Renderosity Forums / Bryce
Welcome to the Bryce Forum
Forum Moderators: TheBryster
Bryce F.A.Q (Last Updated: 2024 Oct 30 3:44 am)
Subject: More on Bryce5 render times - probable platform differences.
I know how you feel, but I gave in and bought it too. But hey things could be worse! Bryce could have been orphaned all together. I have to have a little faith, I'm a Carrara user, and we had to wait almost a year and a half for a patch that would allow the program to even run right right, it was crash city for everyone. So maybe Corel will eventually release some patches to improve things a bit. I do like the Tree Lab though. Bryce never was a fast rendering program, so a couple of more minutes won't kill us, but it would of been nice to see a big speed increase. Corel got my $99, so now I'll have to make the best of it. IT's still the best deal for landscape scenes ou there though, World Construction Set would be nice but I don't have $1000 to spend on it. But believe me, I do understand what you guy's are feeling, I'm torn about it, but I bought it anyway, I probobly should of waited,But............ Hang in there! kaom
So you guys/gals think that there's a Corel conspiracy to screw over approximately 95% of its user base to favor the 3% Mac population. Make's a lot of sense doesn't it. There's "mucho" anger on this list and probably most of it's warranted, but come on. Corel may have done some things wrong, but so did Metacreations. Remember 2.0 on the PC? Were I to reply in the same vein I might bitch and moan about the multiple(!) programs allowing Poser animation imports into PC Bryce or the statement by the Mimic developers that PC users utilize Poser as an animation tool and Mac users utilize it solely for illustration purposes (bullshit), thus no Mac port. Hey, all the more reason to save up my coins for a 1.7G pentium:) Let's not turn the failings or benefits of Bryce5 into a platform war. Face it, both PC or Mac, Bryce rendering times were unworkable. Now with the process's new features, they've gone through the roof. Who in their right mind is going to wait hours for one NTSC frame, never mind 30 frames/sec or even worse 60 interlaced fields per second. Imagine further going to high AA to get rid of the everpresent sparkling! Multiple machine rendering isn't going to cut it either, as most users are limited to one or two machines. Additionally, it's a false argument that somehow all this is made better by having a "true ray tracing" engine. Everyone has their favorite engine and we can argue about the quality of each program's product, but Bryce's is hardly the end all in this respect. IMHO one major mistake with Bryce from the beginning was the absence of a Treelab. (As if landscapes were somehow divorced from trees. ) This has been fixed to a degree (no grass). The remaining major mistake is not having a rendering engine equal to the other components of the program. Maybe this will be addressed in the next upgrade, but I doubt it. manowar (This should keep us busy for the next day or so:)
Thank you Man O' War, for your totally misguided post. Not only did you attempt to dismiss my concerns as frivolous, you indulged yourself with your own 'mucho angry' bitching subject. Very entertaining. Pointing out that Bryce5 may have been optimised for the Mac has nothing to do with 'platform wars', and neither does the existence of platform exclusive products. Corel could have released Bryce5 for the Mac only and it wouldn't have caused me to start frothing at the mouth and lambasting Steve Jobs. It's all a question of informed choice. If Bryce5 really is slower than Bryce4 on the PC, yet faster than Bryce4 on the Mac, then this is an important issue for potential purchasers. Whether the speed difference is by accident, neglect or is deliberate, anyone considering Bryce5 for the PC is entitled to take a dim view. Anyone considering Bryce5 for the Mac is entitled to be delighted. As a PC user, I am obviously concerned that I will be buying something that renders significantly slower than my current version, and that the program makers are possibly focussing the program on a computer platform I don't use. It is very early days, and there just aren't enough results to form a valid conclusion as yet. Hence all the 'ifs' and 'maybes', and the waiting for more information before proceeding. So far though, the fact remains that every comparative report I've seen between Bryce4/5 on the PC gives a slower result for Bryce5, while every comparative report I've seen between Bryce4/5 on the Mac gives a faster result for Bryce5.
It may well be that there are unique characteristics of the Mac OS/hardware that allowed the Bryce team to provide faster rendering than on the PC. If so I see no reason why the speed of the Mac version should be deliberately held back to the same as the PC just to make me feel better. That would unfairly penalize Mac users. But the fact is that I use a Windows box, not a Mac. I may feel pleased for Mac users but how fast or slow Bryce runs on their machines is largely irrelevant to me or my Bryce usage. That B5 will render slower than B4 does annoy me, unless it is also doing more work which will provide superior results.
TomD, this isn't an issue about the speed of computers (PC vs Mac), it's an issue about the speed of Bryce. I'm not talking about comparing the same scene file across platforms; I'm talking about comparing the same scene file on the same computer! The experiences reported in various places so far follow this pattern: Mac user renders Bryce4 scene in Bryce4, then renders same Bryce4 scene in Bryce5 for comparison. Scene renders significantly faster in Bryce5. PC user renders Bryce4 scene in Bryce4, then renders same Bryce4 scene in Bryce5 for comparison. Scene renders significantly slower in Bryce5.
Ah, gotcha. Wasn't clear to me. Thanks for the clarify. However, we know that Macs and PCs have different graphics architecture. Its very possible that the Corel engineers were able to squeeze some additional rendering speed out of the Mac architecture that they could not out of the PC hardware. (I suppose it could also be an OS X issue...) Don't know, I've always understood/believed that "Macs were better for graphics", if that's the case then I'm happy for the MacHeads among us since I very strongly doubt that Corel would make a deliberate business/technical decision to optimize for the Macintosh platform (though, as everyone knows, stranger things have happened...) Thanks for the clarification. :-) TomD
Tom: Thinking that Macs are "somehow" better at graphics is a common misperception ... and fantasy of Mac users worldwide. The truth is PCs do it better (or as well), faster, and for less money then any Mac setup. I'm always amazed at the incredulity of some people when they find out we're doing their video comps and other graphics work on all PCs. However, our clients don't seem to mind when they get the invoice.
OS X is the key here. it's Unix based and that is a very smart move for Apple.
You can't call it work if you love
it... Zen
Tambour
Â
Not realy, I don't own any Apple products, I just know that Unix kicks butt.
You can't call it work if you love
it... Zen
Tambour
Â
No Mac vs PC battle here at all. Peronsally, I've used both and maintained both at various times doing various things. There are graphically optimized elements in current Macs, however, that might contribute to an increase. It could also be the OS. It could also be magic. :-) I think blaming it on magic makes the most sense. ;-) Tom D
I think we can all agree UNIX is pretty awesome, and that was definetly I nice move by Apple. It gives me a little hope for them. I've had strong doubts ever since they insisted on having a single mouse button ... what the heck were they thinking? I think they finally joined the 20th century and added dual-button functionality now ... right?
With the reference 'Wenger' benchmark scene: On a 1.3Ghz Athlon running Win2K: Bryce4 - 212 seconds Bryce5 - 239 seconds Difference: 27 seconds - - - - - > Bryce5 over twelve percent slower. On a 700Mhz PIII running Win98SE: Bryce4 - 500 seconds Bryce5 - 546 seconds Difference: 46 seconds - - - - - > Bryce5 over nine percent slower. I consider approximately ten percent slower render times to be significant, and a serious disincentive (for me) to purchase the new release. Compared to the seven to eight percent faster times reported for various Bryce4 scenes rendered in Bryce5 on the Mac, I'm afraid it becomes rather more than a storm in a teacup.
jval, it has already been reported by some that the numbers of ray hits and misses are the same in both Bryce4 and Bryce5 renders, on both the Mac and the PC (read my first post). This implies that the render options are the same in all cases. Refering to your earlier point, it is indeed possible that the programmers have coded Bryce to take advantage of some aspect of the Mac system. In itself, I have no problem with that. But if that optimised coding for the Mac is what is causing ten percent slower times on the PC, then I feel it's something worth talking about.
Thanks PJF. I agree that identical counts imply identical setups but you know about implications... But there are other differing factors that may affect things such as antialiasing, etc. Who knows? For reasons I already mentioned the Mac/PC thing is a nonfactor for me. However, if Corel optimized it for the Mac and could have but did not optimize it equally for the PC, then I would be irritated. Corel is trying to infiltrate the Mac market so I suppose it is possible they have done so to "score points". But such behaviour seems passing strange to me and I do not think it likely. (Seems also strange to me to be supporting Corel here as I have no respect for them at all. In fact, I seriously considered foregoing the upgrade just because it is Corel's.) |
Do you guy's think Corel would tell us exactly what's up? Because if they optimized it for one platform, and not for another and didn't inform buyers, then that would be a serious conspiracy to the buyers of Bryce 5. I'm getting the hang of setting the render options, and for the normal anti-aliasing settings it's not too bad as far as speed goes. I'm not as disappointed as I was last night with iot when I brought it home from COMPUSA. kaom
Kaom, you could always ask them and find out. But suggesting a conspiracy seems a little premature. Aside from network rendering and architectual imaging Corel does not appear to have made any mention of rendering speed. The features they do claim seem to be present. I'm still waiting for my upgrade so I can't comment from experience. But in the end all that matters is does Bryce 5 more or less live up to its advertised claims. If so I will not feel misled. (But I'm still not at all certain the additions justify the rather large price increase. In comparison Eovia gave me a sidegrade from another product that included 2 3d modelers and a fistful of plug-ins that used to cost extra- all for $150.)
do you know how silly it is to say "they conspired" guys, you're being a little extreme arent you? why in the name of hell would a company do something like that? there are program differences between 4&5 and platform differences between mac and win. Did you know that bryce is a "patchwork" program? the code is not "ordered" as a normal program is. Its just patched together because of the additions to code without a solid rewrite. It was NEVER intended for what it is doing now. You, Me, We as the consumer have pushed it into realms it was never conceived of going. Additionally, I 'assume' that this was the same scene with same setup. Rem, b5 scenes are larger in resolution by default than b4 (odd how its ~10%). additionally, if you are going to declare a 10% difference, you dont say "seconds" you count "hours"..then give a percentage from that. I do have gripes about it, but IMHO..this upgrade IS the best upgrade bryce has ever received. It now supports network rendering..this alone means that it life will probably not ebb into inexistence...add trees, metaballs, and user categories and you have a big improvement. There are things that could still be improved. And render times will ALWAYS be one. But , to even hint that they deliberately made the difference between platform? come on... BT
jval, your right Eovia did offer a killer deal on the upgrade to Carrara Studio(with Amapi5 and all the plugins), for original Carrara users it was only $99. I got my Bryce 5 upgrade for $99 which is better than the advertised $159. In my opinion it does live up to what was advertised. All in all, the Bryce 5 upgrade is not a huge disappointment, the render speed could definately be faster. But consdering the Tree lab, Depth of Field, Render options, and increased grid size for Terrains and Lattices it's not bad. Unless I run into any huge flaws or bugs, I can't bitch too much. I doubt there is a conspiracy with the platform issue, but you never know. I think it's just an issue of the possibilty of it truly running better on a Unix based platfor(Mac). I am a PC user so I can't really say one way or the other about the Mac version. At least we have a new version of Bryce, it's better than it being abandoned. I had to wait a year and a half for a patch to make Carrara work at all, and I paid $400 for it, so things could indeed be worse. kaom
posetech wrote: >>>>>>>>>>>>>>> additionally, if you are going to declare a 10% difference, you dont say "seconds" you count "hours"..then give a percentage from that. <<<<<<<<<<<<<<< What the hell are you rambling about? It's a ten percent difference whether the time units are given in nano seconds or centuries. In this case, seconds were clearest and most informative division for the times involved. >>>>>>>>>>>>>>> Additionally, I 'assume' that this was the same scene with same setup. Rem, b5 scenes are larger in resolution by default than b4 (odd how its ~10%). <<<<<<<<<<<<<<< I assume people actually read threads when they respond, but as you have demonstrated this isn't always the case.
PJF: "But if that optimised coding for the Mac is what is causing ten percent slower times on the PC, then I feel it's something worth talking about." I missed this point in one of your later posts the first time around so now your comments make more sense to me. It does seem a rather large assumption though. Do you have reason to suspect this or is it simply speculation?
I agree, jval, it would be a big assumption, which is why I'm not making it ("if"). Glad you're taking the time to read my posts; I wish everyone would. ;-) What I see is a difference in the reported render speeds for Bryce4 scenes rendered in Bryce5, between Mac and PC. There doesn't seem to be an obvious explanation for the difference, so yes, I'm speculating. Bryce is supposed to be written in a code (called something like 'Axiom') whereby the two platforms it is written for get virtually identical versions of the program. If these various benchmark reports are accurate, then clearly that system has not worked for Bryce5. Whether it's the two versions treating default settings differently (somehow hidden from the render report), or whether there is something beneficial in the code for the Mac (or some other reason); the reports so far show that there IS a difference. I've already spoken out in favour of Bryce5 in other respects, and I want to buy the program. But I'm not going to buy until I have a satisfactory explanation for this difference between the platforms. I shall ask on the Corel discussion group (though someone already has and no answer was forthcoming) and hopefully get something meaningful. In the meantime, nothing would be more helpful than render speed comparisons, especially from Mac users (not enough reports so far for reliable conclusions).
PJF: I just downloaded the B5 demo from Corel's site and I am shocked to put it mildly. Here's a post I sent to the Bryce Forum group. I am getting very different results. ================== I am still waiting for my copy of Bryce 5 to arrive. Meanwhile I have downloaded the trial version from Corel. Several people have complained that it is slower than Bryce 4 so this is the first thing I checked. I must say that I am hugely disappointed! While others are complaining that B5 is about 10% slower on the PC it is much worse for me. I'm hoping I am overlooking some stupidly simple point. First the conditions: I am using an IBM notebook, PIII 650 mhz with 192 ram. I tested with 3 B4 files. The first contains a large irregular glass object. B4 renders a 400x300 in 7:57 while B5 takes 16:44. That's a time increase of 110.48%. The second contains very little transparency and is 400x300. B4 takes 1:59 and B5 takes 3:14- a 63.02% increase. The last one has no transparency at 580x480. B4 is 0:19. B5 is 0:37, 94.73% slower. I noted that the reported render statistics are not the same between B4 and B5 outputs. Each was rendered with normal antialiasing. I also noticed that accessing B5 menus is slightly sluggish compared to B4. I am particularly concerned that my time differences are so much poorer than others have reported. Perhaps the demo version, aside from being a demo, is not the same as the shipping version. As it stands even if I use network rendering with 2 computers I will still be substantially slower than when using B4. Anybody have any suggestions or comments?
Has anyone else besides me noticed something interesting. On the side of my Bryce 5 upgrade box, there is an interesting little graphic. Right next to the Quicktime logo, there is a small circular logo that says-"quote", ("Made for Mac OS-X"). Hmmmmm... I'd like to know what other users think about that. There's no logo that says made for Windows(?). It could just be an advertising thing, or it could literally mean that the program is designed specifically to run optimally on OS-X. The story gets more interesting. kaom
Further to my complaint above referring to the extremely long B5 render times compared to B4: As unbelievable as it sounds the culprit was the Bryce 5 trial version. I now have the actual shipping product on my PC computer and the B5 rendering times are about what I expected. They are slightly slower than B4 (maybe 5%) overall but a few of my B4 scene files actually render faster in B5.
Thanks for the info, and thank God for that! I've been using the demo and it's been terrifying how slow it is. Seems like a bit of a dumb move, from a marketing point of view. It strikes me it would have been better to slap a watermark on the image or save disable it or something. At least tell people that the demo is slower! I wonder if this is just a PC thing, too. Maybe the Mac demo is that much faster again. ;-)
I don't know if this is just a PC thing or not, PJF. My test with the demo had many of my B4 files taking more than twice as long to render with no visible improvement in the image (although I did notice other minor differnces in some of the image renderings). Interstingly, the image that rendered much faster in B5 contained an awful lot of transparency and reflection. Yet the speed improvement in B5 versus B4 was better than the improvement people have reported for their Mac versions. I really have not done enough comparisons to offer opinion yet. But it is possible that the speed differences may be due more to the type of scene being rendered than to hardware platform differences.
Privacy Notice
This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.
On the Yahoo-groups Bryce list, there has been some discussion about a difference in render speeds in Bryce5 between Macs and PCs. Over there, Dan Whiteside reported the summary of his investigations: "...I finally got my release copy and loaded it up on my side by side Mac/PC setup at work. After seeing the benchmark results, I wasn't at all surprised when a couple of different PCB5 scenes came up about 4-5% slower then PC B4 while the Mac turned out to by 7-8% faster. Interestingly, all the render reports (PC B4/5 and Mac B4/5) showed the same Ray hit/miss results." I read from Dan's words that for the same scene, Bryce5 is faster than Bryce4 on a Mac, and Bryce5 is slower than Bryce4 on a PC. The logical conclusion is either that Bryce5 is optimised to run on the Mac platform, or that something is different about the way default render settings are configured in Bryce between the two platforms. Given that the ray hit/miss results are the same across platforms (which implies the render settings are the same), I'd have to lean toward the notion that Bryce5 is straight out optimised for the Mac. That's enough to stop me buying it. I was going to order it this weekend, but shall now wait until more information is forthcoming about this issue.