Dear all ,
I am working on a life long dream , to create pictures , night portraits with night scene added to the image. To say it short , taking a night portrait picture with moving light trails in the back ground. The last picture was taken on the spot. The 2nd picture is a gimp experiment that i created and the first picture i used gimp , with twirl and push setting and used many layers ( and time ) where i changed the position of the lights.
So what we look for is , an automated way so whe can use a layer of a picture , move it , copy it to a new layer , and this goes on as long as needed to see the long lights trails. My example has about 20 layers and will be about 20% of the layers it should need to create a very good result. I hope it is not a violation to share the drive location to see the pictures ? black wall with face and party lights.png - Google Drive and black wall with face and moving party lights.png - Google Drive
So you’re looking for R&D for your soon-to-be commercial app?
1 Like
i am just thrown into this by myself , during conversation with grok , i had the incredible idea , to save a picture in two layers ( or more ) , and after the picture is taken , it would be possible to create the stunning pictures using software or programs like photoshop and recently GIMP to create art pictures who would be the same or even better than when these pictures are taken with the app itself ( meaning on the spot ). I used photshop as autodidact , so my knowledge is very basic , but overlooking over the years hundreds of video’s , i know 100% sure , that it could be done. So my search is to meet people who master manipulating pictures and or can write the code/scripts to be used to create these stunning pictures.
It is not for the app itself. But to create a way so with gimp / whitin gimp / or as addon or so , it would be possible to use the settings of the tools , and create something like a drag the layer around ( with mouse , stylus or even use air keyboard , the last would be fantastic as than it could create the movement of the picture layer , as it would be in real live capturing these images ). I dont know if you have seen the 2 pictures i had made ( while learning to use gimp the first time ). But the picture with the moving background lights could be made automated with script , so when some one move the mouse fast and slow , the background lights would follow the movement , and the back ground lights are filtered to remove all other parts of the layer , only the lights , and set on an low opaque level . maybe each 1mm of slow movement and maybe every 3mm when move fast it should need to create a layer , i talk about about 100 layers as bench mark. At the same time some kind of brush ( like a water color brush ) need to go over all these light dots so the space between dots shows a trail from dot to dot ( very transparent , so it looks like moving lights ).
Sounds like you just need to use some pretty straight forward blend modes in gimp or a similar app.
We’re all about sharing results and learning to use tools that are free of change and give their source code away, so while i wish you luck in with life, this forum is not really the appropriate place to discuss an app that you plan to monetize and all that kind of stuff.
Shoot the subject and scene. In the GIMP, select the subject then invert the selection so as to adjust the background. In Filters>Distorts> select ‘Wind’ … adjust as required …
There are several other ways to skin the cat …
‘Whirl and pinch’ might be of interest for the background:
hI Cedric , thanks already for your positive reaction. I will go and give it a try. There is some potential in both options. i made last night next picture it took me about 3 hours … i will include a real live taken picture , the 3 layer picture with face in the middle , and the last night one.
Here i tried the distortion whirl and punch , in combination of many layers who are changed position a little bit. The wind setting is not usable as far as i tried ofcourse.
The idea is , to use this layer of the lights , not the background , and when you click on this layer and move the mouse , the image would follow the movement of the mouse, the mouse movement should leave a trace line on the layer image , so it can be seen and reset or change the line shape before showing all the created layers. Making each time a layer when the mouse is moved a certain distance ( 0.5mm to 5mm ) And with somekind of algorythme / script with settings for light trail between light dots - intensity - transparency - etc… / settings for smoothness of the light trail - when go slow it can skipp some dots , or add more dots ( layers i mean ) - variation of light intensity according to the mouse pointer speed. The function to isolate the brightest part of the light dots and make sure the borders do not have a shade, erase all dark almost black lines or edges , smoothen the edges of these light dots with a minimum of blur ( as i left a few without blur ) and you can see where i have errased parts of the light vail coming from the light dots. Or even not create dots but smooth light trails ( using smudge , etc seems to always merge the colors towards the end of the trail or smudging ) who use the shape of the light point to make the trails without mixing or blending the colors , have settings like transparency - transparency to the center of the light dot - opaque setting - contrast and saturation settings - etc. I was wondering if this way of creating 100 layers would not create super large files , but in fact each layer has only about 5 to10% of light dots of the image size, so maybe this should not create extra large files. I have no idea… just looking for options or solutions. the picture below is about the original picture that i started with.
OK. In my suggestion, I was assuming that the lights would be part of the background image (street, store lighting, stars, cars, etc.). But they could just be in a different layer than the background and be created by any means possible.
It’s not a question of offending one person in particular here. This forum is dedicated to free/libre software. It seems to me (and probably to most of us here) that you just came up with an idea about what you want, but that you have zero knowledge about how to do it. So you’re trying to get guidance from this forum, and hope people will help you for free so you can develop a paid app a get the money despite you came up just with the idea and not with the technical core of it.
I also think that you took a risk by exposing your idea here, because if some developer thinks your idea is really cool and if they have the technical knowledge to develop the app, they can do it by themselves without giving you credits.
So, if you want to make money from an idea, don’t share it publically unless you already know that you can make it happen very quickly.
Hi Cedric , Thanks for your feedback . it is the idea to use the complete background , but a more layer idea could open even more possibilities. For now it would be creating a way to manipulate a picture in the way described. The purpose is not that important.
I have added some more backgound lights and fixed some of the vertical lines in the light patterns. I think for now this is as close as i can get. But it does not come close to an original live taken picture
Well you edited your post to remove the parts where you talked about hiring a developer and monitizing the app, so it isn’t there anymore.
But I think you’re missing the point. We are happy to help people when the result is that everyone gets something they can use and tinker and learn from; that is what free and open source software is about. We are less giving of our knowledge and time when the results of the collaboration will be locked away from us.
Hope that helps.
1 Like
Your hostility is unwelcome here.
3 Likes
Longest paragraph I’ve ever seen anywhere … 
I didn’t bother reading past the third line. I’m done with it now.
1 Like