If you cannot control the artificial lighting (like the lobby of a hotel) then this method is not very useful. However, if you can isolate the ambient from the artificial, then you can mix them in post, and correct one to match the other. An alternative is to shoot the artificial lighting at night, when the ambient daylight is not as strong in the scene. Then you can combine the ambient daylit scene with the artificial light from the night shot.
Again, this needs to be done in HDR (32 bit). Once you have your 32bit data, you can add light, use the Photo filter to tame a cast on a per light basis, etc. You can stack the lighting layers and use linear dodge blend mode plus and exposure adjustment layer per light to mix light color and level, as well as a master exposure layer to control overall exposure. Then export the document as a 32bit TIFF and you can tone it in ACR, or apply the Camera Raw filter in PS CC as a smart object to get the overall look you want.
Obviously, the applicability of this method is not suitable for all environments.
Here is how I have done this previously, from an old post on another forum - I did not mix any ambient light, but the ambient shot would just be another HDR sequence in the layer stack. The end of the write up below deals with a color cast from a CFL bulb in the same scene as other bulbs, correcting the cast from that single bulb to match the other bulbs. You will also see how, when you apply a color filter to the HDR data for a single light, you can relight the scene with that lamp and the contributions from that lamp are rendered accurately for light cast on the scene, as well as reflections of the light source in the scene.
If you have access to the artificial lights, in terms of turning them on and off, you can shoot ambient with them off, then shoot ambient plus each light group. You can use the same approach and balance ambient with artificial.
***OLD POST***
HDR Lighting Tutorial[/color]
So, I did this exercise a while ago and, frankly, I may have already posted this. However, it bears repeating now that HDR techniques are ubiquitous and not everyone is aware of the true power of HDR imaging.
This exercise uses HDR image acquisition to create a scene-referred luminance representation of a typical interior lighting scenario. So what? Well, when you acquire a scene-referred luminance map of each light source (individually) you can mix them together to create ANY lighting combination from those light sources for that scene. It's like having dimmer switches for each lamp and being able to control them all to create any combination you want!
In this exercise, I acquired an HDR image of the scene for one light at a time – three lights in the scene meant three HDR images. The typical HDR image acquisition process applies – you vary the shutter speed with fixed ISO, aperture and white balance to capture full dynamic range for each light. Although it may not be so obvious, you want to do this at night so that there is no stray ambient light affecting your lamp HDRs. That doesn't mean you cannot also acquire a daylight HDR shot with no lamps to add to the mix … the options are endless.
You can also do this for flash photography, by varying the power output of the flash to change the EV of each shot in the HDR sequence – that technique requires some additional know-how and is beyond the scope of this exercise. But, think about it this way – you can use ONE flash, repositioned in a static scene, to simulate a studio of multiple flashes, that you can mix in post to create any mixture of lighting ratios you want. Totally cool.
Anyway – back to the exercise.
Methods – Setting up the scene[/color]
So, Figure 1 is the basic scene with the three lights labeled for reference. This is a screen grab of the 32bit image preview in Photoshop, so there is basic gamma tone mapping applied for display. In this screen grab, Light one is “on” and the other lights are “off.”
Figure 1 – The basic interior scene with the three lamps labeled for reference. This figure is screen grab of the 32bit preview in Photoshop and is illuminated by Light 1 only. incidentally, the orange color i used for the labels is the same for all three labels, but the label for Light 1 appears darker because it is in the proximity of the very bright area of the image from the lamp while the other labels are on darker background and appear lighter - your brain is awesome, isn't it? Light 1 provides enough illumination to see the scene, which is the point of this figure. Incidentally, I placed a Color Checker under each lamp as I shot the sequence for that lamp. Lights 1 and 2 are CFL bulbs and Light 3 is a straight up incandescent bulb. You can white balance for each light source if you choose, or shoot all lamps with the same WB. The nice thing about HDR data is that you can WB in post with no problems – more on that later.
Figure 2 – Our scene (screen grab of 32bit preview) with all three lights “on.” Figure 2 is a screen grab with all three lights at 0 EV exposure. Here is a good spot to explain how the document that contains all the data is constructed. The document is a 32bit file with each of the HDR lamp images brought into it on its own layer. Figure 3 is a screen grab of the layer structure for the document. For each light there are three layers (in stacking order):
1) A Photo Filter adjustment layer*
2) An Exposure adjustment layer*
3) The HDR layer for that light*Note that the adjustment layers are “clipping layers” set to affect ONLY the image layer to which they are clipped. To set an adjustment layer to clip to the layer below it, hover your cursor over the boundary between the adjustment layer and the layer below it while holding down the “Option” key (Mac) or “Alt” key (PC) and then click on the boundary.Figure 3 – The layer structure for creating the HDR manipulation of our scene lighting. Note the background layer is set to “Normal” blend mode, but the subsequent image layers are all set to Linear Dodge (add) blend mode – this is the key to getting a physically correct result. The critical and elegant aspect of this technique is that when all of your data are in linear, 32bit form you simply add the contribution of each light to get the correct mixture as would occur in real life. That is, the first light (the BACKGROUND layer) blend mode should be set to “Normal.” However, and this is the most important part,
The blend mode for subsequent light layers should be set to LINEAR DODGE (ADD).
This is super important. It tells Photoshop to add the contributions of the subsequent light layers to the background, which is the physically correct method of combining the lighting to get an accurate result. NB: ALL of the adjustment layer blend modes should all be NORMAL. Figures 4, 5 and 6 are screen grabs of the 32bit preview for each individual light on its own in the scene.
Figure 4 – Light 1 only in the scene – 32bit preview.Figure 5 – Light 2 only in the scene – 32bit preview.Figure 4 – Light 3 only in the scene – 32bit preview. Methods – Controlling the lighting[/color]
So, now we have our scene set up – time to turn the lights on and off! To control the “brightness” of each light, you simply increase or decrease the Exposure for the specific light. You can see the general effect as you mix the lights together, but it is easier to think in terms of lighting ratios – that is, Light 1 is twice as bright as Lights 2 and 3, etc. I also stack an Exposure adjustment layer at the very top of the stack to control overall exposure of the scene – you can use this, as well as the 32bit exposure control slider in the lower left of the window to control the exposure in the preview.
That’s it really. You are still operating in a 32bit world, so with all of the adjustments you make, the lighting is being added together correctly to create the final mix. If you had the patience to leave your camera set up on a tripod all day, you can even add in that daylit HDR scene to the mix if you choose.
When I get the mix I want, I Duplicate the image and then flatten the duplicate and then use HDR toning in PS or whatever HDR application I feel like to do the tone mapping, etc.
Extra Bonus – Get funky[/color]
So, you may be asking yourself, what is the Photo Filter adjustment layer for? Guess what?! Not only can you adjust the light’s brightness, but you can gel each light. This may be useful for subtle color balance or for creative awesomeness. Figure 7 depicts a purplish gel applied to Light 2, for example and Figure 8 turns my ho hum living room into a disco.
Figure 7 – Light 2 with a purple gel.Figure 8 – Woot! Freak out. Note that in the disco image, the reflection of the orange gelled lamp (light 1) in the glass of the framed painting to the left of the fire place mantle is correct (see Figure 4 to get an idea of the reflection). The colored lighting falls on the objects within the scene and mixes correctly. This kind of power gives you the ability to turn day into night, or plain into awesome.
So, that’s that. HDR is more than making dramatic tonemapped images. HDR imaging provides you with the ability to control light after the fact. When you shoot an exposure sequence of a single lamp and combine it into a single HDR data set, you have captured the scene-referred luminance map of the environment for that light source. You can manipulate it to relight the scene to your heart’s content and change the look of the environment in subtle and creative ways.
Boom, POW!
Have fun,
kirk
Here are some tone mapped examples. In both cases, I saved the 32bit flattened file as a TIFF and used ACR to tonemap - yep, in case you were unaware, ACR 7/LR 4 can tonemap 32 bit files, as long as they are TIFFs.
kirk