PhotoShop Editing Tricks

Anonymous

Sorry to bring this topic up again, because I know stuff like this has been posted before- and good stuff at that! However, it looks like a lot of those older posts have been removed. I finally have Photoshop and some okay pics from my new Canon Digital Rebel... Now its time to figure out how this program works. Anyone willing to give advise again (with pic examples if possible), and a quick note on what they use to edit--- I know myself and many others would really like the advise! Plus the monsoon here in AZ is kicking up again later this week, and I'm hoping to catch my fist money shot of some good lightening :wink:
Thanks! :D
 
Photoshop

That's a pretty open ended question considering the scope of what Photoshop can do.. and what people might do with it.

Cropping, resizing, colour adjustments etc are fairly easy to master, however you may need to do some reading to get into any production type work. One of the best tutorials to use is to have a peek at the quick start guide that comes with the program; lots of good tips in there for some nice work that will lead you to more advanced techniques.

What are you trying to do with the program?
 
Just editing overall pics. Color, etc. What do people use in Photoshop to generally clean pictures up?
 
Learn to use layers

IMHO Layers is the key to Photoshop. That and never adjust a jpeg directly. Convert it to either TIF or PSD (Save as file type) keeping your original as a back up. Yes, this will increase your file size, but it will keep you from mistakes.

Workflow. The Canon camera (10D, Digital Rebel, etc.) seem to need some type of adjustment right off the bat. Usually in the area of sharpness. Never use the sharpening tools. Use Unsharp Mask instead. Somewhere in the neighborhood of 95%.

My general workflow on basic adjustments will look something like this:

Open the file -> Save as 1234a.psd -> Adjustment layer Levels -> Adjustment layer Color (If needed) -> Adjustment layer Brightness/Contrast (if needed) -> Flatten image (merge all the layers) then Image, mode, LAB Color, click on the channels tab. Channel A apply the guassian blur filter about 1.0 to 3.0 radius (depending on what it takes to help smooth the grain) and the same thing on the Channel B. Click on the lightness channel and apply the Unsharp mask filter to about 94% to 97% depending on how much it needs to be sharpened up. Click on the Layers tab and then convert the image back to RGB and save it.

That's my basic workflow.

There are "Actions" available to help reduce the amount of manual adjusting you may do. Some are free, some are available for a small fee. I suggest looking at www.fredmiranda.com for some decent atomation actions.

I hope that helps get you started. That's just a fast overview of some basic photshop skills.
 
While you're a Photoshop noobie, I would suggest you become familiar with the Auto-Levels command. It will fix brightness, contrast, and color to a pretty reasonable degree with the minimum amount of effort. With practice, you can get good enough at using Curves to actually do a better job than Photoshop at it. This trick will, however, take quite a bit longer to master. :)
 
I actually just got that program. For removing grain, it is indespensible. I couldn't believe the images I was able to salvage. The latest version can be used as a Photoshop plugin, making life all the more convenient. My favorite thing about Neat Image is that it allows you to shoot with a high ISO setting without the image degradation that normally accompanies it. When your source image has a low ISO setting, you can use Neat Image to clean it up to the point where you'll find things that you didn't even know where there.
 
Levels can make or break

Hi John G.

Personally, I don't like using the Autolevels command. In very few cases, I've seen improvement. It's just easier to adjust the levels manually. While curves will do roughly the same thing, It's way more subtle in how it works.

About the only time I use curves is on large prints and then very sparingly and usually to help pan out what I couldn't get in levels.
 
http://www.luminous-landscape.com/tutorial...photoshop.shtml

http://www.luminous-landscape.com/tutorial...workflow1.shtml

The first link is "quicker". And as for the monitor calibration constantly mentioned in there. It's not AS needed as he makes it sound. If you have the time and money go for it. Otherwise you can work with your photolab to get the image like you want and they'll save the file with the specified settings.

I myself have NO flow, which I should by now. I just tinker so much I never took the time to set up some set of steps. After a while you just see the image and know you don't need every step. Curves is really really nice, but most images won't need it. Then again, if we are talking sky and foregrounds, many WILL need curves. I've been using the burn tool to balance the sky down with the foreground, but that tool has some unwanted effects, like offing the color and burning darker mid-tones(or whatever) too much(but that burn tool is GREAT on a blue sky). Curves so far is sort of touchy, but I see it being the most usefull tool on there(as far as I know photoshop elements does NOT have curves). Always sharpen LAST. Use unsharp mask. Radius I think depends on your resolution. When I go to 300dpi I usually up the radius to 3.0 or more and then adjust the percent as you like(there is no certain percent). DSLRs I guess leave the image soft on purpose to give more control, so expect to sharpen, especially if you shoot wide open and/or don't have a great lense. And nothing makes an image suck quicker then it being oversharpened to where you can see the white outline/halo. I'd rather see a soft image then any of that. Yet I catch myself going too far many times.

RAW is a whole bunch of fun too. http://www.luminous-landscape.com/reviews/...pture-one.shtml

The canon ap that comes with the camera for converting really sucks. I don't use it at all except for just converting them to TIFF. I don't adjust them with it at all, which probably ain't the best thing to do, but it's steps are LEAPS and that is all you have...little control. Apparently capture one is good, but reading on there it's $500 just for converting RAW files..lol. Biggest thing to do if you find yourself not happy and ready to spend money is get a better lense if you don't have one already.

I constantly find myself having to adjust selected colors from the kit lense just sucking some serious ass. This is especially true with storms as the cloud you are pointing your cam at is likely in the "low light" department, often rather backlit. A cheap lense will just give you back what it thought was there(this is with center point meetering before anyone comments). So, since I'm down this road. Here is what I do to try and fix any colors you can see in your cloud that you know are off. Go to Enhance/adjust color/hue and saturation. Where it says Edit: Master, go down and pick any ol color and then slide the saturation to 100%. Now look down below in your control box there is a horizontal slider. You can skinny up your selection area. Now you can move that bar around and find exactly what colors are off in your pick....you just focus where the color just looks off and move it till that 100% increase "lights her up". Now you have your bad color, go back to saturation and move the bar as low as you like. Bad lenses will throw lots of fun colors at you and that is one way to tame them back down or remove them all together. One many of my images I have to do that step over and over trying to grab that exact color while leaving near colors alone. Grass is one fun one and the sun. You can grab a selection if something like grass or the orange in the sky is the same as your off color in your clouds(or whatever).

A camera shoots one exposure. Something in every image will be 'off', photoshop allows us to get the images closer to reality moreso then it lets us create fiction. I hate those people who say, "this was shot and nothing at all was done to it"....often film people. It WAS developed though??? With digital we ARE the developers, especially in RAW. And....shoot in RAW, you'll thank yourself later.

Read about the red tables down on this page.... Understanding Histogram in Photography (in Plain Language)

BTW if you are new, everything you could ever want to know is on that site somewhere.

http://www.luminous-landscape.com/tutorial...nd_primer.shtml

You know you have gotten good at photoshop when you can bring out an impossible photo, but find yourself not being able to do it again...lol. I'll find I have over sharpened an image or something and saved it, so I'll want to do it over. Then when I go to do it over, I can't. Nothing more satisfying then that....cept being able to do it more then once of course...grrrrrr.

Mike
 
Last edited by a moderator:
[Broken External Image]:http://www.extremeinstability.com/stormpics/stex1.jpg

[Broken External Image]:http://www.extremeinstability.com/stormpics/stex2.jpg

This was not easy...lol, but not terribly hard now. The first image is the converted RAW image with nothing done to it during conversion. Canon's conversion ap is a joke. I will try and remember what all I did to develop it. I think I first tried to use curves but wasn't having much luck. Most any movement was a bad one. I brightened up the foreground a smidge in that. I then gave up on curves with photo impact and opened it back up in PS elements 2. Used levels and raised the mid-tones(middle bar) till the foreground was close. I then took the burn tool and burned the mid-tones of the sky to bring them back down and to bring out the contrast that the RAW obviously lacks. I only use it at 100% and keep going over all but the foreground to make sure it's application is even. Because it will only go so dark until you re-click the mouse down again. So you can be assured of one exact level of adjustment where you use it. Problem is it wants to burn the shadows too much with it set on mid-tones. There was some shadowing there but not as dark as I got it. You can use it more like a brush at less then 100% but I myself am horrible at drawing and it never works well when I do it that way. The burning caused some unwanted colors so I did that thing I mentioned in the previous post to remove them. It was still a bit flat so I adjusted the contrast up a bit more till it felt right. That is close to what it looked like in person but the shadow wasn't so dark. The burn tool can be very usefull but it can be hard to use. I know others on here(like Aaron) can chime in on how to get the job done better. Some use a mask layer to adjust the sky and foreground seperately and then merge them. That would be sweet, but I have no clue how to do it and PS elements doesn't not have masking.
 
Yeah, I don't really have a particular flow I use much either. If anything, I at least go to brightness and contrast first, just to get a feel for the range I'm working in. The way I approach things is that I shoot with the lowest ISO possible, and the slowest shutter speed I can afford at a given focal length. This enevitably produces blurs when there's fast-moving objects in a low-light environment such as under a beefy updraft. But it gives me the least noise. Once I have the image in photoshop, I can slam all the various attributes and try to find a nice medium of information versus artistic composure. I don't necessarily believe that a good picture has to be what was seen with the naked eye at the time it was shot. There's often plenty of really interesting stuff happening that you may not have noticed at the time you took the picture. Here's a small example:

Original Image
[Broken External Image]:http://www.amazingcheese.com/StormTrack/Example_a.jpg

Photoshopped Image
[Broken External Image]:http://www.amazingcheese.com/StormTrack/Example_b.jpg

I never realized there was a nice little funnel in there until I did some work with this. From my viewpoint when I took the picture, it didn't look like much of anything was left. But after some enhancement, it became evident that this tornado still had a couple minutes left in it. I also didn't remember there being that much hail (though the dents on my truck would speak otherwise).
 
Oooo... workflow. Just depends on the picture. Typically I convert from raw to 16bit tiff. I then resize to whatever I am working with. If I am posting on web, I typcally just autolevel and shapern for quickness. If I am going for prints or backgrounds, it is a bunch more detailed.

First I'll convert to a color profile for the D30, then adjust the colors. If some shadoes need to be lightened, I'll do that next. Finally, after the exposure is corrected, I'll make last minute color changes, and then I run a complex sharpening action from www.fredmiranda.com. This makes sure only the details are sharpened, and iskips all the pretty background colors, reducing noise.

Aaron
 
This is the topic of countless books, very long books. I think there are some good suggestions above.

As for a RAW converter consider Capture One Phase One DSLR (if you have a DSLR). Cost is about $100 and you get superior conversion for way less than the regular C1 software. A better choice might be Breezebrowser. It is only $50 and converts many different RAW formats. It also creates decent web galleries.

 
Back
Top