Forum Coordinators: RedPhantom
Poser - OFFICIAL F.A.Q (Last Updated: 2025 Jan 24 6:22 pm)
Well some of these I don't agree with but for the most part you are spot on. Don't agree with the PNG thing ... Can't save it that way anyway S Though I don't mess with jpg's until the final stage. And the Mat thing is beyond a lot of people. Myself included. Even with the tutorial I saw I can't figure out how to leave in 20 lines of gibberish from a thing that has well over 4000 lines of pure nonsense S Just gives me a headache. And 3000x3000 maps crash my machine. Or used to. But otherwise I like most of it S
On the MAT pose files I cheated bigtime, I just edited a MAT file I got from the Daz textures. Simple and fast... it's pretty obvious once you look at a small one :) I know what you mean about the maps - thats cool - there is room for "lo-rez" textures so to speak :) As for the PNG - I just hate the loss of jpg's - but with a high quality setting (9 out of 10) they compress OK and have a limited number of errors :)
I never actually used a PNG, but I just ran a test, wow, a 1536 x 1536 test file was 600k as a high quality JPEG, 300K as a low, PNG was about 220, and quality was much better than even the high quality JPEG, none of those pesty artifacts. What is the downside? Poser 4 on the Mac can use PNGs, have all versions of 4 been able to use them? Is is going to bother people not being able to open (and edit) the PNGs on paint programs that don't support them?
I find this interesting because almost all of your rules seem to point at techniques I used when making Odyssey and Vickybuster. Haha, do you not like those textures? ;) Good ideas though, even though I don't agree with four of them. :) I'll try to keep these in mind the next time I start on a texture anyway. Take care! SnowS Hoping his pictures are worth 1001 words.
my DeviantArt page: http://snowsultan.deviantart.com/
Â
I do not speak as a representative of DAZ, I speak only as a long-time member here. Be nice (and quit lying about DAZ) and I'll be nice too.
To be honest, I haven't looked at them really close :) I'll go do that. I am not sure what the PNG downside is, all the tools I use can handle them and there are many free utilities out there that can convert them for those folks who can't use them. textures are pretty much the wrong type of image for JPG. They have non of the high degree of variance it was designed for (photographs, usually outdoor scenes with lots of chaos but little sharp detail) and thet need accuracy at small levels (< 3 pixels). Thansk for running the experiment Jim, and posting the results!
I can't even save my textures as PNG S I hate JPG and the loss that it has but I normally work In Photoshop as I working on then thing and then tif at the end. As far as opening it that would bug me as well ... I rarely use a texture as it starts. As far as the lo res ones ... the only place I have seen a difference is in extreme close ups which I normally don't do or see much point of S And there isn't that many utilities to convert the PNG for some folk S
While my working progress in texturing i'm using the Photoshop internal .PSD- format with all layers for rendering. In this way i take control over the seams and be able to correct it at once, save it and render again and so on. The final Texture i save as TIF using LZW-compression. It's smaller than the PNG and it takes the same time in rendering.
Paint Shop Pro 5 opens and saves png like a dream :) The other thing you may not know (or probably do s) is that you can use the native Photoshop file format '.psd' on your texture maps :) Poser loves 'em. When creating transmaps, I always save them as psd (yeap, in Paint Shop Pro LOL) so I can fiddle with layers as I'm viewing the work. It's way cool. s Xena
PNG was not intended by its authors to compete with JPEG and the PNG specs are clear on its inadequacy to compress as economically as JPEG. PNG was specifically created to bypass Unisys' insane patent on GIF's LZW compression. A replacement format for small images was needed as JPEG isn't suitable for those. Curious Labs' decision to switch to PNG for thumbnails is borne of the fact that if they had chosen GIF compression each copy of Poser would incorporate a licensing fee to Unisys. I've read the Unisys licensing contract, it's 27 pages long, and it sucks. Aldus/Adobe's Tag Image File Format (TIFF) is indeed rich. It's also data-bloated; this format was designed to carry enough data about 24-bit images to make color separation easier from PageMaker (the sole reason TIFF was invented). Whether compressed or not, the uncompressed version in memory is huge. Also, there are mutually incompatible compression schemes for TIFF which can plague users who choose the wrong one. In general, good JPEG compression (e.g., 60%) should be without visible artifacts, and the files should be suitable for renders. The high res textures DAZ is offering for Vicky2 are nice, but you don't really get skin realism for your buck (unless you think your skin's texture looks like the Ground bumpmap) and the memory requirements are unpleasant. Start hurling rotten fruit at me now.
Ok... a few things.... 1) The PNG format is not simply a JPEG patent protest - it includes many features JPEG does not - including alpha channel support. 2) PNG - like TIFF is a loseless format, so it will not degrade the images over time. 3) Jpeg was meant to be suitable in loss for viewing of images with the naked eye assuming the image has little small detail. When you use a Jpeg as the input to another approximation process (rendering a texture map is a process that has to take liberties with the input map - it is an approximation) those flaws are magnified dramatically in the final render. In other words, you want to use Jpeg to store a family snapshot? cool. As a texture map it is the death of high quality renders. 4) TIF is cool. I can live with it if you all ship textures as TIF :) anything LOSELESS is fine by me. 5) The daz high rez maps are, so far, the most real I have seen. i am happy to see something more real for close up renders :) 6) Misconceptions of the Unisys patent are rampant. Poser would not have had to pay a fee for each copy shipped - they would have had to pay a license fee for the library they used to manage the GIF format. As an example, adobe does NOT pay a per copy fee for Photoshop. 7) the use of PNG for thumbnails in poser was driven by something other than GIF license fees - if that was their concern they could have used JPG. 8) using PSD's directly is a cool idea! I'll try it though I doubt it will work for me - I usually use the latest layer effects and they are rarely supported outside of PS itself. So poser can see PS 5 layer options but not always 6, for example. (try a layer effect glow and let me know :)! Thanks for all the feedback! Cool thread!
JPEG is an optionally 'lossy' file format; quality is rated on a 1 to 100% setting, with 60% being a reasonable standard for photos. What it does is (based on your setting) look for similar rectangular chunks in your image, average them together, then replace the similar chunks with the averaged version -- THEN compress the image, which now has more redundant data in it than the source image. At settings below 60% patchiness becomes visible in the output image, especially around text, and the image degrades because A) the patches (the averaged chunks) don't blend in smoothly with the backgrounds any more and B) the free hand you've given the computer to decide what the term 'similar' means results in a blurry-looking image. What most people don't know is that the JPEG standard permits assigning different levels of lossy to different regions of the same image, which would permit text portions of the image to be crisp while the rest was handled at a regular lossy setting. Unfortunately Photoshop doesn't support this yet in its export; you have to buy a third-party plug-in to handle that. The JPEG2000 spec is arguably the coolest yet. Among many improvements, its compression scheme stores the same image at multiple resolutions, which means better gradual image loading and more importantly, the opportunity for web and print software to pick the appropriate resolution for the task at hand. The implications for choosing render resolutions for prototyping and final renders are obvious. Unfortunately JP2K is only implemented in a few experimental science lab applications and no browser vendor is going to support a format no image editing software can export. I've skimmed the Unisys contract. While what you say may be true, I do recall the contract specifying an estimated number of licenses in a year to help define the rates -- a method they got from BMI/ASCAP and the rest of the publishing industry.
Interesting thread. Strongly agree with "let the geometry do the work" concept. (Goes back to the old debate on which is more important in making a "looks-like" face - modeling vs. mapping.) Strongly disagree though, on not bump mapping skin. I love what a lightly applied random noise map will do to break up the edge transitions on skin highlights.
Kageboshi, jpeg is a safe and useful format, as long as you use it properly. It only degrades over multiple saves, because each time more data is thrown away. That just means that you should keep your original/master files (such as your Victoria textures) and to adjust the texture always go back to a fresh copy, not a second or later generation that's been saved in jpeg. When I make a change to an original jpeg, I always save it as a layered Photoshop/PSD file, since I'm sure to want to go back and tweak it more. If you do that with the jpeg saved file, it's gonna get ugly!
Incidently, when I did the lingerie for Supermodel Vickie I wanted the customer to have the option to make stocking overlays on other Vickie texture maps, I came up with a way to do that using a JPEG, assuming they have Photoshop or something similar that supports alpha channels. They can paste the jpeg into a new channel and make that a selection, then just fill the texture map with the desired color (black is tradional!). I don't know why I hadn't thought of that before. In the past I've emailed Photoshop files with the stockings in a layer, but they are huge files. I have no problems with medium to high quality JPEGs for body and face texture maps, if anything the artifacts help by giving it more randomness. And I think 3000 x 3000 body maps are silly, unless you plan to render close ups of fingers or something!
I always found that to reduce artifacts and get good quality out of a jpeg you had to save it at about 95% or greater, so that it compresses only the black and dark grey pixels, it still saves you a fair bit, but it's nowhere near as economical as setting at at 50 or 75% but if image quality is your main criterion it makes no sense to use anything less. later jb
Well, we render "final fantasy" stule eye close ups and so on for clients, and both the JPEG flays and lo-res map flaws are fairly obvious... but then, we are competing against renders from firms using orders of magnitude better software :) I don't worry about saving space - hard disks are cheap and bandwidth is not a primary concern so we stay with lossless formats as much as possible. But thats just us :)
This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.
I have been working steadily on getting good, realistic close ups. This combines with the recent thread on what makes a good texture led me to these thoughts - take them for what it is worth :) Soulhuntres RULES OF TEXTURE (ROT) version alpha .1 :) RULE 1: let the geometry do it's job! it is 'easy', pretty and seems like a good idea to force detail on the model by drawing right onto the main texture. Things like shadows on wrinkles in hands and feet and highlights on lips to suggest gloss. DON'T DO IT. If you paint light and shadow onto the texture, you create problems with many types of renders. The renderer will do highlights and shadows and gloss for you - in a way that is correct for the scene.. if you follow the other rules :) RULE 2: Speckles bad! that nice dot pattern effect that looks great in Photoshop to give the skin some texture gets distorted all to hell when the character morphs or moves... and the seam between it and the head is way too harsh. Take a tip from Daz and use a more organic color pattern - blur and smear my friends, can help you :) RULE 3: use the other maps Not just for eyelashes anymore, transparency maps can be used for lots of things - a missing tooth or adding some character to your next creations smile (fangs, anyone?). EYEBROWS on the brow geometry, not painted onto the head :) Bump maps are the appropriate place for detailing the hands and feet, not to mention getting a nice effect to the skin. >the renderer will do the right thing if you help it< Other maps can help you add gloss to your 'whet look' lips - the right way :) RULE 4: Test with the default lights This is important - not everyone is using your ligting setup - and it is too easy to use lights to cover up or compensate without knowing it. If your skin color isn't what you want in the default lights or close to them then you need to play with the map. You cannot predict the ligting someone will use with your maps... so try and be general. RULE 5: resolution, resolution, resolution 3000x3000 at least please.. we can make smaller ones ourselves. RULE 6: jpeg is dead - Use PNG for your texture maps, they work in the pro pack and they don't suffer the same detail loss as jpeg. If I am going to have to load your texture into photoshop and add makeup or something to it I don't want to lose detail because of that so I have to save as PNG anyway. RULE 7: give me MAT or give me death lip and eye color, skin tones and makeup styles .. all nicely matched out please and available with one click. Anyway, there they are - so far :) I am not really trying to lay down ;rules' - it was just more amusing to write it this way :)