Adding to Cart…
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.You currently have no notifications.
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
I have done some window shopping the past few days, and noticed a few things worth discussion I think.
I've had so much bad luck trying to find close fitting stuff for G3F figures that actually fit, I decided to give some generation 6 swim suits a try. This is exactly why it dose not work threw Auto-fit. The sides of the pelvis end up weight-mapped to the legs, causing odd distortions that just are not natural, lol.
Despite the first impression, that is not 'Photo shopped' to hide Poke threw, that is the mesh UV stretching because the leg is bent outward. I've seen that on the hand full of things I tried a few months back. It is particularly bad with fringe or hem-lines in that spot. In the Promos for that same figure, I noticed something else That is a big question mark.
That odd shift down on the sides of the hip. I can't tell if that is the Mesh shifted down there on the figure, or if it is artifacts from contortionist pose presets. G3F has quite a bit more adjustability with all the extra bones, and yet for some odd reason some PAs appear to be persistent in making Pose sets that cause the best made outfits to turn into pretzels. I have had no issue at all leaving all the 'Limits' turned On and posing G3F my self for some rather realistic poses, If you can't get a joint to do something with the Limits On, look at the positions of the neighboring joints. I have three Pose sets that I'm going to remove from my run-time because there that bad, I can do better from scratch without breaking outfits.
Oh, and then there was that Lace Shader set (Fbric Basics Lace for Iray). I was going to use them along with another set for some testing of render times vs shader settings. I was also going to use a simple black and white checker patron as well, with a 25% and 75% variation as well. I just have not gotten to it yet. One thing I am sure of, is that using incredibly high detailed On/Off opacity maps is pointless, as I'm sure the maps will get 'Blurred' down in size for many renders. So any improvement in render times you may have gotten from using On/Off only maps is 'Rendered infective' by TDL make reducing map sizes for small renders, lol. If your that worried about render times, use the Daz Default or the Omni shader, and avoid the AoA shader at all costs.
Yea, I think Livia is kind of, 'Cute'.
Well, seeing as I'm on a borderline 'eh' mood, and I brought up figures. Sturge Weber Syndrome. I had seen a rather attractive figure in one of the promo emails, and went to look at the Promos for her. Incredible looking mats, the eyebrows are not over-plucked, and the overall shape looked good.
Then I looked a tad closer at the face, and got Laurie HD (V6) and Stella HD (V6) flash backs.
There is absolutely no way at all to reduce that exaggerated ridge on the upper lip without turning the head shape completely off. It's kind of a shame, as everything else on the figures is damn near 'Perfect'. I'll need to check the other figures I have from that PA, tho I suspect the PA is fond of Lip ridge like JJ Abrams is fond of obstructive Lens Flare Effects, lol.
Honestly, from what I've seen, the movie is not as bad as Star Trek. They had the CG department removed a lot of the Glare.
I keep hearing about that "On/Off" theory improving render times, but has it actually been tested?
Either way, isn't it better when a texture pack includes _huge_ tiles? Downscaling is trivial, but upscaling isn't always possible.
The colour render looks quite nice to me actually, very much like, y'know, myself, especially in a slight smile =) The b&w screenshot is over-the-top, I agree.
Been under the weather the past few days, and fighting an Odroid as well, grrrr. lol.
Opacity maps. I'm not sure, and at the moment I don't have the time to devote 100% of my CPU to do some tests without other windows open (I wish that Odroid Xu4 just worked, lol). As for the opinion based on "Waiting for the water to boil" experience, it is hard to compare because it is not apples to apples.
Near the beginning of this thread, I was fussing with trying to get some light from below the floor. I did that using an opacity map with the glass gray in the map (not Black/White). Needles to say the render times was unbearable for even spot-renders to set stuff up.
Then when I was making my "Test Chamber", I did it again for the RF absorber like wall panels. Except this time it was strictly Black/White opacity maps. UE2, three DZ-spot-lights, and three Uber-panels, with a rather incredibly fast render time.
Unfortunately, there are so many drastic differences between the mat diffuse wall panels in my test chamber, and the reflective back light glass floor from my earlier experimentations, that a direct comparison just is not possible. There are so many rogue differences between them, that the opacity method alone may be dwarfed by other stuff, lol. It's almost as bad as trying to explain why some discrete component amp is better then the BUF634/OPA2134 based on there being no newer op-amps that perform better the OPA2134 in actual tests, lol.
You want to know what is the saddest thing about that graph, the OPA2134 was made in the 1980s. It operates on a much wider voltage range then most of the newer ones, and the newer ones are not significantly better then it in actual performance. Did Burr Brown really set the bar that high, that 30 years later we do not have a better opamp that can operate on 13.8VDC, lol.
Just create a distance light pointing straight up and don't have it cast shadows, then it will shine through an opaque floor. (That's in 3DL, if using Iray make the floor material slightly emmisive instead)
I didn't even think of doing that. After seeing what "No shadow" lights do to figures, I never gave no shadow stuff a second consideration.
Accept shadows, and Cast Shadows is two separate things. So stuff would still cast shadows onto the floor, so that would work. Just don't try to make intricate light patrons with the floor grating, lol. Chalk that down on a sticky note to try later on that Test chamber floor (how dose it effect reflections, hmm).
Set the light to not cast shadow, it doesn't change the floor in any way.
It's true that no-shadow lights can mess SSS up big time - not always, but it may happen. The biggest problem with a distant light, though, is that is has no falloff, which means it will evenly illuminate everything, regardless of how high it is situated. It's not that realistic for a glowing floor.
I would suggest using a floor-size plane with an area light instead, like UberArea - if it is set up to use no shadows, it will affect render times minimally, and it has configurable falloff, which can really help a) realism; b) potential SSS issues.
The On/Off opacity maps theory seems to be true: http://www.daz3d.com/forums/discussion/comment/1035308/#Comment_1035308
Lights-vs-SSS. That is why I was thinking more on the floor not casting shadows rather then the light. Also, If I just duplicate the upper Uber lights in mirror like locations below the floor, I don't need to ditch the outer box ("Outside1") in the test chamber and replace it with five or more planes. If I use a distant light, it kind of needs to get into the test chamber, that would be much more involved.
It is much to consider, along with the other dozen things I'm looking at, lol.
Opacity Maps. That is very interesting results, and I have honesty not even made the maps I had in mind yet to try. I was thinking of something like a 0/255 checker patron vs a 64/192 checker patron, just so TDL make cant just cheat with a poly-face setting with no map (evil laugh). It may be the extra cycle or two accessing a map pixel from ram, or it may be more involved calculations. The results I got for time vs opacity percent dial settings was in itself not expected (100% at the top, 0% at the bottom, in 5% steps), so I honestly don't know how a map will effect the times in the same scene.
Your test appears to break it down to the simplest it can be for purely a surface setting test and quite valid. I do have a small question I'll ask over there, hmm.
I was thinking about a single plane - take a look, I made it red for the screenshot. You can move it just 5 cm below the floor (-5 units in "y" translation). It has UberArea base applied. The shadows are off. The falloff settings are in the screenshot on the right.
The render is the chamber with no other lights, but the samples were increased from 8 to 128. When you have other lights, the noise may be less apparent, so fewer samples could work.
And of course you can play with the falloff.
That looks good, and I honestly cheated with the samples of the lights in the 'Chamber' exactly as you describe. Relying on the combination of lights to make up for quicker rendering lower sample lights. I still have some reservations with having most of the light going straight up, when the goal is to mimic the light bouncing off the floor from lights in the corners.
Well, I just got the "Thing" talking to the network after figuring out that the factory install was corrupted (I thought it was a screen res setting). It works.
I now have something to listen to music with, without it being my workstation, so that opens up some potential options now. I have read some posts that claim the fan is a tad noisy, tho I don't find it quite as obnoxious as some describe... Nvidia NV30Flow FX
Honestly, it's almost entirely a distant air moving sound rather then a GPU fan winding out in your face, lol. I'm still going to work out a heat-spreader for the thing, however that heat pipe cooler in that box is a bit much I think. So, its time to make a 3D model of that thing.
You could paint a simple square map for the light colour, darker in the middle and brighter in the corners. Floor bounce happens anyway, especially from a reflective floor like you've got there.
The thing with area light that the light isn't all going straight up actually... it goes out in all directions (okay, a hemisphere) from a point, kinda like the real deal. It's not that noticeable when it's all one intensity/colour, but should be much more interesting with the map.
I kind of knew about that spherical emission nature of "Uber surface" plane thing (Edit: AreaLight not 'surface', oops), however I thought it was based on many points on the surface to mimic the lobe-like emission of the flat surface. I vaguely remember a map in the ambient channel only controls apparent luminosity of different points on the surface when viewed directly or in a reflection. is the Illumination the same, or is it a genuine let-long map like UE2?
By the way, I've not played with DS in some time (starting to go threw withdrawals, lol), as I've been trying to get the temps on my Odroid XU4 under 70c (158F) with different types of cooler setups. That is important for another reason as well, from what I've experienced on the Odroid XU4 so far with Ubuntu Mate/Linux. I will not be moving to window 8, or 10 at all on my x86 workstations, there just not useful for a workstation. In fact, if it was not for hacks to turn the "Favorites" menu into a real "Programs" menu, and adding a "Quick launch" thing to the task bar, windows 7 would be equally as useless for a multi-display multi-window work station interface. I have decided that when windows 7 is killed off my M$, I will be moving to Ubuntu Mate, as the desktop is far superior to windows 7 and actually useful for getting work done rather then unproductive playing with a wannabe cellphone interface. It's a painful decision, however it comes down to it being useful or not, and nothing after win7 is useful to me. The only two things keeping me from ditching windows on my workstation today in favor of Ubuntu Mate, is Sonar X3, and Daz Studio 64bit. Eagle CAD, Fldigi, GIMP, and SPICE works on Linux. And there are good alternatives to Autodesk for Linux already.
Studio 32 bit works well...and theoretically 64 bit should, but it's not quite there yet. Although it's getting close...
Yea, I keep hearing rumblings about 32bit DS on Linux in other threads here at Daz, tho I'm not there yet and I don't think M$ is going to 'Kill' windows 7 tomorrow. From what I had seen on the seriously-lacking Raspberi PI2 and the much Much more capable Odroid XU4, Ubuntu Mate is an incredible workstation interface. I should have moved from XP to Ubuntu Mate, not Windows 7. Except most of the stuff I use isn't ported to linux yet, and the ARM processor on those micro-computers is not exactly the same thing as the x86 CPU in my workstation (they don't speak the same language). It will take some time.
I am going to use the XU4 as a "Support computer" for my work, doing things like show PDF files, brows web pages, and listen to tunes while I work on stuff on my workstation. I'm more then likely going to get a second one for making an Entertainment PC to replace 'LAB1' on the other desk (the caps on the motherboard gave out a few years ago), depending on a few things I have not gotten around to trying on the XU4 yet (streaming vids and music off of HPSS).
You mean this?
Yes, exactly. That looks so good. It was in a 'Beta' thread, and something didn't quite work on Linux with a beta DS, and it was kind of 'Asked' to not discus that there. So I'm not sure how fond of Linux Daz is, tho I'm very happy with what I've seen on the Odroid XU4 so far.
I don't have Linux on my primary work station (for many reasons), so testing DS on Ubuntu Mate is not going to happen here (it's x86 not Arm RISC to start with, not to say the lack of ram and other stuff on the Odroid XU4). For the moment, the Odroid XU4 is the guinea pig, I'm not messing up any of my other computers yet. I'm still just a toddler-linux-user giving folks at the Ubuntu Mate and Odroid forum lots of N00b grief, lol.
http://forum.odroid.com/viewtopic.php?f=97&t=15252&p=131570#p131570
http://ubuntu-mate.community/t/raspberry-pi-2-clones-vs-ubuntu-mate/4019/8
So, much being discussed else where, and my DS experiments and fun have taken a slight decline the past few days.
Each point of an area light is considered to emit rays hemispherically (is that even a word? hope you get what I mean). If it was just one direction per point, it would be like a "portal" for a distant light which has all rays parallel.
UberSurface will have ambient (which will light up stuff a bit, but _only_ when using GI), but it's not an area light; area light is the UberAreaLight shader (it's in the "Light Presets" folder by default, I think); it's applied to the surface like UberSurface is.
A colour map for an area light is simply based on the UV of the underlying mesh. So if it's a square plane emitting light, just use a square map.
Yes that was a mistake, more then a typo.
yes Area light Plane. I fussed with trying to make the over-sized "soft boxes" in my test chamber look like there made of six different color LED's (RGB & CYM) some time ago. In any case, it was mapped like a regular Studio Primitive Plane, and I don't think I messed with the "Color" setting under 'Light'. There is a story behind that, lol.
I had picked up some 13.8vdc LED lights that were supposedly white, and they put out a good enough amount of light for me to see what I was tripping over at night. So I set one up in the kitchen by the coffee pot, and it looked good at first. Then when I used the light some time later while making some coffee, I noticed that not only was the coffee grounds black, the coffee can was also much darker then anything else over there and I had difficulty seeing the level of the grounds in the can of coffee. I kind of knew instantly what was going on, and just turned on the regular light. After making coffee, I grabbed a prism and looked at the spectrum of the LED light on a paper, and the red spectrum was indeed almost completely gone in a few key spots. That's it, I'm done with these junk white LEDs, I want a true full spectrum White LED light...
Michael Fincke holding a General Luminaire Assembly. (not me)
each point emits rays hemispherical-ly, or hemispherical emission of light from each point on the surface, same thing in a way. There are many suffix conjunctions that are spoken to make sentences shorter to say that are not in any dictionary, lol. I still struggle with them. I've gotten use to speaking in a particular way only for a spell check to indicate it is wrong when typed out, lol.
Awesome story actually, in the sense that it really shows how our eyes adapt to light and we think it's "white" until we run into something that specifically shows that it's not. Also why the "auto" white balance setting in a camera will rarely if ever give you a picture that looks similar to what you saw with your own eyes...
Yeah we replaced our burned out overhead bathroom light with a "white" LED bulb.. it's rediculously blue. Won't make that mistake again.
I have not posted much the past few days, weeks. I've been chasing down an 'Opportunity' of a sorts. Going back to the beginning, I picked up an SSD to put in my work station for DS, and I had already run into the 5V 23 amp limit on this Power supply before. So I decided to hook up an oscilloscope to the beast to be sure it was ok with the extra SSD, and that's when things went south. I figured the new motherboard had crummy USB ports that put a 10kHz background noise into my speakers, possibly a grounding thing or other, and it didn't happen all the time either. So, with the computer on and looking at the scope with the side cover off, and that's when I noticed a coincidence. The same exact pulsing tones were coming from the speakers, and the motherboard at the same exact time, and it was the same exact timing of a wave I was looking at on the scope. Ah ha, it's a singing inductor on a V-reg, with insufficient filtering around it.
I traced it down to the CPU V-reg in short order. So I tossed a few poly-film caps on the CPU power cord to see how well it would help, and that's when I noticed something rather condemning with this otherwise top-end PSU. No it's not the electrolytic cans in the PSU, I checked them.
It's the wires it came with. The HDD and PCIe cables are all copper, and all four of the included CPU power cords are all aluminum, WTH. While it will not make inductors sing, it certainly will not help matters at all (it will make it worse, a lot worse).
So, needless to say, my workstation has not been on for any length of time the past few weeks, and I dread the work I still need to do. I scarred up my hands trying to pull the pins out of one of the CPU cables to replace the wire last night, and I gave up and just ordered a replacement cable with copper wire.
Also, I need to pull out the motherboard, and some where under that blue heat sink is an inductor that needs to be re-potted (if not all of them).
It has been a very long few weeks.
P.S. and for those that are courius, yes that none other then a Seasonic X650 PSU (the litle brother to the X850), powering an ASUS M5A97 Motherboard and ten heard drives (soon to be 11).
http://www.hardwaresecrets.com/seasonic-x-series-650-w-power-supply-review/7/
it's a good PSU.
Computer drama, day something-teen, fans.
This is a "Priceless Artifact" at this point. It is a 120mm fan with a plug for a thermistor to directly control the fan's RPM. Your eyes are not deceiving you, it dose not require a watt-hog computer to control the speed of the fan. Now before you freak out over me not wanting such advanced fan control features as 4-pin PWM control, not everything that has a cooling fan includes a PC to control the fan speed. My Mirage 2-meeter and 440MHz amplifiers don't, My UPS backup inverters don't, the Samlex SEC-1225G power supplies don't, and the Kurzweil K2000RS and K2500RS certainly dose not. The only plug they all have is a 12VDC 2-pin power plug for a fan. So the A2029 and the original A1357 fans were perfect for them, and my computers.
As you can probably guess by the residue marks on the fan, I was just cleaning this one out in an alcohol bath (I ran out of Isopropyl, and need to get more). It's bearings are on the way out, so I figured it could not hurt. It has been cooling the chip-set and PCI cards in my workstation since 2006, and I now need to replace seven of the 120mm fans that have already got the "Lucas Treatment". The problem, they don’t make anything like this any more, I've been looking. I've found one from a 'shady' supplier that dose not come close to the top-end RPM or CFM ratings of the fans I need to replace (500RPM - 3500RPM, 95CFM or more).
Now, I'll be honest. I have managed to find just enough 80mm A1357 fans to get me by, however I have not found any A2029 fans at all, and I need seven of them, sooner rather then later, lol. Some rather keen of you will notice the white 120mm "Silverstone FM121" (2400RPM, 110CFM) fan from the original 2005 case-mod back in service on my CPU cooler, it was pulled out of the mothball fleet because I could not find a new 120mm A2029 fan anywhere. In fact, none of the 120mm fans I've seen reviewed lately have a thermistor plug or can muster a minimum 90CFM, that is just outright pathetic.
So I guess it is up to me to 'Make them', grumble.
Hi ZDG, long time no see! ... not pretending to fully understand what your problems are. But even if it is not 3D related it is always interesting to follow your work. :-)
Hope you get things figured out and come back to 3D soon. ;-)
Oh my.
Hope you succeed. Whichever route you're going to take.
MN-150374, I am slowly working threw it. It's kind of like a puzzle, where you know the blue pieces are probably part of the sky, and it is just a matter of trying to peace them together into something that works. Tracking down noise is kind of the same, especially when you can see the wave shape on a scope. Different things produce different shapes just like different parts of a puzzle are different colors, and it becomes a game of tracking down the locations they belong. With a puzzle it's just a mater of putting the pieces together once you've found where it goes, electrical noise is a tad more work to fix it once you found where it's coming from. My computer is still a tad bit of a mess, tho I think I found a way to eliminate the noise getting into my speakers, lol.
I just need to get it all back into the computer and looking good, lol.
Kettu. My life has been the very definition of multiple routes, lol. Just because I spent the past week digitizing cassettes for a friend, dose not imply that I did not at least look at the Daz store, or check out the latest progress at LHC. I thought I would have the "Daz" Content path on the new drive and up and running Sunday, and it's been non-stop challenges exactly like the past month, lol.
Morning Eve, lol. Sunday morning came, and the (not so) smart tab database was still trying to use the old drive path. I gave up and just wiped the database clean, and rebuilt it from scratch, And that essentially lead to Monday morning...
It's been a long month...
(EDIT) Thank you so much to whomever got rid of that nagging 'You Must log-in" screen in Studio 4.9.
Tad bit of an update. I had tapped 5V from the lower HDD rack for a USB port for the sound to attempt to get some clean power to my external sound thing (UA-25EX), and that didn't entirely solve the noise issue from the CPU V-reg for whatever reason. Yet when I added the sixteen 1uF Film caps for the SSD drives down there, that did cure the noise issue. So I guess I just needed more poly-film reinforcements on the 5V rail. Whatever, it appears to work.
I ordered a SSD rack-thing, and not only did it arrive late on the very day the 30-day return expired on the drive I needed it for, it arrived a mangled mess. So the two SSD drives are just kind of hanging lose in there for the time being. I'll just take a day to head to the machine shop to make a new one for the SSD drives.
So, I have just gotten to looking at a figure that I wanted to install for a few days now (purchased back on Saturday).
http://www.daz3d.com/wild-spirit-bundle-hd-character-jewelry-and-outfit
The mat options are quite different from what I've gotten use to.
There is no more Iray and 3delight folders, just some options in the 'Materials' folder. I'm guessing the 'Legacy Iray' mat is for the older Iray version that had an off color thing going (or maybe the new Iray is off color, whichever way it is). And there is a "Quick Shader" for 3delight, and it is indeed much, much, much better then some former renditions, lol. If only I could see the faces when I asked a particular HPC computer company if they could build a machine that would "start rendering this today", lol. Gasps, "Good God" was exclaimed, Chins hit the table and the other three turned white in fear. I think they can rest easy now, as a better solution has been found, using some settings in the shader, rather then throwing a few hundred CPUs at it, lol.
I'll need to do the spot render again, as it was somewhere around three minutes, and I wasn't even looking at the screen, lol. Close enough to call it "About three minutes", that's a full order of magnitude and change faster then it was on some other figures. Thank you.
(Edit) that was just the hand. Full body, well, it's still thinking about it.
As you can probably guess from looking at my task bar, I need to close some stuff to do a proper test, tho it looks like more subsurface zones dose indeed add more time to the pre-compute delay. Hmm. I killed it at 18 minutes.
18:55 giv or take. Looks like I killed it just before it got it's GUI out of it's back-end, lol. and at about 22:41 the face starts to render, lol. Spot-render done in 24:01.
So, that's about 19 minutes for the computer to think about doing it, and about five minutes for it to actually do the spot-render once it got something resembling 'motivation' or a "round-2-it". If there was a faster workstation available, this useless thing would have been $#!+ canned already, lol. I need at least one-hundred-and-sixty cores running at 4GHz minimum in order for this to be usable, because this just is not going to work for me.
"3Delight shader" as in "a real new shader" or "a shader preset for UberSurface/AoA/whatever"? Basically I'm curious as to what the surface tab says right after the word "shader" somewhere at the top.
Without giving away 'Trade secrets' regarding map settings, it's not remarkably different from other PA's figures. And it is not just FW or SA figures that use the AoA that have such horrible face plant times. Further more, when trying settings on a simple primitive, I've never had times remotely as bad as the figures regardless of what I do to the primitive, it just dose not add up.
(edit) seeing as I'm on a bit of a nitpick trend. Augusta Hair for Genesis 3 Female(s)
http://www.daz3d.com/augusta-hair-for-genesis-3-females
Is there a color expansion for that anywhere out there in the Luminiferous Aether?
http://www.daz3d.com/colors-for-augusta-hair
https://www.renderosity.com/mod/bcs/touchable-augusta/109328