$$$rightbound:358.8$$$split:3$$$splitnr:4$$$tilt:auto$$$tiltsplit:false$$$elexagg:1.2$$$range:750$$$colorcoding:true$$$colorcodinglimit:500$$$title:test 1$$$description:sd$$$email:$$$language:en$$$screenwidth:1920$$$screenheight:1040

This is one of the reasons why I say that it is one of the best software to analyze long distance photographs
So first of all, thank you for your help up to this point.
Second, thanks to JTolan, for proving once again that the Earth is spherical.

If by any chance you have a discussion with a flat-earther and this same video is mentioned, as far as I'm concerned, there is no problem using the images shown here.
Cheers

 

[Mendel]

Although, just out of curiosity, I am wondering if there is any way to create an interactive 360 ° panorama from an image created by Panorama Maker ... If that were the case (and if I am right), a panorama calculated for 40 kilometers high could be projected in Panorama Maker to emulate a rectilinear view similar to Google Earth, where the curvature corresponding to 40 km appears.

http://panoramaviewer.1bestlink.net/ works for that.
I just don't know how to best export an image from PanoramaMaker, it seems to generate lots of vertical slices?

 

[Sunsignature]

http://panoramaviewer.1bestlink.net/ works for that.

Thanks for that link!

I just don't know how to best export an image from PanoramaMaker, it seems to generate lots of vertical slices?

Yes, I also noticed that problem: there is no option to export the image. Chrome also does not recognize the panorama as an image. Captures may be a good idea, although you would have to be extremely careful. I'll see what I can do.

 

[Mendel]

Thanks for that link!

Yes, I also noticed that problem: there is no option to export the image. Chrome also does not recognize the panorama as an image. Captures may be a good idea, although you would have to be extremely careful. I'll see what I can do.

What do they send if you request the panorama by email?

 

[Sunsignature]

¿Qué envían si solicitas la panorámica por correo electrónico?

"Your request has been received. You will receive one or more emails with links to download the panorama. Please, be patient. It may take an hour or two."

Sounds good, I ordered one, I hope it is sent in JPG or similar format.

In case it doesn't work, I prepared the following 360 image by joining screenshots, but have bad definition:

I tried to put it on the web you sent. As for the 2D projection, it works, and you can tell that it is a 360 ° panorama. However, it is not what we are looking for since all it does is repeat the same image again. The 3D projection, meanwhile, looks quite strange and distorted. I'll keep looking for other websites or apps for the same purpose that allow us to project panoramas and see if any of them work this way.

 

[Ray Brown]

The panorama maker also shows a thin blue horizontal line that indicates eye level of observer. This is useful in showing that at relative distances and elevations the top of a structure or land feature that is higher above sea level than the observer is actually below the observer’s eye level on the “real” curved earth.