2:56 amOff
Apertures for high-contrast ratios
It's been said that in order to be able to setup lights in high-contrast ratios (say, 16:1), the aperture f-stop used should be large enough to allow the fill to be mesured. The rationale behind this is that if the main light was setup for an aperture of say, f/4, then we wouldn't be able to read the fill, which would be four stops down.
In this case, this is flat out false, as f/4 -> f/2.8 -> f/2 -> f/1.4 -> f/1 is enough to do a reading (the light meter cannot display anything below f/1, a feat achieved only rarely in the history of photographic lenses in any case — altough you may be among the few who can afford a Leica Noctilux f/0.95).
But what if we wanted to use f/2 for the main, to produce a very shallow, dreamy depth-of-field, yet still maintain a 16:1 ratio?
The fact of the matter is that light intensity has no limitation in "dimness" — and the depth-of-field we want to achieve is more important than some mathematical restriction, after all! Although we must use an aperture our lens will ultimately be able to yield (for the exposure), it doesn't matter what "aperture" corresponds to the fill light, as long as it is enough stops below the main so as to maintain the desired ratio.
Since the light meter cannot display apertures below f/1, we can simply raise the light meter's (not the camera's!) ISO setting a couple of stops higher. For example, if we raise the light meter's ISO from 100 to 400, just for the sake of calculating the ratio, we gain two more stops that can be displayed. The main would register as f/4 (two stops higher than our desired f/2 at ISO 100, just like we expect), and our fill would register as f/1. No problem!
(Just remember to set the ISO back to 100 when you are done calculating your ratio...)
5:35 pmOff
Direct reading using a Sekonic light meter
The default setting for how readings are displayed on the Sekonic Flash Master L-358 is in full shutter stops, with 1/10th aperture stops, as it was traditionally.
Various reading display settings on the Sekonic Flash Master L-358
For those who have learned their f-stops in the digital era, it usually makes more sense to talk of thirds of stops. If you want your Sekonic Flash Master L-358 to display direct readings, enable the DIP switch #3. Switch #4 determines if the reading will be in half or thirds of a stop.
Here's how my unit is normally setup:
DIP Switches on my unit
4:49 pmOff
LightRoom Print Job
The Print Job panel of Lightroom's Print module needs some explanation, because as it is, it looks fairly simple and straightforward, but it is also quite opaque about what each parameter is actually doing.
Lightroom's Print Job panel
Print to. In this drop down menu, you can decide to export the layout to a JPEG file instead of sending it to a printer. Instead of simply sending the image content to a file (as the Export function would do), this exports the whole page—that is, the image as well as the white border, the watermark, etc. This might be useful if you're sending the file to a professional print service—check with them about the appropriate parameters to use for resolution and color profile (a common requirement is a 300 ppi sRGB).
Draft Mode Printing. This mode is mostly self-explanatory, yet what needs to be said is that the image data used for the print is not extracted from the original files, but rather taken from the (lower quality) image previews contained in the catalog. Therefore, depending on the size of the previews you have configured Lightroom to generate, the quality of the draft print will be affected. (This might not be a problem if you are printing contact sheets, for example, since then the images are pretty small to begin with.) The advantage though, is that you can print in Draft Mode even if the original files are offline—all you need is for the images to have been cataloged.
Print Resolution. This parameter defines whether to resample the image data to a given resolution or to use the native resolution of the file (whatever is necessary to print the image to the desired size). Indeed, according to Adobe's documentation, checking that box and setting a resolution will "resample the image data if needed, depending on the print resolution and the print dimensions". It then goes on to say that "To use the native resolution of the photo (as long as it isn’t lower than 72 ppi or higher than 480 ppi), deselect Print Resolution."
As it's been said before, the general wisdom tells us that we do not want to resample our images if we're outputting the job to a printer (unless the resolution reaches extremes), so we should usually leave this box unchecked.
Now, this is a very important parameter that has seen its behavior change with each new version of Lightroom. Back in the original Lightroom 1.0, there was no way to disable the resampling, and this left many users unsatisfied—you should definitely update to at least 1.1 if you are in this situation, since Adobe swiftly addressed the issue and it was now possible to disable it.
But then in Lightroom 2.0, the whole output sharpening apparatus was rewritten as a result of a collaboration with PixelGenius (the guys behind PhotoKit Sharpener, the popular sharpening solution based on Bruce Fraser's seminal work on sharpening), and it was later found that the optimal result was not necessarily obtained by systematically using the native resolution.
Here's the crutial part:
- If you've performed the output sharpening on an image outside Lightroom (such as in a round trip to Photoshop), you definitely will want to use the native resolution—you do not want to resample an image that has already had its output sharpening carefully applied.
- If you're working on a raw file (or any other file that has not yet had output sharpening applied) inside Lightroom and you're using a version prior to 2.0, you also do not want to resample the image, as you will likely lose some quality, for reasons that have been discussed before.
- If you're working on a raw file inside Lightroom version 2.0 or more, then tests have seemed to indicate that you could achieve an arguably better result by upsampling the image by about 50%—that is, if the native resolution is, say, 240 ppi, you'll want to resample the image to 360 ppi. This remains true only if you are combining this with output sharpening (which will be discussed next) and only if you are upsampling—it remains true that you should use the native resolution rather than downsampling the image.
Print Sharpening. This parameter defines if Lightroom is to apply output sharpening to the image (which is not the same thing as capture sharpening, applied elsewhere in the Develop module, nor creative sharpening, applied locally). If you have already performed output sharpening on a file, you won't want to use additional sharpening and you'll leave this box unchecked. Otherwise, you'll definitely want to add some output sharpening.
According to Jeff Schewe, in relation to the output sharpening in Lightroom 1.x: "Low and Medium are useless—I can't see the result—and High is almost worthwhile." In other words, when using Lightroom 1.x, you'll likely want to use the "High" setting to apply a meaningful amount of output sharpening—or, better yet, perform your output sharpening in Photoshop before going to the print module.
In Lightroom 2, on the other hand, the output sharpening algorithm has been considerably improved, so as long as you've applied an appropriate amount of capture sharpening in the Develop module, using the "Standard" setting will likely be a good choice most of the time, making things really, really simple.
Naturally, sharpening is applied differently depending on the paper type used (if the paper has a coating or not), so you'll need to specify it under Media Type. According to the documentation: "Matte includes watercolor, canvas, and other nonshiny types of paper. Glossy includes luster, semigloss, photo gloss, and other shiny types of paper."
To finish off, don't forget to disable the printer color management and select the appropriate profile for the printer/paper/ink combination you're using. The "Relative" rendering intent is appropriate most of the time, but for critical work, you might want to go perform some soft proofing in Photoshop before deciding.
As for the 16-bit output, if your printer driver supports it (get the latest version), you should certainly use it. Even if the driver doesn't support it, if you're not sure, using the 16-bit mode won't do any harm—it'll just be useless and might slow down the spooling a bit.
2:19 amOff
Resampling images for print
Regardless of the time an image takes to be spooled to the printer, resampling an image (i.e.: changing its pixel size using interpolation) is to be avoided, as resampling an image is detrimental to its quality.
The bottom line is this: if the native size of the image allows you to print using a resolution between 180 and 480 ppi, and if proper sharpening has been applied (considering the destination of the print, its resolution, etc.), you should let the printer do whatever interpolation is required.
Only if you intend to make a print so large that you have less than 180 ppi available, or if you intend to make a print so small that the resolution is above 480 ppi, should you resample the image. If you need to resample, keep this in mind:
- It is arguably preferable to resample by doubling (200%) or halving (50%), rather than using intermediate steps.
- For up-resing, the preferred algorithm is "Bicubic Smoother" — although you will then have to add a significant amount of sharpening.
- Output sharpening should always be applied after resampling.
For a detailed discussion on the topic of sharpening, refer to the Real World Image Sharpening book series. The Luminous Landscape also has video tutorials on the topic, notably chapter 12 of "From Camera to Print".
1:57 amOff
Cancelling color casts
Color casts (overall color drifts) are avoided by using the appropriate white balance setting at capture time (the most accurate of which is a "custom" white balance), or by including a neutral reference in the picture for adjustment in post-production (such as a WhiBal).
Unfortunately, we don't always have time to follow this procedure, so we must often correct color casts arbitrarily in post-production. The problem is that it is not always possible to find parts of the subject which are of a neutral color, with which we can cancel the color cast. And even when we think something is neutral, chances are it is not really (who says this white wall was really pure white — it probably was yellowish!)
The quick and dirty solution is to pick something "close enough" as the reference for setting the white balance, and to adjust, subjectively, until we get something that looks more natural (completely neutral images often look unrealistic, too cold or too warm).
There is a trick to fix color casts in Photoshop which involves:
- Duplicating the image's layer;
- Applying an "Average Blur" filter on that layer (this reveals the overall color dominance);
- Inverting that layer (to obtain the opposite color of the cast, the one that neutralizes it);
- Changing the "Blending Mode" of that layer to cancel the color cast of the image.
Obviously, this trick can only get you so far — you still have to judge by eye what "feels right", by adjusting that layer's opacity, to apply more or less of the cancelling.
That being said, the procedure can be optimized to be executed faster and to render smaller files. Here's the trick. With a Smart Object of the image:
Apply an "Average" blur Smart Filter:
Create a "Curve" Adjustment Layer and use the "Gray Point" eyedropper anywhere on the image:
In a single click, you have set the three color curves to perfectly neutralize the average color of the image!
Color cast cancelling Curve Adjustment Layer
To finish up, simply delete the "Average" blur Smart Filter, which you don't need anymore:
You end up not with a new full-size layer of a fixed color (taking up disk space uselessly), but with a much more flexible "Curve" adjustment layer.