why always jpg, why not .png? higher quality?
Mistara
Posts: 38,675
in The Commons
DS, CARRARA, and Poser can use png files. not sure about Blender.
bumps, displacement, makeup/tattoo layers be higher quality!!
submitting to your consideration 
Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
PNG also use a larger amount of memory and resources. This would become a major issue on machines with less memory or if you have limited memory like using Iray with a GPU.
todays pcs have so much more memory. noiseless be sooo much better bump and displace
8.5 carrara can use the png's alphamask
Iray converts the textures for rendering. If anything, jpegs slow things down. Uncompressed image formats seem to work better.
It's not just the resources that it would use up on a home system, it's also the server storage space that such files require. A model using jpg textures takes up about 1/4 the space as the same model using png's or another "lossless" format. Jpg's compress better at save and also compress better when being zip'd for upload and sale. Many sites (specially smaller ones) require that you use jpg textures to save on storage space on thier end as well. I used to do png textures but after getting complaints from various sites about file sizes I was forced to switch to jpg...
As much as I agree on the quality of non jpeg files, I can't imagine the file sizes of characters using them. I already cringe at characters I go to d/l at hundreds of MB of file sizes. Not everyone has a fiber-optic network, LOL.
Image compression is pretty good these days, and if a jpg is compressed at a quality level of 90% or more, you would need the most amazing set of eyes to spot the difference.
I took a typical 2048x2048 pixel texture file from my library ".../Genesis2/BaseMale/M5PhillipEyes05.jpg" and converted it to several other formats. Not exactly a true test because the original is in JPG and already exhibiting some loss, but it shows the difference in storage needed to save an image. Even one with lots of solid color space which GIF is supposedly good at.
It would be interesting to take an original texture in a lossless format and then output it in several JPG accuracies to see at what point differences are noted.
PNG needs about 5 times the storage of the original. Multiply that by the sometimes large number of textures sent with a complex model and the storage would quickly be come huge and so would the corresponding download time. Not saying it can't be done but the difference is significant.
PNGs are still larger than jpgs. When space and memory is an issue, jpgs are a better choice, especially when you're not doing extreme closeups to warrant the increase in detail.
Working, I use the highest quality. Posting/final image, I use jpg.
PNG and JPG images, decompressed into memory, use the same amount of RAM (and only slightly more if the PNG has an alpha channel). The physical file size does not dictate the amount of RAM used in memory when the image is converted to a matrix of RGB data. Programs like Photoshop will tell you how much memory a file takes in RAM. Check the Docs File Sizes indicator; the value on the left is the flattened size, which is how a JPG or PNG is decompressed into memory.
That said, a PNG is likely to be 4-8X the size of equivalent JPGs, assuming a JPG of reasonable compressed (not too much, not too little). The larger PNG will require more storage space for the assets (why we have 2TB drives these days), and will take a bit longer to load. However, once in RAM, the two types of images are equavalent.
JPG has always been a poorer choice for textures because of compression artifacting. To avoid that artifacting, the file has to be saved with a lower compression ratio, and pretty soon, it's better to use lossless PNG or TIF. The artifacting is not too noticeable unless you move in real close, but on some models, this is desirable. As is usual with this sort of thing, the true answer comes down to "it depends."
Forgot to mention that texture compression in Iray is likely to be less efficient if the texture contains JPG artifacts, as the artifacts introduces noise. The more noise, the less compression. The "cleaner" the image, the better the Iray texture compression will be.
Several years ago, I ran a series of render tests (3Delight) and TIFs were actually the FASTEST to render...the reason is that 3DL didn't need to spend time coverting them, just rewriting the header info. So even though they were the largest files, they rendered the quickest. A few seconds to a few minutes depending on the number/size/shot etc. Every renderer has a 'native' format for texture files and if the images are already in that format and don't need to be converted there will be a speed increase. Most of them are using some uncompressed format. The conversion process is going to result in some amount of quality loss...whether it is acceptible or even noticeable is the big question.
That said, for most things, the end results were often not that different to the naked eye. Yes on closeups there were noticeable differences, but 'average' shots, nope.
Also for working with the images...either editing/customizing or just creating different resolutions the uncompressed (tif/bmp) files are much better.
I'd rather ask Why not Webp lossless ??
http://cloudinary.com/blog/flif_the_new_lossless_image_format_that_outperforms_png_webp_and_bpg
yeah. I can't imagine my products with tifs and more pngs!! I'm already at a half a gig per girl for a FINAL product.
What I wouldn't mind is to have the OPTION to download a tif version of the files...especially skin tectures.
I chalk it up to the "we need a better standard, there's now one more standard" effect... Unlike things like giphy/webm, replacements for JPGs are solving problems that 99% of the populace do not have. And even then, formats like webm are still taking a long time to displace well-established formats like aGIF.
Back in the stone ages, Poser files used to come with BMP images.
we're growing used to the hdr file sizes and their hugemongous d/l sizes.
havent found a way to un-noise a jpg.
artifacts gut rection is to blame it on the render s/w first, then eventually need to drill down to the texture folder and hafta apologize
BUM !
I was going to say, as much as I love Lyoness' girls, the file sizes are HUGE. I want to spend my time playing in DS with them, not downloading endlessly. Besides which, I've never erally had an issue with jpgs for textures, since that's pretty much all I've ever seen used. Tif and png formats were always reserved for things like bump maps or specularity maps, layers we don't actually see in rendering.
and Tiff
I think what we have is fine; the main complaint I'm used to seeing is that they are usually higher resolution than what most folks want, and some users end up downscaling them
downscaling is the easy part
is the filterforge resource jpg or something noiseless?
I remember there being some comment when a skimpy bikini arrived on the market some months ago, whose download weighed in at a hefty 2GB. Not surprisingly the bulk of that was texture files, not the geometry of the bikini. Imagine how big it would have been if the PA had shipped the product with png maps.
There isn't much need these days for uncompressed BMP or TIF. Many (especially free) programs used uncompressed TIF to avoid paying patent royalties. Those patents are expired, so LZW compression can now be included without exacting a cost. Although there are compression options for BMP, most use those files uncompressed, which is unnecessary. Both TIF and PNG compression is lossless.
I've seen numerous attempts at alternative file formats that certainly best TIF of PNG on size. But they tend to concentrate mostly on storage size, and not other aspects important to graphics professionals. For example, TIFs can contain layers, so you can use them with appropriate software (like Photoshop). Keep the layers intact while designing; there's no need to produce a flattened version if the decoding software (which includes D|S) can read the layered TIF, and respect the visible/hidden layers . This gives you the benefits of working with layers, yet the flexibility of the rendered image only showing the layer (and layer effects) you've elected to make visible.
I played around with a few of the newer formats -- which are intended mainly for the display of Web graphics via a browser -- and found various issues with things like alpha channel quality. I've not tried FLIF, but it's rather pointless until the software we use in our workflows support it.
Yes, the "bump" maps were ".bum" which were actually bmp's if you dissected them to try to modify them. I can't remember the details but they were strange greenish beasties.
yes. my files are huge and I recommend that people downrez them if their machines can't handle it. I try to leave as much information in them as possble, not converting them to JPG until the very end.
Actually if you think about it they were similar in some ways to normal maps.
I haven't worked with normal maps much. I stick pretty much to abnormal maps.
For the quality, I'd be happy with that.
+1