Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Add frozen object from short exposure frame to long exposure frame
#1
I am going on a trip to the Yucatan in May and want to take good pictures inside some of the Cenotes (sinkoles that expose groundwater underneath). Some of these cenotes are underground with not much light (Beams of light from the sun come through from the holes at the top of the cave, and some stationary artificial lighting is provided from the people who work at these caves as well).

So my plan is to use my Rokinon 12mm f2.0 with my Sony A6000 on a tripod and set it to max aperture and long exposure, with maybe ISO of about 400 to 800. Do those settings sound about right to get some good low-lit cave pictures? Would there be any reason to lower the aperture? Keep in mind, I do not plan on going any further and setting up external light sources, etc. The cave should be lit enough to be able to grab a decent picture with long exposure.

Anyway, the more important question or concern I have comes into play when I want to get some action shots. Long exposures and people moving don't mesh, so I have to take a short exposure frame with flash from the same angle with the person, and then insert that person into the long exposure photo. For example, freeze the person when jumping into the water (in the air mid-jump). I think the only way to do that is with a computer program using layers or masks, correct? I don’t want to purchase anything, and I think this can be accomplished with GIMP on my Macbook pro.

Last night, I tried masking myself out of a picture (or whatever the term is), but it was very hard to do in terms of being precise. I tried to fiddle around with the lasso and the different detections (color/shape/contrast), to no avail. Maybe part of that was because I was on a small laptop without using a mouse. But I think most of it was because I am an extreme beginner and this is my first time playing around with photo editing tools.

Can someone please point me to some detailed Gimp tutorial videos either showing how to accomplish exactly what I want from start to finish (Add the frozen person from the short exposure frame to the long exposure frame), or if you don’t know of a video that explains all of that, then maybe some tutorials on something similar, like how to precisely mask out people.

UPDATE: I have included a picture of what I am trying to accomplish. The cave in question is pretty dark, so I am almost positive he used the method of layer masking to insert his self into a long exposure shot of that cave, from a short exposure flash induced shot with the same angle


Attached Files Image(s)
   
Reply
#2
You are on the right track. Your long exposure image in the lower layer, a cut-out of a person in the layer above. I agree that working on a laptop is not easy, a mouse makes a big difference.

The first thing is to understand the selection tools. There is some info here:
http://www.gimp-forum.net/Thread-Gimp-se...-beginners

Another method is to use paths. More difficult for a beginner to use, but work well once you get used to them. There is some info on paths here:
http://www.gimp-forum.net/Thread-Paths-Basics

If you can't get it right you can post a link to an image you want to cut out. Much easier to explain with a real life example.
Reply
#3
Im not sure what you mean with masking out.

The standard way of fusing parts of two images is using a layermask. Depending on the image you can also use resynthesizer, a plug-in that "maqically" removes stuff from images and replaces it with the surrounding, but blending everything real well.
There are other similar tools (G'MIC for example has an impainting filter), but i find resynthesizer is the best.

Other ways of touching up images is the clone tool and the healing tool.

When it comes to tutorials, im a fan of 'Phlearn' on youtube. Its for photoshop but most of the techniques can be reproduced with Gimp.
Reply
#4
Quote:Blighty
You are on the right track. Your long exposure image in the lower layer, a cut-out of a person in the layer above. I agree that working on a laptop is not easy, a mouse makes a big difference.

The first thing is to understand the selection tools. There is some info here:
http://www.gimp-forum.net/Thread-Gimp-se...-beginners

Another method is to use paths. More difficult for a beginner to use, but work well once you get used to them. There is some info on paths here:
http://www.gimp-forum.net/Thread-Paths-Basics

If you can't get it right you can post a link to an image you want to cut out. Much easier to explain with a real life example.


I have edited my original post and attached the type of photo I want to take. Check it out. Thanks for the tips, I am going to read up on these selection tools and try to work on this with a mouse and see if I can do any better. If you know of any more helpful videos, post them here. Thanks. Hope I can progress at this, and I will post my updates

Quote:Espermaschine
Im not sure what you mean with masking out.

The standard way of fusing parts of two images is using a layermask. Depending on the image you can also use resynthesizer, a plug-in that "maqically" removes stuff from images and replaces it with the surrounding, but blending everything real well.
There are other similar tools (G'MIC for example has an impainting filter), but i find resynthesizer is the best.

Other ways of touching up images is the clone tool and the healing tool.

When it comes to tutorials, im a fan of 'Phlearn' on youtube. Its for photoshop but most of the techniques can be reproduced with Gimp.

My apologies for a lack of better terms, as I am all new to this. But see the attached picture in my original post, as I have updated the post. Also see Blighty's reply, as he may have explained what I want more eloquently.

After reviewing those 2 items, do you think resynthesizer would work for what I want? Or G'MIC? All of what you are saying is over my head, so you may have to show examples of these programs or explain in a dumbed down way for me to understand. Or post to some videos relating to what I am trying to accomplish
Reply
#5
(02-24-2017, 06:50 PM)mikej411 Wrote: So my plan is to use my Rokinon 12mm f2.0 with my Sony A6000 on a tripod and set it to max aperture and long exposure, with maybe ISO of about 400 to 800. Do those settings sound about right to get some good low-lit cave pictures? Would there be any reason to lower the aperture? Keep in mind, I do not plan on going any further and setting up external light sources, etc. The cave should be lit enough to be able to grab a decent picture with long exposure.
If you camera is only half as good as what Sony promises, you can use 1600 or 3200 ISO easily, especially if you shoot in Raw mode and are willing to spend some time denoising the pictures. The problem with f/2 is that your depth of field may be too short, even on a 12mm, but a f/2.8 you start getting a usable hyperfocal.
(02-24-2017, 06:50 PM)mikej411 Wrote: Anyway, the more important question or concern I have comes into play when I want to get some action shots. Long exposures and people moving don't mesh, so I have to take a short exposure frame with flash from the same angle with the person, and then insert that person into the long exposure photo. For example, freeze the person when jumping into the water (in the air mid-jump). I think the only way to do that is with a computer program using layers or masks, correct? I don’t want to purchase anything, and I think this can be accomplished with GIMP on my Macbook pro.

Last night, I tried masking myself out of a picture (or whatever the term is), but it was very hard to do in terms of being precise. I tried to fiddle around with the lasso and the different detections (color/shape/contrast), to no avail. Maybe part of that was because I was on a small laptop without using a mouse. But I think most of it was because I am an extreme beginner and this is my first time playing around with photo editing tools.

Can someone please point me to some detailed Gimp tutorial videos either showing how to accomplish exactly what I want from start to finish (Add the frozen person from the short exposure frame to the long exposure frame), or if you don’t know of a video that explains all of that, then maybe some tutorials on something similar, like how to precisely mask out people.
IMHO you are possibly doing it the wrong way. You may want to replace parts of the dark background in the action shot (insufficient flash power) with light versions taken from the long exposure shots. This may require less accuracy (and maybe the lighten only/darken only modes will help).
Reply
#6
So if i understand you correctly, you have two images, one of the cave alone and another with a man jumping into the water in the same cave ?

If so, your approach is correct, you render the person and then put it on top of the other image.
There is no magic technique to render something perfectly, but using the path tool and a layermask is the most professional.
Especially because its non destructive and editable at all times. Something the lasso tool is not.

Photoshop has a few advanced tools like 'Refine Edge', but that isnt a guarantee for a better job.

What you want to do is not exactly a beginners task, so be prepared for a bit of a learning curve.

An alternative might be a G'MIC filter i have found to work quite well for face swapping. Here is a tutorial:

https://www.youtube.com/watch?v=2e6FikWMkaQ
Reply
#7
By the way, the photo you show wasn't taken as you think. There are large shadows on the jumper's body, which are a hint that this wasn't taken just with the camera's flash. So either:
  • He belongs to some body painting-sect
  • There was a big flash on the top right
  • The diver's picture was taken outdoors and Gimp'ed in.
Reply


Forum Jump: