Color Channel Clipping


Today’s Question: I read your response to the question of “clipping” [in the September 15, 2017 edition of the Ask Tim Grey eNewsletter] and again you outdid yourself in the depth of the explanation. The only thing that I did not understand and you only touched on peripherally is the question of clipping on the color channels.

You used the red rose as an example but did not go into how to address clipping of a color channel, only blacks and whites. I have had some specific examples of a photo of a red rose that I cannot get the red color right and I think this may be the issue. Can you address this question as well in the topic of clipping?

Tim’s Quick Answer: Clipping of an individual color channel in a photo can indeed be a problem, primarily from the standpoint of a loss of texture and detail in areas of the image that are dominated by a single color.

More Detail: When we talk about “clipping” in the context of a photographic image, we are typically referring to the white point and the black point. If the whites are clipped than some of the bright areas of the image will become pure white, losing all detail. Similarly, if the blacks are clipped then the dark shadow areas of the image may lose all detail. This is obviously an important thing to keep in mind when adjusting the white and black point for an image, so that detail is not sacrificed in areas where you want to retain that detail.

Similarly, there is an issue as it relates to individual color channels. A red rose is a good example. If we assume a photo that is a close-up of a red rose, in theory the image might be comprised of information that is only present on the red channel, with very little information on the green or blue channels. In actual fact the distribution of values is not quite so extreme, but I think it can be helpful to consider this theoretical example.

Remember that clipping of the white point requires that all three color channels reach the maximum brightness level for specific pixel values. In other words, you can’t produce a pure white unless all three color channels (red, green, and blue) have a white value for a given pixel.

Of course, if you think about a red rose, you can imagine that there would not be any pixels that are pure white within the image. That might lead you to assume that it is impossible to clip any detail, since you can’t easily produce a pure white value. You might therefore assume that you will retain texture in all areas of the image. But this is not necessarily true.

Just as you can over-expose an image so much that you lose information on all three channels, so too can you over-expose (or over-adjust) so that detail is lost on one or two (but not all three) channels.

What this over-exposure (or over-adjustment) would translate to is not so much an issue of accuracy of color, but rather of texture in colored areas of the photo. This is why I highly recommend evaluating a histogram on your camera that includes the individual color channels, if your camera offers this feature (and most cameras now do include this option). Furthermore, when applying adjustments to overall brightness levels (such as with Exposure, Whites, and Blacks adjustments) it is important to evaluate all of the channels, rather than only the areas that might be clipping to pure white or black.

In other words, when using the clipping preview display available with many adjustments (such as Exposures, Whites, and Blacks in the context of Adobe Camera Raw or Adobe Lightroom), you’ll want to try to avoid clipping of individual color channels if detail is important in the applicable areas of the photo. So when holding the Alt/Option key while applying an adjustment that includes support for the clipping preview display, you’ll want to watch for individual colors that indicate clipping for one or two channels, in addition to black or white areas that indicate clipping to pure black or white.

Smart Sharpen Setting


Today’s Question: Which option should I be using for the Remove popup in Smart Sharpen in Photoshop? I’ve been leaving it set to Gaussian Blur only because I don’t know what any of the options do. I seem to be getting good results with Gaussian Blur setting, but could I get better results with a different choice?

Tim’s Quick Answer: I recommend the “Lens Blur” option as the best setting for the Remove popup in the Smart Sharpen dialog in Photoshop. While the difference is relatively minor in most cases, for photographic images the results with the Lens Blur setting are generally more pleasing than with the Gaussian Blur setting.

More Detail: The Smart Sharpen filter is my preferred filter for sharpening images in Photoshop. And when it comes to using the Smart Sharpen filter, as a rule I recommend using the “Lens Blur” option from the Remove popup.

The Remove popup controls what type of blur you are attempting to remove from the image. The Motion Blur setting is obviously aimed at compensating for camera movement during the capture. However, I find that this option doesn’t provide a significant benefit, and so I don’t find it particularly useful. You can get slightly better (but still not especially impressive) results using the Shake Reduction filter instead.

The Gaussian Blur option will cause the Smart Sharpen filter to essentially provide the same overall behavior as the Unsharp Mask filter. This provides perfectly good results in most cases, but the Lens Blur option is more sophisticated and will generally provide better results.

To be sure, the visible differences between the Gaussian Blur and Lens Blur options will often be quite subtle. In general you will find that the effect with the Gaussian Blur option will be a little harsher than the Lens Blur option. With Lens Blur you are able to maintain smoother transitions with less risk of visual artifacts, while still achieving a significant improvement in the edge contrast that provides the sharpening effect.

So, as a general rule I recommend the Lens Blur setting for the Remove popup in the Smart Sharpen dialog when sharpening photographic images. But I certainly encourage you to experiment with different settings to see what your impression is, and to determine what works best for your images.

Discard Originals?


Today’s Question: Do you see any reason for me to keep my original RAW (Nikon D810 NEF) image files after I’ve converted them to DNG using Adobe’s DNG Converter application and have backed up the DNGs?

Tim’s Quick Answer: If I’m being completely honest, in most cases it is perfectly reasonable to discard the original proprietary RAW capture files if you’ve chosen to convert those captures to the Adobe DNG format, and you have ensured the DNG files are safely backed up. This isn’t something I’m entirely comfortable with on principle, but I do have to admit it is a reasonable approach in most cases.

More Detail: As the image sensors used in digital cameras have gotten more advanced, new features have been made available. In many cases, those specific features can only be accessed by using the proprietary RAW capture format for the particular camera model, along with the software provided by the camera manufacturer.

For example, many Nikon cameras (including the D810 mentioned in today’s question) include support for an Active D-Lighting feature. This feature helps to balance out overall tonal values in the image, helping to provide greater detail in the dark shadows, for example. However, you can only retain the advantages of the Active D-Lighting feature if you use the Nikon software to process the original RAW captures.

There are other camera manufacturers with other special features that require software from that manufacturer to take advantage of the feature. Processing the images with Adobe Camera Raw or Lightroom won’t provide access to those features. In addition, converting to Adobe DNG would cause you to lose access to those features.

In theory there might also be other capabilities that are only available by using the proprietary RAW capture files. The idea is that since the camera manufacturer has the most knowledge of the specific image sensor technology employed in a given camera, they can best exploit the information contained in the proprietary RAW capture.

Many of these potential benefits aren’t critical to most photographers. So unless you’re taking advantage of special advanced features of your camera (such as the Active D-Lighting feature on many Nikon cameras, or the relatively new Dual Pixel technology from Canon, among other features with various camera makes), it is probably perfectly safe to convert your proprietary RAW captures to Adobe DNG and then discard the original captures.

Again, this isn’t something I’m particularly comfortable with, but that’s mostly a matter of being a bit more paranoid than I really need to be.

I’ll also add that converting to the Adobe DNG format creates a workflow slowdown if you have adopted my preferred approach of using a synchronization approach to backing up your photos. Metadata updates to DNG files require a full backup of the (rather large) source image file. By contrast, with proprietary RAW captures your metadata updates will be written to a much smaller (and therefore faster to backup) XMP sidecar file.

Keyword Painter Undo


Today’s Question: One of your quick Lightroom tips was adding keywords using the painter tool. So, after adding a keyword how do you commit the word to the image(s) without accidentally selecting another image and thereby inadvertently adding the keyword to the wrong image?

Tim’s Quick Answer: When using the Painter tool in Lightroom to quickly add keywords (or other metadata) to your photos, the metadata is updated as soon as you click (or drag) on a photo. If you add a keyword in error, you can use the Undo command, or hold the Alt/Option key while clicking (or dragging) with the Painter tool to remove the keyword (or other metadata) from the applicable images.

More Detail: One of the unique things about Lightroom as compared to many other software applications is that you never actually need to “save” your work. As soon as you apply an update to an image, whether that is a metadata update or an adjustment in the Develop module, the change is applied immediately.

As a result, when you click on an image (or drag across multiple images) with the Painter tool in Lightroom to add a keyword or apply a different metadata update, the change is applied immediately. When working with the Painter tool, a small white border will be added around each image you have applied an update to. In this way you are able to see which images have been updated. Of course, that also means you might quickly see that you’ve made a mistake.

Fortunately, it is easy to undo such a mistake. First, you can use the Undo command. When you realize you accidentally applied a keyword to an image, you can choose Edit > Undo from the menu or press Ctrl+Z on the keyboard on Windows or Command+Z on Macintosh to take a step backward.

In addition, you can use the Painter tool to remove the current keyword or metadata update from an image. Simply hold the Alt key on Windows or the Option key on Macintosh while clicking on an image with the Painter tool, and the current metadata value you’ve applied to the Painter will be removed from the image you click on (or from the images you drag across).

Note, by the way, that the “quick tips” referred to in today’s question are part of my “Lightroom Quick Tips” video series. If you’d like to gain access to the archive of existing tips and get a new video tip each week, you can get the details here:

Compression Options for TIFF


Today’s Question: Can you discuss the merits of LZW compression, versus ZIP [when saving a TIFF image]?

Tim’s Quick Answer: Both LZW and ZIP are lossless compression options for saving TIFF images. Both of these are lossless compression options, meaning image quality will not be degraded. The ZIP compression option will generally provide file sizes that are smaller than those with LZW compression, especially for 16-bit per channel images. In fact, with LZW compression a 16-bit per channel TIFF has the potential to be larger than the same file saved without any compression applied.

More Detail: While the compression used for JPEG images is always “lossy”, meaning image quality will be degraded even when you use the highest image quality setting, when saving TIFF images you have a couple of options that provide lossless compression.

An uncompressed TIFF file will be quite large compared to a JPEG image. A TIFF image saved with LZW or ZIP compression will still be considerably larger than the same image saved as a JPEG, but generally smaller than if compression was not used for the TIFF image.

For 8-bit per channel images you will generally get about the same degree of file size reduction when using either LZW or ZIP compression with a TIFF image. For 16-bit per channel images in most cases the file size reduction will be relatively modest with ZIP compression, and LZW compression may actually cause the file size to be larger than without the use of compression.

In theory all of this translates to a suggestion to always use ZIP compression for TIFF images, and to avoid the use of LZW compression. However, it is worth noting that not all software applications that support TIFF image files support ZIP compression for TIFF images. These days it is generally safe to use ZIP compression for TIFF images to help reduce overall file sizes, but you may want to confirm compatibility with the software you employ in your workflow before changing the settings for how you save image files.

Clipping Confusion


Today’s Question: I read the following in one of your recent answers, and am trying to get my mind around it:

<<The white slider lightens or darkens the lightest tones. This allows you to clip the whites.>>

What does it mean to ‘clip the whites’ and why would you do it? What function would that serve?

Tim’s Quick Answer: To “clip the whites” means to increase the brightness of the brightest pixels in an image to the point that detail is lost. In most cases you might want to brighten right to the point before clipping occurs, although in some specific cases you may want to create a small degree of clipping. The same basic concepts would apply in “reverse” when it comes to the darkest values in the image, where you could potentially choose to clip shadow detail.

More Detail: The “true” feature of the Whites slider is to set the white point. In other words, how bright should the brightest pixel in the image be? In theory, the brightest pixel in an image should be white, and the darkest pixel should be black. Obviously there are exceptions to this, but naturally this “rule” holds true much of the time.

So, in a typical workflow for a typical photo you might increase the value for Whites until the brightest pixel reaches true white (or thereabouts), and then reduce the value for the Blacks slider until the darkest pixel reaches true black (or thereabouts).

The challenge is to accurately find that position for each slider, and then to fine-tune as needed. This is where the “clipping preview” display comes in. Which brings us to your actual question here.

Clipping refers to a loss of detail when the brights get too bright or the darks get too dark. A simple example would be a gradient that goes from pure black to pure white, with a smooth gradation in between. Let’s assume initially that we are dealing with a simple linear gradient, where the bottom row is black and the top row is white. As you move up a row, the pixels get slightly brighter. As you move down a row the pixels get slightly darker. So we have a nice smooth gradation with detail at every step along the way.

If you were then to increase the value for the Whites slider for such an image, the brightest values would get brightened. Of course, that top row of white can’t be made any brighter, so it remains as it is. But the row below becomes pure white. That’s our first level of clipping, because there is no longer a difference between the top row and the second row. At this point that clipping is probably not especially problematic. But if you keep increasing the Whites value, more and more rows of bright gray pixels will become white, and there will be a larger area with no variation in tonal values. Eventually you’d have a big section of the top of the image that is pure white, with no detail at all.

The same basic concept could apply to darkening the shadows, causing a large area of pure black (with no detail) at the bottom of our gradient image.

Applying that concept to a photographic image, obviously we’re dealing with areas with more random shapes than with a gradient. But the same issues are at play. If, for example, we increase the value for Whites too much with a photo that includes a cloudy sky, those clouds could lose detail by virtue of being “clipped” to pure white.

So, again, the clipping refers to the loss of detail in an area that has gone to pure white or pure black. To a lesser extent it is also possible to have clipping on an individual color channel (red, green, or blue), so that detail represented by those channels is lose. For example, with a simple photo of a red rose, you could have no clipping for the white point, but still have clipping for the red channel. That could result in lost detail for the rose itself.

JPEG Capture Settings


Today’s Question: This is just a simple question, but one that confuses me. I have a Nikon D7100, and have always used an image size of “Small/Fine”, as I do not need any prints larger than 8”x10″. Someone told me that increasing the image size to Medium would give me a better print, as it gives more information. I’m not sure if I received the right information on this. Does increasing the image size to Medium actually give me more information and better colors?

Tim’s Quick Answer: As a general rule, if you’re going to be using JPEG capture instead of RAW capture, I recommend using the highest resolution and quality settings available for your camera. This will ensure both the highest quality and the largest potential print size.

More Detail: I should hasten to add that I highly recommend using RAW capture to ensure the highest image quality and greatest overall flexibility in your workflow. That said, I realize there are situations where JPEG capture may provide benefits that are more meaningful for a photographer than the potential benefits of RAW capture.

In terms of the quality setting, I highly recommend always using the highest quality option available when employing JPEG capture. This setting directly relates to the compression being applied to your images right from the moment they are captured, and with JPEG images the compression always causes a loss of detail and quality in the image. Therefore, I consider it very important to use the highest quality setting to minimize the negative impact on the quality of your photos.

The size option relates to the overall resolution of the photos being captured. In theory you can use a lower resolution setting if you tend not to print your photos at a particularly large size.

For example, the Nikon D7100 that is the subject of today’s question provides a native resolution of 4,000 x 6,000 pixels. At an output resolution of 360 pixels per inch (a common output resolution for a photo inkjet printer), that translates to a potential output size (without enlarging the image) of about 11 inches by 16 inches. The lowest resolution setting in this case is 2,000 by 2,992 pixels. At 360 pixels per inch that still provides a print size of about 6 inches by 8 inches.

As you can see from the numbers above, the “small” size for a JPEG image in the case of the Nikon D7100 provides fewer pixels than are actually necessary to produce an 8×10 print. The “medium” size is well suited to an 8×10 print, providing output dimensions of about 8×12 inches.

However, I think it is also important to keep in mind that you may want to crop your images from time to time. Therefore, I consider it critically important to capture at the highest resolution available in most cases (and most certainly in this example). Unless you are using a camera with an extremely high resolution that far exceeds any output size you’ll ever produce (taking into account potential cropping), I would not recommend using anything other than the full resolution of your camera when capturing photos.

PSD versus TIFF


Today’s Question: Does what you say about photos saved as TIFF images [with respect to file sizes] also apply to files saved as PSD images?

Tim’s Quick Answer: Yes, TIFF and Photoshop PSD images will produce files that are of similar file sizes, all other things being equal. In both cases the file size will be significantly larger than the same image saved as a JPEG.

More Detail: Today’s question is a follow-up to a prior question that addressed TIFF images producing significantly larger file sizes than a JPEG image.

In the edition of the Ask Tim Grey eNewsletter that focused on TIFF versus JPEG file sizes, I made reference to the option to save TIFF images with compression applied. JPEG compression is always lossy, meaning some level of detail or quality will be lost in the process. However, with TIFF images you have the option to save the image with lossless compression, such as the LZW compression option that is generally available in software that enables you to save TIFF images.

The Photoshop PSD (Photoshop Document) image format is actually a TIFF container, and lossless compression is applied automatically when you save a PSD image. In that way, saving a PSD image is very similar to saving a TIFF image with the LZW compression option selected.

However, the actual compression used by PSD versus TIFF images is different, resulting in different file sizes. The actual results will vary based on the composition of the images, such as the number and type of layers used within the image.

The bottom line is that both TIFF and PSD images will produce a file size that is significantly larger than a JPEG image. Lossless compression is automatically applied with PSD files, while it is an option with TIFF images.

Custom Sort Unavailable


Today’s Question: I have been processing photos from a recent trip to the Canadian Rockies in Lightroom CC. After completing the post processing, I tried to change the order of some of the photos in the filmstrip at the bottom of the screen. When I try, I get a screen that says, “The currently selected source does not support custom order. Cannot reorder photos.” I have changed the order of photos in the past, but don’t know what has changed. Can you help?

Tim’s Quick Answer: In order to apply a custom sort order in Lightroom, you need to be browsing a single folder or a “normal” collection. You can’t be viewing the contents of multiple folders, and you can’t sort in a custom order when browsing a smart collection.

More Detail: Lightroom enables you to define a custom sort order for any individual folder or “normal” collection, simply by dragging and dropping the thumbnails into the desired order on the filmstrip or in the grid view. You can then return to the custom sort order later, after changing the sort order to another option. This can be done by choosing the “Custom Order” option from the Sort popup on the toolbar below the grid view display.

However, the custom sort order is only available for an individual folder or for a “normal” collection.

You can’t use the custom sort order for the collections found in the Catalog section toward the top of the left panel in the Library module. You also can’t use a custom sort order for a smart collection. If you have selected multiple folders from the Folders list, or if you are browsing images from subfolders in addition to the current folder, you also won’t be able to use the custom sort order. That means you may want to turn off the “Show Photos in Subfolders” option found on the Library menu on the menu bar. You also can’t select multiple collections at one time if you want to sort in a custom order.

So, the key is to make sure you are only browsing a single folder or “normal” collection (not a smart collection), and that you aren’t browsing images from more than one source location.

Retroactive Sidecars


Today’s Question: After working on images for a while [in Lightroom] without sidecar files, when I change over to utilize the sidecar files will all of the already adjusted images have sidecar files created?

Tim’s Quick Answer: Yes, when you enable the option in Lightroom to save metadata out to the images themselves, Lightroom will immediately start updating all existing images. Thus, XMP sidecar files will be created for all RAW captures, and the metadata within the image file will be updated for other file types.

More Detail: Enabling the option to have Lightroom save metadata out to the images (instead of only within the Lightroom catalog) provides a couple of benefits. One, it enables you to see standard metadata updates (such as keywords and star ratings) with other applications used to browse outside of Lightroom. Second, it provides a backup for most of the metadata updates for your images beyond the Lightroom catalog.

When you turn on the “Automatically write changes into XMP” checkbox in the Catalog Settings dialog, Lightroom will immediately get to work creating (or updating) XMP sidecar files for RAW captures, and updating the image files for other file types. To enable this setting choose Edit > Catalog Settings from the menu on Windows or Lightroom > Catalog Settings on Macintosh. Then go to the Metadata tab and turn on the “Automatically write changes into XMP” in the Editing section at the top of the dialog.

If for any reason you want to make sure the updates were applied, you can also “manually” save the updates so you’ll see an indication of the progress on the identity plate. To do so, go to the All Photographs collection in the Catalog section of the left panel in the Library module. Choose Edit > Select All from the menu to select all photos in your catalog, and then choose Metadata > Save Metadata to Files from the menu to initiate the process of saving metadata to your files.