[Sheik]

Has anyone had time to really test Vue7?
You could create HDRI spherical “sky domes” with Vue6, but the dynamic range of the HDR was very small, so it was not really HDR. Is that better in Vue7?
Vue6 was supposed to have “real scale”, but in practice it didn’t. Can I now make a scene in something like SketchUp, and then easily import it so that the scale is correct and you can start putting ecosystem materials on it? Are trees and environments now all adjusted to fit the “real scale”?
In Vue6 the render time / quality ratio was imo rather low. They say it is better now, is it?
Sheik

[Mattia Sullini]

As soon as the guys from DHL will bring me the pack, i will post first impressions!

[Mattia Sullini]

Here i am: hands on! First test was outputting an HDR map.
It did the work...took 2 hrs on a 2,66Ghz Quad but i've managed to output a 2262x1131 map from a Spectral 2 atmosphere.

Here's a downscaled 8bit image of it, hoping i am not breaking any EULA...the only bad thing are those stripes in the map...maybe something related to the supporting geometry i've chose for export, have to test more.


Tested on a simple scene to see if it works fine for environment. The answer is positive, and the results are really close to the ones i could get with the most similar map from the Dosch collection

here's Vue


and here's Dosch


@Sheik: you can get a real HDR if you call camera's settings and uncheck "Natural Film Response"

[Sheik]

That is looking good Mattia.
From this angle it is hard to tell how sharp shadows you got.
The problem I had with HDR from Vue6, even with "Natural Film Response" unchecked, was that the sun just wasn’t bright enough. I checked the HDR in PhotoMatrix, and it turned out the dynamic range wasn’t much bigger than on a simple JPG. I did make a request for improving this on their forum, and probably they did for Vue7.
So how is the other stuff?
Sheik

[Mattia Sullini]

I'm back from second testing stage!

Here are the results. The dynamic range of the otput image is not as wide as i would have expected, as you stated. not checked with photomatix, but when playing with exposure triggers in photoshop, it seems that the burn point for the sun is only 1-2 steps higher than the one of the brightest part of the sky itself...far less than the average ratio in realworld that should be around 20:1.
This said, i have to admit that in any case i've never found an .hdr map so perfect that i could get hard shadows from the hdr sun...the Dosch collection for examople with the same emirical method i used to check the Vue map, shows only a slightly higher range, like 3:1, not enough to show hard shadows as you can see in the images.
A workaround could be to use the output map for background , and an adjusted one with altered gamma and exposure for reflections and lighting. The result of this kind of quick adjustment is shown in the image also. I have to admit that the altered map looks quite unnatural, with oversaturated colors, and noisier when used in maxwell... but i believe there could be a way to weaken these problems



For what the other matter is concerned, the answer is positive. The picture below shows a 1mx1m grid imported into Vue with a x10 resizing.
Camera is positioned at 2m above the ground and aiming with no pitch. As you can see, the 2nd bar of the grid cuts the image right in the middle, and is cohincident with the horizon, meaning that the bar effectively marks the height of 2m as expected. The plant has been loaded and positioned without the need to scale it to fit the grid's dimensions....it seems that at least this problem is solved!




Talking about the quality/speed ratio, well, first impressions are that the situation hasn't cahnged much...

[Mattia Sullini]

Last update, then it's time to go back to work! I tried with an alternative tecnique. I rendered a panoramic spherical view (is an option for the camera, as you know), then i saved images with different exposures as if i was shooting in bracket mode with a realworld camera.
Then imported all the images in Photoshop and merged as hdr setting manually the ExposureValues, according to the ones set on exporting.
I needed 1/10 of the time to get the hdr i would have needed if i had exported the sky, but the dynamic range of the resulting hdr seems unvaried. Sun aura is burned at -3EV, while the brightest parts of the sky at 0EV, too narrow again.
I have to deduct this is not a problem of the export... maybe since the CG sun is much more controllable than an hdr one, developers tend to split the environment contribution between sun and sky, and then the hdr you can get from from the sky only covers the dynamic range of the sky itself only (it seems to be something around 6 steps), while sun contribution is left for the CG sun. Doing this way is avoided the possibility of getting an over-bright sun.
If this is the way things are set, it would be cool then to have a "merge CG sun to hdr map" option...maybe not effective but cool!

[Mattia Sullini]

I know, this is becoming a monologue, but maybe i reached the core of the matter. I cheated a little bit, and then set to -15EV the image corresponding to -4EV, duplicated it, painted pure white inside the sun (at -4 brightness was not 255,255,255) and saved as a fake -10EV.
Finally i recreated the hdr now on a range of 19EV. Cool. BUT.... when used in maxwell what i got was an image that looked as if i hadn't changed anything except for very sparse bright pixels.
If you look at the preview window i composited with the rendered image, you can easily see it shows much denser spots that now reveal their nature, beeing the "rays" cast by the hdr sun. In fact the preview window shows the hard shadows i am seeking since 3 hours ago.
I can predict that on the rendered image these shadows would appear clear when humanity will have colonized alpha centauri system, then the conclusion could be this: even if we had a full-range hdr map, we would need eons to see the results, then is much better to narrow the range and to use the CG sun for hard shadows and bright highlights.



Here's the result of the Maxwell sun matched to Vue's one (maxwell's is the bright spot on the underexposed image)...meaning the standard way in maxwell since even when you choose to use the image as lighting source, sun remains active until you don't uncheck it in the phisycal sky tab (i am specifying this not for you since i know the quality you get in your images, but for anybody that could be reading this post beeing a beginner )

[Sheik]

Great tests Mattia, very useful info.
Ho did you match the Maxwell sun position to the HDR?
So we should be able to lock location and time to match the HDRI, or at lest specify that information for HDRI skies if posted in the Maxwell Sky repository.
I remember testing some night skies from Vue. They were lit by stars and the moon, but I didn’t get much of a result, only darkness. Your explanation about the pointed light sources would also explain why that didn’t work. Do HDRI maps work differently in other render engines? I remember seeing sharp shadows with HDRI…
I don’t have Photoshop, but would it be possible to mask the sun and increase to dynamic range (ie make the sun brighter in HDR)?
It is nice to see they finally got the “real scale” implemented. I asked for it a few times over the years, and they usually the reply was roughly; “What ever for, why are you so obsessed with the real world?” .
Sheik

[Sheik]

I played with some more HRDI exports from Vue6, and I am confused...
I now got a HDR with a dynamic range of 19:1. There are so many parametres that affect this that I can´t say what the difference is.

Mattia, you mentioned an average "real world situation" would be 20:1, where did you get that info? I am not questioning it, just surprised because some HDRI sky maps I have colected from various locations seem to have huge dynamic ranges (like 12 000:1, (still doesn't create sharp shadows in Maxwell)).

Is there no standard calibration for HDR? I mean should a HDR not represent the real lighting levels at the location where the image was taken? Couldn't there be a setting to use the HDR at 100% intensity, which would again represent that same lighting in Maxwell. I don't realy understand the relationship between the dynamic range of the HDR and the Intensity setting in Maxwell, and then the use of the sun simultaniously... it becomes totally experimental for me. I wonder if there is more info on this somewhere? Maybe I finaly will look at the manual .

[Mattia Sullini]

Concerning the question about incrementing the range, you can actually cheat and do it. What i did was to tell Photoshop that the image saved at -4EV was saved at -15EV, and then created a fake -4EV one by painting white inside sun's disc.... i do not know exactly if this was you were asking for.

Regarding the ratio, you are right.
I mentally transformed the EV steps into a linear scale, while each EV step means twice the light.
Shame on me!

An hdr can theorically represent all the real lighting, but practically you need to record that range. Bracketing can only catch a limited range.
Googling the only device that could be able to catch all the range is this cheap camera here http://www.grafixgear.com/HTML/PDF/SpheroCam_HDR.pdf
Price isn't specified, but i've read elsewhere it should be around 50.000$

Last, increasing hdr intensity, i think will only shift the covered range, while keeping the maximum difference unaltered...i mean if the hdr is somehow clamped and then shows the same intensity for the sun and the brightest part of the sky, increasing its intensity will increase light emitted both from the sun and the sky. What is missed cannot be recovered...think at the RAW files of cameras: it's true you can adjust exposure after-shot, but if shadows or high lights fell out of the range capability of the sensor they are lost.
On the other hand, if you could stretch the range to fill all the scale, you would completely alter the caught relations between the parts of the image and then get weird results again!

[Leonardo]

Here's the result of the Maxwell sun matched to Vue's one (maxwell's is the bright spot on the underexposed image)...

Did you match it, or did maxwell found what was the brightest part of the HDRI and position the sun by itself... (I could only wish - that would be a cool feature)

[Mattia Sullini]

...let's try to be realistic and hope first for the sun to rotate at the same speed of the quadrant...

[Leonardo]

...let's try to be realistic and hope first for the sun to rotate at the same speed of the quadrant...

[Maximus3D]

You guys who got Vue 7 now, can you create some impressive scenes with it and show them off here for the rest of us poor sob's i'd love to see more of what all you can do with it.

/ Max

[Mattia Sullini]

Do you mean Maxwell+Vue or Vue only? If you meant vue only the website is full of it, and there are some really impressive ones
www.e-onsoftware.com

Otherwise these are some pics i can show...nothing impressive as you requested, just a taste!