Depth Of Field

[auri]

This is the first thread in the tutorials section that is not a tutorial, hurray!

[14:35] <tp^auri> Sj can you give a short overview over depth possibilities in mme? Somebody wants some answers in the forum and I haven't once worked with it
[14:39] <CaNaBiS> euhm mme_saveDepth 1; mme_depthFocus 4000; mme_depthRange 2000 or something should give something probably. make sure to mention that you can't have motion blur enabled
[14:39] <CaNaBiS> or probably lower depthfocus
[14:46] <tp^auri> Okey thank you. Bit short for a tutorial but..

mme_saveDepth enables saving of depth screenshots. The depth is no layer in the standard screenshots, the depth is saved in separate tga files.

mme_depthFocus defines the spot that the depth will take as sharp point in distance from the camera

mme_depthRange defines how much the environment is faded / blurred related to the "sharp spot"

And remember, have mme_blurFrames set to 0 when trying

This is an overview about the few cvars you have to create depth of field screenshots. You have to try values for yourself. But ask if you don't understand anything.

Greetings. John

[alextj]

Thanks John for taking your time to help me out, I really appreciate it.

After numerous tries I have finally managed to get something with it!
Here is screenshot from one scene:
Depth Mask
Final Image

Used values:
mme_depthFocus 100
mme_depthRange 200


Though, the result doesn't look very good in screenshot. Have to play more with values and the mask itself. This is very first time I try this method, so obviously it's not the best possible. But I will put here how I did this anyway:


A quick tutorial on how to apply blur using depth mask in Photoshop

*Open your screenshot
*Duplicate layer, select the new layer (or the one that is on top)
*Add gaussian blur (Filter -> Blur -> Gaussian Blur), I used around 8, but it is maybe too heavy, maybe 5 is good.
*Open your grayscale depth image, copy it to clipboard
*Go back to your original image, add layer mask to blurred layer (Layers)
*Select mask, goto Channels, there should be 5 channels (RGB, Red, Green, Blue and BlurLayer Mask). Make mask channel visible and paste your depth image (press Ctrl+V) (Channels)

That's it!
You can adjust Opacity of blur layer, or select mask, goto Image -> Adjustments -> Brightness/Contrast and change contrast and brightness of mask to adjust depth of field effect.


Finally, when you want to add this effect to your movie, you can use Batch feature in Photoshop to automatically add effect to all frames.
I can post Photoshop Batch tutorial if anyone needs it.

If there is any better way to add depth of field, please tell us!

[auri]

Great, thank you for this help.
I always get the feeling that not many people care about this forum.
But I enjoy this much, thank you!

And I think the screenshot does not look so bad. I like the left part. It's realistic that the background of the main room is unsharp as if the picture was taken with a normal camera. Maybe if depthrange is higher and it's only slightly unsharp in the background it even looks more realistic and not disturbin while watching a movie, fixing the watcher's view at the sharp spot and so on

Sleep well

[bladee]

use the afx plugin "frischluft" to add you blur.
download it from ->here<-

[alextj]

Thanks bladee for suggestion.
Too bad I am all running out of time with my studies (this probably sounds common ) and it looks like I am going to have no time to play around with quake and movie making for a while now. But I will be back later for sure.

[Deleted]

alextj: can you please post Photoshop batch tutorial for adding depth of field effect to all frames ? I'll be gratefull.

[biggz]

hey ladz

alextj you example image link's no longer work =[

using a depth mask in photoshop is possible, i've used it for stills;

example


As you can tell I work with rtcw: enemy territory and I'd like to find out if it's possible to replicate/use the same cvars for ET (mme_depthfocus; mme_depthrange).

If there cannot be an "ET mme" i'd just like to be able to use the above cvars as it is much more efficient than editing every single frame in photoshop (if you want accurate results).

Cheers
biggz

#ubik, #sql-gaming & #quaky.net @ QuakeNet, GameSurge and AusIRC

[foksie]

not sure if you can use mme in ET, but there is a tool for ET as far as i know (dont know if it has z_bufer frames).

as for doing it faster i recommend imagemagik/graphicsmagik

the allow for relativly fast automated z-frame blur creation.

here is one example, click to enlarge (a wallpaper i made):

[foksie]

a z_buffer tutorial comming soon in my production

[auri]

When do you come back to irc you infidel?

[foksie]

i am there!

[Deleted]

Strange thing have noticed I... :>
Find differences between this two screenshots :]




No Mr Bones on depth screen... What that means ? This...



Mr Bones chest is blurred like the background...

It looks like that "z-buffer" isn't taking everything from the demo to consider
Not only Mr Bones. Blood, smoke from rocket etc.
And it's not helping me with making my movie...

[auri]

Ah damn it. I shall report to the .. system admin of this world.

[foksie]

anything that is transparent isnt drawn on z buffer. that is standard, cause transparency doesnt work with z buffer properly.

thats why rail core, plasma shots, bfg shots, etc are not drawn on z buffer.

that includes some models like bones, angel, etc.

[Deleted]

God damn it... I mean Allah damn it..... I love Bones model...

[foksie]

render the z frames using the non transparent bones model

and render the normal frames using the transparent bones model

you can change what you render by using

mme_savedepth 0/1
mme_saveshots 0/1

[auri]

That is damn much work..

[foksie]

its not, it takes about 20% more time, you just make two projects, one renders the zbuffer one renders normal frames.

[auri]

Two projects instead of one sounds like 200% percent work.
Kinda.
I know what you mean, it would surely work. But I know that all people rather want a solution.

[foksie]

Unfortunatly there is none.

I reported this to canabis in the early days of 1.6 beta (arround version 4) he tried to fix it, but it simply cant be done properly.

And its not 200% more work, since z buffer rendering is seperated from normal rendering. So rendering it all together is almost same as rendering each of them seperatly.

[ebola]

is it possible to change depth field format?
I would like .tga

[foksie]

i suggested it to Canabis a while ago, expect it in next version of beta.