Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
filter advice: removing people from landscapes.
- Thread starter jwt
- Start date
- Sort by reaction score
I don't know about actual in-camera stuff to do this, but you could just take multiple shots of the scene, layer them in photoshop, and then use layer masks to brush out the people.
http://dsphotographic.com/index.php/articles/how-to-remove-tourists-from-your-photos/
http://dsphotographic.com/index.php/articles/how-to-remove-tourists-from-your-photos/
Photoshop can do it more automatically than that. There is a video tutorial describing the process here: http://frederickvan.com/blog/2008/04/06/photoshop-tutorial-removing-crowds/
Thanks for your comments, but I already know how to do it in Photoshop. I'd like to take one shot rather than spend a bunch of time in Photoshop. Also, I'd like to expand my photographic capabilities with this filter technique, so I'd like comments to be focused on this approach. Thanks.
This is precisely the recommendation I'm looking for, and based on the lighting information I mentioned in the thread-starter, I've given you all the information you need to make this recommendation.
Anyhow, the darkest filter I can find is a 10 stop ND 3.0. I'm just not sure if it will be enough, or if it'll be too much. I need someone with hands-on experience to guide me. I've also heard of people coupling a linear polarizer with a circular polarizer for an adjustable ND filter. I'd consider this route if I wasn't so concerned that there will be color shifting.
If you are looking to shoot long exposures during the day, then a ND (Neutral Density) filter is probably what you need. They come in various stops, depending on how much you want to increase the exposure. You can stack the filters if you want, but if you are shooting FF and have a WA lens on, you can experience vignetting at the corners. Also you may not be able to achieve AF either, so you will have to manually focus your lens.
Yes, that's probably the biggest concern besides vignetting - with the low light getting through the lens your AF will be clunky and slow..
But when you're shooting landscapes your almost always set to 'infinity', so it shouldn't be THAT much of a bother.
But when you're shooting landscapes your almost always set to 'infinity', so it shouldn't be THAT much of a bother.
Funnily enough, my "day time landscape shots" tend to have a very wide range of exposures, different DoF requirements, different HDR or focus stacking, how well your camera does long exposures, long exposure noise reduction comes into play...
Plus, I'm aware I'm shooting with a Nikon D2x, which makes f/11 the limit of what I can shoot without running into diffraction limits. So no, you really haven't given enough information- nor have you defined what sort of people, what sort of movement those people make... But *you* have the data you need- just look at the apertures and shutter speeds you normally shoot at, then do the math based on subject movement- or just pick up a couple of ND filters and stack them up and see what you get. Aim for ~5-10m exposures as a starting point. Obviously if you're shooting in blue sky at noon at 1/500th of a second, 10 stops is only going to get you to 2s, but ~7-8 stops after that you'll be in the ballpark. If you've got an aperture or two of headroom that changes things. You may actually be better off combining multiple exposures off the same tripod position though.
You can stack NDs, obviously your camera's light-handling characteristics are going to have some say in how well that works- you might try some unfiltered long exposures to see what your camera's sweet spot is time-wise.
Precisely, but as far as I know, we both live on planet earth, so I'm asking for your judgement call based on the range of light intensity around peak daylight. This is not a scientific process. We are not in a controlled environment where I can quantify the amount of light and report it back to you. There's no way I can tell what the intensity will be a week or month from now when the shot presents itself. There's no way I can say 4 pm, because 4 pm in San Diego is going to have a different light intensity than London. I said "daylight" because that is the best information I can give you.
Sorry, Rebel XTi f/22 and 30" exposure limits unless I go to bulb.
What sort of people? What kind of movements? Is my guess any better than yours?
3 hours before sunset at f/22 required a 1/13 exposure, which means to max my 30 second exposure limit, I'd need an 8.6-stop ND filter. So, the ND 3.0 should work. Math is not the problem for me. The problem is that I don't have the practical knowledge to know if a 30 second exposure is going to remove random people doing random things from a wide variety of scenes. I'll be shooting in Washington D.C. this weekend. Who knows what the lighting conditions are going to be?
Some of you are trying to help, but I'm getting the feeling that none of you have done what I'm trying to do. Nobody has posted a picture of their work or described hands-on experience. I feel like I'm asking this question in the wrong forum. Hopefully I'm wrong about this.
Daylight is intentionally vague because I'll be shooting in a wide-variety of light intensities, and because that's all the information I have. I'm asking for help so I can prudently deal with this variety. If that requires multiple filters, then that and the type of filters is the advice I'd like to receive.
You could say apparent lack of knowledge and that argument would carry more weight. I'll tell you what--I'll start caring what you say when you post a photo you took using the technique under discussion as credentials to participate in this discussion. If you can't do that, and don't have anything constructive to say, don't contribute to this thread.
Your attitude is just crap. People tried to help and you just threw up on them. Granted, you may not have wanted a specific suggestion but so what? You don't treat people who are attempting to help with acidic responses.
I'd suggest working on your approach and thanking people for at least offering help. Maybe then someone will take the time to explore the topic to your satisfaction.
I'd suggest working on your approach and thanking people for at least offering help. Maybe then someone will take the time to explore the topic to your satisfaction.
You're doing a very basic photograph. Essentially, you are stopping the camera down far enough so you will have an elapsed time on the shutter enabling all motion in the frame to not register on the sensor/film. I have photos like this on 4"x5" and raw files, but the process isn't important enough to warrant me actively searching for them. Besides, your attitude warrants no such activity.
And I assure you I have credentials. Don't worry about that. And for more than just a simple time-lapsed photo. But if you keep finding that no one has the answer you're looking for, realize that either you're not asking the correct question or perhaps you don't understand the correct answer.
Long exposures aren't all that tricky with a digital camera, and yes I've done them. Several people have described the technique already - it's just not that complicated. But experience is your best teacher; and what specific filters let me do what I want may not let you do exactly what you want.
But in any case here's specifically what I own that lets me take this sort of shot:
Singh-Ray Vari-ND filter
B+W #102 filter (4x)
B+W #106 (64x)
Or that I'm asking the wrong people. I bet almost nobody in this forum could meaningfully answer any question I may have on the versatility of Rh(I) as a hydrosilation catalyst.
Instead of posting something that gives me confidence in your abilities, you attack my intelligence twice. Instead of showing me some credentials you say "I assure you I have credentials. Don't worry about that." But words are just that until you back them up. And until you do, I'm not responding to anything you say unless it's constructive.
Cool. What kind of exposure times do you use to get people out of your scenery? Is 30 sec. enough, or do you require something on the order of minutes? Do you notice any color shifting with the Vari-ND filter?
See, now "peak daylight" is different than "daytime." "Bright sun is different than "heavy cloud" too- Even at "peak daylight," you can have a 10-stop difference in base exposure.
But you _can_ quantify the light- you obviously have exposures already, and you can cull the EXIF from those exposures to see what your average, low and high ranges are.
And I'm disagreeing- you've already qualified to "peak"- which is hugely different. A lot of photographers are "sweet light in the morning" or "sweet light before sunset" landscape photographers- plus, with any new photographic endeavor, it's a good idea to minimize the variables until you gain enough experience- so I'd recommend starting at as consistent a light base as you can for your location until you get things "right" under the most repeatable conditions you can. Anything else will cause enough variation that you won't know which variable changed the equation.
Are you saying you don't have diffracion effects on an XTi at f/22?
It should be- unless you're trying to photograph places you've never been in conditions you've never photographed before. For instance, some landscapes have roads, some have paths and some have water- those variations are all different and again, shooting for one first is a good thing for consistency.
See- new information. For DC in late May, I'd say you're going to need 15m or better of exposure. FWIW, an ND3 isn't 10 stops an ND4 is two stops, an ND64 is six stops.
http://en.wikipedia.org/wiki/Neutral_density_filter
I haven't done ultra-long exposures on digital, for sure- but the lack of reciprocity failure means it's actually easier. While I've got a list of ND filters to purchase, I haven't yet because in small format cameras, my lenses aren't anywhere near one-another diameter-wise, so I've been pricing square filters and holders versus a couple of different round filters.
That right there is a perfect invitation for me to bow out. I'm done with this thread. Good luck.
To be honest, I've never tried to remove _everybody_ from a photo (and I like the various degrees of motion you get from people with, say, a 4-5 second exposure). I can guarantee that, in most settings, 30 seconds won't be enough. There's always one or two people that aren't going to budge.
The height of vanity is quoting oneself, I realize... but I wanted to clarify something.
Even a 15 second exposure will get rid of most people completely. It's just that there will always be an exception - someone sitting there, enjoying the sun; a pair of folks having a conversation; etc.
Edit: Another pitfall, when quoting oneself, is that you can accidentally hit the "Edit" button when you meant to hit the "Quote" button.
The answer to that is "It depends". The the person a jogger or some one sleeping on a park bench. It the latter then even an hours long exposure may not be enough. The required exposure time depends on how many of their own "lenghts" they move per second. Looking at my shots I'v taken I most just get "blurred people", not "gone people"
If you want numbers the person needs to not be in some loaction about 16 times longer then they were there. So if the person is walking 16 body lenghts per second you'd need a one second exposure (note lenght is in the direction of movment as project onto a plane normal to the lens axis.
The number "16" is pulled pretty much out of the air. 16 will give you a person exposed four stops below the scene. If you like 6 stops then use 64. You are going to need "way long" exposures to the point where you will have LOTS of problems with thermal noise in the sensor. You would be much better off doing multiple shots and compositing them together. The other problem with long exposures in landscape photography is wind. It will cause the gress and leaves to move and blur ad it auses clouds to move. You are best off taking a serioes of short exposures during lulls in the wind.