AbnRanger

Yeah, if Intel is able to get their CPU's to OC to 5Ghz on their 14nm chips, I don't see it being that tough to do for AMD on 7nm processors. Looks very promising. 16cores at 5Ghz would be amazing.

Glad it helped.

Wow. That's some serious price gouging.

When using Auto-Retopo you have to keep the K.I.S.S. method in mind. It's an algorithm and you are throwing the kitchen sink at it with thousands of undulations and telling it to spit out a rather simple mesh from it. The best option is to create a dummy duplicate of the model, hit SMOOTH ALL quite a few times and try it, then. If the result is good on your dumbed down version, then you can simply re-snap that to the higher detail version.

I'm in total agreement with you, there. Truth of the matter is, Andrew has only started working in the Sculpt Room after a long hiatus from it. The whole app would be about 2yrs ahead if Andrew had not been convinced by a few people, who probably had never used 3DCoat before, to create offshoots. This is why I'm a bit upset to hear of yet another one....right in the middle of Sculpt Layer development...which seemed to stall for a few months. Now, we know why. For a SHOE Design version? Really? This cannot be done within 3DCoat EDU or Professional?

You don't see ZBrush being split up into disparate parts, nor Mudbox. Why does 3DCoat, the smallest company of them all need to waste precious development time and resources to do so? If you only want to use part of an app, don't ask or expect the developer to create an offshoot to cater to your needs. Just use the app for what you need. 3D printing could easily be done with the EDU version. Why then does it need to be split into an entirely different app? It doesn't. This has me really upset because it means all these requests for features in 3DCoat, over the years, go ignored for a few more just because of this needless diversion. My goodness, 3ds Max has users from various industries, and it's not split up. I've never heard of such nonsense.

This is just my personal opinion, but Andrew has already spent needless time acquiescing to similar requests to split the app up and It's a horrible idea, as far as I'm concerned. I don't mean to step on any toes, but the reality of it is that what may seem like a good idea, may actually turn out to be a terrible one. My contention is that for Pilgway, the latter has been the case. I don't think any of these offshoots have paid off. For every new offshoot, they would need marketing, training and separate support. The company is just too small to handle it and as I mentioned before, it isn't financially viable for them to bring more staff resources to do so. It has utterly killed a few years of development on the app because Andrew had to divert his attention from 3DCoat, towards the different versions. PlastyCad, 3DC Printing, 3DCoat Modding Tool, and now 3D4Shoes (I just heard about this). Enough already. This explains why Sculpt Layer development has seemingly stalled. A few years ago, Andrew put Sculpt Layers off because he said he promised users he would redo the Layer Masks in the Paint room. That never got done. You know why? Because he was working on 3DC Printing...all because someone convinced him it would be a great idea. It wasn't. There are a number of things I've repeatedly asked for over the years, and it gets put off because of diversions like this. Performance in the Paint Room is one. Including sliders. They are basically useless in 3DCoat because they are so slow. I've asked Andrew probably 20 times about that over the past 4-5yrs. PRIORITIES. 3DCoat users should get it. Not people who don't even own/use 3DCoat yet. That's my 2 cents worth.

After some further testing on an older Intel i7 970, it seems the behavior is roughly the same as on the AMD CPU. I used the same model and same texture size (4k, not great on 2k, either). Unless you keep your brush confined to a small to medium size, performance around pixel dense areas like the face, is just terrible. It's not something you would expect in a professional texture painting application. For this reason, I really do hope a GPU brush engine is under current development and not shelved. 3DCoat really needs to step up it's game in the Paint Room. The Sliders for DEPTH, COLOR, GLOSS in the layer panel are so sluggish they are literally, not figuratively, useless. This has been an issue for 3DCoat from the beginning.

True, but that also means performance will still be crippled on AMD CPU's in the rest of the app, including sculpting where multi-threaded performance is critical. He really needs to change this, as there are many non-intel compilers available. Him choosing Intel's compilers means users with AMD CPU's, like myself, are getting hosed (in terms of performance). This severely limits one's purchasing options when building a new system or upgrading one's CPU. Even a new Threadripper cannot benefit from all the extra cores and threads..because Andrew chose a compiler from a company that plays dirty tricks.

That's about what I thought. AMD CPU's are pretty much crippled in 3DCoat, wherever Intel TBB is used for multi-threading. Andrew really needs to do something about that.

Andrew removed them a few months ago, thinking they were redundant, since you can LMB + SPACE bar on a widget to enter value numerically. But some of us here asked to have it put back. He did in the latest build, it seems. By the way, can anyone with a relatively new Intel CPU do a quick painting test? Can you take a model with a 4k UV texture map bring up Windows Task Manager and see if it's using all cores. I've got an AMD Ryzen 2700X, which is a pretty beefy new CPU that beats the Intel 8700k in most non-gaming benchmarks. However, Andrew used INTEL TBB (THREAD BUILDING BLOCKS) for the Multi-threading code and in a test just yesterday, it was awfully slow painting in the head regions and only one thread ran close to peak. Total utilization of all cores was roughly 20%. That's terrible for such a common task in 3DCoat...that is supposed to be multi-threaded. This was the case several years ago, but I thought AMD won a case against Intel for this very practice, of crippling AMD CPU's in apps....like 3DCoat...where their compilers were used. This is the problem of using proprietary compilers of Intel. They play dirty and have been sued relentlessly for it.

PhoenixFD uses only the CPU for simulations, as does 3ds Max's Bifrost fluids. Of course, Arnold also is currently only CPU based (but they are still developing a GPU engine for it), so it can help make rendering with that engine much more tolerable and feasible. As for 3DCoat, it's heavily multi-threaded, too, but I don't know how much that many cores will benefit. There is probably a threshold of diminishing returns, involved...and I am still concerned a little, that with 3DCoat's multi-threading code using Intel's TBB (Thread Building Blocks), there could still be a bit of Intel "crippling" of AMD CPU's. IIRC, AMD took Intel to court to stop that, but who knows to what extent Intel ever complied? There are still many calculations that are single-threaded, by necessity I think. I'm using a Ryzen 2700X, now, and it's a stout performer. I use Vue from time to time, and it's main render is CPU based, so it's hard to get away from the need for more CPU performance

I'm totally happy with the 1080 for now, and I usually don't upgrade to a new card unless there is a double performance gain. Seems a waste of $$$ to upgrade for just 10-25% improvements. Having said that, I stumbled on a Newegg Black Friday deal with a Gen 1 Threadripper 1950X for $450 USD. I had to jump on that. Was going to wait til next year to see what transpires after AMD hits the 7nm mark on the CPU. But this was too good to pass up. I just hope 3DCoat is able to leverage the extra cores. I remember Andrew stating in a recent build that memory usage in the Paint room is much better, so it did a little test and it seems 3DCoat isn't really putting more than one or 2 cores to work in the most recent build. Will have to do some more testing to see if it's an anomaly or what.

Yep. If AMD could get their new models out soon, they could take advantage while Nvidia is vulnerable.

3DC Printing is largely the Sculpt Workspace in 3DCoat Professional, but it has a unique viewport and a Supports tool, that is needed for 3D Printing. I am trying to see if I can get Andrew to roll the Supports tool into the 3DCoat Professional version, so users who already paid $380 don't have to spend another $50 just to get that one feature.