From: "Chris Zwar" Received: from mail-pg0-f43.google.com ([74.125.83.43] verified) by media-motion.tv (CommuniGate Pro SMTP 6.1.0) with ESMTPS id 6438744 for AE-List@media-motion.tv; Fri, 20 Apr 2018 07:36:19 +0200 Received: by mail-pg0-f43.google.com with SMTP id i29so2051209pgn.12 for ; Thu, 19 Apr 2018 22:41:58 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chriszwar-com.20150623.gappssmtp.com; s=20150623; h=from:mime-version:subject:date:references:to:in-reply-to:message-id; bh=7175se+WhCdKdeBqArn0/6fjtlvt4QbRRCoGaqmnW6U=; b=CDU7w/S0FtTn8SS6+G6fMltcUjY5QYiFH61iobj9ApqiAjZfasO8zQQkXSWdkiMtog l+XSdMVcWAjiryFBRMYvnH2ZYuYAIBFlXn/GoS/5+tqgkYsJcxRpn0aVNnfPgMSrZMyE 043DGIWbfI9kJWmFRzUyHBfyqPKHVjjPApeIwj7YmZcWWt3kikBD2eoVIPFq4xXjMWKf 2j4QTDQhp+uHeDD4ByypIpyTlgC15BsqeHhDiZkqTn4CBaRoeNHSsTWjBuhXEAEdBce3 ecRZtzLiSb2qPZ6Q07g4l9XSytKpNgIpscjXx7b9R4qCK0hfor7AWgpxKhdxvN+EEMt9 wA9A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:mime-version:subject:date:references:to :in-reply-to:message-id; bh=7175se+WhCdKdeBqArn0/6fjtlvt4QbRRCoGaqmnW6U=; b=AD+0+iogzFITA7VeJoTnwBg3MeLT0yNG3NHLodAsDrW/R+Ofh8oqKiO81/AaNyYfXd HbkXuToEHIl1AN8yptfdYZJhgZZsl4l0UVeLwcg7RSK+we4bDT6Qu/MbiLEX63dNVIt4 me3AewO79us545ykdsCgaURZL9QC+AJt20AOGq0slIjUrb8F8dS9f6t1oBlbJ/vr2PpJ ITwj3v5azDkGW8wG43fCtYseuQnOfU3EBjxUapfpEgBBRN4uByiHMYh6/QD5XtnKAADI j5kbk0ruCSBjEc3xnpTkMUptArrxlSKLiUD3P+L1gIwIK31NNZUCTiaBLWS3Fw8Pzr/H TevA== X-Gm-Message-State: ALQs6tDSYIGh4KtBMGncKD1GR43IvhItJFCYLs0ge5wvP0GAmMozsGkf uDppi46ZIz9HSxQ09dlcX6KiqYpWS8M= X-Google-Smtp-Source: AIpwx4+34rjoyZNs3iH1Z1HBshh+wNYhHKE3H4vQZsxhtuUv3IdwgMfbeG5n9lEvOGBt4qNvBTmKdQ== X-Received: by 10.98.47.2 with SMTP id v2mr4712387pfv.239.1524202916151; Thu, 19 Apr 2018 22:41:56 -0700 (PDT) Return-Path: Received: from [192.168.3.137] (220-244-139-34.static.tpgi.com.au. [220.244.139.34]) by smtp.gmail.com with ESMTPSA id e190sm8905208pfe.171.2018.04.19.22.41.54 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Thu, 19 Apr 2018 22:41:55 -0700 (PDT) Content-Type: multipart/alternative; boundary="Apple-Mail=_3A0259F2-9AB8-4145-8B14-C56AB274FCA8" Mime-Version: 1.0 (Mac OS X Mail 11.3 \(3445.6.18\)) Subject: Re: [AE] Inverse Square Law for glows Date: Fri, 20 Apr 2018 15:41:50 +1000 References: To: After Effects Mail List In-Reply-To: Message-Id: <9BB1D1FD-2F86-430C-8B38-3F5B49A0A76F@chriszwar.com> X-Mailer: Apple Mail (2.3445.6.18) --Apple-Mail=_3A0259F2-9AB8-4145-8B14-C56AB274FCA8 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 That=E2=80=99s a really cool tutorial, it=E2=80=99s basically the same = way I=E2=80=99ve been doing glows except it never occurred to me to = stack them all up on the one layer. Thanks for the link. However two things spring to mind. Firstly, I make things glow A LOT = and I usually find that different colours need different blur settings = IF you want them to look the same. Linking all the blurs for the same = layer with an expression is clever and a great starting point, but I = think that if you had elements that were all unique colours, you might = want to manually adjust some of the numbers if you want the glow to look = the same on them all. Secondly - and I really don=E2=80=99t know, just thinking out loud - = while multiple layers with different blur settings is undoubtedly more = accurate and better looking than the default AE glow, I=E2=80=99m still = not sure that it=E2=80=99s a true inverse square because the blur = profile itself is gaussian. So you are stacking up multiple bell-curves = on top of each other. Sure it looks good but I don=E2=80=99t think = it=E2=80=99s 100% mathematically perfect. -Chris > On 18 Apr 2018, at 4:47 pm, Mathias M=C3=B6hl = wrote: >=20 > Here is a great Ae tutorial about building your own glow preset based = on inverse square law: > https://www.youtube.com/watch?v=3D--LsJ0DErgM = >=20 > Cheers, > Mathias >=20 >=20 >=20 > mamoworld.com > simplify your cgi > Dr. Mathias M=C3=B6hl > Marlow 52 , 18551 Sagard, Germany > Phone +49 (0)38302 - 88 75 61 > info@mamoworld.com >=20 >=20 >> On 18 Apr 2018, at 01:37, Chris Zwar > wrote: >>=20 >> If you want to do it mathematically then it=E2=80=99s more complex = than you think, as you need to consider the gamma of the workspace = you=E2=80=99re using. >>=20 >> Making something glow by stacking up multiple layers with different = blurs is faking it, but probably the best way to do it in 8 or 16 bit = modes. It=E2=80=99s the way i do it too, although I use the screen mode = (not add), and I usually have a desaturated layer near the top set to = classic colour dodge to adjust the =E2=80=9Chotness=E2=80=9D. =46rom a = mathematical perspective, you=E2=80=99re not just looking at the blur = profile but also the gamma of the workspace so this multiple layer = approach is going to be difficult to calculate. >>=20 >> The alternative approach is to comp in 32 bit linear mode, where you = don=E2=80=99t need to stack up multiple layers. You just have your = source layer and make the colour super hot. Then you blur it and the = blur naturally spreads out the light. The falloff will depend on the = type of blur you use. I did ask a while ago (for exactly the same = reason as you) if there=E2=80=99s a blur with an adjustable profile = including inverse square, which would be very useful for comping glows = in 32 bit float mode. I can=E2=80=99t remember the answer, but = there=E2=80=99s something for you to look at. >>=20 >> So basically - multiple layers using add/screen and blur is an 8/16 = bit fudge for a gamma working space like sRGB/HDR. A single, super hot = layer with a single blur in 32 bit linear float is the more natural = approach. >>=20 >> Or you can just use a plugin that does it, as lloyd suggests :-) >>=20 >>=20 >> -Chris >>> On 18 Apr 2018, at 4:28 am, Byron Nash > wrote: >>>=20 >>> So, if I were applying that to say, the blur of a layer, what would = be the "distance" variable? I'm stacking several layers with Add = transfer modes and variable amounts of blur. >>>=20 >>> On Tue, Apr 17, 2018 at 2:17 PM Walter Soyka = > wrote: >>> The inverse square law states that intensity is inversely = proportional to the square of the distance from the source. = Symbolically: >>> intensity =3D 1 / (distance^2) >>>=20 >>> Doing a little math, when the distance is 1, the intensity is 1. = When the distance is 2, the intensity is 0.25. When the distance is 3, = the intensity is 0.111. When the distance is 4, the intensity is 0.0625. >>>=20 >>> walter soyka =E2=96=BC keen live=20 >>> walter@keenlive.com >>>=20 >>> On Tue, Apr 17, 2018 at 1:46 PM, Byron Nash > wrote: >>> Always the consummate salesman Lloyd! ;-)=20 >>>=20 >>> I think this was more of an experiment to see if I could do it = myself and understand the math a bit better.=20 >>>=20 >>> On Tue, Apr 17, 2018 at 1:05 PM Lloyd Alvarez = > wrote: >>> Oooooorrrr you could just use Real Glow = ! ;-) >>>=20 >>> Cheers, >>> Lloyd >>>=20 >>> On Tue, Apr 17, 2018 at 12:16 PM, Byron Nash = > wrote: >>> I'm trying to roll my own setup for a better glow. If I'm trying to = follow the inverse square law to do the math, would I half/double the = amounts at each iteration? Or square root? >>>=20 >>>=20 >>=20 >=20 >=20 --Apple-Mail=_3A0259F2-9AB8-4145-8B14-C56AB274FCA8 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8 That=E2=80=99s a really cool tutorial, it=E2=80=99s basically = the same way I=E2=80=99ve been doing glows except it never occurred to = me to stack them all up on the one layer.  Thanks for the link.

However two things = spring to mind.  Firstly, I make things glow A LOT and I usually = find that different colours need different blur settings IF you want = them to look the same.  Linking all the blurs for the same layer = with an expression is clever and a great starting point, but I think = that if you had elements that were all unique colours, you might want to = manually adjust some of the numbers if you want the glow to look the = same on them all.

Secondly - and I really don=E2=80=99t know, just thinking out = loud - while multiple layers with different blur settings is undoubtedly = more accurate and better looking than the default AE glow, I=E2=80=99m = still not sure that it=E2=80=99s a true inverse square because the blur = profile itself is gaussian.  So you are stacking up multiple = bell-curves on top of each other.  Sure it looks good but I don=E2=80= =99t think it=E2=80=99s 100% mathematically perfect.

-Chris

On 18 Apr 2018, at 4:47 pm, Mathias M=C3=B6hl <AE-List@media-motion.tv> wrote:

Here is a = great Ae tutorial about building your own glow preset based on inverse = square law:
https://www.youtube.com/watch?v=3D--LsJ0DErgM

Cheers,
Mathias



mamoworld.com
simplify your cgi

Dr. Mathias M=C3=B6hl
Marlow 52 , 18551 Sagard, Germany
Phone +49 = (0)38302 - 88 75 61
info@mamoworld.com


<= /div>

On 18 Apr 2018, at 01:37, Chris = Zwar <AE-List@media-motion.tv> wrote:

If you want to do it = mathematically then it=E2=80=99s more complex than you think, as you = need to consider the gamma of the workspace you=E2=80=99re using.

Making something glow by = stacking up multiple layers with different blurs is faking it, but = probably the best way to do it in 8 or 16 bit modes.  It=E2=80=99s = the way i do it too, although I use the screen mode (not add), and I = usually have a desaturated layer near the top set to classic colour = dodge to adjust the =E2=80=9Chotness=E2=80=9D.  =46rom a = mathematical perspective, you=E2=80=99re not just looking at the blur = profile but also the gamma of the workspace so this multiple layer = approach is going to be difficult to calculate.

The alternative approach is to comp in = 32 bit linear mode, where you don=E2=80=99t need to stack up multiple = layers.  You just have your source layer and make the colour super = hot.  Then you blur it and the blur naturally spreads out the = light.  The falloff will depend on the type of blur you use. =  I did ask a while ago (for exactly the same reason as you) if = there=E2=80=99s a blur with an adjustable profile including inverse = square, which would be very useful for comping glows in 32 bit float = mode.  I can=E2=80=99t remember the answer, but there=E2=80=99s = something for you to look at.

So basically - multiple layers using = add/screen and blur is an 8/16 bit fudge for a gamma working space like = sRGB/HDR.  A single, super hot layer with a single blur in 32 bit = linear float is the more natural approach.

Or you can just use a plugin that does = it, as lloyd suggests :-)


-Chris
On 18 Apr 2018, at 4:28 am, Byron Nash <AE-List@media-motion.tv> wrote:

So, if I were applying that to say, the blur of a layer, what = would be the "distance" variable? I'm stacking several layers with Add = transfer modes and variable amounts of blur.

On Tue, Apr 17, 2018 = at 2:17 PM Walter Soyka <AE-List@media-motion.tv> wrote:
The inverse square law states = that intensity is inversely proportional to the square of the distance = from the source. Symbolically:
intensity =3D 1 / = (distance^2)

Doing a = little math, when the distance is 1, the intensity is 1. When the = distance is 2, the intensity is 0.25. When the distance is 3, the = intensity is 0.111. When the distance is 4, the intensity is = 0.0625.

walter soyka =E2=96=BC keen live 
walter@keenlive.com

On Tue, Apr 17, 2018 at 1:46 = PM, Byron Nash <AE-List@media-motion.tv> wrote:
Always the consummate salesman Lloyd! ;-) 

I think this was more of = an experiment to see if I could do it myself and understand the math a = bit better. 

On Tue, Apr 17, 2018 = at 1:05 PM Lloyd Alvarez <AE-List@media-motion.tv> wrote:
Oooooorrrr you could just use Real Glow! ;-)

Cheers,
Lloyd

On Tue, Apr 17, 2018 at 12:16 = PM, Byron Nash <AE-List@media-motion.tv> wrote:
I'm trying to roll my own setup for a better glow. If I'm = trying to follow the inverse square law to do the math, would I = half/double the amounts at each iteration? Or square root?






= --Apple-Mail=_3A0259F2-9AB8-4145-8B14-C56AB274FCA8--