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
Now that you have the basics of how to navigate and configure Blender to make it very Carrara-like, I'll take a short side trip to discuss one area that is very important to consider when jumping to another app like Blender.
And that is NODES...
A node-based application can be extremely intuitive and flexible and efficient. I think it's safe to say that once you've become familiar with nodes you'll never go back... :) :)
And Blender makes use of nodes in a wide range of features, including
1. Configuring materials and textures
2. It's internal node-based compositor
3. Working with render passes
4. etc...
Blender actually uses a node-based approach for configuring materials/shaders and textures. And personally, I find it FAR more intuitive and efficient than Carrara's shader room. Honestly, I never quite figured out how to do more than just simple shaders in Carrara's shader room. To me the "shader tree" concept is very confusing when you start to get layered shaders...
I don't think I'll go into much detail on nodes, I'll leave it up to anyone interested to pursue on their own. My only recommendation for those jumping over to Blender from Carrara is to keep in mind that if you're not sure how to do something, nodes might be the answer. Especially with materials and the stuff I mentioned above.
And to give you an idea of how awesomely nice the nodal material/texture features are, along with the physically based and realtime rendering of the Cycles renderer, together with the user-configurable "rooms" are, here is just one example of making a nodal texturing "room" (Screen Layout) that combines the node editor, a 3D View, and a realtime rendered view. Much much nicer than the shader preview in Carrara that always gives me fits and frustration...
So next time I want to apply nodal textures I just jump to my pre-configured Texture screen layout and I'm all set.
Oh, and I almost forgot...
One user preference you MUST set is under "File"....make sure the "Show Thumbnails" is checked, so when you load a file you see the thumbnail image instead of the default list... :)
For me, the realtime preview - and the highly customizable UI bit - are the most exciting thing about Blender. Not because these are ground-breaking, but just because these are 2010-era CG basic essentials that Carrara still doesn't have.
Good intro on Blender btw. I look forward to elaboration on realtime render and animation render workflow.
For me, the exciting things about Blender are the following:
Flames
Smoke
Fluids
Soft Body/Cloth
Dynamic Paint (eg., waves)
Node based compositor
Unbiased rendering
Physically accurate materials
Animation/rigging tools (expressions, etc.)
Modelling/sculpting tools
Camera tracking
Video editing
For the most part, with minor exceptions, none of those features exist in Carrara. And what is more surprising to me is how easy some of the more high-end features, such as smoke, flames, fluids, dynamic paint, etc., are to use.
So I suppose the next step is to consider why you would want to integrate an app like Blender with your Carrara work, and how would you do it?
Well, I just gave a long list of "high-end" features that Blender provides but are not available in Carrara. Personally, I feel that Carrara's strengths are in two areas: characters and content. But for the most part, I tend to generate my own content, so for me at least the big strength is characters.
So how do you integrate characters with all that other stuff you can get from Blender (or other apps)?
Well, one way is to do what the professionals do: film your actors in front of a green screen, then composite them later into the environment. The idea is to do your character work in Carrara, then using various techniques, export those results into a compositing app, and integrate those other elements which you generated in Blender.
In the other thread I generated on Compositing I give a number of ways you can do that using various techniques.
In thinking about where next to take this discussion or if I should even bother, I realized I overlooked a key strength of Carrara that is a relative weakness in Blender. And that is the extremely useful shader/object/etc. libraries, where you can just drag and drop saved stuff you've made before.
Blender doesn't have that directly, but that functionality certainly does exist.
The concept in Blender is that you can easily pull stuff from other existing "scene" files (aka ".blend" files in Blender) via what's called the "append" function. So one option is to save your materials in a "library" (.blend) file of materials, and append that library file to whatever scene you're working on and just pull those pre-made materials as needed.
And since your materials are generally each going to be a group of nodes you've configured, you can save that group and place it in the library file.
And the same concept applies to anything else you want to save in a library.
Now there are some plugins out there that can help a bit with that process if you want, as well as some existing materials libraries that are available so you can start your collection. But for those interested in making Blender as easy to use as Carrara I'd strongly suggest you consider making a library of materials/shaders, objects, etc.
I started to post this in another thread and then thought maybe this was the better place for it.
If anyone is thinking about trying out Blender I highly recommend the starting videos by BlenderForNoobs on YouTube - he has one called "Learn Blender in an Hour!" and it really will get you up and running and comfortable with the interface in an hour. He has a series of 10 or so introduction to modeling in Blender videos that introduce you in steps to most of the basic commands and shortcuts too.. they are very basic but if you follow along and try to anticipate his next steps you can get familiar with the basics of Blender in an afternoon.
A great tip related to this, and one keyboard shortcut I think you should memorize: you can use Control + Right/Left Arrow to instantly cycle through your layout presets. No matter how complex your scene is this switches to the next layout right away with no delay, so you can flip from your modeling layout to your texturing layout to your animation layout and back in seconds. I stumbled across this in the footnotes of a book I was reading and use it constantly now. No more fiddling with the layout - just set up a couple you like in addition to the default ones and you are good to go. :)
great thread, thanks JoeMomma2000.
Joe,
Great thread. You've done a great job pointing out the similiarities. The methodolgy and layout of Carrara was very approachable for me from the beginning. It drew me in and it's where I began cutting my teeth - learning 3D. When people say that they actually enjoy using Carrara - I get it. That said, I've downloaded the last version of Blender and I am jumping in. It really does have it all. Maybe with a bit of customization I may be able to make her look a bit like Carrara. One of the great and attractive draws is that the community plays such a vital role in development. I really believe that Carrara would enjoy the same level of involvement if things were done differently. It's strange that the conversation of Carrara always goes to the lack of $$$ for development. The fact that Blender is free - you can see that money and resources with surround a great product. Blender is free and I'll bet quite profitable for everyone involved.
Great thread! I appreciate the time you took to do this.
Realtime you have an interesting point, whether Blender, even though free, is profitable for everyone involved.
I checked their website and it seems they are saying that they have presently about 282 donors, donating a total of about 4,200 euros ($4,700 USD) per month. I don't think that's even enough to pay one full time software developer for a year (ie, less than $60k...).
So I'm suspecting that Blender is developed by volunteers who don't do it for the money, they do it for the enjoyment. Though I may be wrong about that...
Which is a completely different world from most private companies who own and develop software. They do it for the money.
I suppose DAZ could release Carrara to the world as open source, but then they'd lose a valuable asset, and also lose control of the software. They want software that will support their content, and if they give it away they can no longer control that. And you also have to ask yourself what benefit would DAZ gain by giving it away? They spent a lot of money on it over the years, why just give that away?
Now if they could find someone to buy it, then that might be a reasonable option. Or they could continue developing it.
As Evil says, who knows?
Okay, well I said I wouldn't go into detail about nodes, but I'm bored, and I'm sure there are folks who haven't yet jumped into node-based applications, so I'll try to explain the differences and relative benefits. And Blender has a very nice nodal system, used for many different things, so if you're going to use Blender as part of your workflow with Carrara, you'll probably want to start understanding nodes and using them. It will make your Blender experience a whole lot easier and more flexible, IMO.
BTW it's been a long time since I've used layer-based apps other than for some quick image modifications, so forgive me if I miss some stuff.
Let's take an example, and compare say Photoshop (which is "layers and history based") to a nodal compositing app which is, well, node-based.
Let's assume you want to perform the following operations on an image:
Now, in Photoshop you open the image, then apply a blur filter, then make another layer and load the tree image, then apply a color correction layer, and save the image.
That's a "layer-based" application. You stack image elements and operators as layers. And every operation you perform is listed in the "history". It's a bit like making lasagna. Just stack stuff in layers. And if you're ready to put the lasagna in the oven, then decide you forgot to add ricotta cheese in the middle of the stack of lasagna, how do you fix that? Well, you unstack the layers of pasta, add the cheese in the middle, and re-stack them. A layer-based approach is kinda-sorta like that. But not quite.
Now, lets say you don't like the blur you did in step 2. How do you undo it? Well, you step back in your History to step 2, which is the Blur step. However, that undoes all of the steps after that (like loading and color correcting the tree).
Of course, you can get around that, but the basic concept of a layer-based approach is for relatively simple modifications on individual images. It's not real flexible.
Now, let's look at a node-based app. How would you do the same operations?
Well, a key concept to understand is that every operation you can perform on your image (or sequence) is an individual, modifiable block, also known as a "node". And you can connect each node/operation to any other node just by dragging a connecting line from the output of one to the input of another. The "flow" of operations is defined not by a rigid, step-by-step history, but by an easily modifiable and flexible diagram showing the connections of all the nodes. And it's very easy to modify how the operations "flow"...you just reconnect the nodes to perform the way you want.
So how would you do the same thing with nodes?
Now, while both processes might seem fairly similar so far, let's now say we want to get rid of the blur on the background. With the layers based approach you have to step back in History to before the blur, and thereby you lose all operations after the blur. But with a node-based approach, all you do is select the blur node and disable it with a single keystroke.
Or let's say you want the same color correction to apply to both the foreground tree AND the background mountain range? With the node-based approach you merely connect both images to the color correction node by dragging a line from one node to the other.
Using our lasagna analogy, a node-based approach is a bit like having an "add ricotta cheese" node placed inside your stack of lasagna. You don't have to unstack those messy layers of cheese and pasta and meat sauce, you just access the "add ricotta cheese" node and it does it for you.
Here are the final image, and the nodes used to obtain the final image. You can ignore the "premult" and "transform" nodes if you want...they are just there to allow me to move and resize each image for composition purposes, and the premult is to, well, premult the image.
As with many things in 3D, the difference between apps may not necessarily be in how much better your final image looks with some apps compared to other apps. Often, the difference in features is more about how quickly and efficiently and intuitively you can perform the tasks you need to perform. And for the most part, that's the difference between a layer based and node based approach.
When doing complex stuff that requires many operations and many images, you want an app that makes that stuff quick and easy to perform.
So is a node based approach "better" than a layer based approach? Well, for some the answer is definitely YES. For others, a layer based approach is fine for the work they do.
But in any case, like I said, with Blender a node based approach is used for a number of tasks, so it's probably a good idea to at least become familiar with the approach so you can be competent in both.
I can't argue with your main point but to nit-pick a little bit, it's obvious here that you haven't used Photoshop in a long time or were not very familiar with it to start with because your example is a "bad" way to do it. Normally when retouching a photo (or a render) you would end up with dozens (sometimes as many as 50-60) named and grouped layers in order to avoid exactly this problem. That's just good technique.
You wouldn't blur your image, that's a bad habit - you'd make a copy of that layer and blur that and if you later wanted to change it you wouldn't have to undo everything, you'd just go to the blurred layer and redo just that layer (by deleting it, making a new copy of the stil untouched layer below it and blurring the new copy) without touching anything you applied above it.
The advantage of node-based editing in this case would be that you don't have to have dozens of layers for each image, which can lead to very large file sizes. Your overall point about flexibility stands though - there are definitely cases where changing an intermediate layer would make things above it not work correctly and that almost never occurs (in my very limited experience so far) with node-based editing.
If you're using the latest versions of Photoshop, you can go one step better and convert the layer for Smart Filters, so that you can, say, apply a Gaussian Blur. All you need do is click on the Gaussian blur in the layer to bring the panel up and re-adjust the amount of blur. Wherever you can work non-destructively, that is the best way to go.
Yikes, I have Photoshop in my trusty antique CS3 which is almost 10 years old now, when Smart Filters were introduced to use with Smart Objects from the previous version.
Sorry for the poor example. I hope some folks can still benefit from my postings in spite of the Photoshop errors.
And yes, I'm guilty as charged. I admit I'm a Photoshop moron. My brain got rewired to nodes maybe 15 or so years ago, and I never looked back. Personally (and this is only my personal opinion, though many others feel the same) I find the layer-based concept to be antiquated, inefficient, and inflexible. And I'm sure some of that is just personal bias because I haven't kept up with the times with layer-based apps. And a lot depends on what you're working on. But anyway...
I think when people get the hang of nodes, and realize you can change the operations on your images merely by dragging and dropping a line to a different location, it opens a whole new world.
And when you start using a node concept to generated materials/shaders for your surfaces, rather than the convoluted (at least in my opinion) system that Carrara uses, it might make a lot more sense and be a lot more intuitive.
And by the way, I'm curious...are there any layer-based/Photoshop gurus out there who know of some ways in which the layer-based approach is superior to nodes? Maybe I need to re-think things...
Thanks.
I think, like you said, it is more about what you are doing with it. I think the layer style works better if you are drawing/painting- I can apply a layer style to a layer and start painting and see live what it is looking like, or paint light and shadows on screen/multiply/overlay layers and see right away what it is doing. I don't know if nodes allow you to do that? If I am painting on a figure in Blender for example, can I see live, as I paint, the effects of the compositing nodes? I haven't gotten that far in learning Blender yet but I am under the impression that it applies after the fact.
The thing I like better about Carrara's material system is that all my available options are right there where I can see them and I can quickly flip through them by just changing a setting in a drop-down and see more or less what they are doing, while in Blender I have to hit Shift-A, search for the node I want (assuming I know what it is called), and figure out how it is supposed to connect to/plug into other nodes before I can start playing with it's settings and see what it is doing. That is almost definitely just my lack of familiarity with all the different nodes, but a system where I don't have to know which ones need to be connected to which inputs and which outputs and which factors because the program does it for me seems more intuitive. It hasn't stopped me from starting to learning Blender's system (and I know you are not just talking about Blender, but about node-based systems more generally, Blender is just the example I am currently struggling with).
I'd love to see a basic primer on using material nodes that builds from simple to more complex materials so I could learn how things are supposed to go together - even something as simple as adding a bump or displacement to a material seems incredibly complicated in a node-based system.
I have been reading this as I did try out Blender in 2012, but struggled with the best YouTube tutorial that I'd found IMHO by a high school teacher. Then I found a glitch and could not go any further. Maybe it's time for a re-visit. It's a good topic, Joe.
Found this when I was trying to search for material that evaluated node vs layer based software to educate myself. It's from 2010, and it is centered on video compositing. It's long, so you might want to skip to the conclusion on page 22 of the doc, or page 25 of the pdf. It's a computer science thesis by a Swedish student. It's an analysis and not a "how to".
http://www.diva-portal.org/smash/get/diva2:327273/fulltext02
For me PS is an old friend... 2D graphic design is my job, 3D a new hobby with a HUGE learning curve that I hope to turn into illustrations for publications and museum displays.
Mark, I am going to try to find a primer myself as I struggle with the node tree order in Carrara, when or when not to apply an effect, and at which level. So I end up going back and doing it in PS then re-importing as a texture map. Time to move on and learn proper node procedure.
I think I first tried Blender around then too and was scared and confused. :) Sometime since then they seem to have done a major overhaul, as just about everything can be found in a menu now without too much trouble, instead of hidden behind just a mystery keyboard shortcut,and I am finding it much friendlier now. I suspect much of the program's reputation for being impenetrable to newcomers may be due to the older interface. I'm to the point now, after just a month or two of messing about with it for a few hours a week, that I am thinking about creating my entry for the upcoming Halloween contest over at rendo entirely in Blender just to see if I can.
Mark, that sounds cool and look forward to seeing this. OK, I will be taking a new plunge into the new and improved Blender during autumn. Oddly, I still am subscribed to Andrew Price's mailing list, and have saved them, maybe I should actually read them!
I think I mentioned it earlier in the thread but I didn't have the link at the time - I really recommend BlenderForNoobs "Learn Blender in an Hour!" video on YouTube
It brings you up to speed on the current interface in, believe it or not, about an hour (you really need to follow along to get the effect) and it's very painless. He has a series of beginner modeling videos too,each of which is about 20-30 minutes that introduce you gradually to a lot of the modeling tools. He's got some more advanced modeling videos and some texturing ones too, but I haven't gotten to those yet.
Nothing I love more than spending a lot of time on tutorials, and the response is "hey, I found this cool tutorial on the web !!".
I get that a lot with the tutorial threads I start. Very litte of "hey, can you answer these questions for me or help me do this", just folks talking among themselves about other tutorials they've found. I dunno, maybe my stuff is so bad that nobody can understand it. And add to that the jabs and personal stuff that always gets thrown my way, and it reinforces the fact that I must be insane to keep doing this. Which I am.
Ha. I love your intro to the topic. You are either a professional diplomat or quite possibly Canadian.
For your effort and explanation, I immediately respect your opinions and advice. Tell me more.
I need to bookmark this thread, Joe, it's excellent. But I am not a Carrara modeller. I am very new to creating models from scratch as presently I use a lot of content/characters and work with customising those. However, there are nature and scene objects that may work up better in Blender than in Carrara for me as the tools seem more precise from what I remember and what you and others are saying. So I need to get back to the very beginning with Blender. Hence wanting a really noob tutorial. If I can spend some time with it over the autumn and winter, then I can apply some of the extremely useful techniques you have demonstrated. And I like that they can be integrated into Carrara. First, baby steps for me.
Cheers for the topic!
Silene
I don’t know if Photoshop or Blender is better for compositing images. Here just to explain or maybe to help out: you can manage image composing in photoshop more or less like working with nodes.
Take example A. Every picture-layer is a smart object. Big advantage of smart objects: you can size and resize layers as you like, no quality loss here. As for the effekt-layers those are only effecting the layer they are assigned to. Every effect can be revised again and again. Even blurring is no one way here. You can stack adjustment-layers. So its as effective and simple as nodes for everyday compositing tasks.
Example B is for those who don’t know how to manage smart objects or have older photoshop versions. Connect the adjustment-layers to the picture-layer you want to adjust by alt/left clicking in the layer panel between the picture-layer and the adjustment-layer. In this way the adjustment-layer is only effecting the assigned picture-layer, the picture-layer is masking it off. The tree (I made it black and white so the effect stands out) is still green, it is only effected by the adjustment-layer assinged to it. And the background stays clear of this manipulation. Here too every effect can be stacked and revised again and again. You can come close to the example with smart objects, but there are limitations. E.g. sizing and resizing is no good idea. Blurring is one way – but just duplicate the original effect group and everything is fine.
Coming from Carrara I’m still fighting a bit with nodes too, but it will definitely improve my understanding of effective workflow.
I am trying to get back into Carrara at version 8, and noticed this in a google search. It is preeminantely easy to manipulate objects in blender. The change from right clicking to left is set in File > User Prefs > input tab, on left of screen as well as the preset selectability.
Then when you left click on an object, it may be moved within world, local or normal x,y, or z by typing "g" for Go and typing appropriate arrow keys. That is WAY easier than the neccessity of clicking in margin on directional icons. Also that object may be Rotated by typing "r" then axis "x", "y", or "z", in addition to simultaneously displaying the numeric values on the right of the 3D viewer by typing "n" when arrow is inside of that window.
Comparatively in Carrara would you be able to explain here or to me how to do the same ( with keystrokes ) to objects in Carrara ? I would like to move the camera around in the assembly room with as great an ease.
Also is it possible in Carrara, to convert a procedural material texture into a simple texture and save it externally as a .png file ?