AbnRanger

It seems a lot like Mudbox during it's initial release...powerful in one key area, but very starved of many features you come to expect from an application in it's category, and thus I don't see it immediately replacing Bodypaint, ZB or MB anytime real soon, as it's benefits are largely aimed at a relatively small percentage of the overall market (film work).

The problem with relying on Windows swapping to pagefile is, you know how that bogs things down. It isn't hard to reach the physical RAM limit in 3DC. And when you approach that limit, like any other program...things get mighty slow. However, if we can cache hidden layers/parts to a temp file (perhaps in the 3DC directory) then we can better plan to stay well below the RAM limit, and avoid the lag of pagefile swapping. Otherwise, it's like hooking up a boat to your Pickup, and towing it everywhere you go...to work everyday...to the mall or theater. It's a colossal waste of resources. You tow it only when you need it. It may take a few seconds longer to load from disk to RAM, once you unhide a portion or layer, but that's a much smaller performance hit than having a hidden layer needlessly hording precious RAM and CPU resources, while you're working. Perhaps a small icon at the bottom of the layer/voxtree panel, that lets the user voluntarily offload the layer to cache, but automatically cache to disk immediately whenever something is hidden using the "Hide" or "Cell Hide" tool.

You can google search for those kinds of tutorials, but I think one thing you might find helpful, is using the Bucket FILL tool (in the Paint Room) on a few different layers, and using your blending modes (Normal,Screen,Multiply, etc) with opacity to get a lot of the subtle freckles and skin tone variations. In the FILL tool's parameter rollout, you'll see a section for "Modulation"...there you can see a number of different noise patterns and colors to choose for the patterns, plus the different parameters to adjust. You might glean something helpful with these tutorials:http://www.youtube.com/user/Fugazi3D#p/u/32/aSKrMO-132U

I try to save on a regular basis, and I noticed the big increase in speed/responsiveness immediately after a save. Things seem to really bog down over time, but returns to normal with just a regular save. You can use that file I sent you last week to test. Is it possible to have hidden parts and non-visible layers cached to disk, instead of letting it continue to consume RAM? I noticed when hiding and/or spliting parts to another layer and turning visibility off...the Physical RAM consumption does not change any. Caching those to a temp file on disk or the pagefile would help immensely. Thanks, Andrew.

Andrew, I noticed that when working in Voxels, the brush response gets ridiculously unresponsive after working in it awhile. Then, once you save the file, and resume work, the difference is like night and day. Is that a big memory leak or something?

I thought I was the only one who thought Wes (at Kurv studios) consistently drags his feet getting web training to DVD. 3 months have transpired since Leigh first broadcast this series.His answer is to pay him ahead of time for webinar re-broadcasts he knows many can't attend...and when he's good and ready, he'll provide a secret download, that you can't purchase after the broadcast. I don't know of too many people that are interested in floating their cash out there, waiting weeks or months on end for someone who has shown a persistent habit of not delivering or doing so very, very late. Then he'll try to milk the re-broadcasts til they're dried up and then maybe release a DVD when he gets a wild hair and decides to get around to it (months and months later). Leigh's stuff is top-notch. Too bad the distributor he chose, isn't. There isn't a worse training resource, IMHO, than Kurv. The content is fine...it's just the delivery. There are multiple "free" training broadcasts (covering 3DC) by both Leigh and Javis that are not available on the Kurv site...even though he said he would make them available. My name is Don, and I approved this message.

Why? I noticed when I was trying to do some quick mid-detail work on a model before retopologizing it and moving it to Mudbox (for high detail), that even after I split the model into parts, thinking that it would help...I realized that 3DC was just as slow with layers hidden or parts hidden. I had Windows Task Manager open and noticed that the amount of Physical RAM consumed hadn't changed. The question I have is...Why? Why isn't all hidden portions cached to the pagefile? Even earlier versions of Mudbox would allow you to offload sculpt levels to cache (I suppose a Temp file in the Mudbox Windows Directory). This, along with non-multi-threaded merging calculations, is so in-efficient.

I think it's perhaps a matter of you having to manually open the 3dconnexion Control Panel when using 3DC. When I use Max, it automatically recognizes the device and turns on program running in the background (you can see the little icon in your system tray in the lower right hand part of the screen. If you don't see that when 3DC is running, then it's not running. Although 3DC supports the devices, it doesn't have the deep connection that the major 3D applications do.

Open the 3dconnexion Control Panel > Button Configuration (tab)...and check to see that they are correctly assigned. They should be by default, no matter what application you use. If they work in other apps, then you know it works, mechanically...just something on the software side is funky.

Very insightful interview with folks from Disney about Ptex... http://media.fxguide.com/fxpodcast/fxg-031710-Ptex.mp3

I guess not, then...Thanks for the update Andrew, and also for fixing the "Frame Object In View" command in 3.2.08

+1. As cool as Ptex may be, it doesn't amount to a hill of beans if the rest of the application still has show-stopping slow-downs and tools that just don't work that well (ie, Pose tool...just spent a few hours wrestling with it, trying to get decent results, and finally gave up trying). Performance and stability should be PRIORITY No.1 for the next 6-12 months, in my opinion...here's my vote for no more wiz-bang features until those areas have been fully addressed.

In a somewhat related issue, the "Frame Up" or "Zoom Extents" as it's called in Max, no longer works at all. Not sure if this is a bug, as I've never been able to get anything remotely satisfactory since you added the feature months ago. If you lose the model in the viewport (sometime with the SpacePilot, the sensitivity might be a bit high, and the model goes zipping out of the viewport), hitting SHIFT + A makes it worse as it zooms wwwaaaaaaaaayyyyyyyy out and off in a corner of the screen. This is a sore nuisance and I hope you have time to look into it sometime soon, Andrew. Thanks... Win 7x64 w/CUDA Also, Andrew, I noticed on your Twitter that you mentioned you couldn't test Renderman since you don't have license for it...could you contact them and let them know what you are working on? I would think they would be happy to issue you some form of license for such purposes, as it's a big benefit for their users.

I wonder if 3DC is ready for us to install CUDA 3.0. I tried a while back, when 3.0 was still in beta stage and had issues, to which I had to uninstall and load the old version again.

Got it...thanks. I've seen a few tutorials where the artist did full retopo work after coming out of ZB. That's modeling twice, essentially, and seems like it's taking the long way around the block. I'm curious to see where Andrew goes with sculpting, after he finishes Ptex. Surface tools in the Sculpt room would allow him to create some brushes that are hard to do in Voxels, such as the Inflate brush (one of the Weta artists recently requested it and Andrew said it wasn't easy to do in Voxels...I guess without any true normals, it would be).

Cool stuff

So, what is your normal practice, then? Voxel mid-detail > manual retopo that is optimized for ZB (evenly spaced quads) > ZBrush for detail > 3DC for 2nd trip to Retopo, for mesh that will be animated?

I bought 3DC as a replacement for Deep Paint 3D (for all intents and purposes, the program is dead and has been for years now). I wasn't serious about sculpting then, but have become more so recently. Having hit those limitations, I realize what you're saying. If, in the near future, you had the same capability in the Sculpt Room that you now have in the Surface Tools, does that change your workflow any, or would you still feel that you needed to take your rough model over to ZB for detailing?

I can't tell you...I have you on my ignore list, so I'm officially ignoring you....or something. Seriously though, in the upper right hand corner of the webpage, click on your screen name and at the bottom of the drop down list, you will see "Manage Ignore List." That's where you want to go to add Splooge to the list.

Yep...an overhauled Sculpt Room (move or mirror Surface Tools there + Multi-Res/SubD levels found in ZB & MB) as a next on the "To Do" list would be great. Allowing users to avoid Voxel merging issues, and to be able to do all their high-detail sculpting there.Voxels are highly useful and are very promising in many respects, but they aren't the ultimate sculpting paradigm we all expected. A Hybrid approach would perhaps be the best way forward, me thinks.

Nice find. What I would hope to see is Andrew stumble upon a clone of himself (figuratively), to whom he can outsource development on things like Ptex, while he continues to "hammer the rock" (Core elements of the application). You know what they say..."Everyone has a twin somewhere in the world."

Look...folks here are the ones who suggested Ptex in the first place. Andrew has been working on the Ptex format, and as he stated, he's close to being finished on that. My suggestion was thereafter, to suspend any further development efforts on it for a while, so he can focus on the core aspects and impediments of the application. 3DC will not be judged in the industry by it's Ptex performance. But it is being weighed chiefly by it's sculpting performance. Word gets around in the industry when other users voice the limitations it currently has. That's why I believe those concerns deserve a higher priority than fringe goodies, and yet unsupported formats.I made no demands, just voicing my opinion (albeit a frustrated one) and offering suggestions like everyone else. I understand what you're saying, but nevertheless every software application improves from release to release either by addressing it's weaknesses, offering new features, or both. All I'm suggesting is that we temper our desire for new goodies until those weaknesses currently within the application are addressed. They deserve a higher priority, in my opinion.Mudbox was quite slow compared to ZBrush for it's first year or two in the industry, and had no painting capability. Those were it's glaring weaknesses...and in MB 2009, they were soundly addressed. With proper focus, I'm sure Andrew can develop solid solutions as well...we just need to stop requesting things that will distract him and hinder his progress in those areas.

You have more control in a true sculpting environment than having to rely on switching to image-based sculpting to get where you want to go. I'd much rather sculpt all the detail I want in Voxels or a overhauled Sculpt Room and simply use the Paint Room to...well, Paint. When I initially started with this character, I wasn't getting the fine control I wanted in the Paint Room, that I could get in Voxels/Surface tools. Again, having to leave Voxels to do high detail is not what Andrew intended, I'm sure. It's just that the problems Voxels introduce force a convoluted workflow. What's more is, there is no multi-level/multi-res workflow in Voxels, so once you increase the resolution, you're pretty much stuck with it....or lose much of your work.

BTW...on the merge issue (lack of Multi-threading for the task)...I restarted on my little Hobbit character, wanting to showcase something of note for 3DC, and after starting fresh from surface mode and only a few brush strokes, I take Andrew's advice to hit ENTER...guess how long I had to wait? 3hrs had passed and it was still in full choke mode. I decided to kill 3DC in Task Manager. There is no way to get mid-high level detail on a character sculpt done in Voxels with a tight deadline. You are asking for trouble if you do. You have to exercise a litany of workarounds...including hiding most of the object you're working on. Voxels choke well before either MB or ZB will...and waiting hours for surface mode to either update or merge is a major no-go.I think the best answer right now is to drop Ptex development for a spell, as there is scant support for it (according to Artman, it already works for regular obj exporting)...mirror the surface tools to the Sculpt room, and focus on improving that toolset until sculpting is competitive with Mudbox and ZBrush. Voxels are great for what they can do, but they are also a major impediment to 3DC for what they can't do, or struggle to do....and that is to sculpt a high detail model with a deadline in mind. Merge operations are as lengthy as rendering out an entire animation sequence. Even if Multi-threading was introduced into the equation, the wait times would still be unreasonably slow. I think we all thought Voxels was the Holy Grail of Sculpting, but as it turns out, it is not...not even on a beefy system.

As much as I love 3DC...it really does have issues....show-stopping issues. That is not the case, currently with either ZB or MB....not when we're talking about sculpting hi-detail. On my little Hobbit character, I've been stalled repeatedly (after having to switch to surface mode for mid-high detail, and after just a few strokes, I hit ENTER and it takes HOURS and HOURS before I can resume. I've had to go to bed waiting on it or just kill 3DC in the task manager and lose all the work I did.Progress has been stopped because once I get to the level of resolution in Voxels I need to get a decent amount of detail, it really starts to bog down...and that's on a fast quad-core with 8GB RAM and an NVidia GTX 275. I'm really growing more frustrated by the day. If this was on a tight deadline, there is no way I could use 3DC to get it done, the way things have been going. I wanted to do something to showcase what 3DC could do, but it looks like Mudbox is going to have to take over. I personally think Andrew should drop what he's doing with Ptex, once he has the current issue ironed out (sounds like he's close), and make priority No.1 the hideous merge calculation times. As I stated before a number of times, Voxels do present some great possibilities, yet it also presents its share of headaches and bottlenecks. That is why I wanted Andrew to try and mirror the surface tools (hopefully with the same speed or better) over to the sculpt room. This way, those of us who would rather forgoe the benefits of Voxels in favor of less hassle with the long merge times, have a good option to turn to within the application....instead of having to leave it and use ZB or MB to get the job done. You can offer all the workarounds you want, but at the end of the day, they are mere band-aids over a sucking chest wound. The focus of development truly needs to get back to sculpting. Mudbox isn't that much older than 3DC, and it overtook ZB in raw sculpting performance (although you need a fast system to do it). 3DC, with Andrew at the helm can do the same, I'm sure. We just need to quit asking him to divert his attention to fringe feature requests, and let the man focus.