robcat2075

Both of those CPUs have some overclocking potential. I don't know how much one might surpass the other with that.

Here's the way I see it... I can already map and texture any A:M shape I want. Or at least i thought i could. What does this unwrap program do that i can't get done now with A:M? Post me an example of something that won't work in A:M. I'm not being sarcastic or dismissive, I want to see what it is that I'm not seeing. Now, looking at this with serious intent... what would it cost to entice a programmer who is either familiar with that code and would do the work to familiarize himself with coding for A:M... OR... is familiar with coding for A:M and is willing to do the work to familiarize himself with that unwrap code? What would it cost to hire someone who could get this task done?

She carries it well.

Looking more closely at this chart and comparing the single core performance of the i7-3770K to the FX-8150 (chart doesn't include an FX-8350)... (Cinebench tests a CPU's ability to render an image in the 3D program C4D, probably about as similar to A:M rendering as a common benchmark will be) For our single-core Netrender purposes the i7-3770K ($300) looks to be quite a deal compared to the i7-3960X ($1000). We can regard these numbers as a comparison of frames rendered per unit of time. If the FX-8150 core can output .97 frames per minute then we could expect the i7-3770K core to output 1.66 per minute of the same scene. I could extrapolate the FX-8350 score to possibly be 1.07 based simply on the increase in clock speed. 8 AMD cores x 1.07 = a NetRender total of 8.56 (frames per unit of time) 4 intel cores x 1.66 = a Netrender total of 6.64 turning on hyperthreading in an intel CPU gets about a 25% increase in throughput for 8 logical cores over just four physical cores based on one forum member's benchmark results. 4 intel cores x 1.66 x 1.25 = a Netrender total of 8.3 So it's getting close to a draw between the Intel box and the AMD box However, if I were building a computer for me to work on (model, animate, test render) with A:M, I would go with the intel CPU since it would be about 50% faster for the single A:M process.

Good thinking... inside the box!

There's always going to be something new in the future, we hope, but what about now? What if we need to make a box NOW? Here's my initial try Intel: i7-3770K four/eight core CPU + CPU fan + ASRock Z77 Pro3 Motherboard combo $390 16GB DDR3 RAM $100 ATX Case w/ 500watt PS $60 Video included on MB $0 500GB HDD $30 Windows 8 OEM $100 total: $680 AMD: FX 8350 8 core CPU + CPU Fan + ASRock 970 Extreme4 Motherboard combo $319 16GB DDR3 RAM $100 ATX Case w/ 500watt PS $60 Basic video card $50 500GB HDD $30 Windows 8 OEM $100 total: $659 In both cases I'm presuming a DVD drive, mouse, keyboard and monitor will be borrowed from another computer long enough to install the OS and other software. i figure the rendering throughput would be similar between the two boxes because four intel cores can do about what six AMD cores can do. If hyper threading (8 logical cores) were used with the intel CPU that would probably bring it up to what 8 AMD cores do. The intel CPU has the advantage of being about 70 watts while the AMD is 125 watts Both these motherboards have room for an additional 16Gb RAM beyond what is spec'd above.

You can simplify your angle numbers by changing the rotation "driver" to Euler instead of the default Quaternion. Quat is the default in A:M because it avoids gimble lock in complicated rotations but for simple one-axis moves Euler numbers will be easier for you to understand. To change: Select the bone, Open the properties window, Object Properties>Transform>Rotate on Rotate and choose Convert Driver To>Euler Here is a bone being rotated on one axis from 0 to 45 to 90 to 120 degrees The three-number groups are, from left to right... interpolation of the curve through that CP. 1="default" (one of the below previously set as default. "Spline" is the default default.) 3=hold 5=linear 7=spline 9=zero slope Time. I believe this counts 30ths of a second. Any other FPS setting, such as 24 fps, will create fractional values. Angle in degrees 1=

Is there an open source version of this algorithm out there? One that doesn't require A:M to become open source itself, preferably.

About every year we speculate on the most powerful renderbox one could make for the money. One thing I find hard in scoping this out is that it's difficult to find up to date benchmarks of CPUs that test single core performance. That would be useful for judging Netrender usefulness since each Netrender node is a single-threaded process. At the consumer level it seems like the best AMD cores take about 50% longer to do something than the best Intel cores. Hard to find comparisons of server-lever CPUs like Xeon and Opterons that can work on multi-CPU motherboards. Anyone want to spec out a box for 2013? Lets say under $1000 for a box (or boxes) that can run Netrender nodes and be networked with your regular A:M computer. Overclocking is fair if that is easy for the user to do.

I give some details in this thread http://www.hash.com/forums/index.php?showt...48&hl=punch In a paint program I make a blank bitmap with some pattern to identify the edges. I wrap that on the mesh and I can see from the partttern where the edges landed (usually the back). I make the bitmap to be approximately the proportions of what i imagine the unwrapped mesh would fit in.

I can shed some light on a few of these... Although not the same, I find it convenient to do a cylinder or sphere wrap on a mesh, fine tune that in the UV editor and then take that to a paint program. After you have made at least one key, select any or all curves in the PWS timeline pane, then >Curves>InterpolationMethod>choose type This will make the current and all future keys default to your preferred interpolation method. You can give different curves different defaults with this method if you wish. You can still choose any individual key later and change its interpolation with >InterpolationMethod Interesting. You coudl make a feature request at hash.com/reports. I'll note that it is possible to make poses for models that hide/unhide any set of bones. You can manage clutter easily that way. Check "Enable OpenMP" in Options>Global. This helps speed up a few things in A:M like particle simulation and mangement of large numbers of Patches. However, not much of A:M lends itself to parallel processing. Martin's team gave it a serious try in the past and Steffen has been looking at it seriously in the present but, so far, it has not gotten great gains. Much like multi-core, this has been seriously looked at but no reliable results yet. There are people who have used A:M's MDD export feature to go to other GPU renderers but there are substantial limitations to that. For example other renderers don't understand A:M hair or materials. Steffen is looking into adding some fast OpenGL effects like Screenspace AO in v18.

That does sound very cool. I'm not aware of a formal release of the file format but like Sebastian said it's text readable. Start with one bone, look at that, and that will tell you most of it. If you get stuck, ask.

Happy happy birthday to Lloyd and Jason, dependable forum regulars, both! wfSkqMP-L-4

Can you embed that and put up a PRJ that does that shot?

That's looking great. I like the hair, too!

That resembles a bit some of DaVinci's "grotesque" caricatures which I had sometimes thought of trying to do in 3D but he didn't make matching front and side views.

Looks like he walked into the telescope one too many times!

There may be an empty folder left behind after you do this delete. Delete that manually in the PWS

In the properties window for the model there are "User Properties". Expand those, then you can >Delete on the ones you don't want.

The brochure refers to it as a "gated community."

Cool video, marcos!

I'm going to say he just wasn't renaissance-picky about perspective. He wanted to suggest some perspective but perhaps he had a bit of the cubist in him and wanted to reduce things to their most recognizable form. Those supports that jut out on the top roof would all have to be very carefully drafted to be perspective correct and i don't get the sense that he was into that thing.

As soon as I saw this I thought Rodney would be the ideal candidate.

Welcome back to A:M! We look forward to your future work!

The advice to break it into sub-models is the best path, but I'll note that if you have a multi-core CPU, checking Tools>Options>Global>EnableOpenMP should get you some performance increase on large models.