Why is so painfully slow?

[Per-Anders]

If you are finding you're eating into virtual memory then you'll want to increase your memory to the point where cinema when rendering does not have to use any virtual memory, these days i'd say a gigabyte is the bare minimum so you may need more, however given the scene you mention that is likely not the problem (unless you're using many hundereds of thousands of photons).

9.6 is actually most of the time much faster than 8.5 for rendering, however rendering is always a slow process depending mostly on the speed of your CPU, caustics depending on how many you have can take a very long time to render correctly and while setting up only use a single thread afaik (so multiprocessing wont help in that situation) even with the fastest machine, but if you want faster rendering then fundamentally a better CPU is the way to go.

[lllab]

...and a semperon is a very slow chip...
cheers
stefan

[schum121]

You need more RAM. Not cache memory, not a better cpu, not anything else... RAM.

It's rendering in RAM and your graphics card is displaying it. You basically need more capacity (IE more memory), doesn't matter how fast your computer is if it can't hold more.

You need at least 2 gig RAM for 3d software editing. I'd recommend 3.

[LucentDreams]

You need more RAM. Not cache memory, not a better cpu, not anything else... RAM.

It's rendering in RAM and your graphics card is displaying it. You basically need more capacity (IE more memory), doesn't matter how fast your computer is if it can't hold more.

You need at least 2 gig RAM for 3d software editing. I'd recommend 3.


Hahaha depends on your type of work I usually do fine with just a gig, and I've done a lot of complex projects with only 512 in Recent times too. Not to say that I don't doubt ram would help, but they guy is using a semperon, I mean a low end CPU, thats going to add a serious speed hit compared to someone using an athlon, I mean rams only going to hurt when you actually start hitting its peak and going to VM.

[schum121]

you're right... a new cpu is going to help in processing the data. a new cpu will cost shite loads, then ull hav to fit it and worry about mother board and heatsink compatibility, or worse get some1 else to fit it and charge u a bucket load of cash.

more ram will do the trick

[dann_stubbs]

I am wondering compared as to what it was in Cinema 4D 8.5 the renderign in Cinema 4D 9.6 is a P.I.T.A.

If you specially try to use caustics only it takes like 5 minutes to render a 720x480 frame with only one object and a glass shader with fresnel activated. And only two lights.One spot with soft shadow and one directional.

I am using a AMD Sempron 3000 processor 64x with 1 gig of DDR PC400, Nvidia quadro FX 500-600 128 mb and Windows XP Pro.

Any comments respect to what to do to accelerate C4D would be appreciated.

Thanks in advance.

the default render settings were increased a lot in the 9.x compared to 8.x and earlier

compare the render settings for rays, shadow rays and reflection rays between your scenes and lower the 9.6 if they are still at defaults - it should bring the render times much closer together

dann

[Kuroyume0161]

I agree with everyones' responses here. ;)

More memory - yes!

Faster cpu/bus - yes!

Render settings check - yes.

On the cpu, it really depends on the mobo manufacturer and the processor. I just updated an AMD64 3500 to an AMD64x2 4800 (socket 939) without any problems whatsoever. Worries about reinstalling Windows x64 or other requirements were just that. The BIOS complained - saved the settings. Windows 'found new hardware' - rebooted. Now I have a dual core that is much faster than previously. I tend to stay far, far away from the 'el cheapo' processors (semperon, duron, et al). These are excellent for someone who surfs the net and uses Word. :) For 3D CGI, use the best, forget the rest.

Robert

[bobtronic]

Actually the later revisions of the Sempron are not that bad. They have more cache than the earlier revisions and that helps alot. I have a Semprom 3000+ (2 GHz) too and so far I am quite satisfied. On the memory issue, more memory will only help if you are working with large scenes which hit your memory limit and the OS has to switch to VM. I am doing fine with 1.5 GB RAM.

cheers,
Bob

[georgedrakakis]

I am wondering compared as to what it was in Cinema 4D 8.5 the renderign in Cinema 4D 9.6 is a P.I.T.A.
i am really sorry to hijack this thread, but what P.I.T.A means?
once i knew a site that explained acronymns, but i've lost the bookmark.
tia,
george

ps: don't bother to answer, just figured it out, but unfortunately i had already pressed the submit button

[Kuroyume0161]

But how much does the Sempron cost for what you're getting - about $100? The AMD64 4800 used to be over a $1000 - but I just picked one up from EggHead for $280. Can't beat that very easily for a dual-core with all of the extra power. :)

I do agree about the memory - and there is the hard limit for 32-bit of 2GB so more memory won't help beyond that. My feeling is that it's processor power coupled with render settings more than anything if the memory isn't being pushed to VM.

[TheCore]

Time is money, and when a still is treathening to use 3,5 hours of rendertime on my comp - I just send the job to Rendercore.com. They got a renderfarm with something like 500 Mac Pros ready to burn some asphalt. Those comps are hyperventilating. I did a test - rendered a c4d-file for 3,5 hours -urkh, and then sendt it over to Rendercore - and had it done in 4 minutes. I mean... FOUR MINUTES!!? Thats insane. The price was aprox 10 USD, very nice I think.
Use the renderfarms, it pays off. Not all of them supports Cinema 4D 9.x - however, Rendercore does - and have great support.

[Primitiv]

I figured it means Pain in the Axe.

[chromecity]

I think Dann is onto the real issue here. The person who started the thread said 9.6 was much slower than 8.5. That's the key. The way this issue was described, it doesn't matter which processor, how much memory, or any of that. If it's the same for both, then 9.6 should not be running significantly slower than 8.5. Check those render settings, indeed.

And for the record, 2GB is not a hard limit for Win XP Pro (SP2). If you have 3GB, you can use the 3GB boot.ini switch (referenced several times elsewhere in this forum) to make more memory available for C4D.

EDIT: I now see that Kuroyume0161 knows the benefits afforded by the 3GB switch from his participation in some of those other threads - and he even has 3GB (okay, actually 4GB) as well, so I don't know why he would have said 2GB is a hard limit. Have you now encountered something that prohibits using the 3GB switch, Robert?

[dann_stubbs]

I think Dann is onto the real issue here. The person who started the thread said 9.6 was much slower than 8.5. That's the key. The way this issue was described, it doesn't matter which processor, how much memory, or any of that. If it's the same for both, then 9.6 should not be running significantly slower than 8.5. Check those render settings, indeed.


i was surprised too - how the clear description of 9.x being slower then 8.5 suddenly turned into hardware arguments.

i remember when R9 came out there were quite a few threads along this line due to the increase of default render settings (which from what srek said were increased due to the complaints about lower quality from the previous lower render settings)

now with R10 having changed it's default GI render settings i guess we'll get another round soon except GI related.

dann

[LucentDreams]

i was surprised too - how the clear description of 9.x being slower then 8.5 suddenly turned into hardware arguments.
dann


One would assume when comparing rendertimes one would be using the same file, thats how they got confused. Gee my million polygon sphere renders slower than my 100 polygon sphere.

First thing anyone should do is use the same file, then settings are guaranteed to be no different, His scene coudl have easily have been faster as you and I both no that betweeen 8.5 and 9.5 there were several speedups all over, so the only possible slowdown woudl be related to the known shadow maps maps issue. Any other differences woudl be because one isn't using the same file.

Its like your calling some people dumb for misunderstanding someone elses dumb mistake.

[vancliff]

I was surprised and delighted too with all the answers here. Some are pro and other are cons, but anyway my debate was about the difference in rendering with 8.5 compared to 9.6 in its defaults.

As one of you suggested and i wasnt aware of that was because of the many complaints in rendering quality between both. But my thought was and is still valid, why making the rendering slower than early versions, it should not improve with better versions?.

I have now a better pc than before. and tough it was athlon based it was in the 32bit and a sempron 64 i guess, correct me if i am wrong, is better than an athlon xp+ 32 bits. And 1 gig of DDR PC400 is better than DDR PC133.

Thanks to everyone who participated in this thread, i have learnt a lot.

[chromecity]

...but anyway my debate was about the difference in rendering with 8.5 compared to 9.6 in its defaults.

As one of you suggested and i wasnt aware of that was because of the many complaints in rendering quality between both. But my thought was and is still valid, why making the rendering slower than early versions, it should not improve with better versions?.What some of us have been trying to get across is that the rendering time with the same render settings has indeed improved in most cases. It is merely the default render settings which have been altered so that a default rendering produces a better quality rendering. If you alter those settings in R9.6 to closer match the settings of R8.5, you should indeed get faster rendering in most cases. It's trivial to save your render settings, so it should be quite easy to experiment with this.

[LucentDreams]

I have now a better pc than before. and tough it was athlon based it was in the 32bit and a sempron 64 i guess, correct me if i am wrong, is better than an athlon xp+ 32 bits. And 1 gig of DDR PC400 is better than DDR PC133.

Hmm looks to me like the CPU assumptions may have been more accurate than some though.

64 bit can add some speedup, especially in memory heavy scenes, but the fact that you've gone form athlon to Semperon means you''ve sacrificed a lot of speed too. Semperon's and Celerons are not ideal for 3D work, not saying they cna't be used, but they are very inefficient.

For an example, a Semperon 2600+ runs at 1.83 internally and 333 externally, while an athlon 2600+ runs at 2.08 and 333.

I'm also not sure your semperon is 64 I haven't heard of a 64 bit semperon, and you'd only see the benefit running XP pro 64 qith 64 bit version of Cinema 4D. Running 32 bit windows on XP64 woudl be slower and running 32 bit cinema on 32 bit windows with a 64 bit processor will see no speed advantage.

[chromecity]

Yeah, this particular comparison really seems like apples and oranges - there are several variables at work here that all have an impact (okay, no real Apples here, but you know what I mean). I think at some point he is going to have to either try both versions of software on the same machine or else just take people's word for it that things have generally sped up in 9.6 in terms of overall render speed. Of course, if you jack up the quality settings (as has occurred with some of the defaults), there's an understandable trade-off for some of that speed.

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.