So... these comics?
What is taking so long? Every time I made a comic in the past, I learned something about production and saw where my workflow was slowing me down. As a result the transition to a new comic took long for me to re-adjust. I need to set my environment up so that I can keep producing. A while back I bought two new machines to exclusively take up the heavy task of rendering all the images of my upcoming comics, and so that I can continue on other computers while they were busy. But before I could even think of rendering my new batch of comics, I needed a better figure to work with. Although Victoria 4 served me well and gave me a lot of pleasure, it was time to move on to a better, more detailed figure, tailored to my requirements. I spent the last part of 2014 modelling and rigging Project E, and up until today she still undergoes constant refinements on a weekly basis. I can now safely say that Project E is ready for my comic. (But not yet for the market for this, like promised, I will try to make a simplified version for the store, right after I do some improvements to the mesh... and figure out a way for you to transfer textures!!)
But by far one of the biggest most time-consuming parts of making a CGI comic, outside of making a base figure and rendering hundreds of images, is the posing and animating of the characters in the scenes. When you're talking about a typical Erogenesis comic, you're looking at a different complex pose per image per character (and I always have several in one scene!), and in my upcoming comics we're easily looking at well over 300 images per comic (and 500+ for the next black desk!). That's thousands of unique poses! (I don't use stock poses, they suck) Although in the past I enjoyed working on the figures in such detail, it slows production down immensely!
Enter the Kinect! Poser's latest software solution is called 'Game Dev', which is essentially a pimped-up version of Poser Pro 2014, and it has a feature that allows you to apply motion capture to your figures. Now I've heard that Poser's MoCap feature isn't the best and most accurate in the world, but I am pretty sure it will at least do 50% of the work for me from what I've seen online. So that's one thing I am excited about!
True 3D!
I'm quite sure that most of you have heard about some of the developments taking place in the realm of Virtual reality these days? In the US they've been developing a cheaper way of experiencing Virtual 3D Reality for the consumer market, called the Oculus Rift, and although the device is not yet market-ready, the developer kits have sold like crazy and are sending shockwaves throughout the online community. Youtube is filled with crazy tech-fanatics showing us the possibilities of this Oculus Rift and its got me very interested in joining the hype. One of the first things I'll be trying out is making true 3D images of Lali, and if time allows, I'd love to make an entire comic in 3D!! BTW there are other devices getting developed like the Samsung GEAR and I believe Microsoft also has plans for Augmented Reality.
Can you imagine Lali in 3D?? or Sen for that matter!!!! :D
Genesis 2 attempts
FYI, ever since the release of Lali's Bits for V4, I've been asked numerous times if I'm going make a version for the new Genesis characters by DAZ. Although I was reluctant to do this at first (mainly because of the amount of work involved) I've been thinking about it more these days. Now don't get your hopes up, trust me when I say this: I have tried.
Although DAZ Studio is not my preferred choice of Software, and I do have strong opinions about its design and usability, I am not by default anti-DS. Its still very good figure development software and on top of that it is free! I do keep track of their new releases and I do install it from time-to-time to see what the new features are. I have strongly considered developing Project E in DS like Hivewire did with Dawn, but its DS's learning curve that daunts me, especially considering all the other work I need to do. But who knows, maybe one day I might take up the challenge. In any case, I would be stupid not to keep track of DS's developments considering what I do.
I bought Victoria 6 and Victoria 6 High Definition a while back with the intention of perhaps trying to convert the bits to them. After playing around with them several times, one thing is for sure: whatever development I would wanna do on V6, is not happening in Poser! If I would develop them in Poser, the developments wouldn't successfully translate back to DS at all. And just for the record, the DSON bridge is not an option. Development with DSON lodged inbetween is impossible. What I did was export V6 as a CR2, that seems to let V6 work in Poser without the DSON bridge. Also, V6HD is actually not HD at all, since the base mesh is no different from the regular Genesis 2 models. The HD part is apparently just a DS feature.
So yeah not developing in Poser makes any hope for the bits for Genesis a bit slimmer. And no I cannot just migrate to DS; DS is good for figure development but Poser is way better for comic-making. Like I've said many times, Lali's Bits is a very complex system that can only be developed as I use it. Its like with Project E or what the guys of the Oculus Rift do, we develop these things in an applied environment, fine-tuning them according to what is required of them. Project E is another such an intimately complex system that responds narrowly to the movement of the limbs. You cannot just develop that in a in a separate environment than you actually use it. Within Poser, I am constantly refining her when suddenly I see her doing a pose that causes a problem around some bodypart. I make a morph to correct it, or change the Weight-Mapping or the ERC programming. You don't wanna know how complex Project E is under the hood. I cannot do this for Genesis unless I start using Genesis as my base model, and therefore also DS... and that ain't happening considering how limited Genesis is in so many areas, and how not suitable DS is for rapid internal image production (as you know I only use Poser's inbuilt Renderer to speed up production).
One thing I will try is the GenX plugin whenever I have time. I still considering trying to make something for Genesis since I could use the money right about now. But I think I'd better take my chances with Project E rather than wasting my time with Genesis. We'll see though, we'll see. the point of this story is just to let you know that I have tried!
New Website
Jim, I still need to email you haha! Busy busy!
No problem, busy is a good thing. I will be here when you are ready.
ReplyDeleteHey there
ReplyDeletejust a quick note on the mo-cap topic
I am eagerly following this little piece of awesomeness:
https://www.kickstarter.com/projects/1663270989/project-perception-neuron
on the oher side, if you still want to use kinect for mocap, place a large, non reflecting plane of cloth behind you and wear something completely different in contrast to that
Deletealso place it a bit to your side, or use two kinects^^
I tried that with another project, but not with poser
hey Knight! sorry I didn't reply! I tried what you said and its definitely important to have that contrasting thing happening. Also, does two kinects work better, do you know that? And does Poser accept multiple kinects?
DeleteThanks for your tips!!!! I've got a blog coming up with my findings. Its not a great as I thought it was, but its a start I suppose.
Considering how much work is still ahead for you with PE, I'm glad to stick with the Lali Bits for as long as it's needed. I even pointed some friends to your product on the Rotica shop and encouraged them to buy the product (hinting that PE would be even greater a product, too :3). I have a question too: have you ever considered an adjustment morph for the pussy prop so it matches the Elite morphs? It seems many characters these days use the Utopian morph. Just an idea ^^
ReplyDeleteyeah I need to give people a breakdown of how to deal with Elite morphs and others. Basically the more an FBM fucks with the hip region, the tougher it is to use the bits. So you need to unfuck a copy of the Elite morphs in that area. This does make all the crotch areas with those FBMs identical crotches but at least it works better
DeleteI have been happy with use of lali on Aiko4 - the main issue is A4 (as perhaps do some of the Elite) changes the curve of the ... crotch base? So V4 is more flat from front to back, and Aiko4 is more curved, which tends to mean I need to fiddle with the prop's XYZ "Y" axis, pushing the front up and/or the mid down. A4's butt also sticks out more, so the ass morphs are tougher to use (what's there is buried deeper inside A4's mesh). For causal nudes shots, it works fine.
DeleteI just noticed about this and although it only works right now for Maya and Katana, who knows? There might be something to keep an eye on there, Ero.
ReplyDeletehttps://renderman.pixar.com/view/non-commercial-renderman
Looks like they are working on supporting Cinema 4D too. That's great news!
Deleteyeah I checked that out a while back and looks exciting, but I'd KILL for their animation suite called Marionette!
DeleteHey Ero, You said that you do all your renders in Firefly...
ReplyDeleteAs a long time poser user I have to ask. How are you able to get such fantastic lighting inside of firefly. I have never gotten one of my renders to look that good lighting wise.
-DP
yep firefly! answer: I wouldn't know how to explain it just like that, I'm just using it on a certain way and I need to explain exactly how. Its definitely not just Lighting its also Materials and render settings. I've been meaning to make a tutorial for a while now but just don't have the time to complete it. I'll try to hurry.
DeleteI personally think that many Poser/Daz users underestimate the power of Firefly and Poser's ability to mix materials etc.
DeleteBut now since "unbiased" render engines are all the rage...
Just a sad heads up Ero - installing PP2014 SR5.2 (10.0.5.28925) seems to break the original Lali for V4 - some weird cross-talk or linking between V4's right shoulder and ass-cheeks. Pose arm up, looks great; pose arm down and she's got a built-in flesh-'bustle'. I am using an old saved PZ3 as base, so I'll try building a new 'A4 Lali' from scratch. When I have some quiet time I'll to figure out what is 'moving' down there when the arm moves.
ReplyDeletehey thanks for the headsup. I haven't noticed any issues with SR5.2 yet, but then again I'm using PE not V4 anymore. I'll check it soon. I'll post it on the forums at rotica to see if others have this issue too.
DeleteMind you, there's much that can go wrong. For example other morph packages, accidentally transferring weight maps the wrong way, removing a bulge modifier. Also copying keyframes, Poser was never able to deal well with copying keyframes and therewith distinguishing between dependency controlled morphs, but that's most conformed figures that suffer from that.
Out of a matter of interest... When you installed the Poser SR did you choose to remove/backup all the config files? The reason I ask is because if you use your existing config it can throw up all sorts of bizarre errors.
DeleteAlso, I have an original Lali that I have updated over time but when I make a new character I make it from that, I have had issues with adding little morphs from other vendors which screw up Lali. If you have an original 'clean' version where you haven't added any extra morphs try that and see what happens.
Actually, good news - I sent you an email direct with some sample images, but the problem is with "Meipe & Xameva's V4 Perfect Complete" ... so not your problem. Use of Lali just MOVES some distortion from the right shoulder down to the crotch area. Anyway, removing (or not using) X&M perfect shoulder solves my problem (good, as I was considering deleting PP2014 and 'starting' over)
DeleteRegarding all these people who have been trying out the Kinect mo-cap option, it seems that most are getting excited without realizing one of the most important issues that I don't think has been addressed: It's the scaling of the data onto the biped mesh.
ReplyDeleteIn the past, using .bvh files for Poser was a crapshoot because, depending on their origin, you'd never know how they'd work on whatever figure you have loaded in Poser. Even if the mo-cap data or .bvh file was recorded off of a biped and then imported on a biped skeleton in Poser, the data was, most often, way way off. Especially in shoulder regions.
The reason is because the conversion scale that imports the raw data is either wrong, or missing skeleton parts. For example, I had a lot of .bvh of different types of walks and turns (walk then turn 90 degrees, walk then turn 45 degrees, etc.). They all did not interpret properly on Victoria 4's skeleton because the data was not interpreted for that skeleton properly. In fact, I think there was an extra bone 'actor' in the spine section (in other words, two separate body parts for what is represented by just the 'adbomen' actor on V4).
Now I could be talking out of my ass here, because I have not tried recording mo-cap using the Kinect and loading up that data via the Poser Game Dev suite (both of which are things that I own). But I think it is the reason no one has come forward since the release of Poser Game Dev with any compelling proof of concept work that the Kinect mo-cap is viable at all for animation. Someone prove me wrong, but I'm pretty sure not all the pieces are there yet.
Ero, if you want to take this offline and discuss further, you know where to find me. :-)
yo Ep. yes you're entirely right, I did get those problems. I wrote a whole report on it on Renderotica but haven't yet come round to posting it here. What people say about the mocap is true, its not great at all. It captures movement nicely, but the pose is always screwed up, even if you use Poser native figures. To be fair to Poser, they did mention it in their webinars that it was for simple stuff, but I'm afraid its not even good for that. Even in those Poser videos (and all other videos) the first thing you see the figure do is lock up their shoulders. you would assume they'd calibrated it properly didn't you? So this might be the scaling thing you're talking about. A simple calibration tool would've helped.
DeletePoser is not being very clever with their new features, they're either very unstable or useless. They're too much in a hurry to compete with DAZ I suspect.
Anyway, I saw your email, I'll reply asap!