Asking for tutorial help

PoobyPooby Posts: 122

Hi friendly folk. I wonder if you could take a look at this. I'm trying to make a simple cage deformer.


password: canvas

I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

Comments

  • AhmidouAhmidou Posts: 179 ✭✭

    Hey @Pooby, can you be more specific about the part you're missing?

  • PoobyPooby Posts: 122
    edited May 2016

    I'd love to be, but I've had so many things go wrong at every stage that I've got to the point where I just want to watch someone do it and copy it or give up entirely.
    Canvas just gives me the feeling that I'm fighting it all the time. It reminds me of Mental Ray, compared with ICE's Arnold. It seems very pedantic about the slightest thing and unless you get it all right, you get nothing.
    I cannot find explanation for things in the nodes like 'max unscaled distance' in cLosest location node so I'm just stabbing around in the dark always wondering if I've not set some parameter that I don't understand, correctly.
    Its made me very grumpy today as you can probably tell.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • AhmidouAhmidou Posts: 179 ✭✭

    I'm afraid it's going to be hard and slow without coding a bit.
    When searching for a node, you can have many returns, and if you don't know a bit about the different data structure offered by Fabric, you can be pretty lost.
    For instance you shouldn't use the GetClosest node but the ClosestArray one and feed it with the GetValueArray output (make sure to set the attribute name).
    Also, the reference frame is not provided, so you have to build your own and that's where you might need to use KL. You can still use the new Execute.For feature from 2.2, but it won't be
    multithreaded.

  • PoobyPooby Posts: 122
    edited May 2016

    I guess if things are hard and slow without coding, it sort of undermines the point of nodal programming for most people. I dont want to pass myself off as some kind of Guru but as someone who knows how to make relatively complex deformers, If replicating my simpler stuff in Fabric is this hard for me, most newcomers are not going to have a hope in hell of working out where to start.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • malbrechtmalbrecht Fabric for Houdini Posts: 752 ✭✭✭

    Hmm ... @Pooby, without joining the (religious?) quarrel on "nodal programming" versus "code programming" ... wouldn't it be a better idea to formulate some "requirements" like:

    • for this and that task I need a node/preset that does this and that
    • how would I reproduce this specific node from ICE (or wherever) in Fabric?
      ... and then go from there?

    Nobody has claimed, as far as I know, that Fabric "replaces" ICE 1:1. So it is not exactly "fair" to expect 1:1 the same solutions to work in Fabric that worked in ICE. But with this community, it should be possible to create a bunch of community-presets that can get people like you, experienced "nodal programmers", into some "comfort zone", which would then result in cool tutorials/presentations that could attract more users. Those users would then have new requirements (or bring in new ideas), so everyone would benefit.

    I hope I was able to bring this across diplomatically enough ... :)

    Marc


    Marc Albrecht - marc-albrecht.de - does things.

  • PoobyPooby Posts: 122

    Hi. I understand that its not a one to one with ICE.
    However, I know that Fabric can do this, so I am frustrated that its so hard. Concepts such as locations are similar, and I'm only doing the very basic part, no Reference frame etc and my tree looks totally logical in Canvas but it does weird stuff.
    I'm not struggling with the concepts of how to make the tool. I'm just finding that its not working, with no explanation of why and I cannot find the answers anywhere. Look I'll make a video and show you what I mean.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • Aaron KentAaron Kent Posts: 66 ✭✭
    edited May 2016

    Paul - all I can do is empathize - wish I could do more. I keep getting friendly requests for more tutorials on vimeo but Canvas is a hair pulling teeth nashing experience for the code challenged...

    AK

  • PoobyPooby Posts: 122
    edited May 2016

    Yes its a pain. I managed to get the first bit working, but not replicate the reference frame part yet. Although I have attempted that and I'm not sure why its not working.
    My tree ends up full of nodes where I'm just trying to recreate higher level nodes that are in ICE by default. Its probably using 4 times as many. No wonder people go on about this spaghetti thing. Its no fun working with visual programming on such an atomic level and defeats the purpose. You want to be able to bang out stuff fast, and have the ability to dive in to the atomic detail only where necessary.

    Now, over time I could probably replicate at lot of these handy ICE nodes for myself, such as reference frame or just things like randomise by range, etc etc but that doesn't help when I'm doing tutorials. People don't want to stop and download a bunch of extra stuff just to follow me. It needs to be there in Canvas as defaults.

    One thing that compounds the problem is this doubling up of nodes for arrays and single values. Its a shame that nodes can't by default, internally work out what they are based upon by the input, without the user having to explicitly choose them. ICE manages it, so It would need explaining to me why Canvas couldn't. If that was the case, It would simplify matters a lot.

    I think the Fabric team need to make a judgement whether they really want to appeal to Artist TD's in the same way that ICE did. If so, I think a push in this area is needed to make it more friendly to visual program, because the 'solution' that keeps coming back over and over is 'learn KL', and I see that a cop out that bypasses the visual programming altogether and certainly wont help it improve.

    The point is that with a Visual programming system designed for speed iteration like ICE, I can make a cage deformer in well under 10 minutes and be able to explain how it works so that anyone would understand.

    Even if I knew Canvas really well, I think I'd have a lot more explaining to do and it would take a lot longer than 10 minutes in its current state.
    I know that nobody is saying that Canvas is finished. I'm not having a go at anyone and I still think Fabric is awesome, but It would be good to know what the plan is about making Canvas more accessible or whether learning CODING in KL is always going to be expected of a newcomer.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • AhmidouAhmidou Posts: 179 ✭✭

    IMHO in it's current state Canvas is more intended for for mid level/advanced TD that know how to program, and I believe that if you don't start to learn basic coding at some point, there will always be cases where you'll be frustrated with Canvas and Fabric.
    Once you get the basics 10 nodes ca be converted to a few lines which is way easier to manage.
    Maybe it's time to give it a try ;)

    PS:Block which is the ability to build multi threaded loops with nodes are supposed to come later this year, maybe around Siggraph if we're lucky.

  • PoobyPooby Posts: 122
    edited May 2016

    Let me explain my standpoint because this isn't about me and this is easy to get misconstrued.
    I'm not throwing my toys out of the pram because I don't want to learn coding. I'm happy, for my own sake to learn KL. I've already written some stuff in KL, I even made a tutorial about starting in KL when I started in Splice a couple of years ago.

    here, if you're interested.

    But I'm interested in going out there and getting people excited about using Fabric Engine as I did with ICE. I get regular mail from people thanking me for teaching ICE for free and I hope to do the same in Canvas and I don't thing learning to code in KL as a start point, is going to interest people at all. I had a very muted reaction to the splice video I made. People think Fabric is for techies, and at the moment they are right. I want to help change that.

    I am very aware, from having made hundreds of tutorials, that there are certain things that will attract users and get them interested and other things that make most people run a mile.
    So I see my role here as to give feedback on what needs to happen to make Canvas appealing.

    Me learning KL is not useful in that regard..

    let me clarify. I'm NOT saying that I don't want to learn KL. I will and am. Its a great idea. Can't wait to know all about KL. Totally agree that its a must. KL Rules, etc.

    I must be explaining this really badly because I seem to have this conversation every time I come onto this forum.

    Its that the solution of ME learning KL to solve my own tasks is not useful in regards to helping the development of making the Visual programming part of Canvas better to use for people who I would like to get interested in it through making tutorials to entice them in.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • malbrechtmalbrecht Fabric for Houdini Posts: 752 ✭✭✭

    Hmm ...

    This is totally out of topic and uncalled for repetition, so I am hiding it away in a spoiler:

    If you want to get people to "cruise along the west coast on a Harley", you don't go out and build Harleys that can be driven by a 3-year-old without any practicing. Instead, you teach people how to drive. That includes traffic signs, balance and stopping at the gas station. It's not just the cruising, even if the cruising is what you want to do.

    It's not like people who "want it all, now, without effort" are right. It is not like "artists don't need to learn new tools". Art is not a craft that replaces everything else. Artists keep telling me that "everyone is an artist, everyone can do art". They are plain wrong, they just force their own perspective on everyone else, ignoring that people are different. All those people want, is to see my weep in tears, because I can not live up to their "facts". But I don't have to.
    If learning to accept (not even to write, just to accept) that sometimes 2 lines of code can replace an "artwork of nodes" is out of reach for those people, then, maybe, Fabric is not the best thing for them. Would that be the doom of mankind?

    I said it before and am willing to repeat it: I am the first to try my best to hack presets in order to make life easier for those who don't want to do that. I have helped many people all over the world with technical issues, because they wanted to concentrate on the "art side", but accepted help where they hit a wall. I think that is a good approach. I love working with people who love working together.

    Building bridges is what I am here for. Working together is what I am here for. Making people seeing the world like I do, just because I am right by definition, is not what I am here for :-)

    This is not to say that @Pooby is forcing anyone to anything, this is not what I am trying to say.

    That being hidden, I am not really sure I get what you, @Pooby, actually expect from Fabric: If it was a set of specific presets/nodes that makes things easier for the kind of deformation work you want to teach, I do think that this is possible and just a question of working together to get them (the nodes) slammed together.
    If it is a kind of "representing arrays" or "autocasting array inputs", that is something the Fabric devs may or may not provide, depending on their perspective - but to some degree we (the community) can build workarounds that in fact do work. Like autocasting nodes.

    I do understand that you want to build deformers. But ... me still being stuck on modo, I could not do with your tutorials, because I can not pipe in-out geometry from modo to Fabric. So your tutorials would not help me understand Fabric, because I could not put them to use.
    Tutorials, however, that teach me how to address problems, how to find solutions by understanding the basics of "3d development", does help me. Procedural geometry I can do with in modo, so creating graphs for procedural geometry does help me. Such tutorials do not need "high level" nodes, in fact, those would make things over complicated.

    Fabric, in my eyes, is a powerful set of tools, NOT an "out of the box" machine that provides everything on the highest possible level. It is a tool to make tools. It is not an oven, meat slicer and dish washer in one, it is a set of pliers, screw drivers, soldering irons to make ovens, meat slicers and dish washers.

    I am sure that your perspective is helpful to the Fabric developers and that they are taking, what you say, into account for future releases.
    I am not sure that those people you are referring to, who refuse to leave their warm bed of never learning anything new, who refuse to write a single line of "god damned code" just because, are right in their world view. They are entitled to seeing things that way, but I don't think anyone should break his heart over not being able to attract them to Fabric.

    You know ... last week I learned to whistle on two fingers of one hand (instead of using both hands). I always thought that to be totally unnecessary and that I am the only one in the whole world who really understood the secrets of life, because he refused to learn whistling on two fingers of one hand.
    Turned out, I was wrong. It's actually quite helpful.

    Marc


    Marc Albrecht - marc-albrecht.de - does things.

  • AhmidouAhmidou Posts: 179 ✭✭

    I understand your frustration, all the problems you're pointing to are bothering me too, particularly the proliferation of nodes having the same functionalities but have different contexts or datatype, or not having dynamics array ports.
    In the perspective of having ICE type ease of use for Artist-TD, I think that it's probably going to come at some point, but Canvas is just not there yet.

  • PoobyPooby Posts: 122
    edited May 2016

    Malbrecht.

    I didn't use ICE for 3 years after release, because I thought I was an artist, not a techie. I completely understand the newbie viewpoint.

    It took someone to explain it very clearly to me along with some explanation of what vectors were etc and show it was actually pretty straightforward. Once I started dabbling, I found that it was fun to use.
    Once I learnt a bit, I started doing tutorials and found from feedback, many following the same path I had done, but this time, with me as the tutor.
    Now, what I'm saying is that I'm trying to do the same thing in Canvas, but its far far messier and not a lot of fun and I'm not making progress very well. This is WITH 6 years of Nodal programming experience. The first time with ICE I was coming in cold and still I progressed faster than I am now.

    Now, you might attribute that to the fact I'm comparing it unfavourably to ICE, or expecting it to work exactly the same way and having a tantrum when I can't do it. I'm not doing that. I'm just finding that its not very easy to just grab a few concepts, roll with them and gain momentum like ICE was/is.

    I don't know if you know ICE. but that is also a tool to make tools. I am well aware of the concept behind Fabric Engine.
    https://vimeopro.com/pooby/ice Thats some of my tutorials starting at the beginning years back. It might give you an idea of what it does if you're not that familiar with it.

    I'm not encouraging people to be lazy and not learn coding . All I want to do is make some honest tutorials that show Canvas is easy and fun to use and inspire them to give it a go. However first it actually HAS to be easy and fun to use, or I cant make the tutorials look that way.

    The counter argument I hear seems to be that its never going to be easy and fun to use, so deal with it and learn KL. Which I find rather defeatist seeing as I know these things CAN be easy and fun to use, as I have fun in ICE every day.

    As you have mentioned. There seems to be little response from Modo land. This is because they don't know where to begin. I guarantee you that if Canvas is fun and easy to pick up, then they will use it, just as Softimage users eventually started to use ICE. All it takes is for it to actually BE fun and easy to pick up, and then most Importantly, for someone to demonstrate that.

    Showing the usefulness alone means nothing to people. You have to make them feel that THEY can do it. its about instilling confidence and inspiration.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • malbrechtmalbrecht Fabric for Houdini Posts: 752 ✭✭✭

    I hear you, yet I find Fabric fun and easy to use. Took me a couple of hours to get into it and have been having fun ever since. Whatever I have seen of ICE, so far, made me shudder.

    All I am saying is: People are different. I like that.


    Marc Albrecht - marc-albrecht.de - does things.

  • PoobyPooby Posts: 122
    edited May 2016

    What you are saying is to be expected from your perspective, but you are coming from a coding background and I am not trying to cater to people with that kind of previous experience. I'm interest in attracting people who know nothing thus far.

    I'm glad that you can see how unfamiliar things make one shudder though. That's what I am talking about.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • Kevin TureskiKevin Tureski Posts: 126 ✭✭

    Without getting drawn too deeply into the details of this back & forth from two different perspectives, it is fair to say that we agree with your point @Pooby that Canvas is not [yet] as "artist friendly" as it could/should be and we will be doing things to address that. The addition of Blocks will make some - but not all - of the things that you are finding too difficult now much easier. We also want to make the learning curve less steep and less intimidating for non-coders. Input from these forums helps us figure out what to focus on. Both will take time but we'll get there.

  • PoobyPooby Posts: 122

    Thank you very much.

    This is not aimed at anyone in particular, but what makes sense to me is that, if there is a desire to make canvas easier to use, then it makes sense to encourage artists who actually use nodal programming, to stick with and give feedback on it, rather than persuade them to start coding as an alternative better method.

    It's only by trying to use the nodal system in place that you can actually notice where it falls short ergonomically and can thus make suggestions for improvements.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • malbrechtmalbrecht Fabric for Houdini Posts: 752 ✭✭✭

    coding as an alternative better method

    Oh, hell, I give up :lol:

    (I don't think that anyone here anywhere has ever tried to make "coding" look "better" than the graphical approach. The fact that some people feel more comfortable with it does not imply any kind of qualifying it. In fact, I think that everyone where who suggested not to refuse to accept some small amount of writing code as a helpful thing was actually trying to support the point of view that for a certain group of users "coding" looks like "evil doing").

    That said I will hence forth accept that "one can never please everyone" ...

    Marc


    Marc Albrecht - marc-albrecht.de - does things.

  • PoobyPooby Posts: 122

    But coding IS currently a better approach. I wasn't being sarcastic. It's true.

    You can access the full power of KL with code whereas Canvas has too many limitations currently, and is too convoluted.

    Hence why Canvas needs improving. We are taking at cross purposes it seems.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • PoobyPooby Posts: 122
    edited May 2016

    Can't edit the above on the iPad so I'll add here.

    Currently, to do certain things like Pex operators, you have to make a KL node.

    Now, if you take the current state of canvas as being how it will always be then it totally makes sense that you accept that if you want to parallel process large data sets and use 2d arrays etc, then yes. You need to learn to code and newcomers need to know that.
    It is my opinion that in all likelihood it will put the majority of non coders off learning it as It's going to make them shudder, in the same way ice makes you shudder as a coder.

    However. The current state is a transitional period. In a few months there may well be ways of achieving the same without making that KL node.

    I'd prefer to wait for that to happen before releasing tutorials that involve those processes.

    Cg artists use nodal systems all the time for shading, rigging, etc so it's familiar. It's going to produce less of a shudder than learning a whole new paradigm, syntax and language.
    It's nothing to do with things seeming 'evil' or such like. People simply like things that are accessible.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • cormactioncormaction Posts: 4
    edited May 2016

    First post here but just to weigh in a little on Pooby's side. I'm a stranded Softimage user in the big debris filled ocean of Maya at the moment and sadly no-one has bothered to pack the paddles! I am hoping that Canvas will be a nice big friendly paddle for me to grab onto in the future.

    I'm a fan of Pooby's work and have followed the man for a couple of years in softimage land and he does great work for the community there and has opened many peoples eyes to the possibilities of making their own tools and creating things they probably thought they couldn't. I've tried many times to get my head around programming in general, Python being my most recent attempt, I also began learning KL too. I always get to a certain point and get stuck. My brain just isn't made that way. However, for whatever reason, ICE was created in such a beautiful way that it was accessible to braindead people like myself and allowed us to easily throw together simple or even complex tools or sometimes even cobble together bits of other peoples tools in order to get the job done. It has saved my ass many many times. I was able to show other artists my ICE tree and get them on board extremely fast and with no headaches and have them make any necessary changes themselves.

    As Pooby said, telling the type of artists I'm talking about to learn how to code to do these simple things is just not going to happen. Some peoples brains just aren't built that way but somehow nodes bridge that gap and make a lot of sense to most artists. The fact that Canvas is almost there, even when things seem to be named and based on/inspired by ICE's way of thinking suggests that it can be ICE and even better. So why not? If the possibility and opportunity is there to take it to the next level where people can easily string together graphs from other artists in the community's in a crunch then why not take it there and make it easily accessible to the masses? This could be absolutely revolutionary and has all the potential to be, so it seems to me arguing that people should stop and learn to code is a regression in my very humble opinion. Yes it is better from an efficiency standpoint but artists are messy dirty creatures and efficiency isn't the goal, the outcome and journey to it is.

    Nodes are modular and great fun, I can't imagine trying to sort through snippets of code and debugging them for missing commas, misspellings and semi-colons is going to appeal to the masses. Lego is just more fun than math equations. Just my two cents.

  • PoobyPooby Posts: 122
    edited May 2016

    Thanks.
    I think that people, Once they know how to make stuff in Canvas, will have a much higher chance of wishing to learn KL later, than presenting it to them at the start as a requirement. The important thing is to get them interested in trying anything at ALL in Fabric and to do that, it has to look fun.
    Most people don't have a clue what a vector is. I didn't when I began in ICE. Its important to bear in mind the starting level of the demographic you're aiming at.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • Aaron KentAaron Kent Posts: 66 ✭✭
    edited May 2016

    Here's something to throw out there from left field and this feels like the appropriate thread. Much like Lynda or Digital Tutors I would be happy to pay something if Fabric tasked someone with doing some training for non coder Canvas hopefuls. It could be the tutorial I outlined or just something similar to Digital Tutors ICE reference Library. I honestly feel like just a simple video showing a simple function of each available node within Canvas would help those of us struggling to make the transition from ICE.

    AK

  • cormactioncormaction Posts: 4

    just to add, I was just trying to think of a way to articulate the artist/code conflict a little better for those who advocate for the coding side of things. One of the main reasons, for me at least, an elegantly structured entirely node based workflow works really well is down to frequency of use.

    An artist working on commercials day in and day out is usually a generalist who has to constantly switch between being a modeller, a rigger, a texture artist, a lighter etc. Its all a bit A.D.D. in our world. I find when I learn to code in a specific language it's all well and good for the short period when I learn it and use it but when I come back to it, perhaps even months later its pretty much all disappeared from memory. Unless you work with code every day, which generalists tend not to, it can be very difficult to remember how exactly to format strings of text, which flags go where and in what order. This breaks the flow of work and creative energy.

    This is where the beauty of nodes come in for me. All I have to do is remember the general concepts of what I need to do and hit tab or search for a node with a nice logical name that I think will do the trick and after just a minute or two the flow comes right back. I've been switching between Soft and Maya a lot that past couple of years and even if I'm out of Soft for a while and need to jump into ICE, it all comes back very easily. Whereas if I try to pick up where I left off with Python I'm almost back at square one and forget about it if I have to code up something in a hurry, I'm totally lost! So what I'm basically trying to say is that I think and I reckon many of us nodey type people understand that coding is a much more efficient way of doing things but the power and flexibility of doing things through nodes for a person who only dips their toes in on occasion is something quite special.

    Main reason I push this point a little harder personally is because I feel this view isn't being fully heard by the developers of various softwares. I have been involved in betas of other softwares development and sometimes I feel it's difficult to get this point of view across to engineers and developers who's lives are steeped in code and not in the day to day production side of things. Canvas has so much potential to be all things to all people much like ICE was, I just hope it goes that direction and simplifies the process for the "casuals".

    Apologies for long rambling post, perhaps I'm preaching to the choir here but it's a subject I happen to be a bit passionate about ;)

  • fortonforton Posts: 9

    I've been doing a test with canvas to make a basic strand setup.
    It looks like the perfect theme to make a tutorial about, it's all nodes exept for two cases where I had to make a loop to do stuff with an array.

    Problem is that I never did a tutorial before, but I'm willing to try.
    Can someone give me a tip for what screenrecorder to use? preferably opensource...

    I also have lots of questions, for example about using my beautyfull strands and data in my dcc of choice.
    I have zero experience with python or alambic in xsi and that's where I had to give up.
    You see, I have probably as many questions as tips, but if we all hang on together we will make it in the end. :smile:

    cheers, Wim

    PS: I'm still convinced that a simple ice node would be the perfect data bridge between fabric and xsi (next to the preferred use alambic of course).

  • PoobyPooby Posts: 122
    edited May 2016

    I totally agree that a Fabric data ICE node would be a perfect interchange between Fabric and Softimage. I think there is reluctance due to the fact it wont make any sense except to Softimage, which undermines the central 'deploy anywhere' philosophy of Fabric.

    However, I think that other DCC's need to up their game and be able to deal with that kind of atomic data. At the moment, you can swap data across, but you'd need to convert it to 'pointpositions' and read/decode it from pointposition data at the other end in ICE.. Clunky, but it works.
    For Fabric to be really useful for character work, it needs to at least be able to write weight-maps. I use dynamic maps for skin wrinkling deformation based upon tension etc. It would be a pain to have to 'sneak' this stuff in through pointpositions. We need Canvas to output whatever it likes, and the dcc to be able to read whatever is appropriate. Point positions is the bare minimum required to be honest. Its ok for basic games, but not great for higher end work.

    Please have a go at Tutorials. It would be great to have as much knowledge about Canvas as possible out there.

    Heres a few free screen recording options http://www.hongkiat.com/blog/win-screen-recording-softwares/ I used to use the free Camtasia for screen recording which did the job,(but then bought a commercial one as it had a few more options.)

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

  • malbrechtmalbrecht Fabric for Houdini Posts: 752 ✭✭✭
    edited May 2016

    Moin,

    as promised I'll keep my mouth shut on the discussion, but:

    Can someone give me a tip for what screenrecorder to use? preferably opensource...

    There are a few considerations you can make when choosing "the right screenrecorder". OpenSource may be one (I tend to not use OpenSource more and more due to the fact that I don't like an unknown number of people contributing bugs to a software I want to rely on - that's just me, obviously, just recovering from the worst nightmare: Trying to build a Linux desktop system).
    Another thing is that screenrecorders will take away performance from your system. So it depends on a number of components: What do you want to do, do you need performance, do you have a GPU that can do life-encoding, do you want high quality capturing.
    Last, but not least, the OS is a thing to consider.

    Many tutorial creators swear by Camtasia. Sometimes to a degree that makes one think, it's just because of the high price that the tool gets such a praise. That said, it does have a lot of bells and whistles that can help reducing the time of "post production".

    I have been using OBS Studio lately (which is, coincidentally, open source - and free), to be found here: https://obsproject.com/ - it's cross platform. Go for the studio version. It can encode your stream on the GPU, which, in certain cases, helps with performance of the sytem. Setting it up is a bit complex, let me know if you need help.

    Previously I have been using (free) CamStudio, which comes with a free lossless codec, which also helps with performance - but I found it still takes away quite some CPU cycles, the installer is "strange" (sometimes trying to install nagware) and I did have some stability issues at times. It's Windows only, as far as I know. Because of the installer BS I am not giving a link here, let me know if you want to try it out nevertheless.

    Some video editing software include screen recording features. I used to do tutorials with VideoDeluxe by Magic, but that software is so bugged that I can not even recommend the more-or-less "free giveaway" versions you can sometimes find.

    For editing I am still using VegasPro, but mainly because I am used to the workflow. I want to make the switch to Davinci Resolve (if only it had audio capturing, I'd even pay for the full license!). Blender is being used by more and more tutorial creators, as its editing suite seems to be quite flexible. With Camtasia you could have it "all in one" - but OBS Studio does have a (limited) editor built in as well.

    I hope this helps - we may need to start a separate topic to discuss this further, though :) I am doing a lot of editing (and tutorials) and would love to share experiences.

    Marc


    Marc Albrecht - marc-albrecht.de - does things.

  • Paul DoylePaul Doyle Administrator, Fabric Employee Posts: 229 admin

    @Pooby said:
    I totally agree that a Fabric data ICE node would be a perfect interchange between Fabric and Softimage. I think there is reluctance due to the fact it wont make any sense except to Softimage, which undermines the central 'deploy anywhere' philosophy of Fabric.

    There is that, and also the harsh/sad truth that Softimage is less and less present in customer studios. We simply can't justify engineering time on an application that cannot be purchased anymore. We have committed to supporting Softimage for as long as we have paying customers using it, but it's very much in the context of serving as a bridge to other DCCs.

    @Aaron Kent said:
    Here's something to throw out there from left field and this feels like the appropriate thread. Much like Lynda or Digital Tutors I would be happy to pay something if Fabric tasked someone with doing some training for non coder Canvas hopefuls. It could be the tutorial I outlined or just something similar to Digital Tutors ICE reference Library. I honestly feel like just a simple video showing a simple function of each available node within Canvas would help those of us struggling to make the transition from ICE.

    Hey Aaron - we are paying attention to this thread and are constantly working to make Fabric more accessible. We are pushing very hard on certain key functionality for the next two (ish) major releases. The high-level item that really matters to you is the Blocks feature. Once we have that in place we will spend time on more entry-level material for people. Doing it prior to that doesn't make much sense.

    One thing that I want to make clear is that we aren't trying to reinvent ICE. There are certain things it might always be better suited for, and that's fine. "In ICE we could..." isn't always going to be something that we strive for, because there were a lot of things that ICE couldn't do that Fabric can do. Hopefully the pluses will outweigh the minuses! That said, quality training material, samples and tutorials etc. is something we will push on as it clearly has value.

    We took a deliberate design approach with Fabric that was to address the high-end, high-performance, R&D-centric crowd first. We're now well into the next phase of making it accessible and easier to use, but always with the caveat that we aren't going to compromise on performance. So sometimes it might not be as easy as another approach, but there will usually be a good reason for that. And sometimes the answer might be "that's going to be a lot easier to write in KL" - we're working to reduce the occasions where that's the case, but visual programming does have limitations and downfalls that are inescapable. Our hope is that once we get into Blocks that people will have enough 'meat' to get somewhere useful without writing code.

    Anyway - we're grateful for the feedback and the fact that you guys are prepared to give up your time to work on tutorials. Please keep telling us what you'd like to see.

    CEO at Fabric Software
    Twitter

  • PoobyPooby Posts: 122
    edited May 2016

    @Paul Doyle said:

    There is that, and also the harsh/sad truth that Softimage is less and less present in customer studios. We simply can't justify engineering time on an application that cannot be purchased anymore. We have committed to supporting Softimage for as long as we have paying customers using it, but it's very much in the context of serving as a bridge to other DCCs.

    Thats understandable, I'm learning Fabric as a replacement for ICE and Softimage in future, but the point I went on to make was that I think Fabric needs to be able to feed in more than Point position data when it comes to meshes and at least send scalar per point information as weightmaps and preferably any kind of data.

    At the moment, compared to other DCC's Softimage may be where this Fabric data is most useful in a scene because it has ICE and the two would work together seamlessly, but on the other hand I don't actually need Fabric whilst I have ICE and Softimage. I only ever work on the standalone Fabric version anyway for learning.
    However I'm looking toward this future time when i may need Fabric in a DCC, and it would be a shame to have any kind of regression in ability from what I'm getting now with ICE and Softimage.

    I'm primarily trying to help the visual programming development, in my own small way, through providing feedback on Canvas.

Sign In or Register to comment.