Comments

Blacklisted
    Show Search Options

    Before commenting, read the how to comment guide.

    post #77277
    33 41 4 Q
    Blacklisted

      Saw these pics round on here of James Harriman skunk girl images .

      Gotta say thanks for doing this .
      I was sad when I found out he passed . Good to know there are modern fans of him

    • |
    • 2
    • sealofharps said:
      Saw these pics round on here of James Harriman skunk girl images .

      Gotta say thanks for doing this .
      I was sad when I found out he passed . Good to know there are modern fans of him

      I've been a fan of his art since 1998, and I'm now in my late forties. My favorite artist is Jim Hardiman. I originally discovered his work in Genus #32, and I've since acquired a large collection of his artwork, including comics, prints, and even some of his portfolios. I have the last drawing he made of the skunk girls, and it’s got his signature on it. It’s an unfinished group shot of them in swimsuits. His mom sent me some of his art through the mail because I was helping manage a memorial page for Jim on Facebook at the time. Sadly, not long after, Jim's friend Herbert visited Jim's mom and walked out with some of Jim's rare art pieces without even asking her for permission. He also started sharing photos of Jim and her online without consent, which really bothered her because she values her privacy. This made her so upset that she decided to cut ties with the fandom.
      I had just reached out to him a month before he passed, trying to offer some encouragement, but it seemed like he had lost hope. He desperately needed a kidney transplant, and his health was getting worse. When I found out about his passing on FA, it really hit me hard.

    • |
    • 1
    • Thanks for the backstory. I first saw images of his back in the day on dialup internet on the USENET feed, fur.artwork.erotica. Such a different time.

    • |
    • 1
    • chastitycoyote said:
      Thanks for the backstory. I first saw images of his back in the day on dialup internet on the USENET feed, fur.artwork.erotica. Such a different time.

      The first time I came across his artwork online was around 2002 in his Yahoo "Skunkworks AMA" group. There was this one photo he shared of himself in a tux at his brother's wedding, red faced and looking all mean at the camera. It was actually me and my friend Derek who talked him into posting that. His mom was not thrilled when she found out, and I kind of regret it now. But you know, you can't go back and change things.

    • |
    • 1
    • post #77265
      261 439 12 E
      Show all comments
      Blacklisted

        mittens893 said:
        I can't even find the words to describe how hot this is 🥵

        If only you saw the cursed generation results of all the other attempts. I didn't expect these clips to garner the kind of favs & attention they have, but glad you and others like them. :3

        farclaws said:
        damn is amazing, i can't wait to see what AI art will look like in 5 years

        It's fun seeing the tech grow in real time. I think it'll be mostly dialed in within about 1-2 years to the current level of ease/editing with images.

        fluff said:
        incredible. please god do this with a lioness.

        I do have some others in the works.

      • |
      • 4
      • dirtypig-2 said:
        wow so great quiality. Can you explain me how i can make awesome animations like this? Please

        Thanks. It's quite a few steps, but I can for you and others curious with how this works:

        You first need to generate a source image. The best way is to do it locally on your PC, since most online image generating programs do not allow any kind of NSFW content. You will need a decent video card capable of rendering image generation, and you can google where and how to download Stable Diffusion with an AI interface to run it. Most like Comfy UI, or a111. You will also need a model suitable for the content you want to generate. The main one for furry content is Yiffymix, which you can find on Civitai. This image was generated locally using the a111 interface with Yiffymix v52 as the model, and it took several dozen attempts before I narrowed down what I wanted, and then cleaned any minor errors using the inpainting tool.

        Once you have an image you like, you then need to use an image to video generation site. I use Kling AI, but you can try others like Hailuoai. There are trial periods for both of these sites, but you will eventually have to pay for credits. Both sites have a strict NSFW filter that scans the image you upload, and you have to be creative with certain elements to get the image accepted. Exposed boobs and genitalia are difficult to get by, but in the case of ferals, it is less restrictive since their "bits" are always out in a SFW manner. Even still, it may not be accepted if it thinks the image is naughty. You can sometimes get around this by turning the saturation of the image up to confuse the detection software, and then edit the color back to normal levels later with a video editing software.

        I'm using Kling as my example, but once your source image is accepted, choose the duration of the clip (5 or 10 seconds), use the motion brush if you want to give specific movement to certain areas, and give it a simple prompt of what you want to see, then let it attempt to generate. About 80% of the results you get will not be what you want. The character will move in a cursed manner, the laws of physics get thrown out of the window, body parts and attire will randomly appear or disappear, or it will just morph and shimmer a still image with no movement. There is no way to edit any specific errors, and you have to just retry the whole attempt if it doesn't work. You are very much rolling the dice, and the odds are rarely in your favor. Also, keep in mind that after your trial credits run out, a 5 second video clip (using the better 1.5 model) is about 35 cents (usd), and a 10 second video is 70 cents, so you will be burning some money on bad results, and they don't refund unless the attempt completely fails. It takes me about 20-40 generation attempts to get something like this and my previous videos. I have much better results generating 5 second clips and merging them together using Davinci Resolve, which will also allow you to color correct the video if you had to blow out the saturation of the source image.

        It's still early and restrictive for video generation, and I'm sure a guide like this will sound primitive in a couple of years, but image generation was in a similar boat not that long ago. In time, we'll be able to generate videos locally when it's more open source and optimized, so if all this sounds daunting, just give it a year or two. I'm quite excited to see what we'll be capable of generating.

      • |
      • 6
      • "ocillet":
        ...In time, we'll be able to generate videos locally when it's more open source and optimized, so if all this sounds daunting, just give it a year or two. I'm quite excited to see what we'll be capable of generating.

        Its going to be a heck of ride till then for sure.

      • |
      • 1
      • ocillet said:
        Thanks. It's quite a few steps, but I can for you and others curious with how this works:

        You first need to generate a source image. The best way is to do it locally on your PC, since most online image generating programs do not allow any kind of NSFW content. You will need a decent video card capable of rendering image generation, and you can google where and how to download Stable Diffusion with an AI interface to run it. Most like Comfy UI, or a111. You will also need a model suitable for the content you want to generate. The main one for furry content is Yiffymix, which you can find on Civitai. This image was generated locally using the a111 interface with Yiffymix v52 as the model, and it took several dozen attempts before I narrowed down what I wanted, and then cleaned any minor errors using the inpainting tool.

        Once you have an image you like, you then need to use an image to video generation site. I use Kling AI, but you can try others like Hailuoai. There are trial periods for both of these sites, but you will eventually have to pay for credits. Both sites have a strict NSFW filter that scans the image you upload, and you have to be creative with certain elements to get the image accepted. Exposed boobs and genitalia are difficult to get by, but in the case of ferals, it is less restrictive since their "bits" are always out in a SFW manner. Even still, it may not be accepted if it thinks the image is naughty. You can sometimes get around this by turning the saturation of the image up to confuse the detection software, and then edit the color back to normal levels later with a video editing software.

        I'm using Kling as my example, but once your source image is accepted, choose the duration of the clip (5 or 10 seconds), use the motion brush if you want to give specific movement to certain areas, and give it a simple prompt of what you want to see, then let it attempt to generate. About 80% of the results you get will not be what you want. The character will move in a cursed manner, the laws of physics get thrown out of the window, body parts and attire will randomly appear or disappear, or it will just morph and shimmer a still image with no movement. There is no way to edit any specific errors, and you have to just retry the whole attempt if it doesn't work. You are very much rolling the dice, and the odds are rarely in your favor. Also, keep in mind that after your trial credits run out, a 5 second video clip (using the better 1.5 model) is about 35 cents (usd), and a 10 second video is 70 cents, so you will be burning some money on bad results, and they don't refund unless the attempt completely fails. It takes me about 20-40 generation attempts to get something like this and my previous videos. I have much better results generating 5 second clips and merging them together using Davinci Resolve, which will also allow you to color correct the video if you had to blow out the saturation of the source image.

        It's still early and restrictive for video generation, and I'm sure a guide like this will sound primitive in a couple of years, but image generation was in a similar boat not that long ago. In time, we'll be able to generate videos locally when it's more open source and optimized, so if all this sounds daunting, just give it a year or two. I'm quite excited to see what we'll be capable of generating.

        Thank you for taking the time to describe the steps involved! Clearly the results just keep getting better!

      • |
      • 1
      • post #77248
        47 96 1 E
        Blacklisted
          post #77243
          111 159 2 E
          Blacklisted
            post #77238
            3 11 1 E
            Blacklisted