The Lift & Carry Forums - View Single Post

fredbloggs1969 · Today, 08:11

I have been looking into a slightly longer scenario-type AI video. As The Wolfe suggests, a bit of a story makes for a better experience.

I have uploaded a "proof of concept" 2 minute video to see how easy it would be to make such a thing.

https://www.mediafire.com/file/z7ej6.../0417.mp4/file

I appreciate that it will not be everyone's "cup of tea". It does lean very heavily into my tights/pantyhose on women fetish as well. (But hey, it is my story, so I do what I want)

The story moves a bit "quick", skips important plot points and is never going to win an award, but it shows a "direction" and, as I say, proves that something more complex can be attempted.

As I appreciate I have particular tastes, I thought it might be useful to share my "process" so those interested in a similar scenario can try their own with their particular ingredients.

1) I create the characters. I use Flow for all this work. I created both the women and the kitchen where the main action takes place. Flow is free for image creation. Example below.
https://www.mediafire.com/view/aqt71...61244.png/file

2) I "storyboard" the scenario. (This video was based on the first story I put on the old Yahoo groups 25 years ago, using Poser to create the characters). I use flow to create static images. Like the link below. You can add the characters and the background as "ingredients" to get the image.

https://www.mediafire.com/view/68888...%2529.png/file

The good thing about flow is that you can ask it to change the picture. This is very useful when trying to animate a lift, as you can get a carried image and one where the liftee is not lifted. Work backwards. Get the carry first, then ask Flow to show the liftee being put down.

3) I then use Kling AI to create the video. This is where it starts to cost you. Using Kling Video 3.0 is great, as it gives multiple angles, and despite the speech being a little robotic at times, it is fair. For the lifts, I use the before and after I have created in Flow. Sometimes I will extract the end frame from one video to start the next. I let Kling decide the angles and cameras in many cases, with only basic direction from me. I find that the more complex instructions give "variable" results. You may not get the right result, and a costly reshoot is sometimes required. The video cost me around $30 to "shoot"

4) I then put the completed shots together in the free version of Capcut. Cost prohibited me from doing complex scenes showing the 2 women moving from one lift to another, so I cheated and used transitions in Capcut. It is only a trial scenario for me, and I do not charge anyone for downloading, etc. (This will not be on my DeviantArt page). Maybe I will add more detail later.

Today, 08:11	#7
fredbloggs1969 VIP Join Date: Mar 2009 Location: Southwest UK Posts: 163 : 0 For This Post 334 Total	Proof of Concept Video. I have been looking into a slightly longer scenario-type AI video. As The Wolfe suggests, a bit of a story makes for a better experience. I have uploaded a "proof of concept" 2 minute video to see how easy it would be to make such a thing. https://www.mediafire.com/file/z7ej6.../0417.mp4/file I appreciate that it will not be everyone's "cup of tea". It does lean very heavily into my tights/pantyhose on women fetish as well. (But hey, it is my story, so I do what I want) The story moves a bit "quick", skips important plot points and is never going to win an award, but it shows a "direction" and, as I say, proves that something more complex can be attempted. As I appreciate I have particular tastes, I thought it might be useful to share my "process" so those interested in a similar scenario can try their own with their particular ingredients. 1) I create the characters. I use Flow for all this work. I created both the women and the kitchen where the main action takes place. Flow is free for image creation. Example below. https://www.mediafire.com/view/aqt71...61244.png/file 2) I "storyboard" the scenario. (This video was based on the first story I put on the old Yahoo groups 25 years ago, using Poser to create the characters). I use flow to create static images. Like the link below. You can add the characters and the background as "ingredients" to get the image. https://www.mediafire.com/view/68888...%2529.png/file The good thing about flow is that you can ask it to change the picture. This is very useful when trying to animate a lift, as you can get a carried image and one where the liftee is not lifted. Work backwards. Get the carry first, then ask Flow to show the liftee being put down. 3) I then use Kling AI to create the video. This is where it starts to cost you. Using Kling Video 3.0 is great, as it gives multiple angles, and despite the speech being a little robotic at times, it is fair. For the lifts, I use the before and after I have created in Flow. Sometimes I will extract the end frame from one video to start the next. I let Kling decide the angles and cameras in many cases, with only basic direction from me. I find that the more complex instructions give "variable" results. You may not get the right result, and a costly reshoot is sometimes required. The video cost me around $30 to "shoot" 4) I then put the completed shots together in the free version of Capcut. Cost prohibited me from doing complex scenes showing the 2 women moving from one lift to another, so I cheated and used transitions in Capcut. It is only a trial scenario for me, and I do not charge anyone for downloading, etc. (This will not be on my DeviantArt page). Maybe I will add more detail later.