Return-Path: Received: from mail-ve0-f172.google.com ([209.85.128.172] verified) by media-motion.tv (CommuniGate Pro SMTP 4.2.10) with ESMTP-TLS id 5022102 for AE-List@media-motion.tv; Tue, 02 Apr 2013 17:17:56 +0200 Received: by mail-ve0-f172.google.com with SMTP id oz10so636302veb.31 for ; Tue, 02 Apr 2013 08:25:34 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20120113; h=x-received:mime-version:in-reply-to:references:from:date:message-id :subject:to:content-type:x-gm-message-state; bh=lRH1DOpPM25Jpx54ZRsRgs5eEAc1fLsjmJ5pY7Twd+I=; b=UnwipGw+hv0Og4hKmeM+ado+A5OWyMfN84gdflwnarj3NtI3q7urU6t1ZqEh9R/enx VMcS0GLDKVGxqNB7Uo2BX3X1RQtA8hDfn0x6jPDtNeUjEM3uV+iNP7RAGDBAvI6TSpMO vhQ4fTBXK55wHv8JkHdxsGt1iRE3nmmTd52mF/24o0OddmoBNBeI27Od83YGpKkO4VBS jMlUEAEmULWE1SqFqcs+iyTBy4pCVuKJyU5HOoAf+S6qJ1o0DzjZuwuhwFk50gSXsZb0 YPyE+lbXqz0VN47fddUYE6GrDhy8xUAy5iOh3qn9yfsM5BCuzcLoLZ1Uhe8uasdtgUGr 111A== X-Received: by 10.52.30.48 with SMTP id p16mr11235630vdh.118.1364916333818; Tue, 02 Apr 2013 08:25:33 -0700 (PDT) MIME-Version: 1.0 Received: by 10.58.241.161 with HTTP; Tue, 2 Apr 2013 08:25:13 -0700 (PDT) In-Reply-To: References: From: Robert Kjettrup Date: Tue, 2 Apr 2013 17:25:13 +0200 Message-ID: Subject: Re: [AE] spill suppression To: After Effects Mail List Content-Type: multipart/alternative; boundary=20cf307f3b2c39cc7904d96259ef X-Gm-Message-State: ALoCoQn+yndMOw2T+8ZewVrmaadt9PYoLjYkAN0U2F3btkGGjo1/xdhjIKbdPvfD1gsgqc5Zi/hy --20cf307f3b2c39cc7904d96259ef Content-Type: text/plain; charset=ISO-8859-1 Ok... i tried both DarkenColor and Darken and didn't notice any difference and just continued working with DarkenColor without researching what the difference between the two modes was. I got great results anyway, but will keep your advice on best pratice till next time ;-) - Robert 2013/4/2 Darby Edelen > The method I just posted does not despill directly. It gives you the > amount of green that despilling would remove from the image. You could set > this layer to "Subtract" to remove that green from the original image. > > Also I wouldn't use the Darker Color blend mode for despill. Darker Color > works based on the luminance of the pixel where as Darken works on a per > channel basis. The results generally won't be any different (unless you've > modified the R/B channels as well) but I consider Darken to be best > practice. > > > On Tue, Apr 2, 2013 at 1:46 AM, Robert Kjettrup wrote: > >> Hi Darby >> >> On my last two keying jobs i used your tip with channel mixer set to >> GR:-50, GG:0, GB:-50 and setting the layer to DarkerColor with great >> results, much better than the native despill in Keylight, thanks for that >> tip. :-) >> >> Just tried this other channelmixer usage, and that i can also see as a >> good alternative, but it is a bit harsher in its despill in darker areas, >> more like the default despill in keylight. I setup the layer with the >> channelmixer effect as a difference blending mode. Correct? it looks >> correct :-) >> >> - Robert >> >> >> >> 2013/4/2 Darby Edelen >> >>> A different Channel Mixer setup can pull a good spill matte. In the >>> case of a straight up average R+B spill suppression you'd zero out all the >>> fields in Channel Mixer and enter for the Green channels: >>> >>> Green-Red: -50 >>> Green-Green: 100 >>> Green-Blue: -50 >>> >>> Then the green channel will contain the spill matte. Shift Channels can >>> move that information around to use for either an Alpha or Luma matte. >>> Sure it's not as easy as having a one stop solution, but it's not too much >>> of a pain either :) >>> >>> >>> On Mon, Apr 1, 2013 at 6:31 PM, Chris Zwar wrote: >>> >>>> Thanks for these links, I enjoyed the videos. What a shame the native >>>> Cinegobs plugins are Windows only, it would be nice to be able to pull a >>>> spill-matte so easily on a Mac. >>>> >>>> -Chris >>>> >>>> On 10/03/2013, at 9:55 AM, Daniel Griffin wrote: >>>> >>>> I've found the Green Limited by the Average of Red and >>>> Blue technique very useful and has given me the best >>>> spill suppression results so far: >>>> >>>> >>>> http://ae.tutsplus.com/tutorials/workflow/advanced-spill-suppression-methods/ >>>> >>>> >>>> Daniel Griffin >>>> www.kubikafilms.com >>>> >>>> >>>> 2013/3/9 Chris Meyer >>>> >>>>> Many thanks, and absolutely nothing personal, but I have a client >>>>> confidentiality issue on these particular shots that I don't want to tread >>>>> anywhere near. >>>>> >>>>> - Chris >>>>> >>>>> ________________________________________________ >>>>> Chris Meyer | Crish Design >>>>> http://www.crishdesign.com >>>>> >>>>> >>>>> On Mar 9, 2013, at 12:48 PM, Andrew Embury wrote: >>>>> >>>>> I'd love to have a crack at that too Chris! >>>>> >>>>> Cheers. >>>>> >>>>> - Andrew >>>>> >>>>> >>>>> On 2013-03-09, at 2:13 PM, Jack Tunnicliffe wrote: >>>>> >>>>> Chris: >>>>> >>>>> Can you send me a frame. I'd be curious if I could focus in on the >>>>> area you're talking about. Maybe a screen grab with some arrows or pointers >>>>> showing me the area as well as a frame of the original, like a tiff. >>>>> >>>>> Jack Tunnicliffe >>>>> Java Post Production >>>>> 402, 2206 Dewdney ave. >>>>> Regina, SK >>>>> Canada >>>>> S4R 1H3 >>>>> P. 306-777-0150 >>>>> cell. 306-536-4321 >>>>> www.javapost.ca >>>>> ______________ >>>>> >>>>> >>>>> >>>>> On Mar 9, 2013, at 10:36 AM, Chris Meyer >>>>> wrote: >>>>> >>>>> I have found the edge decontamination of Refine Matte very useful in >>>>> some situations, but not all. The two shots causing me grief this week have >>>>> 1) a large amount of spill across the bare shoulder/upper arm of an >>>>> actress, and 2) more than half way across the shirt on the back of an actor >>>>> who is turning. Unfortunately, when I increase the decontamination range on >>>>> Refine Matte, it does something very ugly to the color channel information >>>>> - what looks like a edge pixel repeat, instead of a simple removing of >>>>> color from otherwise detailed pixels that I wish to keep. >>>>> >>>>> - Chris >>>>> >>>>> ________________________________________________ >>>>> Chris Meyer | Crish Design >>>>> http://www.crishdesign.com >>>>> >>>>> >>>>> On Mar 9, 2013, at 12:41 AM, Todd Kopriva wrote: >>>>> >>>>> Check out the Refine Matte effect for the motion blur color >>>>> decontamination. >>>>> >>>>> On Friday, March 8, 2013, Chris Zwar wrote: >>>>>> >>>>>> >>>>>> >>>>>> I don't know if it's technically 'spill', but motion blur is the >>>>>> tricky thing to deal with when keying. If the key isn't set up well then >>>>>> you can end up with footage that looks like it has a spill problem in >>>>>> moving areas - actually the screen is showing through semi-transparent >>>>>> regions. >>>>>> >>>>> >>>>> >>>>> -- >>>>> -------------------------------- >>>>> Todd Kopriva >>>>> 1510 E Alder St >>>>> Seattle, WA 98122 >>>>> (206)437-2296 >>>>> -------------------------------- >>>>> "Ultimately, all knowledge is self-knowledge." >>>>> - Bruce Lee >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>>> >>> >> > --20cf307f3b2c39cc7904d96259ef Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
Ok... i tried both DarkenColor and Darken and didn't n= otice any difference and just continued working with DarkenColor without re= searching what the difference between the two modes was. I got great result= s anyway, but will keep your advice on best pratice till next time ;-)

- Robert


2013/4/2 Darby Edelen = <dedelen@gmail.co= m>
The method I just posted does not despill directly.= =A0 It gives you the amount of green that despilling would remove from the = image.=A0 You could set this layer to "Subtract" to remove that g= reen from the original image.

Also I wouldn't use the Darker Color blend mode for despill.= =A0 Darker Color works based on the luminance of the pixel where as Darken = works on a per channel basis.=A0 The results generally won't be any dif= ferent (unless you've modified the R/B channels as well) but I consider= Darken to be best practice.

On Tue, Apr 2, 2013 at 1:46 AM, Robert Kje= ttrup <robert@stvmayday.dk> wrote:
Hi Darby

On my last two keying jobs i used your tip with channel mixer set to GR:-5= 0, GG:0, GB:-50 and setting the layer to DarkerColor with great results, mu= ch better than the native despill in Keylight, thanks for that tip. :-)

Just tried this other channelmixer usage, and that i ca= n also see as a good alternative, but it is a bit harsher in its despill in= darker areas, more like the default despill in keylight. I setup the layer= with the channelmixer effect as a difference blending mode. Correct? it lo= oks correct :-)

- Robert



2013/4/2 Da= rby Edelen <dedelen@gmail.com>
A diffe= rent Channel Mixer setup can pull a good spill matte.=A0 In the case of a s= traight up average R+B spill suppression you'd zero out all the fields = in Channel Mixer and enter for the Green channels:

Green-Red: -50
Green-Green: 100
Green-Blue: -50=

Then the green channel will contain the spill matte.=A0 Shift= Channels can move that information around to use for either an Alpha or Lu= ma matte.=A0 Sure it's not as easy as having a one stop solution, but i= t's not too much of a pain either :)


On Mon, Apr 1, 2013 at 6:31 PM, Chris Zwar <chris@chriszwar.com&g= t; wrote:
Thanks f= or these links, I enjoyed the videos. =A0What a shame the native Cinegobs p= lugins are Windows only, it would be nice to be able to pull a spill-matte = so easily on a Mac.

-Chris

On 10/03/2013, at 9:55 = AM, Daniel Griffin <dagrial@gmail.com> wrote:

I've found the Green Limited by the Average of Red and Blue=A0technique= =A0very useful and has given me the best spill=A0suppression=A0results so f= ar:



Daniel Griffin
www.kubikafilms.com


2013/3/9 Chris Meyer &= lt;chris@crishde= sign.com>
Many thanks, and absolutely nothing per= sonal, but I have a client confidentiality issue on these particular shots = that I don't want to tread anywhere near.

=A0- Chris

________________________________________________
Chris Meyer | Crish Des= ign
http://www= .crishdesign.com

On Mar 9, 2013, at 12:48 PM, Andrew Embury wr= ote:

I'd love= to have a crack at that too Chris!

Cheers.=A0

- Andrew


On 2013-03-09, at 2:13 PM, Jack T= unnicliffe <jack@j= avapost.ca> wrote:

Chris:
Can you send me a frame. I'd be curious if I could focus= in on the area you're talking about. Maybe a screen grab with some arr= ows or pointers showing me the area as well as a frame of the original, lik= e a tiff.

Jack Tunnicliffe
Java Post Production=
402, 2206 Dewdney ave.
Regina, SK
Canada
S4R 1H3
______________
<= /span>

=

On Mar 9, 2013, at 10:36 AM, Chris Meyer <chris@crishdesign.com> wr= ote:

I have found the edge decontamination of Refine Matte very useful in some = situations, but not all. The two shots causing me grief this week have 1) a= large amount of spill across the bare shoulder/upper arm of an actress, an= d 2) more than half way across the shirt on the back of an actor who is tur= ning. Unfortunately, when I increase the decontamination range on Refine Ma= tte, it does something very ugly to the color channel information - what lo= oks like a edge pixel repeat, instead of a simple removing of color from ot= herwise detailed pixels that I wish to keep.=A0

=A0- Chris

________________________________________________
Chris Meyer | Crish Des= ign
http://www= .crishdesign.com

On Mar 9, 2013, at 12:41 AM, Todd Kopriva wrote:

Check out the=A0Refine Matte effect for the motion b= lur color decontamination.

On Friday, March 8, 2013, Chris Zwar wro= te:


I don't know if it's technically 'spill', but motion blur i= s the tricky thing to deal with when keying. =A0If the key isn't set up= well then you can end up with footage that looks like it has a spill probl= em in moving areas - actually the screen is showing through semi-transparen= t regions. =A0


--
--------------------------------
Todd Kopriv= a
1510 E Alder St
Seattle, WA 98122
(206)437-2296
---------= -----------------------
"Ultimately, all knowledge is self-knowledge."
=A0=A0 - Bruce Lee



<= br>





--20cf307f3b2c39cc7904d96259ef--