Realistic Zoom Setting

[rhagen 0]

Some of you may have read a tutorial on California Classic's web site, the topic of which is related to this discussion. For those whom haven't read it I offer it here as "food for thought". :)http://www.calclassic.com/sci_tutorial/index.htmlRon

[FalconAF 213]

From that tutorial:"Flying with a false ZOOM to misplace the scenery at false LAT/LON is the silliest mistake any flight simulation user can make."That about covers it. The parallex distortion caused by this will never give the simmer any sort of "real world" display representation in any view chosen.

[Art_P 6]

Wow! My head just exploded.ArtBiostar TF560-A2+ | Athlon 64X2-6000+ | 4GB PC2-6400 RAM | Geforce 8800GTS-320MB | 500W PSU | 250GB HD | FSX(SP1-Acceleration) | UTX | Audigy SE sound | Vista Home Premium 32 bit | CH Yoke (modified) & Pedals | Logitech Extreme 3D Pro Joystick | 22" WS LCD monitor

[Guest jprintz]

>From that tutorial:>>"Flying with a false ZOOM to misplace the scenery at false>LAT/LON is the silliest mistake any flight simulation user can>make.">>That about covers it. The parallex distortion caused by this>will never give the simmer any sort of "real world" display>representation in any view chosen.>>They are right in the above quote, in my opinion, about the folly of flying with false zooms. What I don't get, though, is that they go on to contradict themselves by suggesting we stray from zoom 1.0 in FSX because "FSX has additional and serious design errors by default." The "errors" of which they speak are not errors at all, though, just the result of a design decision to make the default zoom in FSX narrower... more realistic for more people, actually, for more of those "average users with average sized monitors placed at average distances from their eyes."Why do they call FSX's zoom settings "serious design errors by default"? Well, I have tremendous respect for the work those guys have produced, I really do, but after reading their tutorial, they've pointed out **nothing** fundamentally wrong with FSX's zoom settings. Instead they've just pointed out that FSX misaligns the view parameters for their airplanes, planes (and views) they admittedly worked so hard to get EXACTLY right for FS9. And they DID get it exactly right for FS9. (I used to love flying some of their stuff!) But just because a user has to adjust FSX zoom so that FS9-designed planes show up properly in FSX... this does not mean FSX is fundamentally flawed. It means it's different, with a different zoom "fulcrum," that's all. And it's what I believe is an objectively BETTER default zoom, actually. Hopefully from the discussions above, some can see why many simmers believe that FSX's default render of ~34 degrees horizontally is more accurate as to perspective, for most people, than FS9's default ~47 degrees.Also, this statement:"The default VC view must deliver the necessary head up and head down views simultaneously." ... is completely incompatible with some of the other design principles they enunciate, especially the ones regarding zoom... UNLESS we take pains to sit only a foot away from our monitor, that is. Personally, that hurts my eyes, though I guess some might tolerate it.

[Geofa 108]

Seems to me this gets a little intellectual.Fact is you want to find a zoom that gives the correct perspective from the cockpit and the outside view. This is often a diplomatic compromise, and frankly does not match reality yet. I didn't find fs9 got it right anymore than fsx-and make constant adjustments depending on aircraft.Trig, and all that is great. I want to take a sim flight that reminds me of a real flight-pure and simple. If the view from the cockpit is not right it will not make it. If the view from the outside is not right it will not make it either. I don't use a slide rule to verify-just find what seems to get me to what is familiar.Fact is you want to find a zoom that gives the correct perspective from the cockpit and the outside view including size of objects, range of view and cockpit. This is often a diplomatic compromise, and frankly does not match reality yet. I didn't find fs9 got it right anymore than fsx-and make constant adjustments depending on aircraft.When the sweet spot is found however, my senses say yes!Trig, and all that is great. I want to fly a sim flight that reminds me of a real flight with the limitations of a non motion monitor-pure and simple. If the view from the cockpit is not right it will not make it. If the view from the outside is not right it will not make it. I don't use a slide rule to verify-just choose setting to remind me of reality. Certain aircraft just hit it-the Real Air Marchetti continues to amaze me in this department.Which brings up my pet peeve-despite outside views of aircraft being amazing down to the last rivet and oil stain, and the outside view while a far second is still pretty good-the cockpits views are eons behind-both in instrumentation look, viewpoint, and functionality.....period!GeofaMy blog:http://geofageofa.spaces.live.com/

[curt1 79]

This is a great thread, and yet there is another point in this. I always see screen shots and videos from sim pilots where they are seated way back from the yoke and instruments. They are a good 6-10 feet back. I realize with this setting they can view all the gauges at once, and if that is what they prefer that's fine. However, a real world pilot is looking out, and then must look down to see the instruments. I do this and it does require more work, but it is more realistic is it not? Curt

[michal 44]

One can lay an egg trying to read this complete tutorial. *:-*There are many dubious statements there.Michael J.

[Guest jprintz]

Geofa:I couldn't agree more with some of the things you say, your last paragraph especially. In the end, I suppose it's true that any combination of settings will be some kind of compromise, so people may as well pick whatever FEELS right to them and makes them happy.Scott

[january 19]

-------"I always see screen shots and videos from sim pilots where they are seated way back from the yoke and instruments. They are a good 6-10 feet back.-------------Curt- I'm not sure whose screenshots you refer to, but any I have posted, are from my setup where my eyes are about 26" from the monitors. Screenshots distort the situation in several ways- as you mention distance to screen appears much greater than actual, especially when they have to be reduced in size to fit within AVSIM limitations. In addition screenshots show monitors in a flat plane where my side monitors are angled toward me, and they remove the monitor frames- thus creating misalignment between monitor images.It is for these latter reasons that good multi monitor airborne screenshots are difficult to impossible.And because of camera focal lengths, digital camera shots DO have to be done from further away. The best Wide Angle on my camera requires it to be about 48" from the monitors.Regards Alex Reid

[=Hollywood= 0]

>>>Do you agree that there's considerable visual distortion and>>warping at zoom 0.5? How about at zoom 1.5? >>Sorry, no visual distortions for me. Optical cameras give you>distortions but not MSFS - like the famous camera>obscura there are no distortions to speak of since there>are no lenses here. By the way FalconAF has good basic>information on the zoom ...>>Actually try and take some screenshots for a panoramic view, there are definitely distortions at the edges of the images at all zoom settings even 1.0 and you're not able to easily just stitch a panorama together like you'd expect to be able to...This is regardless of the widescreen=1/0 setting or not in the fsx.cfg file...

[=Hollywood= 0]

>Wow! My head just exploded.>>ArtSo can we now discuss which came first, the chicken or the egg??? :D

[michal 44]

>Actually try and take some screenshots for a panoramic view,>there are definitely distortions at the edges of the images at>all zoom settings even 1.0 and you're not able to easily just>stitch a panorama together like you'd expect to be able to...Sorry, but in my vocabulary these are not distortions. Something that matches what a human eye would see in identical set-up is not a distortion. Obviously every object viewed at highly oblique angle doesn't look the same as the same object when viewed directly ahead - this is a simple geometry. FS creates an image equivalent to the camera obscura which is known to be one type of "lens" which is distortion-free.Michael J.http://img142.imageshack.us/img142/9320/apollo17vf7.jpg

[sprintz 0]

>>Actually try and take some screenshots for a panoramic>view,>>there are definitely distortions at the edges of the images>at>>all zoom settings even 1.0 and you're not able to easily>just>>stitch a panorama together like you'd expect to be able>to...>>Sorry, but in my vocabulary these are not distortions.>Something that matches what a human eye would see in identical>set-up is not a distortion. Obviously every object viewed at>highly oblique angle doesn't look the same as the same object>when viewed directly ahead - this is a simple geometry. FS>creates an image equivalent to the camera obscura which>is known to be one type of "lens" which is distortion-free.>>Michael J.>http://img142.imageshack.us/img142/9320/apollo17vf7.jpgI think that the only thing that in all aspects matches what a human eye would see (or comes the closest that computer monitors can) is what was described before, the panoramic, zoom 1, multi-monitor setup.As to the theory of an ideal, distortion-free lens, I guess I'd just say that we may be making this all way too difficult. Before offering a little hypothetical, which I think is extremely illuminating, let me say first that I trust my eyes, and also the eyes of many others who agree that there is distortion in the image at extreme zoom settings. Frankly, I was very surprised to find people who disagreed with that. Personally, I've looked on several different computers now, different aspect ratios, etc., and the image at extreme zooms such as 0.3 is always distorted. It is especially noticeable while rolling the aircraft. I can see how someone who's used to flying at zoom 0.3 or whatever doesn't notice or has gotten used to it, but that doesn't mean the distortion's not there.What "sealed the deal" for me is this: if we could get, as some are implicitly saying, 110 degrees of scenery into our monitors withOUT any geometric distortion, then... well, what's the hypothetical limit as to the # of degrees we could accurately squeeze in? (110 being an approximation of # degrees in zoom 0.3) Not what we can force the sim into rendering in actuality, since FSX is limited to zoom 0.3 (for reasons related to down below), but again, hypothetically. If we can find a way to get 110 in there, why not 111? If 111, why not 112? How about 180 degrees? Could we squeeze that amount of scenery into our monitors without any geometric distortion? 270 degrees? Or, could it even be possible to get all 360 degrees of scenery onto our monitors without obvious geometrical distortion? Of course it's not possible. And mathematically, I think it makes sense that if the rendered image is distorted when 360 degrees of scenery are forced in, then it's also distorted, but a little less so, when 359 are forced in, same with 358, and so on.... Until you keep zooming in and reach zoom 1.0 that is, which is the "best case." It's not "1.0" for nothing. Tinker with any virtual planetarium programs like Stellarium or whatever and you'll see exactly what I mean. And beyond what my eyes tell me, beyond what's obvious to me when I'm rolling the aircraft while in zoom 0.3, there is the math...The number that is the zoom setting is just a constant multiplier used in the trig equations that help draw every pixel of scenery that appears on your screen, bunches of data whizzing through several sine and cosine functions many times per second. Pause the sim and hit + or - and consider what's happening. It's just trig. There's no way to avoid it. And multiplying those trig equations by anything other than 1.0, the zoom factor, distorts the shape of the results, just as 2sin(x) IS DISTORTED in comparison to the "default" sin(x). Not just moved or shifted, NOT just bigger or smaller (as needed for no distortion), but distorted and its shape fundamentally changed -- a normal sine wave vs. one that's been stretched up and down. The stretched one is...well, stretched! And so will be the image. Because it's nothing more than well-ornamented sine and cosine functions that tell each little piece of scenery at which pixel it needs to appear. At varying zooms, except 1, there will be more or less "stretch" distortion, but always some. 1.0*sin(x) vs. 0.3*sin(x). Simplified, yes, but it makes the point. When you change the zoom, the relationship between nearby pixels is changing in a way that is non-linear. Scenery objects are not just moved. They are warped.Last thing: It is true that the image at zoom 0.3 may be close to geometrically sound at 0 degrees bank, and some of these kinds of images have been presented as "evidence" of 0 distortion... but, keep rolling the aircraft, when the trig really comes to life, and the difference between where that tree would've been rendered(at zoom 1) and where we're now asking it to be rendered (at zoom 0.3) is compounded. Why are the right angles in that building no longer looking "right," and why does that cloud seem to bend and warp when I roll the aircraft? It shouldn't be surprising. You're telling the math to do that. There really IS something "special" about zoom 1.0. It's the baseline, presenting the most pure version of the results. NOT the only useful version, I think we should keep in mind, but definitely the "best" version in terms of real-world perspective and geometry.

[Geofa 108]

Someone mentioned that the zoom of 1 has changed from fs9 to fsx-for fps...?!I've submitted pictures above comparing real to different zoom levels-shouldn't we be comparing that? Who knows what criteria was established for the setting of 1. I know on the Baron the viewpoint is incorrect at that setting.I also know that going to a setting less than 1 sharpens up the textures and makes autogen a more realistic size....Perhaps those with the "blurries" and those who have never had them comes down to a zoom setting?! I can get real "blurries" going to a zoom of 2.GeofaMy blog:http://geofageofa.spaces.live.com/

[sprintz 0]

The number of degrees of scenery displayed at zoom 1 HAS changed between FSX and FS9. This bolsters what I'm saying. It doesn't detract from it. There is a range of realistic and possible values for the field of view at zoom 1 that the program could be made to take, but the programmer must choose ONE value to be the baseline for the projection. That is the zoom 1 value. All other values get progressively more distorted. The image we see could never be truly 100% distortion free, but the sim can be programmed to be *least* geometrically distorted at a setting of X degrees. And X might be 45, or 35, or 25, but the programmer has to pick A standard that is set equal to zoom 1. This is why I've called that setting the "fulcrum" of the display engine. There is nothing inherently magical about "33.5," the number of horizontal degrees displayed in FSX at zoom 1. There IS something important about the zoom of 1.0 though. And 1.0 and 33.5 are fundamentally intertwined in FSX. So, though 33.5 is not inherently special, it is also not arbitrary, and I've mentioned why I think the change was made from Fs9 to FSX, and also why, for most simmers, it's probably superior to the FS9 zoom "fulcrum" of (IIRC)about 45 degrees.I think that what is most telling here is the answer to this question: Is there a lens or a scenery display engine or whatever that could get all 360 degrees of scenery onto the monitor without massive geometric distortion? If there's not, and I don't think there is, then why not one that could do 350, or 340, or 270, or 180? I think it's because the value chosen needs to be at least *close* to the field of view that the typical monitor occupies for the typical simmer. Otherwise right angles would rarely be displayed as right angles.But Geofa, you're right. There are good reasons to stray from zoom 1 while flying. I think I said that in my last post. But this had turned into a discussion about accurate perspective and zoom, and that's what I was trying to answer. It should not be a secret. Perspective and distortion are "best" at zoom 1... not because I have some kind of twisted affection for it (wild man that I am, I'll sometimes go down to zoom 0.8 ;-), but because of the math.