Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.Daz 3D is part of
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
Here's a very simple "directional blur" applied only to the spaceship (via a mask). No velocity passes, just drag and drop a directional blur tool, connect the mask, make a couple of tweaks, and voila
And here's the result of the EXACT same flow as the directional motion blur, but the ONLY difference is reversing the mask of the spaceship so that the directional blur is applied only to the background.
And that change consisted ONLY of clicking a single checkbox to reverse the mask, and the new BG motion blur updated in real time.
And as I showed previously, the cause of a real motion blur is the motion of an object AND the motion of the camera at the same time. Especially in the air, photographing a spaceship.
So you might want to consider applying one directional blur to the background, one to the foreground spaceship, and maybe even some camera shake effect.
I also wanted to cover a couple more aspects of smoke and fog which I think are real important.
Like I said, often you see a fog or smoke effect that is almost like you're looking thru a thin veil of uniformly light gray fabric. Or, if real depth info is used, it's like you're looking through a thin veil of uniformly gray fabric, except it gets more opaque with distance.
Which is fine if you're simulating, say, a forest on an overcast day.
However, what's important to remember is that generally, smoke and fog have a height, and a thickness, and they often hang close to the ground. Or at least relatively close. Which you realize if you fly in an airplane on a foggy day. As you climb through a low altitude, suddenly the sky clears and it's sunny.
Which means that often, fog and smoke aren't uniform effects with distance.
As you can see in the attached image, the fog varies by location and height. It's not uniform. And that means that any lights that are behind the fog (sunlight, highlights, reflections, etc) can make an interesting glow in the fog. And that glow can also vary by location.
And if you study the attached image, which I think is an absolutely GORGEOUS image of a forest path with some fog and some gorgeous colors and highlights and occlusions, you can see all the incredible variation in lighting and color.
The fog is there, but the variations are wonderful. And it transmits the sunlight in a bright burst of light that glows in the distance. And notice how as you look thru the trees you see small variations in the apparent color and thickness of the fog.
And on a side note, study the leaves and the trees. I'm sure the image has been Photoshopped to death, but the image is absolutely gorgeous and can be very instructive. And notice how your eyes are drawn to the path and the glow in the distance. And the incredible variation in colors.
Well, after over 8200 views, I figure SOMEBODY must be following this thread. :) :)
I'll just assume the participation is so light because many folks here are afraid to show what they don't know.. :) :)
Anyway, I briefly touched on what's called "relighting" in your 2D composite using the normals pass. You add some 2D lights that use the normals information to determine what parts of an object's surface are pointed in their direction so they will light that surface.
However, there's a much cooler but more involved procedure, and that involves in using the 3D features of your compositing app. If it has them, that is.
A good example of this is the fire video I posted earlier in this thread. It was a composite of a flame sequence generated in another app, a Blender render comprised of just a textured floor and wall, and a smoke sim sequence generated in Fusion.
https://www.youtube.com/watch?v=1pJ-8a5v1iU
One of the things that was incomplete about the composite was that the fire wasn't integrated with the surroundings. There was no glow on the floor and wall from the light of the fire. And that's a great situation for using the 3D relighting features of your compositing app.
What does that mean? Well, if you look at the video, you'll see a big fire that flickers and moves and has various colors. And if it was a real fire it would cast bright, flickering light on the floor and walls surrounding it. And that means that the light it casts needs to be a function of the light of the flames in the sequence of flame images.
Hmmm...how are you gonna make the light it casts depend upon the light and colors in the image sequence?
Y'know, on second thought, I'm thinking I'll end the tutorial at this point. Honestly, it just ain't worth the frustration from the constant negative reactions and lack of interest. It's pretty clear that most people aren't really interested in learning when most of the responses disregard the issues and only focus on ridiculously silly stuff like being offended at being called "hobbyists". Which most here are, but apparently that's irrelevant.
And I'm still scratching my head at the criticisms because I didn't answer each and every question anyone posed over the last 4 weeks. Nobody here can answer any questions? Wow. I'm speechless on that one.
It's sad that the ones who do most of the talking are also the ones who focus only on irrelevant emotional sensitivities, but ignore almost 300 posts worth of what some in the real world consider very good information. "Students"? I don't think so. I can't remember ever seeing anyone in my college classes pack up their stuff and go home solely because the professor didn't say everything nicely enough. :) :)
But like I've always said, it's not about learning and issues here, it's about playing in the sandbox and convincing everyone that you know everything. Even if you're clueless.
So I'll let everyone get back to playing. It's not my loss. :) :)
Sorry to those who might have been finding any of this useful. If you're interested, maybe I can direct you to some different forums on the net where folks are really interested in discussing and learning this VFX stuff.
Great work, would be cool to see it continue, but even if it's reigned in here, it's still a trailer load of useful tips. If you are interested / consider it useful, it might be good to edit everything into a pdf quick guide to compositing (that would take some work though, editing out personal stuff and side banter, ensuring images used are distributable as per rights, etc) and either putting it up on the Daz Store if you're so inclined or as a free download elsewhere and let the stats on number of downloads over time judge on how useful or otherwise people find the material.
Joe, thanks again for taking the time to explain the different passes and their potential uses. At over 8,400 views now, the community must be interested. I know I tend to be more of a 'lurk and learn' person...can't speak for anyone else, but I bet there are more than a few here who also appreciate your efforts.
I took a deeper look at the cool fog effects you created in Fusion using the Position pass. Unfortunately, Fusion is still unavailable for the Mac. I did take a peek at the Channel Boolean Tool in the manual for hints as to what is happening and it looked...impossible to understand at a glance. :-)
I vainly tried to figure out an equivalent in After Effects using its (probably lame) 3D Fog effect, but failed. I don't understand how to "drive" that position pass. Can it, for example, be used as a matte to be used with a photo of smoke or fog, and if so, how to tell the compositing app how to interpolate that pass? I need to experiment more...
Meanwhile, thanks for your most recent examples as well. I am inspired to attempt to recreate that nice fluffy uneven ground fog.
I'm finding this thread to be an eye opening experience in every single page. Your boundless enthusiasm about the subject is infectious and fruitful for us all. I am beginning to see many more possibilities, and your demonstrations and explanations are easy to follow. The value of this thread will grow over time, as people will bookmark this and refer to it so much more than you will ever know. Don't be too focused on the immediate reception, which is far more positive than negative. I think you are doing wonderful work.
In my own case, I would be much more talkative if I had anything useful to add, But as I stated before, this is outside of my area of experience. I'm also for personal reasons limiting my posts in the forums lately, it is not purposeful that the timing of my decision matches closely with the birth date of this thread. That said, if my lack of direct participation gave you the impression there was no interest from me, I do apologize. I've never enjoyed reading posts from you so much as I do in this thread.
Realize that many of the questions you might receive from us are almost child-like, because for me at least, that is where my level of understanding lies. So indeed, having you answer our silly seeming questions in your own insightful words goes a very long way. People are hungry for knowledge, Joe. You're the one that made us hungry in the first place. You made us want food we didn't even know we wanted until you made us taste it. Thank you for that taste, my friend! Don't stop feeding us now! Dont take the desire for more knowledge as a criticism if you can avoid it. People are bringing questions and it is fair for them to hope for you to provide the answers because you are the thread initiator. You have our attention Maestro.
Several of us have a long history of spark inducing interactions between certain individuals, myself included. Many a useful thread has been dismantled for that reason which needs to be changed. This is the main reason why I'm policing myself a bit more. Whenever the conversation is focused on the images and the concepts the discussion is fantastic. As soon as it delves into personalities and how one personality mistreats another...., like I'm doing right this minute, it all goes to crap. So I'm going to take my own advice and leave it at that. Let's stay focused on the imagery, the lessons of compositing and professional workflows. On the images I say...
Preach, brotha!
Following your instructions, I've begun working with render passes in particular the depth pass in Octane. Thanks to you I've finally got something like an atmosphere for my tropical scene renders. Last week for days at a time I had about 10 examples of scenes I'd applied the effect to but I had not saved them. Octane spits them out so quickly and I was playing around with them so much back and forth that I lost track and failed to save them properly. I will upload some new examples over the next days or so.
Below is the only finished composite I managed to save from that series of tests. The depth pass was applied as a separate layer above the original image as a Screen in Paint Shop Pro X6. I tinted the depth mask blue rgb 147 and this was the result. Thoughts?
I think it is an improvement on the original, which I do not still have saved unfortunately so we cannot directly compare them now. I do not think it looks exactly like an "in scene" haze effect would operate, but it is close enough.
Thanks again, Joe. Please keep up the good work. The Carrara community members Evil, Dart, 3dAge and myself as well as all others can see the clear benefit in what you are doing. No one means you harm or disrespect. Quite the contrary. Soldier on.
Rashad Carter, I can see suggestions of an atmosphere there, nice job! I am very interested in pushing these techniques further, especially regarding landscapes. It seems to open up more possibilities of creating punchy, dynamic aerial perspective without all the time consuming atmospheres and GI and stuff...with way more flexibility. Keep up the tests! (And save an original for comparison... :-) )
I had some time to experiment a little more with the Position pass today. After Googling around a bit and plugging in my second brain cell I created this stunning test scene (first image) and rendered a Position pass. In Carrara the camera is facing down the 'Y', so that is my "depth" (Green), left to right is the 'X' (Red), and up and down 'Z'? I think I am reading that right, but not sure.
I then brought both into after effects and started to try to isolate parts of the Position pass using Set Channels. In the second image you can see I have only the Red channel showing, and I applied a Curves filter to exaggerate it a little (...a little too much, I can see some banding...). I Pre-comped that and then applied the Set Matte effect and used the Red channel for the matte. I then used that as a matte for some fractal noise. I tried the same with the blue channel (third image), same concept but changed the settings to blue, and since I wanted to try to create "ground fog" I had to invert the Set Matte effect.
The final image shows the fractal noise layer being applied via the red channel of the position pass and the blue channel of the position pass.
Strangely, for this test, I could turn off the Set Channels effect and the Set Matte effect would look the same. In this instance it acts more as visual feedback?
I also messed with some blurs and stuff.
Even Evil, Dart, and 3DAge have to be laughing at that one... :) :)
Are you new around here or something? :) :) :)
I'm enjoying it, and learning a lot. I have little to say on the subject yet, as my brain is basically empty on this matter and set to "suck".
And Rashad, where are your Fake GI Outdoor Scene notes? I've sat patiently through the indoor stuff (which is not an immediate concern to me) waiting for the main feature, and my popcorn is getting cold... ;-)
Hello there Joe, interested in the post-work showed here, a question regarding something mentioned by you in the above lines, how Carrara gives you a file with layers? I have done some tests and none gives me that result, can you point me please in what I need to set up to get them?
Regards
Otto
Hello there Joe, interested in the post-work showed here, a question regarding something mentioned by you in the above lines, how Carrara gives you a file with layers? I have done some tests and none gives me that result, can you point me please in what I need to set up to get them?
Regards
Otto
The layers show up for me if I save to a photoshop format. Oddly enough, if you use the photoshop format, the 3d layers do not seem to get generated (like the depth map, position map, etc. There may be another file type that supports layers, but I have only tried with the photoshop format.
The layers show up for me if I save to a photoshop format. Oddly enough, if you use the photoshop format, the 3d layers do not seem to get generated (like the depth map, position map, etc. There may be another file type that supports layers, but I have only tried with the photoshop format.
Thanks mate, but this is in the render room right? 'cause I am not getting it at all, here's what I am using as rendering options, I did chose the PS 3 and 7 versions, and nothing:
Thanks mate, but this is in the render room right? 'cause I am not getting it at all, here's what I am using as rendering options, I did chose the PS 3 and 7 versions, and nothing:
Otto,
First you need to click on "Add" and select the Multi-Pass Elements that you want to render.
Then, if you are saving in Photoshop format, the different elements (or "render passes") will appear in Photoshop as either layers or channels.
(Which may be why you are not seeing some of them, ncamp. Look under Channels.)
Hope that's helpful.
-Dean
Yep,
They are in the channels. Thanks for that!
ncamp
I really should be getting on with that. Soon...very soon.
I've been reading this thread almost every day, and have only posted once, I'm sure I'm not the only one.
There's plenty of reasons other than lack of interest that will keep people from posting.
How many airplanes take off and land safely every single day? But we never hear about those, we only hear about the ones that crash.
That's how the internet works sometimes.
Let that sink in for a second and it might explain why some people don't post more often.
Hopefully you'll keep this thread going, but I can understand if you choose not to.
Regardless, I've managed to learn and pick up new techniques from you, so I greatly appreciate the time you've spent on it and thank you for your hard work.
Joe - just to show your words didn't fall on deaf ears:)
I don't have a fancy compositor, so had to learn how to do this in Blender. Really fascinating stuff and thank you for sharing your expertise!
Probably not the greatest, but it is a start.
Before and after
Thanks mate, but this is in the render room right? 'cause I am not getting it at all, here's what I am using as rendering options, I did chose the PS 3 and 7 versions, and nothing:
Otto,
First you need to click on "Add" and select the Multi-Pass Elements that you want to render.
Then, if you are saving in Photoshop format, the different elements (or "render passes") will appear in Photoshop as either layers or channels.
(Which may be why you are not seeing some of them, ncamp. Look under Channels.)
Hope that's helpful.
-Dean
Found it! Thanks much....
Thanks mate, but this is in the render room right? 'cause I am not getting it at all, here's what I am using as rendering options, I did chose the PS 3 and 7 versions, and nothing:
Otto,
First you need to click on "Add" and select the Multi-Pass Elements that you want to render.
Then, if you are saving in Photoshop format, the different elements (or "render passes") will appear in Photoshop as either layers or channels.
(Which may be why you are not seeing some of them, ncamp. Look under Channels.)
Hope that's helpful.
-Dean
Here the screen shots that I wait since the beginning for finally understanding what one speaks.
Thank you Dean, I will re-examine all the thread now.
And thanks again, Joe ! ;-)
Thanks...very much "monkey see, monkey do", but understanding is starting! Really getting to like the node-based shaders, so logical and flexible!
I have a great question: is it composing ? :-)
http://www.dilmaza.com/intro/
Thanks diomede64 for the link - nice info and explanations.
Here is a link to cooking up your own passes along with suggested Blend Mode to use in the composite:
http://www.g-3d.com/Newer_AAC_3d_multi.html
I know it's old, but might spark some ideas in case Carrara doesn't render a certain pass correctly for people, or someone needs a custom pass. Ignore the software titles, should be achievable in Carrara. :-)
I tried another landscape. First image is straight render from Carrara, 1 Keylight and 1 Bounce, both Distant. 2nd image is after all adjustments. I used the Position pass and isolated the blue channel to use as a matte for some fog. The fog is made of a few layers of Fractal Noise with different Blend Modes. Nothing great, but having fun experimenting.
Be careful about applying stuff that works in one software to other software. Different 3D apps and compositing apps handle this stuff differently.
For example, I believe Carrara's depth pass is the opposite of most apps in that far stuff is white and near stuff is black. Other apps use far stuff black and near stuff white.
Nuke calls the "normal" blend as "over", and the "Linear Dodge (Add)" as "plus", while Fusion accomplishes the "add" with a "Normal" mode with Alpha Gain at 0. :) :)
Someone raised a question about some basics of images, so I thought I'd answer here in case others might benefit. The question basically revolved around images and channels and layers, and what it all means.
And I'm sure many have noticed that when you start messing with Photoshop or Gimp or whatever everyone uses, the concepts of what all this stuff means can get kind of confusing.
Anyway, here's my $.02....
Basically, every computer image you see is comprised of what are called "pixels". And that's short for "picture elements". And all it means is that every image is nothing more than a whole bunch of colored squares. So, for example, if you have a 640x480 image, it is made up of a total of 640x480 = 307,200 square pixels.
And what's most important to understand is that each pixel has only ONE color.
Now, as an example, look at the image below of a skyline. Beautiful. Lots of colors, lots of sharp lines. It's a 1280x960 image, which means there are a total of 1,228,800 pixels. And each of those pixels has a single color.
If you don't believe me, I've zoomed into the top of one of the buildings where there is clearly a sharp line of the building against a blue sky. Now, if you look at the zoomed segment you'll see that yes, in fact, the image is made up of pixels. And even the areas that seem to be sharp lines are actually, when you zoom in to the pixel level, nothing more than a bunch of jagged pixels that are fooling your eyes to make you think it's a sharp line.