Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
And BTW, to show the "raw" re-texturing, without the inclusion of the shadows from the AO, is the image below.
Just by multiplying the AO pass with this result gives a decent, although very diffuse, shadows on the dress.
Very interesting results.
Stil digging this thread, I am banging my head a bit on the AO pass. I have always had issues with making it work, I watched you make the changes and tried it myself on a simple scene.
Scene is a plane, with 6 objects, 2 squares, circle, cylinder, cone and a icosahedron (remind me never to have to type that again). Default scene setup. (first image)
I render with the Ambient Occlusion multi-pass, but without Indirect Light and I get a white box :) Makes sense.
I turn on Indirect Light and I get black where every object in the scene is, it doesn't matter if I have Scene Lights on or Off. Scene -> Abient set to 0 or 100% doesn't make a difference to the Ambient Occlusion Layer.
So turning on Abient Occlusion Only, and I get some minor Black lines outlining the object, not enough to even tell what they are. (what you were getting when you turned it on.) Upping lighting Quality, or Accuracy doesn't much, the object accuracy or shadow accuracy to lower pixels also doesn't do much. (second image).
I set it to 10 feet and crank up the settings and I get a very blotchy layer. Interesting even if I turn off the mult-pass rendering I get no ambient in my main render, never seems to merge them. If I lower the settings get more fuzzy blotchy layer (third and forth). I never got a nice clean ambient and nothing seemed to come in the main beauty pass even with multi-pass turned off.
Continued in next Post
Continued. (like you if scene lights are on or off doesn't matter for the multi-pass ambient layer)
I try turning on Skylight and that just clips off some of the shadow area.
So I turn on HDRI lighting, turn all the shaders to white, render with skylight on, then with indirect on and off (first and second).
That looks closer to the Ambient pass you showed, though this is a non-multi-pass (beauty pass). (and not blotchy)
and in the above posting I forgot to show this, which is why no ambient in the render even with no multipass but the settings turned on (third).
So I am asking for help, not sure what I missed or where to look. (if this needs to go in another thread so we don't muddy this one up, let me know and I will do that.
Thanks
Milo
Did you increase the photon accuracy? Did you increase the photon amount? Did you increase the lighting accuracy?
Oops! Pardon me, I think I put Carrara in Zombie mode and I have to do my impression of the Sampsonite Gorilla for my computer.... I may be awhile..... :shut:
Yeah, I learned many years ago that when you're working with 3D software and are unsure what your settings do, just start by cranking any setting that sounds like "quality" or "accuracy" to it's maximum :) :) :)
Then later on when you figure out what they do, you can bring them down to normal ranges.
I tried that just now, 500,000 Photons and Map Accuracy to 100% not much difference, however settings Lighting Quality to Best at accuracy of the smallest 1 pixel produced the result. Even with the Photon Count back to the default. I didn't think about the Lighting Quality settings, that took away the Blotchies and with the accuracy there tightened it up a bit. You also have to have the Object accuracy and Shadow Accuracy set down to a low value, in this case I put it at .5 pixels
Skylight makes no difference but Indirect LIghting checked and only can get the ambient Occlusion pass to come out if you render Ambient Occlusion Only. If I am reading this right I have to render each scene/frame twice if I was wanting the Ambient Pass once with Ambient Occlusion only and once with either full indirect lighting or what ever settings on the beauty pass. Will have to see how that affects shadows.
Thanks EP I forgot about lighting Quality right under the Photon counts
One area I touched on lightly before but didn't do an example for was using the World Position Pass information for a hobbyist favorite...FOG !!! :) :) :)
And in the real world it is also a VERY useful and extremely common effect, although it is often used in a way that is not obvious or really noticeable, but merely to shift the viewer's focus and attention to the main subject of the image.
The first image shows the composite WPP RGB image next to the final output image. And below that is the composite flow. Basically you just merge the WPP info into the beauty pass aux channels using the all important channel boolean tool, then feed that into the Volume Fog tool. And that tool generates a true, 3d volume fog which actually uses the position of the scene objects to show or hide (occlude) the fog effect. It also includes scaling, position, color and many other adjustments.
The second image is the final beauty pass after the application of a volume fog. Note that I changed color and scaled it so it acted more like a "ground fog" or smoke. And notice how it respects the scene objects, and even "hides" under the car object.
Also, as with just about anything in compositing, you have incredible flexibility. In this case, you can even choose to have the volume fog tool output ONLY the fog component as shown below. That way you can modify the fog component to, say, remove or tone down the bright fog on the car windshield or front hood.
I think everyone will agree that the volume fog "occlusion" effect is far superior to a simple depth pass fog. It gives a sense of objects actually "existing" in the fog and occluding the background, as opposed to objects just being covered by a uniform fog. Which is, unfortunately, the fog effect you see in many/most images...
Also, keep in mind that there is a very important aspect of fog that is often/usually/always ignored... :) :)
And that is the fact that fog is actually tiny drops of moisture in the air, and when light strikes it the fog tends to glow a bit. And that volumetric lighting effect REALLY sells the idea that the fog is real, and has depth and volume.
And then (depending upon how cool you are....) you might even consider connecting some other tools into your fog, such as a Noise tool. That will give you a more, well, noisy fog effect with more variations.
In this case I set the additional noise input to simulate light rays emanating from beneath the car.
And here I utilized yet another super groovy aspect of the volume fog tool, and that is the ability to use a simple mask (in this case a circular mask tool with a soft edge) to mask out the volume fog effect on the front windshield and hood of the car. The flow for that is shown below the image.
And keep in mind most of this stuff can be done in seconds...in less time than it takes to open Carrara and set up your scene for a render.
I also want to discuss "cool backgrounds" :) :) :)
Why? Cuz they're cool.
I realize it's kinda of counter-intuitive for many hobbyists, but you don't need to show everything in the scene, clearly. Some times you can just make a cool, kind of out-of-focus background image that looks like it's just an out-of-focus camera, and put your main subject in the foreground. And cool backgrounds like that are real easy to generate.
Below I have an out-of-focus-looking image that can be used for a background. I just took some noise and applied a lens blur to it, and after some tweaking, I got this cool looking "bokeh" image. You can see the effect of the hexagonal shaped blades in the lens.
And I used that image to re-texture the screen in my car/girl image.
And here's another tweak of the fog in the image, but this time using what's called a "b-spline" mask. Basically you just drag and drop the mask tool (circle, image map, b-spline, etc.) and in this case just draw out the mask shape you want, on top of your image, and voila, you have a mask. And you can then move and rotate and re-draw the mask and see the effects in real time.
Far easier than going back into Carrara and trying to tweak the volumetric fog (is there one in Carrara? :) ) to make it match exactly what you want in your final image.
And like I said, you can set up a pre-made and tested composition that you've already perfected, and save it for use later. All you need to do is change the source images and tweak the settings for that particular image and you're all set.
Hi Milo,.
You're trying to get an "Ambient Occlusion" pass,. in a really "open" scene, where any Direct and Ambient light would be able to hit the objects.
Occlusion is where the Ambient light isn't getting,. which is different from Direct shadow
I'm not sure why Joe has avoided answering or helping you figure it out and skipped quickly past to play with fog effects.
I thought this thread was to Help people,. but it's clearly not
still using the term "hobbyist" with an obviously derogatory intention.
Are artists "Born" a Professional ?,. ....or do they learn ?
Student is a much more apt term.
I'm Out of this thread.
Welcome to the club ! ;-)
Although I have MA degree in fine arts I can accept myself as hobbyist considering my short 3D graphic story. I takes time to master the skills especially in the field of art developed on computer science that for an artist is magic on its own. I do apreciate every bit of knowlege that help me to get to grisp with this kind of tools. So thank you Joe for your efforts to make me a bit less lost in that art genre. If there is nobody left just keep going for me! I will use all that as soon as I come back to my renderings.
Greeting Joe.
Marek
Yeah, I learned many years ago that when you're working with 3D software and are unsure what your settings do, just start by cranking any setting that sounds like "quality" or "accuracy" to it's maximum :) :) :)
Then later on when you figure out what they do, you can bring them down to normal ranges.
Hey, he was wondering why there was the graininess or ways to correct it. I'm not just spouting crap off the top of my head. I suggested things to try that may work, and as I recall, I never said to crank everything up to maximum. If you actually have a suggestion, then suggest, otherwise, well, you figure it out....
That's fine if you can't stick to the issues and partcipate to help the other "students", and instead just leave in a huff. But why not consider thanking me for my almost 4 weeks of working this thread? And why not stay and participate and help the other "students" to improve?
Or maybe learning isn't really what you're interested in?
I was agreeing with you.
Wow, thank you Marek. Like a breath of fresh air. :) :)
Anyway, since there are folks like Marek who are still interested in my lengthy ramblings, I'll continue. And I hope others can step in and help answer questions if I don't.
I'll mention a minor issue, but one that might require some clarification. When you work with compositing, you'll hear two terms which sorta kinda sound alike: matte and mask. And you'll hear them used in what seems to be an almost interchangeable manner. But they're different.
The simplest way I can explain the difference is this:
A matte defines areas of transparency. A mask defines areas where you want an effect to apply.
Now those aren't strictly correct definitions, but in general it can give you an idea of the difference.
For example, in the beginning I showed an image of a character in front of a green screen. And the purpose is to later, in post production, remove the green background and make it transparent so you can insert a different background. And the way you do that is to form a matte, to define the transparent (green) areas of the image.
On the other hand, in the previous discussion on volumetric fog and world position pass, I used a simple tool to draw a grayscale image that would block an area of the volumetric fog from appearing on the windshield and front hood of the car. I drew a mask. I defined the area where I wanted the volumetric fog to apply.
Now what we can do with the basics we've learned so far is to try to integrate our 3D scene with a simple background image. It will be instructive to show all the aspects you need to consider when integrating 3D and 2D.
Keep in mind that this is an extremely common and key aspect of compositing. Integrating 3D images/scenes with 2D footage shot in a studio or on location.
So let's take a cool image of Rashad's home town (NYC) and figure out how we can integrate our 3D images with that image.
Below is a photo I downloaded from the web. Now, the first step is to observe and understand the photo. What can we tell about the image?
Well, it seems clear that the image was shot in the early morning on a clear day. How do we know morning? Well, if you're familiar with NYC, you realize the camera is looking north over Manhattan. And it's also clear from the shadows and the light on the building that the sun is located on your right side, which is East. So clearly the sun is rising over Manhattan.
And you can also tell some other things from the light and shadow. The sun is low on the horizon, and the light has the orange tint of a rising sun. There is also a bit of blue ambient skylight.
So we know we can start setting up our Carrara scene with, at a minimum, maybe a Realistic Sky with sunlight, and some distant light to simulate the ambient skylight. And of course, shadows :)
We also know that we want to make sure the sunlight is orange, which is why the Realistic Sky is nice. It simulates the orange-ish color you get during sunrise and sunset when the sun is being colored by the hazy atmosphere. And we also need to make the ambient skylight a bluish to match the blue sky.
Keep in mind that colors are a key element in compositing, in order to make sure your foreground blends with the background.
Next what I can do is to set up a very simple 3D scene in Carrara. I brought in a very old spaceship I had lying around. And I chose that ship because I had given it a very reflective brushed chrome texture, which will reflect scene colors and light very well, and be easier to integrate with a background. It will "inherit" the ambient lights.
I also added only two scene lights. A sunlight set to be low on the horizon to the east, and with an intensity of maybe 120% or something like that. The other light was a blue tinted directional light at only maybe a 50% intensity.
And that's pretty much it for the 3D scene. Just render, then save out all the passes you'll need, and you're done in Carrara.
Next what I did is to bring all of the needed render passes into my compositing app and bring them all together to get a final image.
Below is the compositing flow I used. It may look confusing, but really it's pretty simple. Each pass is brought in, and in most cases I added a color correcter to the pass to modify it as needed. So I have diffuse, ambient, GI, reflection, shadow, and specular passes. And the "merge" tools are basically used to determine how the images will be blended, like you do in PS.
And at the end of the line I use a simple Object ID mask to form a matte to place the spaceship over the background skyline.
In reality I'd probably do it a bit differently, but I think this should be a little clearer to understand.
And the composited image, which will be the starting point for making a final, integrated image, is shown below.
Note how the orange sun to the east is reflected off the tail section of the spaceship. That helps to integrate the ship into a scene which is clearly lit with an orange sunlight, low in the sky. And the rest of the ship is in relative shadows, as are the buildings in lower Manhattan. This is because the sun hasn't reached a point in the sky where it is causing a lot of overhead skylight and bounce light.
Now, the next step is to decide what's wrong with the image. What can be improved?
Well, the first thing that sticks out is that the image is cluttered. Everything is in perfect focus. And especially when the light level is relatively low, we don't expect everything, both close and far off in the distance, to be in perfect focus.
And presumably we want the viewer's attention to be drawn to the spaceship. Why? Because you dont' often see a spaceship from the planet Gooblie 7 circling lower Manhattan. :) :) :)
So we have to decide what we want to do about that...
Now, one EXTREMELY important point about the compositing process and the flow I showed.
What that allows you to do, that is virtually impossible to do in your 3D program, is to individually modify (boost gain, color correct, etc.) every single aspect of your image, and do it in real time, with real time feedback.
People talk excitedly about the benefits of the latest unbiased renderers, partly because they give realtime feedback with changes. Well, this is another method to have far more flexibility, as well as much faster realtime feedback.
In this case, the rendered image I got from Carrara was a bit dark compared to what it should be if it was really in the skies over NYC. So the compositing flow I used gave me the ability to boost, for example, the luminosity of the reflections while at the same time allowing me to give them a more orange hue.
And you can do that with every single aspect of your image.
Someone previously mentioned the idea of "doing it right the first time". While that's a nice idea, in practice it can be extremely difficult and time consuming, if not impossible.
Now, one thing you can do to make the spaceship stand out from the background is a straight and simple blur, as shown in the image below.
This is a lens blur that really isn't a distance lens blur, just a background blur with the spaceship masked out of the blur effect.
It's okay, but the background blur is a bit overdone to show the effect. In fact it looks a bit too much like a miniature.. :) :)
So this is where some artistic innovation comes into the picture...
What can you do to make this look more real or more exciting and to focus attention on the subject?
Well, there are many many things you can do....
For example, take a step back and think about your image. What's the purpose of an image of a spaceship circling NYC? Well, maybe it's because somebody was flying in a 737 from LA to NY and took a photo out the window just as a spaceship passed. :) :)
Which means you need to modify the image to make it seems like it's a handheld cellphone shot taken by someone very surprised to see a spaceship when they looked out of an airliner window :) :)
Or maybe it's in the future and that's the latest version of a Toyota JUV (jet utility vehicle) :) :)
It all depends on your purpose.
Or if you really wanted to get cool, you could build a scene in Carrara of some skyscraper objects and project that background image on them (camera mapping) and animate the camera. And you can even insert some volumetric fog in your compositing app, based on the WPP data from that scene.
And of course, another thing you can do is ask the internet for some images of airplanes over NYC in the morning, and get some ideas for how the spaceship SHOULD look in terms of lighting and color and shadows and background blur, etc.
Or maybe apply some motion blur to the spaceship and background...
Or you could treat the image as if a camera flash was used to take the photo, and light up the spaceship with a bright flash.
Or maybe simulate an explosion of anti aircraft guns trying to destroy the alien spaceship.
The possibilities are endless.