Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
I'll also add links to my Instantiation Via (Poser) Geometry Swapping ? and Towards better grass in Poser... threads (both at Hivewire) which is where the 54-bone geometry swapping figure approach took me, and where the dual-club wielding ladies in pink bikinis totally diverted my attention. But those links are very relevant for the new figure-based approach, because I want to use it for casino chips, playing cards, coins, bank notes, little asteroids, clumps of grass, etc. Of course the geometry-swapping part of it is a no-go as far as Studio is concerned, so that part of it'll be Poser only.
I have Poser, currently can't remember if it's 9 or 11 cause it's not installed, it's saved to disk at the moment. Have to pull that out one of these days.
I'm trying to remember how to do all those things I forgot how to do, like creating curved text in GIMP. I was surprised , how little effort it took to remember the first bit of this. Create the curved path, create the text (ordinary text), make sure the path is selected in the Paths pane, right click the text layer in the Layers pane, select Text Along Path, and Bob's your proverbial uncle.
Now I have to remember how to turn that newly created text-as-path path into a nicely anti-aliased chunk of curved text. I think I did a Stroke Path with an antialised 1 or 2 pixel brush to give me the antialiased edges, then did another Stroke Path on a different layer with a soild, un-anti-aliased 1 pixel brush to give me a solid outline to fill. Then I selected the area enclosed by the solid outlined bit, and then did a bucket fill on the aliased outline layer. I'm sure there's an easier, better, far more sensible way to do it, but where's the fun in doing sensible things ?
I've found that not having software installed is the best way to prevent it from annoying you. I like to keep my mobile phone switched off, preferably in a completely different location from me, for the same reason.
Edit 22 Feb: I've got the template for the Monte Dubbioso one (re)done now. Here's the $1,000 chip. It should simply be a case of changing the colours and the value (plain text) to get the rest. But I need to tidy up the layers in the XCF first.
I've also attached a set of 6 annotated screenshots to remind me how I did the curved text, and the zipped up GIMP XCF file as it stands. It needs a bit tidying up really - maybe later today.
Note: the sandy-coloured patterned ring was simply created by flood-filling a 1024x1024 layer with a tiling pattern and then doing Filters > Distort > Polar Coordinates
Edit2, later on 22 Feb: Whoops ! Just noticed a minor mistake that I'll fix and then do the other value Monte Dubbioso chips (I need to rotate the three inner semicircles by 60 degrees). I'm happy to say that resolving my general antialiasing problem was simply a case of getting the antialiasing right on the layer masks and using solid colour for the layers themselves.
Monte Dubbioso Casino Chip Textures Release 1
1024x1024 JPG images (75% quality) in the attached zip file.
I've also attached the zipped and tidied up XCF plus a zipped exported PSD version in case anybody doesn't like my colour choices. They're set up for the $10,000 chip, but it's easy to change the colours and value. The fonts used are Vladimir Script and Bernard MT Condensed, so if you don't have those I assume that GIMP/Photoshop will pick the closest you do have. Several layers include the word '(HIDE)' - these are the ones you don't want visible when you export a texture (most of them were for creating the curved text, and I decided to leave them in.
That's got me nicely back into the swing of things. So I think I should try doing the 54-bone figure thing next. To allow it to work in Studio I'll start without geometry swapping, so the base mesh will need to be 54 colocated 24 vertex* cylinder disks as the mesh for the (effectively independent) bones.
Thinks: Since each bone can be scaled independently it should be easy to have diffent sizes and thicknesses for the disks, so a variety of coins is possible this way. But coins have different fronts and backs (which is something I though about back there, but then realized on looking at the picture of doubloons tossed in the air that you'll only ever see one side** of each coin in a (non-animated) render...
*Although I may try using the 32 vertex cylinder with 4 control edges.
**except if you have strategically placed mirrors
(attachments scanned an okayed with Malwarebytes and Bitdefender free versions)
Next Step: Start On A 54 Boned Figure With Each Bone Effectively Independent
Abandoning the morphing prop idea (having to redo every single morph for each base mesh really is too much!) I'm going ahead with the alternative approach that I mentioned a few times.
Since I've already got a working Poser version of a 54 boned geometry swapping figure (with a deck of 54 cards as the basic mesh) it seems a good idea to start with that. But since it was only uploaded to another site* and linking to it doesn't appear to be allowed I'll re-upload it here - just for my reference.
For it to be useable in Studio I need to strip out the geometry-swapping stuff. So I started with the 'Instantiation Via Geometry Swapping Test.cr2' in the attached zip, and removed. all the geomChan and alternateGeom sections (which exist for each of the 54 bones), plus a few parmNodes and valueParms related to geometry variants (screenshots for my reference). I was very surprised that I managed to do that without any obvious error on my first attempt. 'Deck54 Geometry Swapping Removed .zip' Poser runtime attached if anybody wants to play. Since the 54 cards load up colocated I've also included a pose to spread them out a bit - I needed that to check that all 54 cards are present and correct. Note that the back of the cards hasn't got a proper texture and simply uses the full 54card texture
Here's a screenshot of the posed deck54 loaded up in DS4.12. From a quick check the behaviour seems to match Poser.
Next step - either create scatter poses A to D based on the offsets in that post, or create a 54 disk (casino chip, coin, etc) base mesh. Not sure which appeals most, they're both such delightful options !
Edit 22 Feb: I'd got as far as the offsets for bone 20 before I remembered that (a) all my offsets are positive integers from 0*** to 51 1 to 54 whereas what I want are offsets from -25.5 to +25.5 -26.5 to +26.5 to keep the spread centred around the figure origin, and (b) bone_1 etc aren't particularly good names and it may be better to change them first. And (c) I haven't got any random rotations. And (d) ...
I think I'd better stop for today - starting to make stupid mistakes.
*it was originally attached to the first post of my "Instantiation Via (Poser) Geometry Swapping ?" thread at the site beginning with an H**, so if you want to read the background you'll need to search for it.
**and ending in an "ivewire3D".
***yes, I know
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
Edit 23 Feb: I've finally worked out how to group the relevant faces with the appropriate bones from a donor CR2 in the Poser Setup room - without going near the group editor. When creating the model in Blender do the mesh for each bone as a separate object, and name each object to correspond to the internal bone name in the donor CR2. My test mesh was a casino chip duplicated three times, so four chips, each to be assigned to a separate bone. Then when exporting the OBJ select all the objects and tick 'Objects as OBJ Groups', not 'Objects as OBJ Objects'. Blender appends the mesh name to the object name to make the group name, so it's necessary to do a regular expression find and replace on the resulting OBJ file to correct that. Then simply import the OBJ into Poser with all options unticked, go to the Setup room, and load the donor CR2 from the library. Since I only used four objects/bones for my quick test I only selected to use the first 4 bones from the donor CR2 when prompted. Exit the Setup room (without bothering to even look at the group editor - that still confuses the heck out of me!) and test - I applied the pack of cards laid out in suits pose from the attached Deck54 Geometry Swapping Removed.zip.
All seems good so it's probably time to create that 54 casino chip OBJ.
Regarding the base mesh to use for the chip, I'd originally said a 24 vertex cylinder (72 faces) for the 54 chip scatter, and a 32 vertex with 4 control edges (224 faces) for a single chip prop. A 32 vertex cylinder without control edges has 96 faces, and a 24 vertex cylinder with 4 control edges would have 168 faces. Multiplying those by 54 chips gives:
24 vertex cylinder x 54 = 3888 faces
32 vertex cylinder x 54 = 5184 faces
24 vertex cylinder + 4 control edges = 9072 faces
32 vertex cylinder + 4 control edges = 12096 faces
So adding control edges to the 24 vertex cylinder increases face-count by a factor of less than 2.5, and is only 1.75 as many faces as a plain 32 vertex cylinder. So I'm thinking maybe the 24 vertex cylinder + 4 control edges should be the base mesh, which would allow SubD (Studio/Poser) and smoothing (Poser) to work on it...
Edit2, later on 23 Feb: Rather than making a snap decision I created a single chip of each of the 4 varieties just mentioned, UV mapped them (Top camera > Project From View (Bounds), flipped the bottom face UVs left to right to prevent it being a mirror image, scaled the UV down to 99% because I did that before for some reason, imported into Poser, applied a texture, checked that all front/back faces look correct and line up correctly with the sides, rendered, added one level of subD and rendered again.
Well, we already know that subd messes up the non-control-edged disks so that's no surprise, but scaling the UVs down to 99% causes the little + on the top of the crown icon to bleed over onto the edge. I'll have to try again without scaling the UVs down
Edit3, even later that same day 23 Feb: Re-UV-mapped all 4 base meshes to 100% of the UV map and the problem with the crown + bleeding onto the edge under subd doesn't appear.
Neither does the similar thin line spotted a long time previously (not sure why it only appeared on some chips), which was the reason for scaling the UVs down to 99%. But that particular problem only shows up if if the textures at UV(0.5,0) and UV(0.5,1), or UV(0,0.5) and UV(1,0.5), are different colours - which they aren't for this set. So I quickly modified the $500 texture to make the top and bottom of the UV mapping different colours and had a look at it in Poser 11. I only see it now in the preview. Maybe it was a PP2014 thing ?
But now I've seen a different problem with Poser smoothing - maybe the control edges are too close (edit: Fortunately I'djust made a very stupid mistake in my render settings!) ? (I think this sort of thing is why it takes me a long time to do things - in trying to fix a minor problem I uncover even minor-er problems and try to fix them...)
Same test in Studio (4.12.1) both 3Delight and iray - seems you have to apply subd to get rid of it
Since I've checked Studio 3Delight/iray maybe I should also check Poser cycles - or maybe there's a way to sidestep this - rotate the UV map a bit so that none of the vertices are right on the edge ? I wonder...
Edit4, later still on 23 Feb: Rotating the UVs by half a sector, i.e. 7.5° for the 24 vertex disks, seems to do the trick - that line doesn't appear any more. Neither in DS 4.12 (3delight/iray) nor in Poser11 (firefly/superfly) , with or without subd. and even though the little + above the crown icon is damn close to the edge it's not bleeding over.
So I think rotating half a sector hits the sweet spot between the 100% UV problem and the 99% UV problem !
And I realized that the Poser Smoothing problem was me - I forgot to increase the crease angle from default 80 to something over 90. Doing that it works correctly
Final thing for this post, hopefully - when I rotated all the UVs 7.5° I didn't take into account that I'd flipped the bottom face, so that's been rotated 7.5° in the opposite direction and is now 15° out of sync with the top face. Easy to fix - just need to remember to do it.
Edit 24 Feb: Rather than start a new post and bump this thread with no real progress, I'll just post these confirmation pics here. Test renders in Studio 4.12 3Delight/Iray and Poser 11 Firefly/Superfly of the four basic chip meshes both with and without subd, and Poser 11 Firefly/Superfly renders of the four basic chip meshes with Smooth Polys. They all seem fine, without any extraneous thin vertical line on the edge (although because the texture images are 75% quality JPG I think that the JPG artefacts around the + of the crown icon are causing a barely visible vertical line when subd is applied - but I only think I noticed that because I was specifically looking for it, so I'm happy to let it go). Also a Poser 11 of the four basic chip meshes above a mirror plane to verify that I've correctly corrected the rotation of the lower face UVs - I have.
Edit 25 Feb: Just uploaded a zip of the latest (very simple) blend file 'NewChips04c rotated UVs +7.5.zip' including the exported OBJs (the MTLs are included too but they don't include anything useful). Next step is to replicate one of the disk objecks 53 more times making sure each object is named to match a Poser bone (I'm thinking 'Scatter_Disk_01' etc - 'disk' rather than 'chip' because the texture for any basic disk shaped can be used, e.g. coins, casino chips, tiddlywinks, etc).
Also thinking about having up to half a dozen materials (once again named in a non-disk-type-specific way, e.g. Disk 01-09 Material, etc) so you can mix different value casino chips, coins, or tiddlywinks.
Why just 'or', why not 'and/or' ? Say $1 and $5 casino chips, cents, 50p pieces (nope, they're not round!) 10p pieces, doubloons, and a green tiddlywink ? Different diameters, thicknesses and diameter:thickness ratios are the main problem with that, although I think some simple ERC controls (master 'Disk 01-09 Diameter', 'Disk 01-09 Ratio' etc dials) could handle that). And perhaps simple PZ2 pose / MAT pose pairs could be used for preset combinations ? (Note to self: remember to use targetGeom rather than valueParm if you want to save ERC dial settings in poses - I think ?)
Why not simply use Studio instances you may well ask ? Well... Poser doesn't have those!
(and as I've said before, where's the fun in doing something the obvious, sensible way?)
Edit 26 Feb: Just attached a backup of my test runtime ('Runtime (Dev pre cutting down to 8disk only).zip'). I've got the basics of the 54 chip scatter working now, although I'm only using 8 chips and 4 materials. I want to create a new post for this, but I don't want to include any extraneous stuff. But also I don't want to accidentally lose anything important when I start deleting stuff from my temporary runtime. And I don't trust my own on-computer or flashdisk backups any more - I've just had a backup flashdisk go tits-up on me!)
LMAO love this "I've found that not having software installed is the best way to prevent it from annoying you."
Ok have to come back later to read the last few posts
Making reasonable progress (if you're wondering what I've been doing since my last post on 22nd Feb see the edits to that post). I'll be adding updates to this post as and when I make them, and won't add a new post until (a) I believe I've made sufficient progress, (b) this post is getting too long , (c) I just feel like annoying people.
Anyway.
I now have a 54 disk scatter Poser figure, with six material groups (nine items/bones per group). I've added ERC controls to the 'Body' so that you can set the Scale (relative size) and/or yScale (disk thickness relative to its diameter) of each group. Thus I can have coins of different sizes (yes, they're all double-headers or double-tailers - but since you only see one side who's going to know ?)...
It can also be used for any other simple disk/cylinders - like slices of vegetables, tin cans, vinyl records, etc...
Both of the above renders were without any subd/smoothing. The later one shows some of the many flaws. Still, that's what you get for throwing badly chopped veggies, records and unlabelled cans in the air.
I've attached my test runtime, zipped as it stands and without any explanation, for any fool who wants to have a play.
Answers to things that would probably become FAQs if people ever asked them:
1) The ERC uses deltaAddDeltas which, as you noticed, isn't the most intuitive way to set the scales. I'll probably change them to deltaTimesDelta (or whatever it's called)
2) Yes, you can import CR2s into Studio
3) Yes, the disks are rather large - 8'7.2" in diameter to be precise. I've found that scaling the figure down to 5% is a reasonable start
4) No, the 'Farthings, Pennies, Shillings' pose doesn't work. It's supposed to set the ERC master controls for the various scatter groups so the three coins are the correct sizes relative to each other (if you want them the correct size relative to Poser figures you need to set the Scale for the scatter figure to 0.969% (1/103.2 - see my signature line). That's the valueParm v targetGeom thing I mentioned a while back (or maybe I didn't?) - I know how to fix that (I think?)
5) Nope, I didn't make the texture paths relative.
Edit (2 Mar): I forgot to include the two PZ3 files for the two renders above, i.e. 'Farthings,Pennies,Shillings.pz3' and 'Stuff.pz3', so I've attached a second zip containing those. Looking through those two files, and recalling that the original DAZ Millenium Big Cat had 'MORMap' PZ2 pose files that applied a texture and morphs (via valueParm lines) I remembered that POser doesn't save valueParms to a PZ2 pose file (which is where the workaround trick of simply changing the word valueParm to targetGeom comes in), but will happily (and more importantly, correctly) read them from a pose file. So simply copying the relevant valueParm settings from the scene file to the correct place in the non-working PZ2 file should make it work (see next two attached Notepad++ screenshots.
So I really want to create the equivalant of those Mil Big Cat MORMap PZ2 pose files to set the master parameters and textures for predefined combinations.
And to allow users to save their own poses with master prameter settings I need to simply change the valueParms to targetGeoms*
And I need to change the deltaAddDeltas to deltaTimesDeltas (or whatever they're called) to make the master scale parameters more intuitive.
And probably a few other things too.
But I should probably do an edited version of the 'Farthings,Pennies,Shillings.pz2' pose file first to be absolutely sure that it works as expected, and the do a 'MORMap Farthings,Pennies,Shillings.pz2' pose file to make sure that works too.
Edit2 (2nd March): I was having problems finding that deltaTimesDeltas stuff - mainly because it's not called that.
What I need to do is use 'valueOpTimes' instead of 'valueOpDeltaAdd' (other options being valueOpKey, valueOpPlus, valueOpDivideBy, valueOpMinus (as used in my Proof Of Principle - Poser ERC Opening Book freebie at ShareCG - I knew I'd seen it done somewhere! (There's also valueOpDivideInto which I didn't use)). I've attached , for my own reference, a Notepad++ screenshot showing a bit of the opening book CR2 that uses valueOpTimes.
Edit 3 (still 2nd March): Doing a global replace of valueOpDeltaAdd with valueOpTimes, deleting every deltaAddDelta line using an RE, and not forgetting to set default ScatterXX-YY_#Scale values to non-zero values I now have master scaling dials that work in a more intuitive way (three more Notepad++ screenshots attached for my own reference). Unfortunately there doesn't seem to be any way to make Poser display theses as percentages, so maybe I should rename them as ScatterXX-YY_#ScaleFactor ?
But after doing the 'replace valueParm with targetGeom' trick it appears that I need to select to save both Morphs and Body Transformations in order to save my master parameters in Poser 11 (the only version I have installed now). Which is a pain in the armpit because I don't want to save body position/rotation as well. So I have to now edit the PZ2 to get rid of that, which defeats the whole purpose of changing valueParms to targetGeoms in the first place (which in case you've forgotten was so that I don't have to manually edit them into the PZ2). Progress eh ?
Edit 4 (3 Mar): One thing I realized when I did the LPs, veggies, cans, etc version - because I scaled the LPs up by a factor of 12, when I applied my scatter pose the LPs intersected each other and the other stuff too, so I had to manually adjust them. Adjusting the scale of the figure makes the whole scattered figure bigger or smaller, but if two scatters are just intersecting at 100%, then they're still just intersecting at 5%, 500% or whatever, like this...
So I think I want a ScatterPositionsScaleFactor (valueOpTimes, default value 1) that applies identically to the x/y/zTranslation of each and every scatter - that way I could have just twiddled a single dial until none of the LPs were intersecting anything.
Edit 5 (just coming to the end of 3rd March): Just starting creating the PZ2 pose files for the 4 pseudo-random scatter sets A to D (adjusted from (1 to 54) to (-26.5 to 26.5) as mentioned previously, and noticed that the X offsets are actally in ascending sequence, i.e. rather non-random. So I need to shuffle the x/y/z triplets about amongst Scatter_1 to Scatter_54 so that e.g. Scatter_1 doesn't always have x = -26.5. Not a problem, just yet another thing I hadn't thought of. Off to bed now anyway.
Edit 6 (4 Mar): Reminder - the pseudo-random number generation I've always used for this sort of thing is a Linear Congruential Generator, Xn+1 = ( (a* Xn) + c) mod m
It produces X values in the range 0 to (m-1). But much trial and error is required to get parameters that have the desired repeat cycle, i.e. here I need m=54 and a repeat cycle of 54. I remember there were some simple rules of thumb for selecting values of a and c based on m... but I forget what they were and can't track them down.
*changing valueParm to targetGeom to trick Poser into saving it in PZ2 pose files - p101 of B.L.Render's "Secrets Of Figure Creation with Poser 5", 2003
Oh wow this is great and love the idea of using on other disk items. Thanks a lot and great work.
(c) Just to annoy people - still waiting for the apocryphal playing cards, casino chips, coins and banknotes...
(still making progress, noted as edits to the previous post)
Edit (4 Mar): Reordered the triplets as mentioned in edit 6 of the previous post (and values are now evenly spread about the origin in all three axes). Now to create pose files to apply them to the scatter figure.
Still need to do a couple of MORMap pose files.
Still need to add a ScatterPositionsScaleFactor control.
But I think I should then be at the point where I can have a first go at DAZifying it.
Edit 2 (4 Mar again): Set A pose done. But this should ONLY set the x/y/z positions and not touch rotations, scalings, or master dials on the body, so I realized that I now need to strip out all the unnecessary crap from the pose files.
E.g. the PZ2 for setting coin sizes for farthings, pennies and shillings should ONLY set the 6 master dials on the body, and definitely NOT touch the x/y/z positions/rotations.
Edit 3 (still 4Mar): The MORMap was quite easy - attached annotated screenshot collage to remind self.
Edit 4 (5 Mar): Got the positional PZ2s for sets B and C done, and the D one 'll be finished soon. While checking them I realized that a new ScatterPositionsScaleFactor dial might not be what I want. To explain - the image below is three disk-scatter figures with the three poses A, B and C applied to them (and a different colour applied to all the disks of a single figure), which ensures that no individual disks will overlap (assuming of course that I didn't mistype any of the coordinates when I set them up). The idea of a ScatterPositionsScaleFactor dial was simply to reduce the amount of space between disks (the default values of the A, B and C poses leave a lot of space) without affecting their positions relative to each other. However, if I used such a dial on one of the three figures in the picture all the disks within that scatter figure would move towards the origin, getting out-of-sync with the other two figures. An alternative would be a DiskSizeScaleFactor dial that multiples the Scale of each actor/bone/disk ?
Ok I don't understand all that stuff but maybe someone else will come in and see it and comment. Love the image...lol
Don't worry, that's just me thinking out loud.
Edit (5 Mar): Anyway, I've finished the set D positions PZ2 (Poser pose) file, and added an All_Scatter_ScaleFactor control to the figure, so I should finally be able to recreate the tests I did a year ago with the original morphing prop here and here...
...well, not quite there yet. I didn't include any x/y/z rotations in the Disk Set A/B/C/D positional poses. But do I want to ? Probably yes. My vague idea was to have a selection of scatter rotation poses that set pseudo-random (not calculated this time, just dialed by hand in Poser) rotations but don't affect the positions. So if the user doesn't like the built in rotations they can apply another set without changing the relative positions. A script would probably be a good idea too, meaning I'd have to do both PoserPython and DAZ Script versions, but it should be very simple (a for loop to do each actor in turn, and simply set random values for x, y, and z rotations).
The "reducing the X/Y/Z from 100% to 20%" that I mentioned in those posts a year ago should be equivalent to setting the All_Scatter_ScaleFactor control dial for each of the 54-disk-scatters to 5. Anyeway, here's four of the 54-disk-scatter figures (so 216 disks) with only the All_Scatter_ScaleFactor control changed between images - 1 on the left, 3 in the middle, 5 on the right.
I can see definite overlap near the centre of the one on the right, so I probably mistyped a few x/y/z triplets - that should be easy to track down.
Next steps - add some rotations to those four poses, declutter some of the other pose files so that they only include the dials I want (i.e. similar to what I did when I created the MORMap pose for the old English small change), add some materials (MC6) for the Monte Dubbioso casino chips, ... what else ?
Then upload a zip of the Poser version (which will still be under /3DCheapskate/TEMPORARY/) for anybody who wants to play.
Then import that into Studio and create a Studio native version (initially just a temporary test version)
But I think I've had enough for today, and I'll maybe take a break for the weekend. (I've also attached a zip of my test runtime as it stands, just as a backup really)
Thanks so much for all your hard work and so far so good.
Things are proceeding quite well, but I've run into a minor problem*.
First a render of 216 non-intersecting* coins** within a very small space, using the Set A, B, C, and D positions and random rotations.
The PoserPython script for randomizing positions and rotations was very simple, so when I get get round to DAZifying the DAZ Script equivalent should be simple too.
Stripping out the unnecessary crud from the Poser PZ2 files is easy, but tedious and laborious. Luckily I've worked out how to reduce the number of cock-ups I make while doing that - only do one a day !
I'm fairly certain that creating a figure for 54 playing cards, spheres, cubes, whatever will simply be a case of changing the geometry file name (two places) in the CR2, and ensuring that the mesh is correctly grouped into 'Scatter_01' to 'Scatter_54'.
*Problem - the observant among you will have noticed that some of the coins are actually intersecting, which they shouldn't be. That, after all, was the whole purpose of the spreadsheet and resulting tables of x/y/z coordinates in that post over there....
...and...
...Oh bugger ! I've just spotted the reason !
I was one row out in the new spreadsheet I created when I subtracted 27.5 from each of the values, so what I have is:
ROFL ! You have to able to laugh at your own stupidity, don't you ! (That's why I'm such a happy chappie most of the time !
)
Edit: Just noting the dimensions of a few objects.
The original idea was to allow four 54-item scatters (props at the outset, now figures) to be loaded at the same location, have four different positional setups A, B, C, and D applied (morphs with the prop version, poses with the figure version), and guarantee that no pair of scatter items would intersect. The values I calculated for A/B/C/D x/y/z positions were all integer values from 0 to 53 for the prop version (values in steps of 1 unit between -26.5 and +26.5 for the figure version), and such that the scatter items were just over 5 times as far apart as they needed to be, meaning that I could bring them closer together by a factor of 5 (setting the prop morph to 20% instead of 100% for the prop version) or make each item 5 times bigger without touching its position (setting All_Scatter_ScaleFactor to 5 for the figure version) and still no two scatter items would overlap.
However, this was based on each scatter item, regardless of any rotations applied to it, being bounded within a 1 unit diameter sphere centred at the scatter item centre, which is why I created the casino chips so that the disks were 1 unit in diameter. Since I'm working in Poser, where 1 unit = 103.2" (just over 8') this means that everythings huge, but that's a simple scale setting - e.g. for casino chips of 39mm (1.535") diameter) as I just need to scale the prop/figure to (1.535/103.2) x 100% ≃ 1.5%.
I made a bit of a mistake with the playing cards by modeling them at 1.4units x 1unit instead of 1 x 0.714 but that's easy to fix - the scale factor to make a 1x0.714unit card into a 3.5"x2.5" card would be (3.5/103.2)x100% ≃ 3.4%
But if I have a mixture odf cards and chips, say sets A and B are cards while C and D are chips, then what scale factor do I use ? Obviously to get the correct sizes I need to use 3.4% for A and B, and1.5% for C and D. But since the scale factor also implicitly affects the positions of each item by the same amount then sets A and B aren't matching C and D.
Confused ? Don't worry - that's all for my own reference and I understand what I'm on about !
Edit 2 (16th March): For various reasons that seem sensible to me at this time I think I've decided to have my OBJs at 100 times real world size, so a standard playing card 2.5"x3.5" will be 250"x350". For my scatters I've been creating the OBJ meshes at a scale of 1 unit = 1 inch. However, because I import my OBJs into Poser without any scaling 1 unit becomes.1 Poser Native Unit (PNU) , i.e. 103.2" (or 8'7.2" if you prefer). So I need to apply a scaling factor of 0.969 (100/103.2 correct to 4dp) when I finish the OBJ so that the OBJ, after importing into Poser, will be 100 times real world size.
Edit 3 (18 March 2021): I've gone back to the cards, which is where I started. I've resized the original mesh to be be 250" x 350" (i.e. 100 times actual size of 2½" x 3½"). I'm using Poser scale of 1 PNU = 103.2", so in the OBJ file each card is 2.4225 x 3.3915 units.
I've made some minor correctuions to the template for 54 cards and posted it as an update back there, but here it is too:
I've got the Set A/B/C/D pose files (PZ2) with the correct numbers now (thanks to a simple PoserPython script and spreadsheet). I've got a basic PoserPython script that randomizes translations and/or rotations of the 54 scatters (I haven't tied to avoid intersections with that - maybe later). Still going well...
**Yes, I know that the farthings, pennies and shillings are all the same size
Looking great and already have a list of images to make with these.
I ran into some problems (nothing major), when trying to use two or more scatter figures together. I was hoping that I'd be able to use the 'Conform to' thingy to make the second, third, etc, figures automatically pick up the master dial setting from the first figure, but that doesn't appear to work as I'd hoped (actually, it does seem to work, but all the scatters of the 2nd, 3rd, etc figures all pick up the x/y/z positions of the scatters in the 1st figure too, which makes it useless for me.) I was in danger of getting disheartened by that, so I used my usual remedy for that - go and do something different. But this time the something different is still part of this scatter thing project. Playing cards. Take a complete break from Poser, Studio, and scripting and play with pictures for a while. Using pictures from the links REIVAX gave me back in Dec 2019 !
French deck created directly from this by David Bellot (GNU LGPL version 3).
I'm working on I've done a Russian deck (still very much a work in progress) using images from the wikipedia Russian playing cards page, specifically this File:Russian playing card deck (face cards) Russian style 1911 original.jpg (public domain) and this File:Russian playing card deck Russian style 1911 original.png (also public domain). (edit - almost there now.
Edit 2 (19 Mar 2021): Russian deck completed to my satisfaction. Also uploaded a zip file containing the XCF (my GIMP original) and a PSD exported from GIMP.
Edit 3 (19 March 2021): That enjoyable little diversion gave me the strength to dive back into the boring stuff. I've imported the Poser playing card scatter figure (CR2) into DS 4.12 and applied a few of my Poser scatter poses (PZ2), then saved them as a DS native Figure/Prop Asset (DUF) and DS native Pose Preset (DUF) files. It may not look like much, but below is a DAZ Studio 3Delight render of four DS native 54 card scatters with the Set A/B/C/D poses applied
I've also attached some cropped Studio screenshots to remind me of the way I did the DAZifying and noted the directory structure of the files it created:
...\My Documents\DAZ 3D\Studio\My Library\data\3DCheapskate\TEMPORARY\Scatter\Playing Cards\
- Morphs\3DCheapskate\Base\
- Scatter Item Scale Multiplier.dsf
- Scatter Item Translations Multiplier.dsf
- UV Sets\3DCheapskate\Base\default.dsf
- PlayingCards2-5x3-5inches 100_2592.dsf
Footnote: I'm wondering about maybe trying I've started to create a French deck using Sönke Kraft's (aka Arnulf zu Linden) photo here.
And here's a thumbnail of the original which I'm using in accordance with the stated licensing "Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2"). I'm simply cutting a rectangle containg the desired card, using the GIMP cage transform to get rid the perspective, painting over any blue background bits, scaling the result to 114x171, and pasting into the relevant slot in my template.
(Note: the deck is a French deck, but with German letters on the face cards B, D, K(onig) as opposed to V, D, R(oi) - Wikipedia's French-suited playing cards page says they're 'North-German pattern, which are usually 32 or 36 card decks, and where 52 card decks have either 3 or 6 (additional to the 52?) jokers. Half the fun of doing something like this is the stuff I learn along the way ! My favourite so far on playing cards is the relationship between Bowers and Jokers on the WOPC Bicycle Playing Cards 1st Edition page
N.B. With 1024x1024 texture maps giving just under 114x171 pixels per card, these textures will be unsuitable for any render where the cards in the image are significantly larger than that. Only noting that because it just struck me. Also in Poser the default 'Quality' setting for the 'Image_Map' node makes them blurred and needs to be changed to 'Quality=None' (maybe the setting is actually called 'Filter' ? I know what I'm talking about anyway. Also in Poser the default 'Filtering' setting for the 'Image_Map' node, i.e. 'Quality', makes them blurred and needs to be changed to 'Filtering=None')
Edit 4 (20 March 2021): Note to self - in Blender my base card mesh has bottom at z=0, top (i.e. card thickness) at z=0.012112 x-size=±1.21125 y-size=±1.69575
Ooo I like that old looking set.
Please note that the Studio version is basically just the Poser version (see next post) imported into DS4.12 and then saved back out in Studio native format. The script included with this Studio version was simply a (bad) rewrite of the main PoserPython script - for me writing DAZ script is a chore, whereas writing PoserPython is slightly less so.
First public test release of the DAZ Studio native playing cards scatter figure
The attached zip file (DS Content (3DCheapskate Playing Cards Scatter Figure Version 0.1).zip) contains everything you need to take it for a test run. Subfolders in the zip are - data, Figures, Presets (Materials and Poses), Runtime (contains only textures), and Scripts.
Simply extract those folders to whatever Studio content directory you want. You may want to keep it separate from your proper stuff - maybe create a temporary folder on your desktop and temporarily map that as a Studio content directory. Then when you're done with it just delete the folder and unmap it from Studio.
You'll find files in each of the red-boxed subfolders in the screenshot below.
The deck loads scaled at 5% (remember that's 5% of its default size, which is 100 times real world size. Have a play with it. Don't forget the 'Scatter Item Scale Multiplier' and 'Scatter Item Translation Multiplier' sliders which are two alternative ways to tighten/loosen the scatter group.
Feedback, both good and bad (constructive preferably, but my skin's fairly thick!) appreciated.
Suggestions, etc welcome
Notes:
- The script is a very simple cobbled-together thing to apply randomish translations and rotations to the cards. There's nothing in it (yet) to try and avoid scatter items intersecting.
- The Set A/B/C/D pose presets shouldn't intersect with each other, even if you slide 'Scatter Item Scale Multiplier' up to 5 or 'Scatter Item Translation Multiplier' down to 0.2. However, that was based on each scatter item having a largest real world dimension of 1"... but these cards are 2½" x 3½"
Things I'm already considering:
- I want to change the OBJ name used by the Poser CR2 as that's used for the Studio data - \data\3DCheapskate\TEMPORARY\Scatter\Playing Cards\PlayingCards2-5x3-5inchesx100_2592.dsf
- Forgot to DAZify the Poser shuffle poses
First public test release of the Poser playing cards scatter figure
The attached zip file ('Poser Runtime (3DCheapskate Playing Cards Scatter Figure Version 0.1).zip') contains everything you need to take it for a test run. Subfolders in the zip are - Geometries, Libraries, Python and Textures
Simply extract those folders to whatever Poser runtime you want. You may want to keep it separate from your proper stuff - maybe create a temporary folder on your desktop and temporarily add that as a runtime directory. Then when you're done with it just delete the folder and remove it from Poser.
You'll find files in each of the red-boxed thingies in the screenshot below
You'll have to use 'File > Run Python Script...' to run the randomizing script because I haven't yet created the dummy pose file to run it from the library
The deck loads scaled at 5% (remember that's 5% of its default size, which is 100 times real world size. Have a play with it. Don't forget the 'Scatter Item Scale Multiplier' and 'Scatter Item Translation Multiplier' dials on the Body of the figure which are two alternative ways to tighten/loosen the scatter group
The Poser version has a couple of extra 'Shuffle' poses that I didn't DAZify (I simply forgot). This would probably be better implemented as a script.
Notes:
- The script is a very simple cobbled-together thing to apply randomish translations and rotations to the cards. There's nothing in it (yet) to try and avoid scatter items intersecting. The 'no UI version simply sets random rotations and translations using a fixed maximum distance for translation. The UI version allows me to randomize either rotations or translations or both, and lets me specify the maximum distance. Other scripts were to help me set up the CR2 and PZ2 files.
- The Set A/B/C/D pose presets shouldn't intersect with each other, even if you slide 'Scatter Item Scale Multiplier' up to 5 or 'Scatter Item Translation Multiplier' down to 0.2. However, that was based on each scatter item having a largest real world dimension of 1"... but these cards are 2½" x 3½"
Things I'm already considering:
- Change the OBJ name (to tidy up the Studio version - see previous post)
- Implement the 'Shuffle' poses as a script
- Improve the 'randomize' script to try and avoid scatters overlapping
Feedback, both good and bad (constructive preferably, but my skin's fairly thick!) appreciated.
Suggestions, etc welcome
I'll download and give it a shot tomorrow. Kinda sleepy and shortly headed to bed here at the moment. Looks great from the images, want to be fresh when I read teh instructions to use this.
I'm not sure how well this asset will work with your cards and chips here, but I created a one piece card deck and poker chip caddy for those hat might be interested in such thngs here:
https://www.daz3d.com/forums/discussion/516406/poker-chip-card-caddy-freebie-by-tjs-designed-for-feralfey-s-poker-play
I sized it for the FeralFey assets, but you should be able to scale things appropriately with different chip and card assets, for those that might be interested in such things...
I also deliberately kept the poly count rather low, at 354 polys, to help with render load a bit, and to help keep the scene file size down. I actually like how this turned out, but if people want a smoother version, I can look at making a higher poly version with more rounded edges down the road.
Enjoy!
TJS