Recently doubled my processor speed. I should be happy. Then on my latest render set for superfine. It states 9 days + on a 1024x768 render. I use to render at 1600x1200 but pics too large for some people. Had to drop to 800x600 just to get it under 5 days. Is there no end to it. At this rate, in 2 more upgrades and a 10gig processor it will take 3 weeks to finish. Not there is no logic. Ok, im finished ranting now. haha.

Best test is to write down an exact time from a render with your old system, then when switched over, time it with the new system. That's how I know I'm rendering about 12 times faster than before. No way can you have a 2 times faster cpu and be rendering slower than before, just not possible, lol. You do have some of the most complex scenes around. AgentSmith

Oh yes it is haha. My last 3 or so pictures are massive glass things with no refraction (or its 4x worse still). About 750,000 polygons each so bryce goes to the default 6 shadow depth on these. Problem with Xfrog is all those little horns merge and bending each other in the center of those models and bryce has to render the 90% transparency each level in. Sadly id like to see what would hapeen if you set it for 20 on the shadow depth, but im sure the render would get exponentially worse. At 6 deep and 90% transparent... what im thinking your seeing is the default color at 53%. Nothing else is resolvable unless i up the depth for shadows. Sigh, sooner or later ill find the money for 2 or 3 dedicated systems. Probly later haha. Btw, did do a rough test... had taken 16 hours for a gallery picture 'orbit over the hive mind'. This p4-2.533 took about 8.5 hours. Hey wish i could take all those rack mounted pc's home from work. Think they had 40 of them in boxes for a project.

Bryce doesn't support dual processors, for Mac or PC.

i just built my newest computer about a month ago, p4 2.4.. msi board too.. my old p3 550 couldnt touch this computer. my new board supports overclocking and has a great app to jack it up... so i did a test at 2.4ghz with an image with a render time of 30 minutes'ish... i re-rendered the same one at 2.8 ghz and lost about 3 minutes of the total render-time. i never use volumetric world and the content wasnt terribly render-hateful but i was going off the total render, not what i was rendering. as long as the same image was rendered, (after a re-boot), i thought i would get a resonably close aproximation of the differences between my two speeds. not bad i guess but i wouldnt be overclocking my processor for two weeks getting some 6 second animation finished i can tell you...

Just curious AMD or P4. I have used both and find that AMD is still superior. I have finished doing some testing with a Intel 1.7 and 2.5 and my own AMD 1.4 and the AMD was just behind the 2.5 but far above the 1.7.

Actually it was a 1.2gig amd before renaming that was about equivalent to a 1.6gig or 1.7. Got a 2.533 gig p4, would have expected 50-55% increase. But my old render took 16 hours on the 1.2 gig and on the new processor 8-9 hours. Its the transparent 1-20 layer thick glass thats doing me in... but once rendered, it gives a xray look as each layer adds to the 90% transparency to make it darker in the middle.

you could use one render with only one layer of glass, then render another with a volume texture that gets less transparent towards the edges and use both of them to composite as x-ray-effect. Or you could turn of refraction/reflection for 19 of the 20 glasses. That should speed up rendertime.

Mmm, well i switched this summer from a Celeron 400 with 64mb ram to a powerhouse with an MSI KT3-ultra board, a 1.8 Athlon and a gigabyte of memory. The renders are about 1200% faster. I know because i had a render on my old wich took 4 (240 min) hours exactly, on my new machine it rendered in exactly 20 minutes. Rendertimes are now on the rise again, but then i am doing things that i would not have dreamt of with my previous machine.

Clay, can't Bryce Lightning use the second CPU? So you essentially get a networked render on one machine.

Erlik: I heard that it is possible to use both processors using bryce lightning. I think you have to run both bryce and bryce lightning at the same machine and give the IP-adress of the PC as client.

Hmm, I think I just deleted my post, or something happened that was odd... anyway... pmoores - don't know if this will help, but have you tried rendering larger on a lower setting and then reducing image size? For example, where you're rendering an 800x600 superfine 16, instead render a 2400x1800 default AA and then reduce the size by a factor of 9; Unless you've got extremely intricate smoke or checkerboard effects, the result will be just as good (in most cases, better) and twice as quick at rendering small in superfine. (You do need to use proper AA'd reduction to get the best results, though.)

It gets close sometimes, the estimated time to remember is the very first estimate you get after the first render pass. `Course if you have any good amount of glass in your scene, forget it, it almost always wrong, just like ya said. Sounds silly, but...If I'm going to render a complex scene that is for example 800x600, I'll change the document size onscreen to 200x150 and render that. Whatever time that takes will be 16 times faster than what it would take at 800x600. So, if it took 30 mintues, I'll know that the 800x600 render will take 8 hours. *(since you would have to tile a 200x150 image 16 times to cover an 800x600 space). Like I said, maybe silly, but it's fairly accurate. AgentSmith

Just for clarification, I deleted my post in order to edit it and found it didn't return... I was saying that Bryce is sometimes way out in it's estimates, hence AS's reply!