Quick Mask Pro 2024 (free) Update is out on Adobe store
Universal masking tips that are software-independent (in blue) and QMP-specific tips (in orange)
TIP 1 - UNIVERSAL: MASKING STARTS IN THE FIELD.
I cannot stress enough that masking starts in the field. I've said it many times whenever I have the opportunity and I keep saying it because very few people truly understand this and live (their photography life) by this rule. Masking doesn't start behind your computer. No masking software can fix your issues if you neglect this rule. What does this mean? Identify your object in the field and ensure there's enough, uniform contrast between your object and its direct adjacent background. If there isn't for whatever reason, you know this will either be difficult to fix, or cannot be fixed, or is in line with how you process that part of the image. The more complicated, intricate and translucent your object, the more contrast you need.
TIP 2 - UNIVERSAL: THE BEST MASKS RELY ON ENOUGH CONTRAST AROUND YOUR OBJECT
Good masks are dependent on the availability of enough uniform contrasts. Not so revelatory if you know that's also how the human eye detects shapes: through contrasts. Colors can help but are usually not as accurate as contrasts since colors can 'pollute' other parts of the image, bleed over, etc. You need to have a unique color to separate a plane. Adobe's AI technology can detect objects, especially human faces, but even though AI can recognize objects, it still can't create a good mask if there's not enough contrast. A lack of contrast can't be fixed. So if you need good and accurate masks, make sure your main object has enough uniform contrast with its direct background. Too many varying contrasts (dark object/light background, dark background/light object in too many areas in your object) will make masking at the very best a huge challenge, but most likely nearly impossible. This rule is even more important the more intricate and translucent your object is (hair, leaves, fur, etc.)
TIP 3 - QMP SPECIFIC: DARK ON LIGHT IS THE DEFAULT
There are roughly 3 situations when you take a photograph of an object and you expose it correctly:
Darker Figure (object) on lighter Ground (background/sky/foreground etc.) - this is the most common situation outdoors. This is the default in the panel. This is the top menu item in the panel (No Sky Behind Subject) and the bottom menu item (Sky Behind Subject).
Lighter Figure on Darker Ground - usually this happens only when you don't include the sky in your photo, take a night shot or when you shoot in a studio using artificial lighting equipment. This is not the default and the panel offers you only 2 options in the menu item Light on Dark: Low Key and LK Studio. These will cover most situations
Figure and Ground have matching luminance values - if they are roughly the same, then you simply don't have enough contrast and masking will be difficult. If there's a difference, albeit small, and it's not immediately clear if the Figure/Ground is darker then try both options.
It's also mentioned in the PDF manual: if you use Light on Dark, which is the 'exception', then Advanced Refinement adjustments can only take place if you invert the channel mask you want to adjust, first. Invert it back when you're finished adjusting.
TIP 4 - QMP SPECIFIC: ADVANCED REFINEMENTS VS QUICK AND OTHER REFINEMENTS
There are various refinement features in the panel. There's a group of refinement features that are only meant to darken/lighten and will not add (restore) details:
+black and +white presets in the top menu bar (quick access menu): these presets load the selection of the mask you're working on, inside the same mask - the so-called mask-in-mask method. Together with the overlay gradient this offers double protection if you need to add more white or black: white will get whiter without affecting the pure black because it is set to overlay and additionally is protected by the selection that's loaded. And vice versa. This is the go-to preset if there's some white or black left (not pure white or black) that needs to be corrected. Note that each time you press the preset it will load the recently adjusted selection of the mask you're standing on and adjusting! So it's a dynamic correction and always a 'fresh' mask-in-mask correction using the latest corrected mask with every iteration.
Quick refine features under Refine Masks: Similar to +Black and +White features above but offers no double protection by using mask-in-mask. Here you need to roughly select the area you want to darken/lighten. TIP: if you need to do this, then make a quick selection, any selection, and feather it with roughly 15 pixels and save this temporary selection. So you can load it again if you need to do several iterations of the adjustment.
Enhance White and Enhance Black presets under Sky Masks: they do the same as +black and +white but only targeted at the sky.
None of the above will add/restore any details that are omitted during the masking process.
Advanced Refinements under Refine masks: this doesn't just darken or lighten. If used correctly then this feature will restore details that are missing in your masks! Tip: since it's not easy to identify the correct zone, start with one of the middle zones (Mid High or Low Mid) as long as it is not the same zone as the zone you selected with the base mask creation.
TIP 5 - QMP SPECIFIC: WHICH ZONE TO CHOOSE
First a bit of brief history. We developed the first Quick Mask Pro panel 5 years ago in 2017, to automate the rather complex Advanced Masking Method that Joel developed. That version and all versions before the recent Quick Mask Pro 2022 depended highly on the selection of the correct zone. Differences in channel masks could be very substantial depending on the zone you selected. With Quick Mask Pro 2022 and Adobe's AI technology working behind the scenes we succeeded in making the selection of the correct zone less critical. The differences in masks from opposing zones High Key and Low Key for example are small, sometimes even subtle. The differences in masks between adjacent zones are even smaller, sometimes negligible. Somehow bigger differences are in the 3 channel masks that each zone produces. But overall all masks have improved dramatically and have come much closer together, compared to all previous versions of Quick Mask Pro.
The suggestion, therefore, is to just take one of the middle zones: Low Mid or Mid High, and compare them and then try the adjacent zone of the zone with the best result and then compare again. Differences are subtle but perhaps important for your type of mask. The only importance of knowing the right zone is in the Advanced Refinement adjustments where you always have to choose one of the adjacent zones of the correct zone of the base mask.
TIP 6 - QMP SPECIFIC: BE AS SPECIFIC AND ACCURATE AS POSSIBLE IN SELECTING THE RIGHT PRESET
Looking at the panel you've probably noticed that there are several situations to choose from when creating a mask. This is done intentionally because you can't have one preset that will cover all situations and at the same time will produce a good result for all those situations. It makes a difference if there's a sky behind a subject. It makes a difference if the Figure is lighter than its Ground, and it makes a difference if you make a mask for a concrete building vs making a mask for an animal with long feathers. Therefore, a good assessment of the image and making the right preset choice can make all the difference in results.
If your Figure is in front of a sky then always choose the presets for a Sky behind the Subject. The presence of a sky will contribute to the accuracy of the selection simply because it is additional information that benefits the mask. The more information available, the better the mask. This doesn't mean that if you select the presets for No Sky Behind Subject, the mask is much worse. It can be just as good, but the chances are a bit lower.
Also, if you don't have a sky but a flat, smooth background like a smooth wall or studio backdrop, then it will often pay off to select Sky Behind Subject as the software can recognize it as a sky and make your mask better.
Conclusion: selecting the correct and most accurate preset for a given case is important, but it can pay off to try other presets and you can combine the results of different zones! This is something I will demonstrate in one of the upcoming tutorials.
TIP 7 - UNIVERSAL: BE PRAGMATIC AND EFFICIENT
Creating masks is important to control shapes and light in a photo. Hence, if you don't need to control a shape or some light in your photo, why spend time creating a mask of that shape or that area with light?
Or, if you can keep it fast and simple, why make it difficult and time-consuming?
If you need to control a very geometric shape, that's easy to select with the marquee tool, then don't use this panel! Just create the selection with the marquee tool. If your shape is more intricate and needs to be selected, but is not an important or critical aspect of the image, then why waste time trying to make the perfect mask, with or without this panel? Just make a rough selection with the lasso tool and feather it, if that's the fastest way. And if you are into B&W and you also happened to have purchased the B&W Artisan Pro panel then you probably know that some shapes and light can be controlled accurately without using a mask. Just by using the restore feature.
The point I'm trying to make is that masking is just a tool, a way to control shapes. It is not an art and it is not the goal of photography. The photograph is the end goal, and if you don't need to make a mask for that, then don't. Just keep it simple and pragmatic and only make the masks you really use and need.
TIP 8 - UNIVERSAL: CRITICAL MASKS, DERIVED AND INTERNAL MASKS
There are three types of masks (besides hard masks and soft, luminosity masks):
Critical masks
Derived masks
Internal masks
Critical masks are the masks that separate the Figure from its Ground, to use Mark Getlein's terminology (book Living with Art). Usually, there's only one critical mask. I'm calling this mask critical because this is the mask that will ensure that the most important object in your photograph, the Figure, will draw the attention of the viewer. Artifacts or fringing will be immediately visible if the mask is not accurate. This mask, the critical mask, therefore, needs to be the most accurate. This is the mask you would spend the most time on.
Derived masks are the masks that can be derived from the critical mask by subtraction for example and share a mask edge with the critical mask. They therefore only need partial accuracy, the shared mask edge. And that's why you need to create the critical mask first and then derive the other masks if possible by subtraction so the shared mask edge, remains accurate.
Internal masks are masks that do not isolate the Figure from its Ground, like a plane in a building, and that don't share a mask edge with the critical mask. These masks don't need the same degree of accuracy or sometimes don't need any accuracy at all. Be pragmatic. See tip 7.
TIP 9 - QMP SPECIFIC: REFINE FEATURES (CONTINUATION FROM TIP 4)
An overview in terms of the intensity effect of the simple Refine features - not being Advanced Refine Features.
+black / + white (Quick Access menu, top menu): Mask-in-Mask correction method with selective adjustments > strength ~ 10% with each pass - will not affect pure white or pure black (L255 or L0) and works on the entire channel canvas
Enhance black / Enhance White in Sky Masks: Mask-in-Mask correction method with selective adjustments > strength ~ 20% with each pass - will not affect pure white or pure black (L255 or L0) and works on the entire channel canvas
Lighten / Darken under Quick Refine > Normal Adjustments: No Mask-in-Mask correction method (no double protection) and no selective adjustments - will affect all tones, including pure white and pure black in a linear way. Requires a (quick and random) selection to affect an area. Intensity 1 ~ 15% up to 75% for intensity 4.
Lighten / Darken under Quick Refine > Selective Adjustments: No Mask-in-Mask correction method (no double protection) but with selective adjustments - will only affect all tones, except for pure white (L255) and pure black (L0) in a linear way. Requires a (quick and random) selection to affect an area. Intensity 1 ~ 15% up to 75% for intensity 4.
Sky masks, therefore, can be adjusted with the Enhance black / Enhance White features in the Sky Mask Menu but also with the +black / + white feature in the Quick Access menu for more subtle and gradual results. It is not recommended to use features 3 and 4 (normal and selective adjustments) unless it is an adjustment area that doesn't need high accuracy.
TIP 10 - QMP SPECIFIC: PARTIAL BLENDING AND NO BLENDING
The presets in submenu item Sky Behind Subject for both Subject detect and Manual Object detect have the option of No Blending and Partial Blending. The default is that there's No Blending. The idea behind this is that there are cases in which you only want to work on and control the Figure (the main object separated from its Ground) while treating the Ground in a global and uniform way. This implies that you consider everything apart from the Figure as one plane with just a few minor corrections or perhaps none at all. This is very straightforward and makes processing less complicated. Street photography or portraiture for example usually requires that type of post-processing: a focus on controlling and processing the Figure while discarding or minimally processing the Ground.
In that case, you need a mask that will isolate only the Figure in such a way, that processing the Ground is not necessary. This means that whenever the Figure is in front of a sky, the edge of the mask intersecting with the sky should be hard and have no blending at all, while the edge of the mask intersecting with the Ground and all the directly adjacent, objects in that ground, should be soft so to allow for a smooth blending, especially when the adjustments in the Figure are pushed to extremes.
For that, you can switch on Partial blending. Quick Mask Pro then detects where the sky starts and ends, and will make those edges 'hard', and where the rest of the Ground starts and ends and will make those edges soft for blending.
TIP 11 - UNIVERSAL: MASKING ISSUES WITH TRANSLUCENT OBJECTS
The Quick Mask Pro panel is capable of creating very accurate masks of trees with intricate branches and leaves. Yet, when processing the image, the seemingly good and accurate masks results in fringing and other unwanted artifacts around the edges of especially leaves.
Let's get one major misunderstanding out of the way: the problem is usually NOT the accuracy of the mask unless the tree is impossible to mask due to poor or varying contrasts, or if you use the wrong masking techniques. The real problem is the translucent character of the tree with its leaves.
And it is exactly that, that causes the fringing. And pushing the contrast to extremes will reveal the problems that are the result of translucency even more emphatically. There is no set way of dealing with translucency, the solution varies just as much as the shape and translucency of a leaf will vary.
There's one good piece of advice to avoid fringing around leaves: if there's no need to push the contrast to extremes around leaves, then leave it.
Here are some other tips often to be used in combination, they're not exclusive and ultimate solutions!
Make the edge masks for leaves considerably softer than usual. This can be done by feathering it more - rule of thumb: feather it with 2 or more pixels.
The Quick Mask Pro panel gives you the option to use Advanced Refine to bring back more details in trees. The translucent character of a leaf can be incorporated into the mask by making it darker gray (more translucent) and lighter gray to white when it's less or not translucent. Specific use of the Advanced Refine feature, by using the correct adjacent zone and by selecting only the tree outline where leaves are most translucent or intersect with the sky, will give you the shades of darker (or lighter) gray that you need. Remember that Advanced Refine will bring back details first and foremost and isn't only about darkening/lightening.
Just avoid extreme adjustments around tree leaves. A tree shot by bright daylight has a different look, especially its leaves: they appear more translucent the more light hits the leaves. Making the sky right behind it pitch black will reveal fringing that is extreme even if you use a highly accurate mask. The mask of the tree should then have a high degree of darker grays all over the mask. How much? That is impossible to tell upfront. Just experiment with it.
If you still want to do extreme adjustments with a tree like a very dark sky, then photograph the tree in darker conditions or in conditions where the leaves are less translucent (no bright and direct sunlight). This will decrease fringing. Again, masking starts in the field!
TIP 12 - UNIVERSAL and QMP SPECIFIC: TO MASK OR NOT TO MASK
What we mask or not depends on one question: do we need to isolate and control a specific area or not? If we don’t need local control and just global control, we simply do not need a mask.
Here is a guide that might help you assess what needs to be broken down in local areas with local control or if a global area will suffice (and hence does not need any masking). It’s a guide to more efficient masking.
1. First assess what you want to do with an image and what that requires. If your conclusion is that areas need just limited local control, then the local adjustment features with auto blending or the restore features in B&W Artisan Pro for example (or any other software and method you prefer), will suffice and there’s no need to create hard masks. Random quick selections will do.
2. When you’re at this step it means you need control over an isolated area and global control won’t do. Now assess what your most important object is, your Figure, and what is less important your Ground. Sometimes there can be a group of adjacent objects that are equally important (1 Figure group) or even 2 separate objects that are equally important (2 separate Figures), that’s up to you to decide. But there’s always a part that’s called Ground. The most important objects are always Figure(s). In the scheme below when I mention Figure, they can also be plural if you decide so. But I would strongly recommend identifying more than 1 Figure be the exception.
3. There are now several possibilities:
You only want full control over the Figure(s) and limited to no control over the Ground.
There’s a sky behind the Figure > Use menu item Sky behind subject with partial blending on to create the Critical Mask. When post-processing, only process the Figure locally and the Ground globally - it will blend subtly due to the mask with partial blending edges.
There’s no sky behind the Figure > Use menu item No Sky behind subject and leave the option to Wide Range when masking includes translucent details, else choose Tight range. When post-processing, only process the Figure locally and the Ground globally
You want full control over the Figure(s) and control parts in the Ground as well but they’re not directly adjacent to the Figure > Treat the parts in the Ground you want to control as separate or sub Figures
There’s a sky behind the Figures > Use menu item Sky behind subject with partial blending on to create the Critical Mask per Figure. When post-processing, only process the Figures locally and the Ground globally - it will blend subtly due to the mask with partial blending edges.
There’s no sky behind the Figures > Use menu item No Sky behind subject per Figure and leave the option to Wide Range when masking includes translucent details, else choose Tight range. When post-processing, only process the Figures locally and the Ground globally
You want full control over the Figure and control parts in the Ground as well and those parts are directly adjacent to the Figure > One group of (sub)Figures (even if not all are equally important)
There’s a sky behind the group of Figures > Use menu item Sky behind subject and treat the adjacent Figures as one group, with partial blending on to create the Critical Mask. When post-processing, only process the group of Figures locally and the Ground globally - it will blend subtly due to the mask with partial blending edges. To treat the Figures separately within the group of Figures, create derivative or internal masks > See masking tip 8
There’s no sky behind the group of Figures > Use menu item No Sky behind subject and leave the option to Wide Range when masking includes translucent details, else choose Tight range. When post-processing, only process the group of Figures locally and the Ground globally. To treat the Figures separately within the group of Figures, create derivative or internal masks > See masking tip 8
This is the main workflow. If the Figure or group of Figures is a manmade object consisting of multiple planes, with defined edges: create derivative or internal masks per plane to create depth and presence per plane > see Masking tip 8. In natural objects, there are no clear planes with defined and clear edges, so this can usually be skipped.
Creating Depth and Presence is a fundamental principle in B&W Fine art processing that you can read about here and here on my website.
TIP 13 - UNIVERSAL and QMP SPECIFIC: COMBINE DIFFERENT MASKS OR METHODS
The panel always creates 3 different masks per preset. If you try out all presets with the different options (wide range - tight range or no Blending - partial Blending) even if they wouldn't make any sense, you can get up to 50 different masks.
I wouldn’t recommend doing this for every mask but I personally would try to limit myself to a range that makes more sense like Low Key with the exact opposite, High Key, or adjacent zones like Mid High and High Key, and then make a selection of masks to combine. Having said that: what’s wrong with occasionally trying the entire range of zones if it’s just a matter of seconds to create them and having more options to choose from?
Whatever the preference, by trying out more presets you can combine masks.
Suppose High Key channel mask 1 has some details that you don’t have in Mid High channel 3 but Mid High channel 3 has the best overall mask and only needs a few more adjustments. Then stand on Mid High channel 3, indicate the area to adjust with a quick lasso tool selection (or marquee tool, anything quick), load and intersect High Key channel mask 1 via load selection > High Key channel mask 1 > intersect selection, and then add white (sometimes black in case of masks created via Light on Dark presets or just to add more black) via Quick Refine > Normal Adjustments > intensity 4. If you want a smooth transition from the area that doesn’t need an adjustment, then feather the selection (anything between 4 to 20 pixels is the recommendation) to indicate the area before intersecting it with the channel mask to use for refinement.
Or:
Especially in the case of foliage, you can combine the panel results with Photoshop’s Color range selection.
Trees in front of buildings are not uncommon and need to be excluded from the mask. And the best way to do this, in general, is to let the panel create the masks of the buildings and then create a separate selection for the trees or bushes with PS > Select > Color range (Select: Sampled Colors - Fuzziness around 40 (my preference)), save the selection as a channel mask.
Then stand on the channel mask you have found to be the best starting point, load the saved channel mask of the foliage, and then fill it with black via QMP > Refine Masks > Quick Refine > Normal Adjustments > Black intensity 4.
Just 2 examples of combining masks but obviously there are more combinations. Just experiment!
TIP 14 - UNIVERSAL and QMP SPECIFIC: MANUAL OBJECT DETECT OR AUTOMATIC SUBJECT DETECT
First a short explanation of the difference between object and subject, a difference also Adobe Photoshop is using.
An object is a concrete tangible shape in your frame. It can be a clear geometric shape, an amorphous shape, an intricate shape, or all of the before in one. It is always concrete and tangible. Light, therefore, isn't considered an object, although it can take on a very concrete shape. Masking light, therefore, isn't done with hard masks but with soft, luminosity masks. There can be multiple objects in a frame.
A subject is usually, in generic photography terms, an object in a frame. One of the objects or a group of objects in a frame. But, and this is something I wrote about here on my official website, it can also be something that's not in your frame: the subject (matter). This is usually the case in fine art photography where objects are merely used as symbols (e.g. Alfred Stieglitz's Equivalents series on clouds) and the real subject or subject matter is the abstract idea, the message, behind the photograph that cannot take on a specific concrete shape but is expressed through the use of concrete objects. In the context of Adobe Photoshop and this panel, a subject is always related to a specific object and is then the most important object in the frame. There can be only one subject (in the context of Photoshop, one important object) in a frame. For this panel and in the context of Adobe, I am using object and subject interchangeably, the same way Adobe and the photography community in general, use it and leave my (strong) opinion as voiced on my official website, aside.
Auto Subject Detect uses Adobe's AI algorithm to detect the most important object in a frame. And in general, AI does a good job but has serious flaws. Many times it detects only part(s) of an object or it detects the wrong object, especially when there are multiple objects in a frame. That's why there's another option in the panel (and in Photoshop) and that is Manual Object Detect. And this is the feature in the panel I would recommend using if you need a very specific object to be selected.
The Manual Object Detect mode in the panel doesn't mean you have to select the entire object yourself manually, that would render the panel useless. But using Adobe's Object Detect you can identify and select easily the object you want to mask. You can also use the pen tool to indicate it roughly or the lasso tool, but Adobe's Object Detect is what I recommend using as input for the panel. Quick Mask Pro needs your manual input to create a mask when working with Manual Object Detect.
If Adobe's Object Detect selects more than you want to select: use the subtract option to subtract those parts from the selection. It is the 3rd option in Photoshop's top menu bar that will appear when you use Object Detect. If it doesn't subtract enough, repeat it. And if it still doesn't subtract what you want, use your lasso tool or rectangular marquee tool for example to subtract it manually roughly.
If Adobe's Object Detect selects less than you want to select: use the add option to add those parts to the selection. It is the 2nd option in Photoshop's top menu bar that will appear when you use Object Detect. If it doesn't add enough, repeat it. And if it still doesn't add what you want, use your lasso tool or rectangular marquee tool for example to add it manually roughly.
More tips will be added over time...