Layer Masking

[Peteroid]

Just a quick reality check (for me)...

In the context of AS7PRO, a 'layer mask's shape is what is made visible (or see-through), not what is hidden, yes?

I'm use to the word 'mask' to mean the part which is covered up, or the exact opposite of how it is used in AS7PRO.

Again, this is just the way I think... your mileage may vary. This is very easy for me to adjust to. But it did make going through the Layering Mask tutorial a little tough to understand at first...

[funksmaname]

I'm not very experienced with masking in AS - i find them quite confusing so have steered clear in general (but should really practice more)

I think the effect of masking can be either/or what you suggest by choosing 'hide all' or 'reveal all' on the group/bone layer...

I highly reccomend Ramon's 'Lost Layer Tool' which lets you very quickly try out lots of masking combinations without going to the 'layer settings pallette/masking tab' to try combinations - hugely improving workflow.

Also, bear in mind that the masking works from the bottom up as far as i can see so for a 'add to mask' layer to take effect on another layer it should be below it in the stack...

hope this helps, and i've not given misleading information

[Peteroid]

Masking seems to be a very powerful feature. But I'll ask once again:

WHERE IS THE TECHNICAL SPECIFICATION FOR HOW TO USE MASKING?

The User's Manual only has a few examples. These examples do not even mention any of the following topics, and so doesn't even touch what I'm guessing is the full power of masking:

* Reveal All group mask setting
* Subtracting from Masks
* Clearing masks
* the importance of layer order in masking

In short, the masking information is ridiculously unavailable. And this is confused by such things as this:

* LOWER layers hide HIGHER layers (huh?)
* To MASK in AS7PRO is to REVEAL (so I guess at Halloween i put on a mask to reveal my face?)

Come on guys. You need a TECHNICAL SPECIFICATION... and not just for masking...for EVERY AS7PRO FEATURE the kind PROGRAMMERS DEMAND... describing its features. What good is a feature if we aren't given the information to use it?

The User's Manual is NOT a technical specification, but this is all we are given. So anything it fails to talk about is something we - YOUR CUSTOMERS - now can't use... EVEN THOUGH THE PRODUCT IS CAPABLE!

Someone at Smith Micro needs to TAKE THE TIME to DETAIL their PRODUCTS and make this info available to US!

Why put in feature you won't give us the information to use? That makes no sense to me...

[gleeful]

I too found the manual to be seriously lacking and/or confusing on the topic of masking. I have been collecting tips from the forum to help me get a handle on it.

Start here:
viewtopic.php?t=12755

Here is some more stuff on masking:
Masking Notes

Best Masking Tutorial = synthsin's Basics of Masking on Viddler: http://www.viddler.com/explore/synthsin75/videos/1/

Masking always works upwards: a lower level masking operation only affects layers above it, except for the
Group Mask layer.

Add to mask, but keep invisible = creates a hole when in a Hide All masking group.

Subtract from mask (this layer will be invisible) = creates a hole when in a Reveal All masking group.


Layers used in a Hide All masking group will be stacked in the opposite (almost) order compared to the same layers stacked in a Reveal All masking group.

SynthSin recommends the Hide All method.


I hope this helps.

[Peteroid]

I too found the manual to be seriously lacking and/or confusing on the topic of masking. I have been collecting tips from the forum to help me get a handle on it.

Start here:
viewtopic.php?t=12755

Here is some more stuff on masking:
Masking Notes

Best Masking Tutorial = synthsin's Basics of Masking on Viddler: http://www.viddler.com/explore/synthsin75/videos/1/

Masking always works upwards: a lower level masking operation only affects layers above it, except for the
Group Mask layer.

Add to mask, but keep invisible = creates a hole when in a Hide All masking group.

Subtract from mask (this layer will be invisible) = creates a hole when in a Reveal All masking group.


Layers used in a Hide All masking group will be stacked in the opposite (almost) order compared to the same layers stacked in a Reveal All masking group.

SynthSin recommends the Hide All method.


I hope this helps.

Thanks, gleeful! That does help a lot...

[chucky]

Hi Peteroid,
Don't worry if masking seems confusing, almost everyone would agree with that.

I think the masking system should have a full makeover, so that it is clearer to understand.
Personally I have to go to the tutes every single time I use it

Ditching the current wording of the masking methods (regardless of how it relates to the programming) would be a start.
The whole upside down nature of the masking method is also weird from a users point of view.
I think a cookie cutter or window frame comparison works to explain why we all find it so confusing.
When we look through a window, the window frame blocks what is behind, when we use a template to cut something out (like a cookie cutter) we place the template over the object we wish to cut out. Surely the logic of that overrides the programming logic that does the actual job.

This is why people are describing the current method as 'back to front', I wish Mike would take heed of this and fix it once and for all.
Something that seems so simple and innately understood becomes confusing, it defies our real world experiences with masking.

I know arguments can be made to defend the current method, but the amount of confusion surely makes those arguments seem a little lacking in strength.

[Peteroid]

Hi Peteroid,
Don't worry if masking seems confusing, almost everyone would agree with that.

I think the masking system should have a full makeover, so that it is clearer to understand.
Personally I have to go to the tutes every single time I use it

Ditching the current wording of the masking methods (regardless of how it relates to the programming) would be a start.
The whole upside down nature of the masking method is also weird from a users point of view.
I think a cookie cutter or window frame comparison works to explain why we all find it so confusing.
When we look through a window, the window frame blocks what is behind, when we use a template to cut something out (like a cookie cutter) we place the template over the object we wish to cut out. Surely the logic of that overrides the programming logic that does the actual job.

This is why people are describing the current method as 'back to front', I wish Mike would take heed of this and fix it once and for all.
Something that seems so simple and innately understood becomes confusing, it defies our real world experiences with masking.

I know arguments can be made to defend the current method, but the amount of confusion surely makes those arguments seem a little lacking in strength.

Hi Chucky! Thanks for the reply!

Yeah, I've worked with masking now long enough that I finally feel I have a pretty good grasp on it (famous last words).

They actually did a good job with it, it just needs to be explained better in the documentation. Though some decisions seem questionable.

The 'Reveal All' option is the way I naturally think (though I use the other method more now that I know how it works). In 'Reveal All' mode everything is by default shown and the the 'mask' is the HIDER... that is, positive space on the mask is what is hidden on the image/layer.

But the documentation tutorial examples only talks about the 'Hide All' method. They probably did this since it is the harder method to deal with (from what we think the word 'mask' means)... since in this mode the default is everything is hidden and the mask is the REVEALER.

Then things like 'add to mask' and 'subtract from mask' make a lot of sense. But they don't make clear that the same terminology means EXACTLY THE OPPOSITE depending on which mode you use! THAT is VERY confusing, and using the less-intuitive method only for examples doesn't help.

Maybe it would have been better not to use terminology like 'add to mask', but rather 'add to reveal' or 'add to hide', and then dynamically change the menus appropriately depending on the mode being used.

That being said, I still cannot for the life of me fathom why 'mask processing' is done from the BOTTOM LAYER UP, since this means you have to put the mask behind the layer to hide it (which from a natural pov seems wrong, sort of like putting a Halloween mask on the back of your head to hide your face). But that just takes some brain tweaking (but, in all fairness, a design that required brain tweaking should be avoided, and they could have avoided it in this case by processing masks front to back)...

[chucky]

I kinda' understand the bottom up thing but it's rather removed from being immediately graspable. It reminds me of using frisket or masking tape when using and airbrush or painting. You put the mask on top , then you apply the paint, the mask cover and blocks to allow the new layer to be applied. The thing is in this example you have to removed the mask altogether to see the end result. so the comparison is incomplete.
It just doesn't seem appropriate in this application, hell, I used to airbrush all the time and it still seams a leap of logic.

Hhhhmmmm

[jstuartj]

Think of drawing on layers of acetate. You draw on the first layer. Then add a new sheet and draw some more. Now add a fresh sheet and draw a mask shape. Then on a new layer, you simply draw only in the area defined by the lower mask shape. When your done add another layer etc....etc...

You can mask any numbers of layers above by simply following the mask shape defined previously. Until you choose to stop masking at which point you can continue drawing as normal.

Think how difficult it would be to work top->down. You would have to lift layers up to draw below, or flip pages to see where the mask was or have all the layers fully pre-drawn and then erase the areas with in the mask shape.

Both Postscript/PDF RIPs and Illustrator work in a similar fashion. where they are called "Clipping Groups". It requires fewer resources, since once you render a lower layer, you can forget about it and move on to the next layer above.

[madrobot]

My understanding is that a mask will work on all shapes within the group, be they above or below the mask? I didn't think it mattered where the mask actually is in terms of layer order inside the masking group?

Am I confusing something here?

[slowtiger]

No, the mask has to be at the bottom of what it masks. The explanation is simple: AS renders from bottom to top, layer by layer. So it needs to render the mask first before it renders any layer affected by the mask.

[funksmaname]

this also means that you can have partial masks in a group depending on their position in the stack.

*poof* i think i just blew a fusezzZZzT*

[slowtiger]

Of course you can have several masks within the same group. That's the nice part: you can combine and animate them as you like.

[jammydoug]

Hi
I use photoshop (7) quite frequently, but lately for some odd reason every time I use layer mask only black and white colors are available. When I go back to the standard pic my colors are available. And to make matters worse I can't inverse while I'm at layer mask. Any help please?

[slowtiger]

1. You really should ask this in a Photoshop forum,
2. when you only see black and white, you're in the mask layer only. Read about that in the manual. (It's in the rpogram under "Help".)