The Wonderful Dynamic Puzzle of 3D (or, What's my Work Flow)

1235

Comments

  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    This example is the beautiful Armoury by Merlin. The first image is how it came out for me using the lights he provided. It produces a nice effect, but I wanted to reduce the candle and firelight for one, and when I did that it had a ripple effect so I ended up redoing the lights entirely and tweaking the surfaces to get the look I was going for. Now note, there is nothing wrong with the first image, it's just that I had a different image in mind. Also, I might have had some settings wrong as this is just the quick 'click scene, click camera, click light set' type image. But, after testing out a few of the lights and camera angles, this is the one I wanted to do and it was enough to let me know I wanted a different lighting effect.

    Render time this image: 18m6s

    01.jpg
    800 x 800 - 132K
    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    So here is the final scene. The fireplace lighting could be adjusted down and the overall lighting could be sculpted a bit more, but I decided as I was doing this image that I wanted to later come back to this and add a character so I wanted to leave the overall ambient the way we see it now for future use.

    What I did with the scene. Well to start, I used Merlin's lights where they were visible in the scene, but replaced them with Linear Point Lights so I would have more control and then reduced the lighting on them so I could get the starting pool of lights I wanted. I then deleted any lights not in the scene opting to replace them with an UberEnvironment to create the ambient lighting that would have come in from them. The fire is made up of three planes which I converted to UberArea lights and the point light I converted again to Linear Point. The UA lights use the diffuse maps for the color of the area lights and make up the light that extends out from the fire. The LP light is adjusted to create the lighting in the firebox. With the UE light, I originally made it a Dark Moonlight but that proved too challenging in this scenario for me to get the light sculpted the way I want and ended up opting for the Orangish preset. Once I had the lights basically carving out the overall ambient and light pools, there were a couple areas that weren't what I wanted from the image. The back corner was almost black, so I added an LP light at about 10% there to bring some ambient into that corner. There was no reflection onto the cealing from the upper wall candle and after much tweaking I finally ended up putting a second lp light up by the cealing just for that purpose. If one notices a subtle lighting on the cealing over the candle, that's it... a bit of work for a very subtle effect but that's what it takes in general to get the lighting I wanted. Finally, I added a tightly controlled LP light right in front of the leather armor to the left as it was totally in shadow and I wanted to create a specific level of lighting there. That light was set to 5%. The candles in the original were set to around 1750%, I reduced them to about 100% +/- 20%. The fire was 3 planes set to 50% with color maps and an LP light that stayed inside the fire box. The UberEnvironment was set to 50%. I almost forgot, the candles etc were all converted in this case to raytrace from shadow map with the associated resulting increase in render time.

    As for the surfaces, they are basically as Merlin set up. He does great surface work. I did tweak a couple settings to get them to react to the lights the way I wanted, mostly I went for a quick cheat and set many of the surfaces to skin, the ceiling and floor to metallic. I use these settings on the surfaces not for what they say they are for but for how I know they will effect the light reflecting off of them.

    The result after spending a good amount of time on it was good enough that I went for a larger render. The final render time didn't make sense to me as it said "1 day -16.09" whatever that calculates out to.

    For better or worse, this is the final image for now, I hope to revisit it to include a character but that will be more complicated as I don't have the great starting point for that I did with the set. I love a lot of stuff Merlin does and believe it is a good source of learning for anyone who takes the time to play with it, dissecting what he's done with things like shaders. I give his stuff a 5 thumbs up ;)

    The_Armoury,_Scene_01.jpg
    1200 x 750 - 102K
    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    Summary: the back left corner I wanted to be in some shadow, but I would probably bump that up a little. The same goes for the light on the leather armour to the left. I was working at night, and during daylight it comes out much darker. This brings up a good point when working with night scenes as this has happened to me before. I think it's a good idea to work on them under day and night lighting to get a good centered lighting to the scene. This might be a good idea for any scene, but it seems especially important to night scenes since the lighting is so subtle.

    As for the lighting on the floor. I didn't originally set out for it to be as well lit as it was and in my original concept it looked like something of a mix between what the original lighting Merlin provided and this lighting. However, night photography is a bit of a hobby of mine and so I know this is within the parameters of how actual night lighting might be in this scene based on other lighting not visible, the particular fire effect (which varies quite a bit) etc... If it was a standalone piece I would consider tweaking this either in the lighting or post. Obviously post would be faster and doing it in lighting would only be for the exercise of it at this point, or.. if I wanted to reuse this scene in the future and figured I'd want that lighting different going forward. In fact I do plan on reusing this scene by adding a character and this lighting actually plays well into it so thus I left it as you see it.

    Enjoy :)

  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    I decided to put up a version with some quick post for the final image of this particular example.

    The_Armoury,_Scene_01p.jpg
    1920 x 1200 - 328K
    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    Well, this isn't an example....

    This is a straight-out-of-the-box render of Merlin's Secret Cloister. All I did was pop in the scene, the lights, positioned the camera and hit render. (I love his stuff.)

    The render time was 17m14s with mapped lights. Not a bad time at all for this size image of this quality. This is also a great prop for 'kit bashing' where we take parts of it (such as the fence and lights) and incorporate them in another scene.

    Yesterday I was playing with Nightshift3D's Fusion Reactor Room. I tend to like 'clean scifi' type textures etc... so I wasn't sure when I got it. Then when I looked at it, there was just the scene file, not lots of individual props.

    Well, I was like a kid in a candy shop when I started playing with it. The Surfaces are so well set up and there is so much detail to the prop I couldn't settle on a single look. I just kept playing, changing surfaces, changing the lighting, changing camera angles and finding so many possibilities in this single prop. I generally like lots of individual props besides the scene file but this prop is amazing and totally changed my mind in a good way and opened me up to how a single well made prop can be really versatile. The other thing that was amazing was that the lights were very usable for being poser lights, which just goes to show, don't discount anything out of hand.

    I find as I play with different props, it helps open my mind up to different ways of working, different frames of reference. For me, working with a prop is a chance to see inside the mind of the artist to just a little bit and helps me expand my tool kit. Props from a given artist will change over time both as they change and the environment we work in changes. Like a cracker jack box, one never really knows what we'll find with any given prop.

    Cloister.jpg
    1920 x 1200 - 119K
    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    I wanted to put some thoughts up on learning that I have picked up from a career as a trainer. We learn in stages. Also, we also learn differently, so any training material we come across may work better or not as good as something that is geared specifically to our strengths in learning. These concepts might not seem revolutionary but the implications can help us work with the learning process rather then fighting it.

    Ok, I will tackle the second point first as it is a quicker topic. We've all heard the 'I learn better from reading' or 'I learn better from videos (show me)' debates. There is also the 'I learn better from doing' etc... The idea is that our brains seem wired to take in data in one primary format and so we naturally gravitate towards that format. While that is the format that we will get the most out of with our time spent and so makes sense to focus on, the other methods are critical in our learning process. It is too easy to focus too much on the one that gives us the most reward the quickest but that is somewhat akin to trying to survive on junk food. Each method of learning has specific things that we learn from them that we don't learn other ways. At each level of our development, we need to actively incorporate the other methods to advance to the next stage. This is the number one reason people get stuck at a certain level of understanding/learning from my experience. They focus on their primary method of learning and don't attend to the other methods. So again, at each stage of learning, we must actively incorporate some of the other methods of learning that are not our primary. If we prefer to sit back and watch videos or read, we need to take the time to do exercises (that one seems obvious.) But also, if we tend to watch videos, we want to take the time to do some reading up on what we are trying to learn. If we tend to read, we should take time to watch some videos, and finally, if we tend to do exercises, we really want to take time to do both videos and reading . It will pay off in the end.

    The first point, we learn in stages. I like to watch videos, but this pertains to either of the other primary methods of learning as well. If we go through material we think is at our level, it makes sense.. but actually, it is a level beyond our current level. That is, we can understand pretty well a level beyond our current level and often think that's the level we are at... until we try to do it on our own. So, what if our primary method is doing? Well it's the same actually, just that it expresses itself in our inability to do related level exercises on our own, that is, we have not fully developed the transferable skills to function the width of the level, our ability is restricted to a subset of that level, thus we still get a misinterpretation of our true level because we can do some exercises at a level above our true level.

    If we go two levels above our current level, we can understand it for the most part but we start missing details and we can only remember the steps involved right after watching/reading/doing something. In a very short time after, our memory starts to fade. This is the point where most people become frustrated, as we think I know this, I just can't seem to remember it. The truth is, we don't know it we simply understand when it's explained to us, without even being at a level of storing any of it.

    If we get three levels beyond our current level, we get the general gist of what is being explained, or at least think we do, but know we don't understand it fully. This level we can form opinions that are not accurate if we aren't careful because we might have gotten a misinterpretation of the information being presented. But it also can work in our favor in giving us a longer term general guidance in our bearings. We want to make use of this level of learning usually, but with a note of caution.

    Each level we learn, shifts the previous levels down one and opens up a new level.

    We must attend to each type of learning in my experience to move up to the next level.

    We are typically resistant to using methods that are not our primary and thus tend to retard our own growth.

    There is another aspect to this that I will explain with an example now. It was my own learning through DAZ's very good videos. Before I even downloaded DAZ, I watched some of their starter videos. This is part of what made up my mind that this was a path I wanted to invest in. They showed me a lot, but when I actually went to apply the information in the videos, I kept having problems with not being able to apply them step by step. As I was totally new to the material, I would miss things like selecting items in the scene tab in order to work with it. That and a million other little things kept tripping me up. So, I watched, I practiced, I read... and over time, I advanced. Note, I watched the videos mutliple times, but I only really absorbed them at the level I was on... not even beginner, but newbie. After I finally got my feet under me with that, I drifted away from the beginner videos and started on more intermediate ones. This reset my learning for that level and I found myself in the same predicament of having to do exercises, reading, over and over. After a bit of this I realized I was missing some things from the previous level so I went back to the beginner level videos and was now able to follow along at an entirely different pace and understanding. I decided to go back through the exercises presented and this time it wasn't a chore. Doing so pointed out various areas I was missing on and filled in some blanks. Going back to the intermediate level videos made more sense. But, although I was watching and practicing intermediate level videos, I was only at beginner stage, having graduated from newbie, since that was the level I could work at without struggling. See, it's easy to fool ourselves we are a level beyond what we are at when we can do some exercises at that more advanced level.

    While I was doing this, I was going through material on things like shader builder, scripting, etc and got the gist of what they were talking about, but this was actually two levels beyond me.

    I have now graduated to what I consider intermediate level, and am still going back to the beginner videos just now. Much of what they are showing, I could show people myself. However, I am still getting tidbits from watching them that are increasing not only my overall understanding but making me much more efficient with little things that make a difference. This appears to be a good way of telling us what level we are at, If we can teach something confidently to others and not miss steps, (even though we might miss some shortcuts in doing something) we are generally a level beyond that in our own status. Of course this doesn't take into account that there is a breadth of material in any one area and we may level equally across the breadth or we may focus on a particular area and level that area beyond others. In some ways, it's like leveling a character in a game, just not so well laid out and defined.

    Well, I hope this made sense, and more importantly helped some. If it all seemed obvious to some of us out there, just remember it's not to everyone ;)

    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    This is an addendum to the previous post. It has to do with learning styles and what we typically get from them.

    With videos, we get a quick, down and dirty, here's how to do something. They are great at giving us a good quick understanding of something in a short amount of time, but the will miss steps more then the other two, so are the most likely to fool us into thinking we know more than we do. Often people who rely on videos can talk a good game but can't apply it very well.

    In hands-on, we generally get the skills necessary to complete a task, but it tends to be very focused on that task, so we think, I can do this... and extrapolate our experience to other related tasks that when we actually try to do them we find have pitfalls we didn't anticipate and our experience doesn't in the end translate as well as we thought it would. More importantly. we tend to progress as if with blinders on, down a path that in the end isn't the most efficient for our goals. Once committed to that path, and having invested heavily in it, we are resistant to changing to a better path even if it presents itself, as we by nature tend to want to protect what we have already invested in.

    Reading tends to be in the middle, and so we might think the best, but this isn't actually true. We read, we do some exercises... but we are often resistant to doing as many exercises as we should falling prey to some of the pitfalls of the people who rely on exercises, and, we tend to have less of an overall perspective that people who watch primarily videos would have. So, in general we have some of the strengths and some of the weaknesses of each of the other two. It may seem like a good compromise, but really we do want to push ourselves to do more in each of the other areas.

    If we find ourselves stuck, or not progressing as fast as we think we should be.. it is almost guaranteed we need to focus on our non primary methods of learning.

    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    One last important note on this topic...

    This information is to help us progress. We want to be careful in applying these concepts to others as in, 'oh s/he learns by xxx so that means yyy' as that is guaranteed in causing us problems in relationship with others. Some information is meant mostly for application to ourselves. This falls well into that category. If we do apply it to others, we need to make sure it is only in a method to help us better present information to them and not to pass judgement.

    Peace out :)

  • SickleYieldSickleYield Posts: 7,622
    edited December 1969

    For me at least, when I decide I want to make something I've never done (let's call it a widget), I do this:

    -Sketch crude picture of widget. Decide what features it should have.
    -Find pictures online that look like the real version of widget, if it exists or existed.
    -Base mesh widget based on sketch, pictures.
    -Google tutorials for how to rig/setup widget. Discover one or more features are impossible with the tech; discard those.
    -Watch or read every one that seems relevant. Skim the really long ones.
    -Experiment with base mesh widget. Discover important obstacle mentioned in none of the tutorials. Reread them more carefully. It's still not there.
    -Experiment further.
    -Forum post query that may or may not be responded to by Richard Haseltine (depending whether he knows the answer).
    -If Richard doesn't know, experiment. Find actual solution buried in features of program. Edit widget base mesh to fit with necessary changes.
    -Build working widget using workflow cobbled together from all sources plus experimental results, resulting in process identical to none of those seen.
    -Edit forum post to reflect new information.
    -Sell completed widget product. Forum post vanishes into obscurity.

  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    ..Forum post vanishes into obscurity.

    Good example :)

    The last part unfortunately is too true. Our current form of forum isn't the most efficient for information that isn't current. Surprisingly one of the best forums I was on was from (I won't say how many years ago ;p ) as it had a very well setup hierarchical structure and search engine that allowed someone to find relevant discussions from years before as well as current ones. The whole breadth of the forums many year history was well accessible and topics would go on for years because of it, simply building on past information.

    [Edit] One thing I forgot to mention about the forum mentioned above is that it had a tree thread view so people could see question/responses in a collapsed format where just the subject line and author showed. People tended to use good subject lines because of this and it was easy to scan this tree quickly for information, going over what would be 100 pages or more of posts in only a couple minutes finding exactly what one wanted. This was really key to it's efficiency, along with a very good search engine.

    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    Never underestimate the importance of playtime when learning. It rejuvenates us, frees us from worrying if this or that is correct. We should purposely not put too much effort into it sometimes... and occasionally we can surprise ourselves with something interesting even if it isn't a work of art.

    If we aren't smiling part of the time we are doing this, ... well what's the point ;)

    Playtime.jpg
    1280 x 720 - 53K
  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    This example uses Merlin's Harp. Both images are using the same lighting, UE white 80%, distant white 36%. The first image is out-of-the-box with this lighting and it is very nice. For my purposes, I wanted to brighten it up a bit, and more importantly, wanted to adjust the (base board?) the strings come from.

    01.jpg
    800 x 800 - 41K
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    Well the problem was, the said (center strip was all I turned up in quick google search) is part of the base surface and so we are limited in what we can do with it. Poly Edit tool to the rescue... this is a very quick and dirty example so we could have done much better but it works for our purposes. Changing the viewport to wire frame and using ctl+click-drag for selection and alt+click-drag for removing from selection creates our selected surface.. some zooming in and out required. Right clicking on the selected polys and 'create surface from selected...' and we are off and running. Name our surface and it is immediately available in the surfaces tab for us to tweak how we want.

    02.jpg
    800 x 800 - 39K
    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    Summary: the bridge and tuning pins might have been better in their original format, they have reflection on in the modified so it would depend on the environment (and I didn't spend a lot of time on them.) Point is, the originals were very nice on everything in this product. Often it's not a matter of trying to make something 'better' but rather tweaking it to work how we want for an image.

    Render time: 29s

    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited May 2013

    This example was originally posted in the May New Users Contest WIP as an example in lighting. Getting good light in a night seen is challenging. There is a lot of balancing of the various lights and much of it has to be tested at some point in full or near full res. The spot renderer is helpful for, well.. spots, but it can fool us with how that spot fits in the overall balance. Spot renders can look to dark/light or just color balanced wrong when not in context of the full image. The net result is, I find a good night scene takes time to coordinate and balance all of the lights brightness, colors etc, so don't get discouraged... the results can be worth it.


    There are 12 area lights, 7 LP lights, a Distant 'moon' and UE.
    The area lights include ambient, opacity, color maps, translucency, custom falloffs (important so the lights don't bleed across the whole image...)
    All lights have raytraced shadows except UE which is AO+directional. (Note, AO+Soft Shadows is much more common for me.)

    Various surfaces were tweaked to respond to the lighting in the desired manner. The water in particular was totally redone.

    Stars and clouds were added in post.

    Setup time was around 20+ hrs due to much of it requiring fine tuning at close to full resolution.

    Render time was 4.5 hours

    [Edit] Image included here (thanks to Jaderail.) Slightly higher res then the dA version. This image, like the Armoury image were meant to be displayed large, they loose much of the effect when sized down.

    [Second Edit] There is a lot I'm not explaining about this image, such as lines in the image, color choice and balance, camera angle & focal length, etc... Suffice it to say that a lot of thought went into all of the various aspects that made up the image. In the end, whether it worked or not depends on if the image stands on it's own without all of that explanation. :)

    There is one mystery I will leave everyone with. There are three characters in the scene. Can you identify them all?

    Hot_Summer_Night.jpg
    1920 x 1280 - 259K
    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    All of these are just exercises so far, and intermediate ones at that. I still have to work my way through intermediate, advanced and into expert in a few tools to reach the freedom to express things at what might be considered 'Art' for me. It is hard to truly express our ideas when we are struggling with the tools we work with. This of course varies from person to person. A truly inspired 5 yr old could create art with crayons and chalk just because they have it in them. For some of the rest of us, we can share tips on our journey and perhaps help each other out ;)

  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    There are some great ideas on creativity, lighting, workflow, etc.. at 3D Total's Free Tutorial section. They are using Maya, 3DS, zBrush etc.. for the most part, but the ideas presented are abstractable to other environments.

  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    A good coverage of HDRI and where it fails here at CGarchitect

  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969

    This video refers to a workflow with mudbox but could apply to any tools and is a good example of a pipeline/workflow using current tools and sculpting techniques.

  • Joe CotterJoe Cotter Posts: 3,258
    edited December 1969
  • Joe CotterJoe Cotter Posts: 3,258
    edited January 2014

    There was a link to cutting holes in Blender here but as I was watching it, some of the methods I had issues with so I removed the link.

    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258
    edited February 2014

    Here's a good article on principles of animation and motion. It is based in user interface design but the basic principles apply in a general manner, and the ui take may actually help as a visual example for some.

    Post edited by Joe Cotter on
  • IceDragonArtIceDragonArt Posts: 12,548

    Lots of good information in this glad someone referenced it in another thread

     

  • RuphussRuphuss Posts: 2,631

    very nice thread

    and a good approach to learning things is to teach them to others

  • Joe CotterJoe Cotter Posts: 3,258

    Ok, so a bit of a hiatus on this thread but I thought this worth posting. Here is an image of part of a product that is quite old. It shows how the image would look ootb (out of the box) in 3Dlight (Renderman.) Of course as previous posts showed, we would rework this, light it, etc... but the point is, this is what we were starting with and still are if working with 3Dlight.

    01---Electro-Apparatus-OOTB-3Dlight-Renderman.jpg
    800 x 800 - 85K
  • Joe CotterJoe Cotter Posts: 3,258

    Then came IRay and ootb with materials autoconverted (just by changing to IRay as our renderer) we get this:

    01---Electro-Apparatus-OOTB-IRay-Autoconvert.jpg
    1024 x 1024 - 122K
  • Joe CotterJoe Cotter Posts: 3,258

    With a little tweaking of the materials and filtering we get this:

    01---Electro-Apparatus-IRay-Retextured.jpg
    1024 x 1024 - 130K
  • Joe CotterJoe Cotter Posts: 3,258
    edited November 2016

    Now understand, I'm not saying we couldn't get much better results with 3Dlight (then the original post) with some tweaking. What I'm pointing out is that the starting point is different. Yes there are costs in both render time and a learning curve, but that's part of what we do as artists is take some base and work it into something new. Should everyone convert to IRay? Not at all, Renderman is a very capable render engine but it takes work to get the most out of it and many of the tools we had to do that ended up taking just as much in resources and learning curve as IRay so you pay your nickel and get what you get as they say. This isn't meant to convince anyone of anything, it's just meant as something to consider for those who don't already have this information for comparison. Peace. :)

    Oh, a side note... I wouldn't stop with this piece right now as it is if were trying to use it in a piece where it would be in the foreground. I would at least add dirt maps and probably rework the liquid to have a meniscus (that curve at the edge where the liquid meets the glass) some foam probably and perhaps some volumetrics.

    * That was just the default HDRi also (along with luminance from the liquid) for anyone wondering why there is no ambient occlusion. A real scene would have actual lighting.

    Post edited by Joe Cotter on
  • Joe CotterJoe Cotter Posts: 3,258

    I almost forgot the other half of this post series. That is, there is a lot of life left in old resources if we take the time to draw out what hides in them. :)

  • Joe CotterJoe Cotter Posts: 3,258
    edited November 2016

    Hey, a shout out to Bluebird3d on her new YouTube channel. It's a great channel and it addresses some issues for some of us who are deep into this, the fact that we need to find community amongst each other since being so involved in something means exclusion of other things. Most of the time the people physically around us don't share the same level of interest or commitment we do but we can't change who or what we are if we want to grow to be who we want to be. So in that vein I fully support her in this and recommend others who find themselves likewise deep in it to check out her channel, interact, see if we can't build a community and support (not lean on but help when and how we can and just share with) each other.

    She has a thread here related to this, check it out.

    Peace.

    Post edited by Joe Cotter on
Sign In or Register to comment.