iMac27" or other wide screens

[bigwade]

Today I visited the Mac shop in Utrecht NL to see the new iMac27"
Wow, it's big....and a NICE MACHINE...
The wide screen is 2560x1440 pix (3.55:2)
The iMac 24" is 19200x1200. (3.2:2)
The problem is FullScreen pano's looks horrible.....
The edges are stretched too much even with a 70 degr. view pano.
Since screens are going the wideview way we have another problem.
The original idea about FS pano's @6000x3000 pix equis is over, at least on this screen.
Now what ??

(crosspost on panoguide)

[Hans-74]

--- In [hidden email], bigwade <celsius@...> wrote:

>
>
> Today I visited the Mac shop in Utrecht NL to see the new iMac27"
> Wow, it's big....and a NICE MACHINE...
> The wide screen is 2560x1440 pix (3.55:2)
> The iMac 24" is 19200x1200. (3.2:2)
> The problem is FullScreen pano's looks horrible.....
> The edges are stretched too much even with a 70 degr. view pano.
> Since screens are going the wideview way we have another problem.
> The original idea about FS pano's @6000x3000 pix equis is over, at least on
> this screen.
> Now what ??

No problem. Just use FPP as viewer,
Check www.panoramas.dk. Do you find any stretched panos even if the view in the browser is sometimes 3:1.
In Real Fullscreen the view adapts to any screensize

FPP uses intelligent FOV which adapts to the proportions of the display.
You set a zoom value instead of a FOV and you will always have a nice view without any distortions at the corners.

Hans

[Jan Martin]

Interesting problem.

However really big screens existed before (e.g. 30" Appel Cinema), but they
are kind of expensive and therefore not that common. The 27" iMAC screen
will be very mainstream in short time.

What we have is the same problem we had with PC games once. At a time all
games were written to run as fast as the CPU could do. Just because the CPUs
were slow and therefore the  limiting factor.

10 years later old games did not work on new hardware anymore, because they
were running way to fast to play, due to new faster hardware.

Now we just have the same problem with the screensize. Old assumptions on
what "fullscreen" might be are no longer true, and the various panorama
players can no longer cope with the new larger monitors.

Whats needed is to adjust the players to the new screensizes. As well as
provide more pixel for new panoramas.
And we need a "legacy mode" for the players to make sure they would't try to
blow up a (rather small) legacy panorama to 27" if the available panorama
data (amount of pixels) just isn't sufficient for that.

Anyone else?

Jan


On Sun, Oct 25, 2009 at 12:43 AM, bigwade <[hidden email]> wrote:

>
>
>
> Today I visited the Mac shop in Utrecht NL to see the new iMac27"
> Wow, it's big....and a NICE MACHINE...
> The wide screen is 2560x1440 pix (3.55:2)
> The iMac 24" is 19200x1200. (3.2:2)
> The problem is FullScreen pano's looks horrible.....
> The edges are stretched too much even with a 70 degr. view pano.
> Since screens are going the wideview way we have another problem.
> The original idea about FS pano's @6000x3000 pix equis is over, at least on
> this screen.
> Now what ??
>
> (crosspost on panoguide)
>
> --
> View this message in context:
> http://www.nabble.com/iMac27%22-or-other-wide-screens-tp26044109p26044109.html
> Sent from the PanoToolsNG mailing list archive at Nabble.com.
>
>  
>

[Non-text portions of this message have been removed]

[Trausti Hraunfjord]

As outlined by Hans; By using Flash Panorama Player (FPP), there is no
problem, and there won't be any problem.
I am in that group of users, and therefore:  I don't see a problem :)



On Sat, Oct 24, 2009 at 8:19 PM, Jan Martin <[hidden email]>wrote:

>
>
> Interesting problem.
>


> .........
>


> Anyone else?
>
> Jan
>


[Non-text portions of this message have been removed]

[erik_leeman]

Hmmm, based on the fact that I am happy with my 1920x1920 cubefaces on a 1024px width screen, I would roughly need:

2400x2400 cubefaces on a 1280px screen
3600x3600 cubefaces on a 1920px screen
4800x4800 cubefaces on a 2560px screen

Stitched at PTGui-optimum size, my 5D/15mm-fisheye cubefaces are about 3560x3560 pixels in size, so above a screen width of 1920 I would appear to need a better camera/lens combination than what I have now if I want to maintain my present level of image quality.

A 5DmkII would deliver the pixels, but would my lens deliver the required sharpness/contrast at that resolution?

Cheers!

Erik Leeman

(www.erikleeman.com)


--- In [hidden email], bigwade wrote:

>
>
> Today I visited the Mac shop in Utrecht NL to see the new iMac27"
> Wow, it's big....and a NICE MACHINE...
> The wide screen is 2560x1440 pix (3.55:2)
> The iMac 24" is 19200x1200. (3.2:2)
> The problem is FullScreen pano's looks horrible.....
> The edges are stretched too much even with a 70 degr. view pano.
> Since screens are going the wideview way we have another problem.
> The original idea about FS pano's @6000x3000 pix equis is over, at
> least on this screen.
> Now what ??
>
> (crosspost on panoguide)

[Keith Martin-2]

Sometime around 25/10/09 (at 08:52 +0000) erik_leeman said:

>Hmmm, based on the fact that I am happy with my 1920x1920 cubefaces
>on a 1024px width screen, I would roughly need:
>
>2400x2400 cubefaces on a 1280px screen
>3600x3600 cubefaces on a 1920px screen
>4800x4800 cubefaces on a 2560px screen

This does assume that 1920x1920 is the threshold point on a
1024px-wide display and that your scale factor is correct. AND that
you always use the same zoom/FoV setting. I'm not challenging your
maths, I'm just observing and considering. :-)

Of course, another factor that shouldn't be ignored is the data size
of these things. Some of us have broadband speeds that rival local
network performance, but most don't. And many people in the world
still have sub-broadband-level connections. Is it *possible* to
deliver optimal panoramic content in terms of both large-screen
appearance AND 'non-exclusive' file sizes?

Maybe one approach is to have 'fullscreen' displays that can be
configured to use less than the full display area for the actual pano
if the pixel dimensions are above a certain threshold?

I could see something like this being possible through XML
configuration in KRPano, FPP and others. The logic could be "show a
black (or other) border of 'n' pixels width if the width (or height)
is greater than 'x'". In fact, I would rather like to have that
option!


>above a screen width of 1920 I would appear to need a better
>camera/lens combination than what I have now if I want to maintain
>my present level of image quality.

This does depend to an extent on the zoom/FoV that you use. Don't
forget, too, that people will tend to sit slightly further away from
a really large screen than from a smaller one. Okay, the difference
won't be major - we're not talking about TV viewing - but it could
swing things just a little in your favour here.


>A 5DmkII would deliver the pixels, but would my lens deliver the
>required sharpness/contrast at that resolution?

Is that a 15mm Canon fisheye designed for full-frame use? You may be
*approaching* the capability limits but I wouldn't expect there to be
a big problem. When you look at 100% (pixel-for-pixel) scale how do
individual full-res shots look to you?

k

[Keith Martin-2]

Sometime around 25/10/09 (at 11:53 +0000) erik_leeman said:

>Compared to those (non-streaming) video files everybody is watching
>on-line, a pano file size of 5 to 10MB doesn't sound excessive to me
>anymore.

True in one sense but perhaps not the best analogy - especially as
most video seen today is actually streaming in the general sense.
With video you only have to wait for maybe a few hundred KB of image
data before you start to see things, and see the entire image of the
first few frames too. It could be said that you've always seen 'all'
of the current video image... that's what's keeping you distracted
while the next bit is coming through. The pano equivalent to this
sort of thing would be streaming tiles ala KRPano etc., not just
large cubes/equirects.

I'm on an 8m/bit ADSL line that does generally deliver close to the
billed speed, which is probably not too bad for UK standards. I doubt
I'd look at many panos at all if each one was approaching 10MB in
size. Not unless it did some very clever tile streaming.

k

[erik_leeman]

There's no point in ignoring 'Things are a'Changing' continuously, they always have, we HAVE to respond to that. Just like those 500kB two-shot Ipix poststamp-sized 360x180 degree images are something of the distant past for us today, 1.5MB 4-shot 8mm fisheye-based images optimized for 1024x786px screens will be obsolete shortly. Considering the success of YouTube etc. the infrastructure for web-display of huge files appears to be here already. Compared to those (non-streaming) video files everybody is watching on-line, a pano file size of 5 to 10MB doesn't sound excessive to me anymore.

Lacking 'hard data', my 1920x1920 threshold is merely based on my own subjective observations, and my math was VERY simple:
1280/1024 = 1.25  -> 1.25*1920 = 2400
1920/1024 = 1.875 -> 1.875*1920 = 3600
2560/1024 = 2.5   -> 2.5*1920 = 4800
It was just to get a rough idea of what would be required to feed those new BIG screens without having access to one for testing.

Erik Leeman

(www.erikleeman.com)


--- In [hidden email], Keith Martin wrote:

> This does assume that 1920x1920 is the threshold point on a
> 1024px-wide display and that your scale factor is correct. AND that
> you always use the same zoom/FoV setting. I'm not challenging your
> maths, I'm just observing and considering. :-)
>
> Of course, another factor that shouldn't be ignored is the data
> size of these things. Some of us have broadband speeds that rival
> local network performance, but most don't. And many people in the
> world still have sub-broadband-level connections. Is it *possible*
> to deliver optimal panoramic content in terms of both large-screen
> appearance AND 'non-exclusive' file sizes?
>
> Maybe one approach is to have 'fullscreen' displays that can be
> configured to use less than the full display area for the actual
> pano if the pixel dimensions are above a certain threshold?
>
> I could see something like this being possible through XML
> configuration in KRPano, FPP and others. The logic could be "show a
> black (or other) border of 'n' pixels width if the width (or
> height) is greater than 'x'". In fact, I would rather like to have
> that option!
>
> >above a screen width of 1920 I would appear to need a better
> >camera/lens combination than what I have now if I want to maintain
> >my present level of image quality.
>
> This does depend to an extent on the zoom/FoV that you use. Don't
> forget, too, that people will tend to sit slightly further away
> from a really large screen than from a smaller one. Okay, the
> difference won't be major - we're not talking about TV viewing -
> but it could swing things just a little in your favour here.
>
> >A 5DmkII would deliver the pixels, but would my lens deliver the
> >required sharpness/contrast at that resolution?
>
> Is that a 15mm Canon fisheye designed for full-frame use? You may
> be *approaching* the capability limits but I wouldn't expect there
> to be a big problem. When you look at 100% (pixel-for-pixel) scale
> how do individual full-res shots look to you?
>
> k

[Jann Lipka]

I have a 30 inch display 2560x1600 , and yes 6x3 K FS is a bit small .


My Canon 5Dmk2 and Tok@12mm gives me 10Kx 5K .

I hope Krpano makes it easier to load  - here comes one example
10x 5K on my server space ( average speed server located in
south of Sweden )

http://www.lipka.se/panos/ukk_examples/level6/index.html


Fullscreen  loads very good here but I'm on fast connection .



Also - I talked / showed some panos  to a couple of
architects and changed my default viewing
from " normal " to " architecual"

Specially on large screens many people say looking at panos makes you bit dizzy ...

[Ian Wood-2]

On 25 Oct 2009, at 08:52, erik_leeman wrote:

> Hmmm, based on the fact that I am happy with my 1920x1920 cubefaces  
> on a 1024px width screen, I would roughly need:

I'd say that's probably overkill, or you are zooming in a LONG way.

According to Ken Turkoswki's pan resolution page a 1920px cubeface  
would let you zoom in to about 40 degrees vertical FoV on a 1024x768px  
screen before losing any resolution.

For that size screen, if you're not going to zoom in past 75 degrees  
vertical FoV, 1024px cubefaces will *just about* do the job.

I generally use 1536px cubefaces and find they are OK up to about 17"  
screen size, regardless of the actual screen resolution.

http://www.worldserver.com/turk/quicktimevr/panores.html

Ian

[erik_leeman]

It looks like we do things differently, and in my view there's nothing wrong with that.
All panos on my website have been set up (in Pano2VR) to have an initial hFOV of 60 degrees. I allow the large versions to be zoomed in to 45 degrees, and out to 80 degrees. DevalVR tells me that at 45 degrees hFOV the image is zoomed in at '100%', and indeed it looks like that is so, zooming in any further only reduces image quality.
I always try to make every pixel count, that's why I usually need about 15 shots from my 5D/15mm combo, using only the really sharp bits. Three-to-four shot panos from a crop sensor camera are a different kettle of fish for which other rules of thumb apply.

Cheers!

Erik Leeman

(www.erikleeman.com)
 

--- In [hidden email], Ian Wood wrote:

>
> On 25 Oct 2009, at 08:52, erik_leeman wrote:
>
> > Hmmm, based on the fact that I am happy with my 1920x1920
> > cubefaces on a 1024px width screen, I would roughly need:
>
> I'd say that's probably overkill, or you are zooming in a LONG way.
>
> According to Ken Turkoswki's pan resolution page a 1920px cubeface
> would let you zoom in to about 40 degrees vertical FoV on a
> 1024x768px screen before losing any resolution.
>
> For that size screen, if you're not going to zoom in past 75
> degrees vertical FoV, 1024px cubefaces will *just about* do the job.
>
> I generally use 1536px cubefaces and find they are OK up to about
> 17" screen size, regardless of the actual screen resolution.
>
> http://www.worldserver.com/turk/quicktimevr/panores.html
>
> Ian

[Hans-74]

--- In [hidden email], "erik_leeman" <erik.leeman@...> wrote:
>
> Hmmm, based on the fact that I am happy with my 1920x1920 cubefaces on a 1024px width screen, I would roughly need:
>
> 2400x2400 cubefaces on a 1280px screen
> 3600x3600 cubefaces on a 1920px screen
> 4800x4800 cubefaces on a 2560px screen
>

These are really overkill.

The best resolution is the one which corresponds pixel to pixel to the screen.
This gives you the best sharpness and the best panning with minimum pixel-shimmering.
Very few people use zooming and in fullscreen there is no need for it.

All widescreens together are 65% of the visitors.
If we look at the current most used widescreen resolutions the most used is 1280x800 with around 22%. Next is 1440x900 with 11% and 1680x1050 with 9%
I only have 4% 1920x1200

All  the above widescreens have the same proportions.

Apart from the new iMac 2560x1440 there are 2 already used resolutions with the same extra width.
1920x1080 and 1366x768. These are the LCD TV resolutions which are used by almost all  new TV. Many people including me use  an LCD TV as extra screen.
I have 3.5% of them.
I guess  this will become even more usual as Digital TV takes over completely.

To estimate the best resolution is very simple.
The standard widescreen gives you a 65x90 degree view and this gives you no distortion.
You can even go up to 70x96 for most scenes.

So that means that you need a cubeface of 1920x1920 for the 1920x1200 screen for a 65 degree vertical FOV .
1600x1600 does it fine with 70 degree.

The new extra wide format  will give you a 60x90 or 70x102 degree view.
Which means 2560x2560 will give you a perfect size for the big iMac

I am sure even a 1600x1600 will look fine on them as the interpolation both in Flash and Quicktime is fine as long as you do not oversharp the panos.

For general use 2560 is also the max as we still have users with just 512 mb Ram and larger flash panos may fail.

Hans

[Ian Wood-2]

On 25 Oct 2009, at 11:53, erik_leeman wrote:

> Considering the success of YouTube etc. the infrastructure for web-
> display of huge files appears to be here already. Compared to those  
> (non-streaming) video files everybody is watching on-line, a pano  
> file size of 5 to 10MB doesn't sound excessive to me anymore.

Which non-streaming video files are you talking about? ;-)

The vast majority of video on the web these days via YT, Vimeo,  
Viddler etc. is *effectively* on-demand streaming video - not only do  
you not have to wait for the whole thing to download (Keith's related  
point about progressive downloading), you can scrub to any point on  
the timeline and the video will start downloading and playing from  
that point, even if downloading hadn't got that far yet.

We're back to the inherent differences between non-linear media  
(panoramas) and linear media (video) - by their very nature, video  
files can afford to be larger in total because 99.9% of the time the  
player software knows exactly what is coming next and will be  
downloading it while you are watching (dependant on connection speed,  
of course) leading to a seamless viewing experience.

Panoramas, in comparison, require a large proportion of the data to be  
downloaded in advance because there is no really reliable way of  
predicting where the viewer will look next. The tiling players such as  
KRPano deal with this by downloading on demand as you change the  
direction and zoom level, like the old Zoomify plug-in for QTVR files,  
but this leads to a far less seamless experience - those continuous  
'loading' icons drive me nuts if I zoom in a bit and then look around  
within the panorama. Badly enough that I will usually go and look at  
something else and not bother to continue looking at that panorama,  
even on a fast connection. :-(

Ian

P.S. In case it's not clear, I'm not arguing against the use of larger  
files for panos, but linear video file sizes isn't a good comparison.

[Ian Wood-2]

On 25 Oct 2009, at 13:04, erik_leeman wrote:

> It looks like we do things differently, and in my view there's  
> nothing wrong with that.

Definitely!

On 25 Oct 2009, at 13:04, erik_leeman wrote:

> I always try to make every pixel count, that's why I usually need  
> about 15 shots from my 5D/15mm combo, using only the really sharp  
> bits.

Where are you seeing bits of image in the panorama that aren't sharp  
from that combo? Are you shooting with the aperture wide open?

I'm really curious as I used that exact setup for several years and  
have never seen the need to go beyond 6-around plus zenith & nadir,  
even when working towards a 12,000x6,000px equirectangular, let alone  
a 6k image for screen use.

Again, to make it really clear, this isn't meant aggressively, I want  
to know what problems you are seeing that you usually need 15 shots!

Ian

[erik_leeman]

No problem loading it here in the Netherlands on a standard ADSL connection. Ha, you nearly managed to stay out of the shot! ; )

Cheers!

Erik Leeman

--- In [hidden email], "jann_lipka" wrote:

>
> I have a 30 inch display 2560x1600 , and yes 6x3 K FS is a bit
> small .
>
> My Canon 5Dmk2 and Tok@12mm gives me 10Kx 5K .
>
> I hope Krpano makes it easier to load  - here comes one example
> 10x 5K on my server space ( average speed server located in
> south of Sweden )
>
> http://www.lipka.se/panos/ukk_examples/level6/index.html
>
> Fullscreen  loads very good here but I'm on fast connection .
>
> Also - I talked / showed some panos  to a couple of
> architects and changed my default viewing
> from " normal " to " architecual"
>
> Specially on large screens many people say looking at panos makes
> you bit dizzy ...

[erik_leeman]

Hi Ian,
I mentioned I set my panos to an angle of 60 degrees (min45 max80) in Pano2VR, and I THOUGHT it was the horizontal FOV, but you and Hans keep mentioning the vertical FOV as a measure of 'zoomness'. Nowhere in the documentation of Pano2VR can I find which of the two it actually is I am setting there, does anyone know?

I am a classical example of a 'pixel peeper', I do all my editing work at 100% or higher. This means I can most clearly see the difference in sharpness and detail in the centre of each image compared to the overlapping areas of the adjacent images. This makes me want to have as many 'centres of images' as I can possibly get. Also I strive towards zero stitch errors caused by moving objects (grass, clouds, people), which is also easier to do when you have a LOT of rough material to work with.
My standard 360x180 series comprises 10 shots around + several straight up and (especially) down. Working aperture for my Sigma 15mm is f 8.0 and I usually set focus at something about 1.5 to 2 meters away from me.

Regards,

Erik Leeman

(www.erikleeman.com)


--- In [hidden email], Ian Wood wrote:

> On 25 Oct 2009, at 13:04, erik_leeman wrote:
>
> > It looks like we do things differently, and in my view there's  
> > nothing wrong with that.
> Definitely!
> On 25 Oct 2009, at 13:04, erik_leeman wrote:
> > I always try to make every pixel count, that's why I usually
> > need about 15 shots from my 5D/15mm combo, using only the really
> > sharp bits.
>
> Where are you seeing bits of image in the panorama that aren't
> sharp from that combo? Are you shooting with the aperture wide open?
> I'm really curious as I used that exact setup for several years
> and have never seen the need to go beyond 6-around plus zenith &
> nadir, even when working towards a 12,000x6,000px equirectangular,
> let alone a 6k image for screen use.
>
> Again, to make it really clear, this isn't meant aggressively, I
> want to know what problems you are seeing that you usually need 15
> shots!
>
> Ian

[fierodeval]

I agree with Hans. Usually all viewers use the vertical FOV to manage the zoom (VFOV=45/zoom), and horizontal FOV is free. But when the screen is very stretched it's better to use horizontal FOV to manage the zoom (HFOV=45/zoom) and leave VFOV free. This way, with other viewers when you increase the wide of the screen, you increase the horizontal FOV and VFOV is always the same. With FFP when you increase the wide of the screen, the HFOV is always the same and the image only is augmented and VFOV decreases.

Recently I added an option to use HFOV or VFOV to manage the zoom in DevalVR. But maybe the best way is to implement in all viewers an automatic selection in the same way that FFP.

regards!
fiero



--- In [hidden email], "Hans" <hans@...> wrote:

>
>
>
> --- In [hidden email], bigwade <celsius@> wrote:
> >
> >
> > Today I visited the Mac shop in Utrecht NL to see the new iMac27"
> > Wow, it's big....and a NICE MACHINE...
> > The wide screen is 2560x1440 pix (3.55:2)
> > The iMac 24" is 19200x1200. (3.2:2)
> > The problem is FullScreen pano's looks horrible.....
> > The edges are stretched too much even with a 70 degr. view pano.
> > Since screens are going the wideview way we have another problem.
> > The original idea about FS pano's @6000x3000 pix equis is over, at least on
> > this screen.
> > Now what ??
>
> No problem. Just use FPP as viewer,
> Check www.panoramas.dk. Do you find any stretched panos even if the view in the browser is sometimes 3:1.
> In Real Fullscreen the view adapts to any screensize
>
> FPP uses intelligent FOV which adapts to the proportions of the display.
> You set a zoom value instead of a FOV and you will always have a nice view without any distortions at the corners.
>
> Hans
>

[Hans-74]

--- In [hidden email], "fierodeval" <fierodeval@...> wrote:
>
> I agree with Hans. Usually all viewers use the vertical FOV to manage the zoom <(VFOV=45/zoom), and horizontal FOV is free. But when the screen is very stretched it's <better to use horizontal FOV to manage the zoom (HFOV=45/zoom) and leave VFOV free. <This way, with other viewers when you increase the wide of the screen, you increase the <horizontal FOV and VFOV is always the same. With FFP when you increase the wide of the >screen, the HFOV is always the same and the image only is augmented and VFOV >decreases.
> Recently I added an option to use HFOV or VFOV to manage the zoom in DevalVR. But >>
>way that FFP.

The only viewer I know which uses horizontal FOV as default is PTViewer.

The problem in QTVR has always been the use of vertical FOV.
Pano2VR Flash also use it.
KRPano  uses vertical FOV also but the documentation does not mention it.

FPP  has the option to use VFov, HFov or Zoom level which is the default.

Zoom level adapts both HFov and VFov depending on the proportions of the display.

That means that what ever you do the image looks good if you chosen a zoom level that  looks good for you initial display. You can drag the image from a view which may be 500 pixel wide and 1000 pixels high to an extreme wide view which can be 1900 wide and 300 high if you want.
Zoom level 1 is default and corresponds to a vertical FOV of 60 on a normal widescreen monitor.

Hans

[erik_leeman]

Does this mean the definition of 'zoom' differs between viewers?
As Fiero mentioned his definition is vFOV=45/zoom, and Pano2VR's settings appear to follow the same formula, so you get this:

vFOV = 120° -> zoom = 0.375
vFOV = 100° -> zoom = 0.45
vFOV = 90° -> zoom = 0.50
vFOV = 75° -> zoom = 0.60
vFOV = 60° -> zoom = 0.75
vFOV = 50° -> zoom = 0.90
vFOV = 45° -> zoom = 1.00
vFOV = 30° -> zoom = 1.50

And in FPP it you say "Zoom level 1 is default and corresponds to a vertical FOV of 60 on a normal widescreen monitor".

Without a clearcut standard we should be careful when we compare the (effect of) settings for 'zoom' between viewers then.

Regards,

Erik Leeman

(www.erikleeman.com)


--- In [hidden email], "Hans" <hans@...> wrote:

>
> The only viewer I know which uses horizontal FOV as default is
> PTViewer.
>
> The problem in QTVR has always been the use of vertical FOV.
> Pano2VR Flash also use it.
> KRPano  uses vertical FOV also but the documentation does not
> mention it.
> FPP  has the option to use VFov, HFov or Zoom level which is the
> default.
> Zoom level adapts both HFov and VFov depending on the proportions
> of the display.
> That means that what ever you do the image looks good if you chosen
> a zoom level that  looks good for you initial display. You can drag
> the image from a view which may be 500 pixel wide and 1000 pixels
> high to an extreme wide view which can be 1900 wide and 300 high if
> you want.
> Zoom level 1 is default and corresponds to a vertical FOV of 60 on
> a normal widescreen monitor.
>
> Hans

[DemonDuck]

I thought all viewers just maintained the aspect ratio of the display surface
whether in canvas (small viewport) or fullscreen mode when selecting the portion
of the sphere to show.

That's the way I do it.  Then there is no stretching of the image.

fierodeval wrote:

> I agree with Hans. Usually all viewers use the vertical FOV to manage the zoom (VFOV=45/zoom), and horizontal FOV is free. But when the screen is very stretched it's better to use horizontal FOV to manage the zoom (HFOV=45/zoom) and leave VFOV free. This way, with other viewers when you increase the wide of the screen, you increase the horizontal FOV and VFOV is always the same. With FFP when you increase the wide of the screen, the HFOV is always the same and the image only is augmented and VFOV decreases.
>
> Recently I added an option to use HFOV or VFOV to manage the zoom in DevalVR. But maybe the best way is to implement in all viewers an automatic selection in the same way that FFP.
>
> regards!
> fiero
>
>
>
> --- In [hidden email], "Hans" <hans@...> wrote:
>>
>>
>> --- In [hidden email], bigwade <celsius@> wrote:
>>>
>>> Today I visited the Mac shop in Utrecht NL to see the new iMac27"
>>> Wow, it's big....and a NICE MACHINE...
>>> The wide screen is 2560x1440 pix (3.55:2)
>>> The iMac 24" is 19200x1200. (3.2:2)
>>> The problem is FullScreen pano's looks horrible.....
>>> The edges are stretched too much even with a 70 degr. view pano.
>>> Since screens are going the wideview way we have another problem.
>>> The original idea about FS pano's @6000x3000 pix equis is over, at least on
>>> this screen.
>>> Now what ??
>> No problem. Just use FPP as viewer,
>> Check www.panoramas.dk. Do you find any stretched panos even if the view in the browser is sometimes 3:1.
>> In Real Fullscreen the view adapts to any screensize
>>
>> FPP uses intelligent FOV which adapts to the proportions of the display.
>> You set a zoom value instead of a FOV and you will always have a nice view without any distortions at the corners.
>>
>> Hans
>>
>
>
>