The future of Bryce?

12346»

Comments

  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    For example, this 1080p image was rendered outside of bryce, with some depth of field effects, 6 ray bounces, some anti aliasing, shadow effects, global illumination, a transparent plastic material, a bumpy metal material, and with an HDR. The model is approx. 58,000 poly's in all. everything you see here, took 7:48 to render. I know that such a render would take a heck of a lot longer in bryce, and would take more effort to setup too. Even thrown together with not much content value, that's a pretty impressive render time considering all the factors.

    Different model, different application, different computer.

    I haven't tried to reproduce your render here, but it does have: Depth of field, transparent glass (with a bump), IBL lighting, HDRI reflections, metal with bump, an abstract shape and shadow effects.

    All this rendered in 11.39 in Bryce on an 8 year old Macbook laptop. :)

    ShapeTime.jpg
    1280 x 800 - 184K
    ShapeALT.jpg
    1846 x 1080 - 299K
  • David BrinnenDavid Brinnen Posts: 3,136
    edited March 2015

    For me what it boils down to is what your aim is. For hobby artist, the tool you want to choose is the one that is nicest to use. Time is not really that much of an issue if there are no deadlines. So long as you are enjoying yourself and you are able to work towards your own goals then it doesn't really matter. I have spent hours in Minecraft making things for my own amusement, just because it is a great environment for that kind of thing and presents interesting challenges that make the process of building stuff fun for me. Though to be fair, I can't see having a zombie chewing on your leg while you are trying to use Bryce is going to add anything positive to the experience - it is a more serious creative tool than Minecraft.

    Commercially. Then it is a different matter, efficiency is a serious concern. At that point, load your PC up with meaty GPU's and your HD with dedicated software. Using Bryce for commercial projects is difficult, I've done it, Dave's done it, I'm sure plenty of others have too, but it's not an easy route to take. Which is why I also have Octane, Modo, Poser, MD, Grome, Genetica, Crazy Bump, UVlayout, Prepper, DS... and so on, plus a £200+ GPU sat in my PC that isn't connected to any monitors but just sits there waiting to render.

    So you pays your money and you takes your choice. For hobbying, Bryce is fine for me. I like it, it is fun to use and I don't feel that it particularly limits me in what I want to create. For commercial stuff, I break out the professional software and get the job done with as little pain as possible. That is work, work pays for the expensive hardware and software and in exchange for that investment, the work gets done quicker. Fun really isn't part of that deal.

    Edit: cross posted with you Dave, beautiful render! And I was going to say, I concurred with you said before.

    Post edited by David Brinnen on
  • Electro-ElvisElectro-Elvis Posts: 842
    edited December 1969

    @Rashad: Well written, though I am a Bryce fanboy, I see its limitation. (But that's might be the reason, why I love it.

  • FONSECAFONSECA Posts: 469
    edited December 1969

    Fishtales, and bigh.

    Thanks for your help with my tumbleweed question.
    To be honest, I haven't used Bryce's Tree Lab much but I will experiment with it now that I have seen your renders.

    Another form of plant/bush I am interested in making a model of are Willow bushes, the kind described in the supernatural story, "The Willows" which takes place in the eastern Danube river area, somewhere in Hungary, if I remember right.

    chohole

    One of those tumbleweed videos was filmed in the Mojave desert.
    I live in the Mojave desert area of southern California, USA, and much of the vegetation and landscape is weird out here. The Quail birds have what looks like a small pirate flag stuck on their heads, and they make a funny, odd, calling sound. This place would be like another planet to you.

  • FishtalesFishtales Posts: 6,038
    edited March 2015

    It sounds as if it might be the Weeping Willow "Most ominous are the masses of dense, desultory, menacing willows, which "moved of their own will as though alive, and they touched, by some incalculable method, my own keen sense of the horrible." "

    http://www.gardenality.com/ImageViewer/?id=2302&IUID=3&RID=273#imageId=2302

    Post edited by Fishtales on
  • ChoholeChohole Posts: 33,604
    edited December 1969

    Fishtales, and bigh.

    Thanks for your help with my tumbleweed question.
    To be honest, I haven't used Bryce's Tree Lab much but I will experiment with it now that I have seen your renders.

    Another form of plant/bush I am interested in making a model of are Willow bushes, the kind described in the supernatural story, "The Willows" which takes place in the eastern Danube river area, somewhere in Hungary, if I remember right.

    chohole

    One of those tumbleweed videos was filmed in the Mojave desert.
    I live in the Mojave desert area of southern California, USA, and much of the vegetation and landscape is weird out here. The Quail birds have what looks like a small pirate flag stuck on their heads, and they make a funny, odd, calling sound. This place would be like another planet to you.

    Probably. I will stick with sedentary, non peripatitic vegetation and fairly normal bird life. Although I have to admit I do find it passing strange that we get birds singing sweetly as they go to bed, but always get woken by the sound of the jackdaws, not the song birds. May be something to do with the fact that the raptors fly early morning reccies. Kites, Buzzards and Hawks a plenty.

  • useroperatoruseroperator Posts: 247
    edited March 2015

    For example, this 1080p image was rendered outside of bryce, with some depth of field effects, 6 ray bounces, some anti aliasing, shadow effects, global illumination, a transparent plastic material, a bumpy metal material, and with an HDR. The model is approx. 58,000 poly's in all. everything you see here, took 7:48 to render. I know that such a render would take a heck of a lot longer in bryce, and would take more effort to setup too. Even thrown together with not much content value, that's a pretty impressive render time considering all the factors.

    Different model, different application, different computer.

    I haven't tried to reproduce your render here, but it does have: Depth of field, transparent glass (with a bump), IBL lighting, HDRI reflections, metal with bump, an abstract shape and shadow effects.

    All this rendered in 11.39 in Bryce on an 8 year old Macbook laptop. :)

    It's actually quite difficult to compare render content+settings with any match-for-match certainty, since either renderer can be doing other things in the background, or have application specific features.

    But I think I've come up with something somewhat fair.

    In this test, I did the same model, at the same basic angle/size, with the same UV space coloring. The only difference is the bryce one uses a default matte shader (i.e. just the color it is imported with, no texture), while the non-bryce one uses a reflective glossy shader by default, same color, even if it may not appear exactly the same.

    Both renders use 6 ray bounces, and the the default amount of antialiasing, and the same exact HDR map, and are rendered at 800x600 resolution. And yes, I made sure to set the bryce render priority to high. I also made sure to disable other background stuff like sunlight.

    The bryce render uses no reflections nor shadow smoothing; both of which would just significantly add to the render times anyways.

    The non bryce render, in all its glory, took 30.3 seconds to render in total.

    The bryce render, in its not-so-glory, took a whopping 37 seconds.....and is indisputably a far worse quality.

    It's important to note that adding stuff like shadow smoothing significantly upped the render time in bryce...just a minor bit of shadow smoothing upped the render time to over 30 minutes! In the non bryce render, there is default shadow smoothing going on as well, and it's quite high quality.

    This is not by any means a precise experiment, but the results speak alot even without precision. The bryce render, obviously wouldn't be able to increase quality or even match it. without increasing rendering time, significantly. It's clear that things in the first image are not only done in a better higher quality manner, but also in an efficient manner.

    The render time isn't even exact for both, but it's easy to see that it takes approximately 20% more render time, for a much lesser quality output. Which means that for an on-par quality output, we can infer it would take 100's of percentage points in render time to achieve that with bryce, (if not 1,000's of percentage points). I later made all of the materials in the bryce render with a 25% reflectivity, and the render time shot up to 2:05, and didn't really look all that much better.

    gfgdfgdf.png
    814 x 617 - 136K
    gfyhgfhgfh.png
    816 x 618 - 136K
    gfhghgfhgfh.png
    809 x 627 - 337K
    Post edited by useroperator on
  • Tim82Tim82 Posts: 858
    edited December 1969

    i think ill still stick with bryce ;) ...it may be slower than other software, but bryce is one of the most effective in the right hands :) ...check out my gallery, i think my results will speak for them selves :)

  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    But I think I've come up with something somewhat fair.

    All you have shown is that you don't know how to use Bryce.
  • useroperatoruseroperator Posts: 247
    edited March 2015

    But I think I've come up with something somewhat fair.

    All you have shown is that you don't know how to use Bryce.

    That's quite a harsh statement. Care to elaborate? The point of the images were not to match quality -- but rather to match render time vs quality achieved in a given amount of time. I also didn't want any discrepancies in material specific shaders, like metallics for instance, because if one is far more efficient than another, then the results would be vastly different -- and that's probably more for bryce's benefit than anything else.

    I am very well versed in bryce. As I said, I first used it in the 90's.....and I've used it on/off since then, I know of pretty much all of its features and settings, what they mean, and how to use them.

    a render like this old one from bryce for example, is going to take a lot longer than 30 seconds.

    fdgfgdfg.png
    562 x 500 - 416K
    Post edited by useroperator on
  • FONSECAFONSECA Posts: 469
    edited May 2015

    Maybe a detailed looking tumbleweed could be made with Wings3d using skinny thin cylinders, etc

    Post edited by FONSECA on
  • useroperatoruseroperator Posts: 247
    edited March 2015

    Concerning the Tumbleweeds...

    I did some fiddling in the Bryce Tree Lab, testing it's tumbleweed capability, but before long the infamous, "Out of Memory", window popped up.
    So, I loaded up the Large Address Aware stuff, but soon the memory usage creeped up to 3.5GB, and then the "Out of Memory" thing popped up again.
    Those tree models use up a lot of memory or I am doing something wrong.

    Another option for this globular weed thing is to use Lisa's Botanitcal "Dead Bush" model pack, which I bought years ago when it was on sale, and fiddle with the dried dead bush models until they look like spherical tumbleweeds.
    How about a morphing tumbleweed for Poser; that would be really helpful.

    Maybe a detailed looking tumbleweed could be made in Wings3d using skinny thin cylinders, etc

    Meanwhile, I found this animation video while searching for tumbleweed files.

    https://www.youtube.com/watch?v=bey8gCF8kVE
    CGI 3D Animated Short HD: "Tumbleweed Tango" by - HumbleTV

    I looked for you, and it is surprisingly hard to find a free tumbleweed model. my suggestion would be to settle for a more sparse tumble weed design to save on the memory issues.

    I modeled this one by hand with thin cylinders real quick...not a bad approximation, especially if you squint ;)

    xcvxcvxcvxcv.png
    389 x 331 - 94K
    Post edited by useroperator on
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    I am very well versed in bryce. As I said, I first used it in the 90's.....and I've used it on/off since then, I know of pretty much all of its features and settings, what they mean, and how to use them.

    And yet you claim that adding soft shadows adds 30 minutes to your render on your super fast, 8 core computer?
    When clearly above, my render has soft shadows, IBL, Reflections, metal and glass.... And rendered in 11 minutes on my 8 year old laptop.

    Yes, I could have set mine up so it took a day to render, that would involve me forgetting everything I have learned about optimisation of material, lighting and render settings.

  • useroperatoruseroperator Posts: 247
    edited March 2015

    I am very well versed in bryce. As I said, I first used it in the 90's.....and I've used it on/off since then, I know of pretty much all of its features and settings, what they mean, and how to use them.

    And yet you claim that adding soft shadows adds 30 minutes to your render on your super fast, 8 core computer?
    When clearly above, my render has soft shadows, IBL, Reflections, metal and glass.... And rendered in 11 minutes on my 8 year old laptop.

    Yes, I could have set mine up so it took a day to render, that would involve me forgetting everything I have learned about optimisation of material, lighting and render settings.

    setting soft shadows to just a setting of 1, added greatly to the render time. So it's not like I added some insane setting to make the numbers out of wack. soft shadows weren't even including in the actual test....and you can't tell me they would have come at no cost, so regardless, the comparison is still very much valid.

    http://www.filedropper.com/render_3


    you're free to attempt to render it 'better' and in a short period of time if you think you can. there is a hollow tower used in this obj model, to frame and lineup the render.

    Keep in mind, your render up above, has a lot less going on when it comes to ray data. look below the render time window and compare; yours has 3.3 billion rays @ 45 per pixel.....mine has 3.5 billion rays at 72 per pixel. Yours has total 8 billion intersection attempts, mine has over 1 trillion! etc etc. ( that's 125 times more intersect attempts 20 times faster) The scenes just aren't comparable. The model in my scene is over 154,000 poly's.

    I compared them about as fair as you can compare different renderers performance based on time and quality, with probably a bit of leeway in the favor of bryce.

    So I think all things considered, the difference between my beastly CPU compared to your 8 year old macbook, is very much justified in the numbers. And considering that, the difference in the same scenes on different renderers, is also justified, when they're performed on the same CPU. I doubt your 8 year old macbook could even render that scene in anywhere near 37 seconds -- perhaps not even in 11 minutes and 39 seconds.

    I'm not fudging settings just to twist results. I think you're being shortsighted here, and the technicalities are going over your head.

    Post edited by useroperator on
  • Peter FulfordPeter Fulford Posts: 1,325
    edited December 1969


    user.operator only came into these forums recently. I remember when he was first posting inquiries about Bryce about a year ago. He had no preconceived biases, but was entirely open minded at the time. He did his homework on Bryce as well as other applications, compared render speeds and quality, and within a short time decided not to continue using Bryce.

    OK, so having made that decision, why not gallop off into the sunlit, grassy uplands using not-Bryce?

    I'm reminded of the kid who, via an accident of birthdate, was able to leave school a month earlier than the rest of us. So he left school, but rather than make a break into life and move on, he ends up hanging around the school gates on his moped so he can show off how he's not stuck at school whilst we were. It took less than a week for even the most impressionable to realise what a sad loser he was. And the smart kids entertained themselves challenging him to do wheelies and tyre burns on his "freedom" machine.

    (Well done, Dave. ;-) )

    .

  • fixmypcmikefixmypcmike Posts: 19,565
    edited December 1969

    Let's keep it civil, please.

  • useroperatoruseroperator Posts: 247
    edited March 2015


    user.operator only came into these forums recently. I remember when he was first posting inquiries about Bryce about a year ago. He had no preconceived biases, but was entirely open minded at the time. He did his homework on Bryce as well as other applications, compared render speeds and quality, and within a short time decided not to continue using Bryce.

    OK, so having made that decision, why not gallop off into the sunlit, grassy uplands using not-Bryce?

    I'm reminded of the kid who, via an accident of birthdate, was able to leave school a month earlier than the rest of us. So he left school, but rather than make a break into life and move on, he ends up hanging around the school gates on his moped so he can show off how he's not stuck at school whilst we were. It took less than a week for even the most impressionable to realise what a sad loser he was. And the smart kids entertained themselves challenging him to do wheelies and tyre burns on his "freedom" machine.

    (Well done, Dave. ;-) )

    .

    Actually, I still use other DAZ software, which is why I'm still around these forums.

    But while Rashad is right about his assessment, he is wrong about other things. I've actually been here much much longer, it's just that I don't have access to my old account -- because I don't have access to the email address that it was tied to. So I can't simply do something like a password request -- so I merely created a new account.

    I have sinced moved on from bryce, but that doesn't mean I can't discuss it. In fact, I find that most people here are being a bit too temperamental and defensive for no good reason, and in the face on honest discussion that others have stretched on into more, that I am merely reacting and responding to at this point.

    As for Rashad, I think he understands the justification behind my reasoning, as he knows the software I use and has seen the advantages of it first hand.

    Post edited by useroperator on
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    2 and a half minutes... It's a bit noisy, but if I had done a 30 minute render, the noise would have completely gone.

    And look at how soft those shadows are.

    Testing.jpg
    800 x 600 - 161K
  • useroperatoruseroperator Posts: 247
    edited March 2015

    2 and a half minutes... It's a bit noisy, but if I had done a 30 minute render, the noise would have completely gone.

    And look at how soft those shadows are.


    Not bad.....but if I counted all of those dark noisey pixels, how many actual pixels would be left of real visual data? Of course it's only going to take 2 minutes, because all of that noise essentially cuts the vast majority of processing needed, it's a lack of processing done. Also keep in mind, you didn't use the same HDR map either (I forgot to include that). if anything, this all just bolsters my comparison, it doesn't refute them in the slightest.

    I mean I don't even know why you're trying to refute it anyways, when it's made pretty clear. I am only doing this for informative purposes because on your prompt....but it seems more like you're filling your end with pure pride.

    Post edited by useroperator on
  • FONSECAFONSECA Posts: 469
    edited March 2015

    user.operator

    Your tumbleweed model looks like it would be a very useful starting base. Thanks for the help.
    What 3d program did you model it in? The next step is to carefully, turn, curve, and twist the main stems to give it the swirly tumbleweed appearance.

    Below are two tumbleweed models from Turbosquid.com.

    tumbleweed by gamewarden01
    http://www.turbosquid.com/3d-models/tumbleweed-tumble-weed-3d-model/711625

    tumbleweed by dennis789
    http://www.turbosquid.com/3d-models/tumbleweed-weed-tumble-3ds/896844

    Two photo examples from the web.

    tumbleweed-on-the-bonneville-salt.jpg
    603 x 599 - 138K
    russian-thistle-tumbleweed_363_600x450.jpg
    600 x 450 - 135K
    Post edited by FONSECA on
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    2 and a half minutes... It's a bit noisy, but if I had done a 30 minute render, the noise would have completely gone.

    And look at how soft those shadows are.


    not bad.....but if I counted all of those dark noisey pixels, how many actual pixels would be left of real visual data? of course it's only going to take 2 minutes, because all of that noise essentially cuts the vast majority of processing needed. also keep in mind, you didn't use the same HDR map either (I forgot to include that). if anything, this all just bolsters my comparison, it doesn't refute them in the slightest.

    Then you miss the point.
    My point is to show that you don't know how to use Bryce.
    You would get a really bad render taking 30 minutes that would probably still have banding within the shadows on your 8 core machine.
    I can get a totally smooth render in that amount of time on my dual core 8 year old lap top using the exact same model.

    And as there is no reflection on the render, it doesn't matter one jot which HDRI was used. If it casts shadows, I can get them totally smooth in the same time it takes you to get banded pseudo-smooth ones.

    I don't really care that other software can do it quicker. On my new iMac, I have C4D... It's cool, but it's not Bryce. :)

  • useroperatoruseroperator Posts: 247
    edited March 2015

    2 and a half minutes... It's a bit noisy, but if I had done a 30 minute render, the noise would have completely gone.

    And look at how soft those shadows are.


    not bad.....but if I counted all of those dark noisey pixels, how many actual pixels would be left of real visual data? of course it's only going to take 2 minutes, because all of that noise essentially cuts the vast majority of processing needed. also keep in mind, you didn't use the same HDR map either (I forgot to include that). if anything, this all just bolsters my comparison, it doesn't refute them in the slightest.

    Then you miss the point.
    My point is to show that you don't know how to use Bryce.
    You would get a really bad render taking 30 minutes that would probably still have banding within the shadows on your 8 core machine.
    I can get a totally smooth render in that amount of time on my dual core 8 year old lap top using the exact same model.

    And as there is no reflection on the render, it doesn't matter one jot which HDRI was used. If it casts shadows, I can get them totally smooth in the same time it takes you to get banded pseudo-smooth ones.

    I don't really care that other software can do it quicker. On my new iMac, I have C4D... It's cool, but it's not Bryce. :)

    Well it's not because I don't know how to use bryce, it's just how bryce is and exactly what I've been pointing out -- bryce's slow rendering speed on par to the average of what's out there today. It is probably by far, the slowest renderer in existence. If you don't really care, then I don't see why you would try to refute it or reply to a comparison? Don't act like I'm missing the point, because with that in mind -- it's more like you're missing the point.

    As I said, I disabled stuff that would have increased the render time, like TA or sunlight. increasing the quality = increasing the render time; that's just a given. I mentioned that stuff to make it clear I wasn't leaving some innocuous performance hog setting on that need not be on.

    it took you 2 minutes to render something I could probably render in a few seconds. You can't just turn down the settings then say "See I rendered it 'fast' ! You just don't know how to use bryce!".

    That doesn't even compare to my bryce render, since it is soo noisy, that noise is effectively a lack of render processing rearing its head. Had my image been soo noisy, it would have probably been well under 37 seconds, but it still wouldn't have been a comparison to the other image rendered in the other software. So, respectfully, I'm not sure what you're hoping to prove here.

    Keep in mind that due to instruction sets, my CPU could be 100x faster on average in other things, but in a program that uses an old instruction set, the disparity would not be the same. I've already stated that perhaps bryce has some slowdown due to not using the latest instruction sets. Not that any of this actually justifies any of the differences in performance you tried to pass off by taking shortcuts that were the result of a lack of processing, rather than an efficiency increase.

    I feel like I've exposed quite a few disturbing trends/actions here.

    Post edited by useroperator on
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    This exchange started because I said all you were showing is that you didn't know how to use Bryce.

    You replied by doing some comparison and mentioned the soft shadows would have taken over 30 minutes to render... Proving my point that you don't know how to use Bryce.
    I showed you that soft shadows can be rendered in a lot less time. I admitted my render was noisy, I know why it's noisy.

    I really don't care if you can do a render in seconds using other software... That has nothing to do with your knowledge of Bryce.

    I'm just doing a render that will take approximately 26 minutes. I'll post it when it's done. I could probably optimise it more but it's 5am here and I've had a busy night.

  • useroperatoruseroperator Posts: 247
    edited March 2015

    This exchange started because I said all you were showing is that you didn't know how to use Bryce.

    You replied by doing some comparison and mentioned the soft shadows would have taken over 30 minutes to render... Proving my point that you don't know how to use Bryce.
    I showed you that soft shadows can be rendered in a lot less time. I admitted my render was noisy, I know why it's noisy.

    I really don't care if you can do a render in seconds using other software... That has nothing to do with your knowledge of Bryce.

    I'm just doing a render that will take approximately 26 minutes. I'll post it when it's done. I could probably optimise it more but it's 5am here and I've had a busy night.

    So enlighten me then, what magical setting must I disable or enable to make soft shadows take much less of a performance hit? Other than half rendering an image with extreme noise, of course.

    Post edited by useroperator on
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    Keep in mind that due to instruction sets, my CPU could be 100x faster on average in other things, but in a program that uses an old instruction set, the disparity would not be the same. I've already stated that perhaps bryce has some slowdown due to not using the latest instruction sets. Not that any of this actually justifies any of the differences in performance you tried to pass off by taking shortcuts that were the result of a lack of processing, rather than an efficiency increase.
    It's OK, you don't have to justify why your 8 core mega machine is wasted when you don't know how to set a job up optimally for the software you're using. My poor 2.2Ghz Intel Core 2 Duo Laptop is weeping.

    I feel like I've exposed quite a few disturbing trends/actions here.


    If that feeling is as accurate as your feelings about Bryce.... We're OK, we've got nothing to worry about. :)
  • useroperatoruseroperator Posts: 247
    edited March 2015

    Perhaps I should be more clear? It’s the cast shadow softness in the IBL that makes it take soo much time. Not the 'soft shadows' in the premium settings. So I fear, you've made a big fuss over nothing. Go ahead, enable that. See what your render time becomes. Of course, none of this was even included in my base comparison, it was essentially just a foot note you made a big focus on.

    Post edited by useroperator on
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    Bryce lied to me... It took 32 minutes.

    Testing2.jpg
    800 x 600 - 285K
  • Dave SavageDave Savage Posts: 2,433
    edited December 1969

    Perhaps I should be more clear? It’s the cast shadow softness in the IBL that makes it take soo much time. Not the 'soft shadows' in the premium settings. So I fear, you've made a big fuss over nothing. Go ahead, enable that. See what your render time becomes. Of course, none of this was even included in my base comparison, it was essentially just a foot note you made a big focus on.

    The shadows in my render are entirely coming from the HDRI. There is no other light source in the scene.
  • Cris PalominoCris Palomino Posts: 11,142
    edited December 1969

    Since the discussion is becoming rather circular and cannot stay civil, we are closing the thread.

This discussion has been closed.