auditioning surface masks...

[fishmonkey]

i'm messing around with surface masks, and i'm a bit discombobulated to find that the masks do not appear in the Audition window. is there a way to make that happen?

i've also found that sometimes the masks persist even when removed from the surface setup, forcing me to add a white non-mask to get video output again...

[Christopher Ashworth]

Hi sharkape,

The audition window intentionally does not reflect adjustments applied to a specific screen output, which includes masks and warping. The audition window reflects the abstraction of the surface prior to adjustments at the screen level.

The mask thing sounds like a bug. If I recall we had a report of that but haven't been able to recreate it yet...

(mobile)

On Feb 20, 2015, at 8:16 PM, fishmonkey <fishmo...@gmail.com> wrote:

i'm messing around with surface masks, and i'm a bit discombobulated to find that the masks do not appear in the Audition window. is there a way to make that happen?

i've also found that sometimes the masks persist even when removed from the surface setup, forcing me to add a white non-mask to get video output again...

--
--
Change your preferences or unsubscribe here:
http://groups.google.com/group/qlab
 
Follow Figure 53 on Twitter: http://twitter.com/Figure53

---
You received this message because you are subscribed to the Google Groups "QLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qlab+uns...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Christopher Ashworth]

Er, Hm, maybe I've got that not quite right. Sean, did I get that right?

(mobile)

[fishmonkey]

fine-tuning a tricky mask has been driving me crazy today.

i'm editing the mask in Photoshop. if i point QLab at the PSD file automatic updating is very sporadic. to force it to happen i need to go into the surface preferences and remove the mask. this doesn't actually remove the mask, but updates it. then i need to point QLab back at the mask file, rinse, repeat.

the automatic updates seem to happen a bit quicker if i point QLab at an exported PNG version of the mask, however this means having to export the mask every time i change it. workflow killer (although i could set up an action in Photoshop to do this).

still, it seems like a bug.

all this is using a two projector setup with edge-blending...

[sam kusnetz]

The PSD file format is... I could find a way to say it myself, but I recommend this fantastic link instead:

http://code.google.com/p/xee/source/browse/XeePhotoshopLoader.m?r=a70d7396356997114b548f4ab2cbd49badd7d285#107

TL;DR is that it's best to avoid using PSD files anywhere except inside Photoshop.

Best

Sam

--

Sam Kusnetz

Figure 53 Field Operative

(mobile)

[micpool]

Am I missing something here? In photoshop if you open a png and edit it it saves as a png.

Mic

[Michael Long]

It's nice to use a *.psd because you can have vector images embedded within. This is especially great with complex, irregular shapes. I don't think QLab allows vector files for masks, right?

--
Michael Long
Projection Designer - Philadelphia

[micpool]

On my system, to update the mask in a surface I just click on the surface mask field... As soon as I click the field the mask updates regardless of format (including psd)

So for me it's working. I know some people have been having trouble with the mask field with it clearing but not deleting the mask? could it be related to that?

Mic

[fishmonkey]

yes, i have a complex fiddly mask of a set piece that i need to shoot around. i'm needing to translate/rotate/warp the mask to get it just right.

maybe it is related to the fact the QLab will not return to a no-mask state for this workspace. clicking on the mask area does not update the mask, only 'clearing' the mask with the cross button updates it immediately.

and as i mentioned the mask with sporadically update on its own, however so far i can see no rhyme nor reason to when it decides to do so.

as for PSD files not being particularly suitable, it would be helpful if that was mentioned on the web page that talks about masks...

[micpool]

I have never been able to replicate the masks that persist after surface mask field is cleared problem.  There are however some oddities  with this mechanism anyway. Perhaps if the code for surface masks were rewritten to address these, the problem might be caught in the process.

Surface masks are not bundled.

If a surface mask is not present when a file is opened no warning is given

If a surface mask is not present, then the surface mask field is blank there is no indication a mask has ever been there let alone a red x

The workspace has still got the mask path stored, even if it is resaved with the mask missing and the field blank.

If the mask is present the next time the workspace is launched it is found again and appears in the field.

On the file format issue I can  find no difference on either my 10.10 mac mini or 10.8.5.MacBook Pro any difference between the behaviours of the updating of surface masks regardless of forma (psd jpg or png). They update when the surface mask field is clicked, immediately. 

I only use jpgs and pngs in qlab and on other video servers, but the reasons for this are probably historical with older video servers and OSs rather for any good modern reason. But I use psds for lots of other things. They open in the finder preview, any software of any printer I have sent them to, and  in tons of really bad shareware apps without problems. Mind you I do always save them with maximise compatibility checked (this saves a hidden flattened version as an additional layer, which makes the file size a bit bigger). They don't seem to cause any problems when used as normal cues in Qlab. Is there any reason why they should be any more of a problem in the surface settings? Presumably when Qlab opens and applies  the surface mask it flattens it to a grayscale single layer representation of the original file straight away.

I always try and minimise the amount of times I need to redraw a mask to finesse it. I generally output the photoshop canvas full screen  to the projector and trace the set directly. I normally find if there aren't too many awkward bezier curves that I can get the mask within 3mm on the first attempt. 

Mic

[fishmonkey]

thanks for the extra info. despite the clunkiness i have got a good mask sorted out for this venue. i will revisit this issue in a week's time at the next venue...

[Michael Long]

I'm experiencing the exact problem that fishmonkey was describing above. It does update the file by clicking on the mask selection field (great tip, Mic, thanks), but I am not able to delete a mask. When I click the x, it tells me there is no mask, but it still appears on my surface. Even restarting the application does not clear it.

--

Michael Long
330.207.7055
mLon...@gmail.com
michaellongfilms.com

[fishmonkey]

i have resorted to loading an all-white mask as a workaround for this issue...

[Chris Ashworth]

Michael,

Could you send a copy of this workspace to sup...@figure53.com?

Although we’ve seen several reports of this, we’ve never been able to recreate it. Hoping seeing the workspace where it’s happening might provide a clue.

Thanks,

Chris

[Michael Long]

Sure! I'm actually experiencing another problem with it, too, which may also be related to the masking issue.

--
Michael Long
Projection Designer - Philadelphia

[Vinny Osborne]

Yes I had the same problem last week.  If you add a mask to the master video window & then delete it, the mask stays.  I had to copy & paste all my cues into a new workspace to solve the problem.

[Sven Ortel]

Great idea. That just saved me lot of time. Thanks.
Sven

[micpool]

On Wednesday, February 25, 2015 at 2:52:45 AM UTC, micpool wrote:

I have never been able to replicate the masks that persist after surface mask field is cleared problem.  

And now I have found out why! If you put a mask on a surface you can clear it and re apply it with entirely the expected results, UNTIL YOU SAVE. After that when you clear, the field clears but the mask persists.

This seems to be the case for every workspace I have previously created .  I can't clear any masks from these archived workspaces.

I'm not sure how I missed this, every time I tested it, I did save the workspace at the end of the test, but didn't reopen it again. 

Presumably this means the bug is far more consistent and universal than previously thought, and in that the masks only persist after a save, easier to track down

Bundling a workspace with persistent masks clears the problem, as the links to the original masks are cleared, and only masks  that have a mask path displayed in the mask field  of a surface appear in the masks folder of the new bundle

Mic

[Michael Long]

Interesting observation. But I'm not so sure it's that simple. I sent my workspace to our friends at Figure 53, and they could not recreate the mask problems that I was experiencing, even though that workspace had been saved, opened, and closed many times. I haven't tested it with bundling, I'll have to try that myself soon.

--
Michael Long
Projection Designer - Philadelphia

[micpool]

If you just sent the workspace then the links to your mask files would break when suppprt opened the workspace so they wouldnt be able to recreate.

If you bundled, and sent the bundle to support then again the links would break.

There is a fundamental flaw in the implementation of surface masks. If an audio or video file is not found by a workspace then you know straight away in the warnings and by the red x.

If a mask file is missing there is no indication that the mask was ever present and everything looks normal.

Mic

[Michael Long]

Ahhh, I see. Thanks for the clarification!

--
Michael Long
Projection Designer - Philadelphia

[Sean Dougall]

Thanks, Mic — with this new description, and some help from our esteemed Christopher Cahoon, I’ve finally managed to recreate the issue on my end. I’ve tracked it down to a problem with how surfaces deal with the relative file path changes that were made a while ago.

This ought to fix it: http://figure53.com/temp/QLab-3.1.10-b6.zip

--

Sean Dougall

se...@figure53.com

[micpool]

Excellent, That version opens old  workspaces and if masks are deleted they are cleared as they should be. Thanks

I still think there are a few issues outstanding with the masks.

There is still no indication that a mask is missing if the workspace can't find it's masks.  Because of the warning systems you can be confident  that the click track for the act one finale, or the radio broadcast, without which none of the preceding 4 hours the audience has watched will make any sense, are present. It's really important to have the same confidence with masks.  You shouldn't be able to run a show oblivious to the fact that a mask  file is missing,  and only discovering it when your image spills all over the audience, or on the elephant you have just made vanish in front of the audiences eyes.

Deleting a mask is not placed in the undo queue.

And finally, the lower third of each  edit button on the Available Surfaces in video settings is not active to the mouse down

Mic

[fishmonkey]

now you mention it, that one has been driving me nuts. i didn't twig that it was just a hit-box issue.

[Joshua Langman]

That's been driving me crazy too! I couldn't figure out what was up, I thought maybe something was wrong with the trackpad.