How about a small tutorial for today. Today I’m going to show you how I use Dodge & Burn to correct problems with layers masks and luminance selections.

The problem

If you use mask, you have probably noticed a situation, where they just don’t select only what you want. It’s mostly visible around the horizon line, around buildings and similar. When you start to blend, it will just create ugly halos around objects. This is because the mask does not fit the area perfectly, and you also effect a bit of the surrounded areas.

Similar, if you for instance try to select an object with a lot of empty space in between (like for instance the Eiffel tower), this also results in a lot of unwanted selections.

In this example screen-shot, I created a darks mask, to select the buildings. But as you can see, part of the selection are also the clouds, which I didn’t want to include. I hope it’s visible, as it is quite faint.

The solution

You can of course just take the brush and paint out the unwanted areas, or try and select just them and fill them in, but using Dodge & Burn is just quicker. First to remember, burn makes pixels darker, dodge brighter. So when correcting a layer mask, one needs to use Burn on Shadows when trying to remove parts from the mask, and Dodge on Highlights when trying to add more to the mask.

Here, I used Burn at 100% strength and just went over the sky. As it was set to Shadows, it only took the darkest areas (the sky in this instance) and made them even darker. The bright areas of the building were not effected at all, or only very gently. If I wanted to do the opposite, select the buildings more than they are, I would go with Dodge and Highlighs. They would get brighter in the selection, so selected to a bigger degree. Doing this few times, can create a very strong contrast between the selected and unselected areas, but the transition will not be the best. If you use this gently, you will keep the original transition between light and dark areas.

As I haven’t taken any new photos recently, I thought I look back into some photos from last year, so there is a little diversity again on the blog. So for today, one from my London trip last year :)

This is a HDR from 5 exposures, created in Oloneo Photoengine, finished in Photoshop.

I posted some tips for most programs I’m using, but never for the panorama software PTgui. So I’m correcting this today with this blog post :) So here goes:

Increase the number of control points

PTgui is great in combining photos into panoramas, but sometime even it can’t create a nice blend. One thing that can really help, is to adjust the number of control points it’s using. By default this number is quite low, and it may be not enough for all situations. I usually change this to at least 50 control points per pari of images. You can go even higher, but too much may result in very bad performance in PTgui.

Open options

Choose a higher number

Load HDR series

PTgui has also support for HDR series, it just has to recognize them. That means, that every series that you want combine into a panorama has to have exactly the same number of exposures, and the same exposure times. If you don’t have this, the HDR series will not be recognized and you will not get the HDR pop-up. In that case, you have to check the exposures, as porobably one or more has a wrong exposure time and this needs to be corrected in the EXIF.

Loaded series

HDR exposure popup

Save blend planes

Another thing, especially for HDR, is the option to save blended planes. This is a must if you want to blend the HDR manually. When you are saving, you can choose if you wnat the blend planes or the HDR photo. Going for blend planes is better, as you can use them in any tonemapping software, and Photoengine or Photomatix give better results in tonemapping then PTgui.

You also can change the quality of what you save by clicking the text that indicates it.

Choose quality

Choose blended planes

Use 16-bit files

If you use 16-bit files (as you should) you may be surprised by the strange colors PTgui show. This is normal. It will show wrong colors, but the saved ones are correct. So no worries :) Also when you are doing a HDR series, dont worry that the colors and the blends don’t look the best. It’s just a HDR preview, and that is not how it looks after saving.

Strange colors when using 16-bit files

Strange look when doing HDR

Try different projections

You will notice that PTgui support different projection types. There are three visible, but you get much more by clicking the small triangle next to them. For most cases, the cylindrical is great, but sometimes it’s good to try other even if they are not what you were shooting for. The reason is that the distortion will be different and you may end up needing to crop off much less. Also play around with the two sliders on the sides. PTgui sometimes crops off stuff, and with them you can expand the area of the photo and so get everything you need.

Equirectangural projection

Rectilinear projection

Mask out problem areas

When combining photos for panoramas, it happens quite often, that you have something in one photo (moving people for instance) and a empty space in another. You of course want to use the empty space, but the program chooses the other photo instead. So in this case, you have to go into the mask tab, and using the green and red masks, choose areas that will always be included and those that wont be included. This will help a lot, but be careful so you don’t remove the same area from all photos, as you will end up with a hole. Also to much masking makes PTgui create worse blends, so use it only when needed.

The mask is copied to all the exposures if you are doing a HDR series.

Choose mask tab

Mask wanted and unwanted areas

And that’s all for these tips. Feel free to ask if you have any questions to them.

If you have a Canon camera, I suggest you think about adding Magic Lantern to it. It just gives so many new possibilities and makes it so much easier to use.

For today I chosen one of the more straightforward edits I did recently. As always, first the finished and the original 0EV exposure. As you can see, I needed only to get detail in the shadows and get the colors I wanted.

Finished photo

Original photo

I started as usual in Lightroom, correcting the chromatic aberrations, lens distortion and the horizon line.

All exposures

Lightroom edit

I synced the settings and exported all files as 16-bitt Tiffs to be loaded into Oloneo Photoengine. There I only changed the strength and contrast and saved the result. In Photoshop I then did the following edits (layers numbered from bottom up):

1. Oloneo Photoengine result
2. +1EV exposure, to recover the colors as the combined image was too red
3. Color balance to tone down the red even more
4. Color Efex Dfine for noise reduction
5. I still felt the image was too red, so I pushed the red hue a little more towards orange
6. Color Efex Pro Darken/Lighten center to create a vignette

Photoengine edit

Photoshop edit

And that’s all I did with this image. To find out more on how I edit, check out the guides and before after categories on this blog, or check out my video tutorial series here: