[tom]

If you remember, once I asked for the same and I say it's useless.
It can be directly grabbed from max's units setup by the plugin.
The only benefit is, when you play with it while keeping actual units,
you can easily change the scale globally for the scene.
So it's just a multiplier...

[Frances]

Ok, so I understand that size had nothing to do with the 'spacial' airy effect in my knot, but that does surprise me and I find it hard to believe, but Thomas proved it.

Size does matter. If it didn't, Maxwell could hardly be considered a physically based renderer. If you model a house to real scale (ie a door is 30" wide) and scale it down to 1/10 (now your door measures only 3" wide), you would expect that it would look and feel like a dollhouse. It has mostly to do with real world light and shadow decay over distance.

The question is whether Maxwell is able to determine the scene units and perform its calculations based on those units - for example, you wouldn't want Maxwell to interpret your 30" wide door as 30m wide if you were interested in an accurate lighting simulation.

[Thomas An.]

So, in that case, what would you tweak to sharpen the image?

Read this page about Depth Of Field

DOF is affected by:
-Apperture (f/stop) (larger number, more sharp image)
-Lens length (smaller length more sharp image)
-Distance of camera target
-Scale

However, you do not need the knots to be the size of a football field.

In the image bellow the book is in normal size (11in or 30cm long) and I used mostly default parameters - this is more or less how it would look like from the human eye (50mm lens).



Camera:
Lens length=50mm (you may also try 35 for more sharp)
Film width,height=36,24 (leave them at default)
Shutter=125
Fstop=11 (try 16 or 24 for more sharp while reducing the shutter to 60 or 30)

Engine: 6 bounces (you can try 2 or 3 for even faster)
Environment: Skydome=10 (white)
Tonesmapping: Burnt=0.8, Gamma=2.2 (leave them at default)
Output: You may only need the alpha chanel if you want to put a background image later on within photoshop or something. Otherwise you do not need any of channels.

-

[Thomas An.]

In other words, I thought Thomas just proved size doesn't matter!

I didn't prove that size doesn't matter. The point I made was that the Maxwell conversion is NOT defective (there is no bug) and you do not have to specifically work in meters.

For example instead of having your model being 1meter you can let it be 100cm or 39.37inches and it will look the same (in both cases you are dealing with the same size... just different units... but the kind of units doesn't matter, only the scale).

[Micha]

... I'm confused from all this. I see, if I change the Rhino unit system it makes a difference. If an object is 10 high, than it 10 milimeters high if I set the unit system to milimeter and it is 10 meter ... .
Here an example with a big emitter. Between the images I only change the unit system of Rhino without scale.

[Thomas An.]

Micha, your image link is broken.

Anyway, if you change your units from meters to microns without scale (1 meter becomes 1micron) you essentially shrunk the image by 100000 but your camera is still huge. It is like trying to take a photo of a molecule with your digital camera... can you do it ? No.

[Micha]

... the image work's here ... and I'm happy like my Rhino works, anything is good - meter is meter and milimeter is milimeter. Maybe, I dosn't have understand right the discussion here. What is the problem?

... but the kind of units doesn't matter, only the scale

Hmm, I only can say the unit does matter here.

[Thomas An.]

Hmm, I only can say the unit does matter here.

Micha,

Is 39.37inches not the same as 1 meter in Germany ?


btw, the link is broken.

[iker]

Hmm, I only can say the unit does matter here.

... you are increasing/decreasing the distance between the Near Focus limit and the Far Focus limit by changing your Units and that makes your scene to focus/unfocus.

When your scene is in mm and you convert Units to m without re-scaling (to not modify the camera) , the distance is been modified, when you were rendering mm the focus distance was so close and in m is meters away. What Thomas says is it doesn't matter if you use cm or m as long as 1 m is 100 cm.


link is broken

[Micha]

What Thomas says is it doesn't matter if you use cm or m as long as 1 m is 100 cm.

Ahh, so simple.

Broken link - I don't know what I can do, it works here. Maybe, it is a temporary effect.

[tom]

what comes from what... chicken from egg or egg from chicken?
i'm afraid these scale and unit issues won't have an end...

[Frances]

what comes from what... chicken from egg or egg from chicken?
i'm afraid these scale and unit issues won't have an end...

Are we boring you?

This is really not a debatable matter. Scale, size and units are different things. Generally people with a modest amount of cad experience can understand the difference.

The issue here is why some renders come out completely blurry. Thomas has demonstrated (by going to far more trouble than I would have) that Rhinoll seems to interpret the units of the model, and therefore its true size, properly.

I suspect that there are times when the Rhino camera gets seriously messed up and the target ends up in the same position as the camera. This would explain why there is no DOF. I've had situations where the Perspective viewport went weird by itself - the lens length was zero. I can assure you that I didn't set the lens length to zero myself.

[tom]

Are we boring you?

You serious
No, I'm just trying to increase my "funny" points

Frances, I see you and others have great effort on telling scale and unit issues. Thank you for these great contributions. The more I read you, the more I learn. Please keep it...

[Thomas An.]

I suspect that there are times when the Rhino camera gets seriously messed up and the target ends up in the same position as the camera. This would explain why there is no DOF. I've had situations where the Perspective viewport went weird by itself - the lens length was zero. I can assure you that I didn't set the lens length to zero myself.

Yup !
I have seen it happen too and have to watch for it. Zooming in and out to extremes, or switching between parallel and perspective projction can trigger the camera/target collision scenario.

Pascal taught me to use the [_zoom _target] command and I think it works well.
(I didn't have these problems as much with V2)

[Thomas An.]

Are we boring you?

You serious
No, I'm just trying to increase my "funny" points

Frances, I see you and others have great effort on telling scale and unit issues. Thank you for these great contributions. The more I read you, the more I learn. Please keep it...

Hey Tom,

I think you are great guy
You are going out of your way, all the way to the Rhino section of this group to offer help, thoughts, advice, etc..
Thank you