ExpoDisc Conversation with Diane Wallace

There seems to be a confusion here. This is what was done in the test posted:

He took a shot with the Expodisc in a white light source, and showed that it was it had average WB numbers, and perfectly matching RGB. No color cast, as the light wasn't tinted.

Then, he set the WB using either the ExpoDisc, the Pringles lid, the socks, etc.

Finally, he shot WITH the ExpoDisc (known to be R=G=B) and checked the RGB.

He USED the ExpoDisc shot as a "known" grey card. Setting the WB with other things threw off the numbers.

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Lee,

LeeBase said:

LeeBase said:

The only problem with this theory (that pointing the expodisc at
the subject is a proper and reliable technique) is that it makes
absolutely no sense whatsoever.

Click to expand...

I still don't understand your objection. The ExpoDisc produces an
image that, for a given light source, tells the camera what "grey"
should look like.

Click to expand...

That's true if what the rig is pointed at is, on average over the admittedly-wide field of view of the Expodisc, gray.

LeeBase said:

Then you tell the camera, use this picture for grey, and the camera
does, which pulls all the colors into balance.

The color of the source light does not make a difference. It works
with the sun, with all the various indoor light sources, with mixed
light.

It's what the expodisc does to take WHATEER the source of light and
come up with a "grey correcting" image.

Click to expand...

The concern being spoken of is that with the camera-cum-Expodisk rig pointed at the subject, the light it sees reflects the chrominance of the scene (much like the light seen by the camera when it does an AWB calculation, although the ExpoDisc takes in a wider "field of view".).

But what we want to do is to determine the chrominance of the incident light, so the camera can, in effect, "back that out" of the reflected light image, leaving us with the chromaticites of the different parts of the image as they would be when illuminated by some "standard" light (or, to put it in "paint colorimetry" terms, the reflection chromaticity of thediffeetn parts of the object).

Now, argues Wallace, although this is conceptually true, nevertheless because of the large angle of acceptance of the ExpoDisc cum camera rig, only part of the light is "colored" by the subject. The rest will be "incident light".

Well, I have news for you. If the subject is standing in front of a tall, wide red wall, then a great amount of the light accepted by the ExpoDisc will not have the chromatcity of the incident light, but rather the chromaticity of the light reflected from the wall.

What makes the "point it at the subject" mode of operation work fairly well most of the time is that, for many of the scenes we aim at, there is a wide range of "reflective chromaticiites" within the wide "field of view" of the ExpoDisc diffuser, and if we are lucky its average reflective chomaticity will be "neutral"; that is, the average chromaticity of the reflected light will be the same as the chromaticity of the incident light. Note that I said, "if we are lucky" - there is no reality of nature that will guarantee that situation.

But I assure you, if we point the camera-cum-ExpoDisc at a human standing in front of a wide, tall red dirigible hangar wall and take our WB measurement shot aimed at the subject, then set the custom WB from that, and use the resulting custom WB for our real shot, you'll end up with a really green-looking image of the subject. And it may show her standing in front of a rather gray wall.

What is silly is to spend a reasonable amount of money to get an "individually calibrated" diffuser, capable indeed of playing a role in a precise measurement, and then aim it at something that does not necessarily exhibit what we are trying to measure.

Best regards,

Doug

... all that marketing-speak, which says almost nothing about how the thing is actually supposed to work.

Lee, you still seem unable, or unwilling, to ask yourself the basic question "what is the principle behind this device?"

I'll have to fall back on examples - what if the subject is principally orange? How can the expodisc fail to report to the camera that the scene is (on average) rather orange? And when it does, the CWB will adjust away from orange, giving the net result that the "corrected" images will be too cold?

Btw, please don't confuse the issue with the snow example, which has nothing to do with WB. It's not helping!

DB

LeeBase said:

LeeBase said:

The snow/grey problem is nothing to do with WB, by the way, it's
simply an exposure and metering challenge. The camera wants to
render near white as mid-grey unless you compensate.

Click to expand...

Use an expodisc and watch white become white again. The problem
with snow is the same problem, the assumption that the camera makes
that scenes are "middle grey".

Sure, you can compensate by increasing exposure.

And you can use the handy expodisc tool as well.

LeeBase said:

Back to my question. When you shoot a scene with a grey card you
have a known target to which you adjust. When you shoot with an
expodisc you are in effect treating the 180 degree scene in front
of the camera as a grey card. My question is, how do you know it is
a balanced "grey card"? Actually, why would it be, except by
coincidence?

Click to expand...

The expodisc diffuses the light to give you an image that is just
as good (or better than) a grey card.

From their website:
ExpoDiscs work in conjunction with built-in sensors and software to
instantly create an accurate white balance reference regardless of
the light source (incident or reflective light, including all mixed
light and flash) or the color of the subject (including close-ups,
and subjects lacking very light objects or having one dominant
color). The multiple facets of the prismatic lens (more efficient
than a dome) collect multiple samples of light from 180 degrees,
scrambling it to create an accurate average (in the correct
proportion) of all light that contributes to subject illumination.
A white diffusion disc provides even illumination across the full
frame, and the appropriate color-correcting filters are sandwiched
in-between during assembly to create units which are individually
color-corrected to within 3% of being perfectly neutral. Each unit
is provided with a card noting it’s precise color rating and
density. No other commercially available product can provide this
level of accuracy.

Lee

Click to expand...

--
DB

LeeBase said:

... The ExpoDisc produces an
image that, for a given light source, tells the camera what "grey"
should look like.

Click to expand...

The expodisc just smushes together all the light it receives. When used properly, it produces an image that's uniform, and which measures the tint of the ambient light. This "tint" is exactly what you'd expect to get if you lit a grey card with the same kind of ambient light. When you pass such an image to the camera's CWB function, it says "thank you for telling me what the ambient light looks like, I'll subtract this 'tint' from future pictures you take".

If you point the expodisc at your light source, then you can reasonably assume that the smushed-together light it passes on will be the color of the light source. If you point the expodisc at your subject, then the light it smushes together will have reflected off of your subject, and will not necessarily accurately reflect the tint of the ambient light any more.

The expodisc is just a diffuser, one which is promised to not impart its own tint to light that passes through.
-harry

Read what their website has to say:

How does the ExpoDisc read reflective light?

The prismatic front of the ExpoDisc collects light from 180 degrees - it is not looking only at what is directly in front of the camera. In fact, to get a reading influenced by a dominant color central subject, you would have to be very close in order to reduce the impact of skylight and other ambient light spilling in from the sides to contribute to the WB reading. When you step away from the central subject, the influence of skylight and ambient light will become more apparent.

The many faceted planes in the front lens provide the most efficient means (better than a dome) of collecting multiple samples of all light sources. The prismatic pattern of the lenticular front element transmits an increasingly larger percentage of ambient light as the source swings more off axis, yielding an accurate average of all illumination for perfect WB.

The near-perfect neutral balance (+ - 3%) of the ExpoDisc provides digital programs the best means for determining the amount of correction required in any lighting. You will download near-perfect WB.

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Good to see someone is reading objectively. Someone replied to me lower down telling me that the thing does "EXACTLY what it is advertised to do". But does that thing really need doing?

regards.

David Barker said:

... all that marketing-speak, which says almost nothing about how
the thing is actually supposed to work.

Lee, you still seem unable, or unwilling, to ask yourself the basic
question "what is the principle behind this device?"

Click to expand...

hm said:

I'll subtract this 'tint' from future pictures you take".

Click to expand...

Exactly. Which is why it works wheter you aim it at the light source or the subject. In either case it'll give you an image from which grey can be calculated and the difference applied to future images such that their colors fall in line.

hm said:

If you point the expodisc at your light source, then you can
reasonably assume that the smushed-together light it passes on will
be the color of the light source.

Click to expand...

In mixed lighting if you aim at one source, you will have the color cast from the source you don't aim at.

hm said:

If you point the expodisc at your
subject, then the light it smushes together will have reflected off
of your subject, and will not necessarily accurately reflect the
tint of the ambient light any more.

Click to expand...

It doesn't need to reflect the ambient light. It just needs to take whatever light it has and calculate the grey. This is why it DOES WORK to aim at your subject and take the expodisc shot.

Just like it works to have your subjects hold a grey card and then use a dropper tool in photoshop to get your color balance.

hm said:

The expodisc is just a diffuser, one which is promised to not
impart its own tint to light that passes through.

Click to expand...

One gauranteed to be manufactured to strict tolerenaces and which is individually measured such that when you receive it, you'll know exactly how far off it is.

Lee

Yeah - I've read that lot as well!! I know HOW it reads the reflected light. What I want to know is why the $^@!\ that sample of light is relevant to the setting of an accurate CWB, except by coincidence, more or less.

Take Doug's example of a person standing in front of a red dirigible hangar as an example, and you'll quickly see what I mean (I hope). The camera will "see" a diffuse reddish scene, the CWB will produce a complementary adjustment, and hey presto! - the final image looks dreadful. The red hangar will be grey (or would be if the camera could adjust that far) and the person will look ill.

I know this is an extreme example, but surely it's in the tough situations that WB is difficult to get right?

DB

pbleic said:

Read what their website has to say:

How does the ExpoDisc read reflective light?

The prismatic front of the ExpoDisc collects light from 180 degrees

• it is not looking only at what is directly in front of the

camera. In fact, to get a reading influenced by a dominant color
central subject, you would have to be very close in order to reduce
the impact of skylight and other ambient light spilling in from the
sides to contribute to the WB reading. When you step away from the
central subject, the influence of skylight and ambient light will
become more apparent.

The many faceted planes in the front lens provide the most
efficient means (better than a dome) of collecting multiple samples
of all light sources. The prismatic pattern of the lenticular front
element transmits an increasingly larger percentage of ambient
light as the source swings more off axis, yielding an accurate
average of all illumination for perfect WB.

The near-perfect neutral balance (+ - 3%) of the ExpoDisc provides
digital programs the best means for determining the amount of
correction required in any lighting. You will download near-perfect
WB.

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Click to expand...

--
DB

There are a TON of pros who agree with you. PS - someone mentioned there are probably a ton who don't.

Anybody have the name of a well known pro photographer (someone with a professional appearing website who makes all of their money from either shooting or teaching photography) who thinks the ExpoDisc is bad for setting WB??? Or even not worth the $100? These people spend $100 on a UV filter for goodness sakes!

The only reason there is such a debate here is because people (including myself) are mostly amateur photographers who are very cost concious about their hobby.

Paul

LeeBase said:

LeeBase said:

I'll subtract this 'tint' from future pictures you take".

Click to expand...

Exactly. Which is why it works wheter you aim it at the light
source or the subject. In either case it'll give you an image from
which grey can be calculated and the difference applied to future
images such that their colors fall in line.

LeeBase said:

If you point the expodisc at your light source, then you can
reasonably assume that the smushed-together light it passes on will
be the color of the light source.

Click to expand...

In mixed lighting if you aim at one source, you will have the color
cast from the source you don't aim at.

LeeBase said:

If you point the expodisc at your
subject, then the light it smushes together will have reflected off
of your subject, and will not necessarily accurately reflect the
tint of the ambient light any more.

Click to expand...

It doesn't need to reflect the ambient light. It just needs to
take whatever light it has and calculate the grey. This is why it
DOES WORK to aim at your subject and take the expodisc shot.

Just like it works to have your subjects hold a grey card and then
use a dropper tool in photoshop to get your color balance.

LeeBase said:

The expodisc is just a diffuser, one which is promised to not
impart its own tint to light that passes through.

Click to expand...

One gauranteed to be manufactured to strict tolerenaces and which
is individually measured such that when you receive it, you'll know
exactly how far off it is.

Lee

Click to expand...

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Yep - it won't work here. Your right.
Do you take a lot of these type photos?

David Barker said:

Take Doug's example of a person standing in front of a red
dirigible hangar as an example, and you'll quickly see what I mean
(I hope). The camera will "see" a diffuse reddish scene, the CWB
will produce a complementary adjustment, and hey presto! - the
final image looks dreadful. The red hangar will be grey (or would
be if the camera could adjust that far) and the person will look
ill.

I know this is an extreme example, but surely it's in the tough
situations that WB is difficult to get right?

DB

pbleic said:

Read what their website has to say:

How does the ExpoDisc read reflective light?

The prismatic front of the ExpoDisc collects light from 180 degrees

• it is not looking only at what is directly in front of the

camera. In fact, to get a reading influenced by a dominant color
central subject, you would have to be very close in order to reduce
the impact of skylight and other ambient light spilling in from the
sides to contribute to the WB reading. When you step away from the
central subject, the influence of skylight and ambient light will
become more apparent.

The many faceted planes in the front lens provide the most
efficient means (better than a dome) of collecting multiple samples
of all light sources. The prismatic pattern of the lenticular front
element transmits an increasingly larger percentage of ambient
light as the source swings more off axis, yielding an accurate
average of all illumination for perfect WB.

The near-perfect neutral balance (+ - 3%) of the ExpoDisc provides
digital programs the best means for determining the amount of
correction required in any lighting. You will download near-perfect
WB.

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Click to expand...

--
DB

Click to expand...

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Doug Kerr said:

Well, I have news for you. If the subject is standing in front of a
tall, wide red wall, then a great amount of the light accepted by
the ExpoDisc will not have the chromatcity of the incident light,
but rather the chromaticity of the light reflected from the wall.

Click to expand...

Ok....whenever I take pictures of people in front of giant red hangers, I'll use the incident method instead.

Outdoors, I'll probably aim at the sun.

But in mixed lighting indoors, I'll aim at the subjects.

Lee

The only circumstances in which using a grey card and using the expodisc for reflected light sampling would be comparable, would be if you were to point the grey card towards the subject, photograph the grey side of the card (to sample the light reflected from the subject onto the card), and use this as your CWB reference. I can't imagine anyone doing this, but who knows?

--
DB

Another fascinating post, thanks to you.

I think I like many have found such a huge leap in quality from AWB (or one of the presets) to the Pringles lid (or coffee filters etc) - that for the outlay involved (ie a couple of cents and a slight heartburn), we're simply enamoured with the reults! The next step probably seems to most (me included at this stage in my ability/requirements) to be a huge outlay ($100) for very small increase in quality.

I'm sure at some stage we may also feel the need for perfect WB in all conditions where nearly close won't do - it's just a shame people can't accept that and (sadly like so many other topics on this forum) will fight and miss-inform to suit their particular corner.

But anyway, until we reach that stage, I would appreciate if you could keep firing the info our way, so that those of us that do try to advance our own knowledge and ability may learn from your learning.

--
Kind Regards
DWBell

David Barker91965 said:

Lee, you still seem unable, or unwilling, to ask yourself the basic
question "what is the principle behind this device?"

Click to expand...

Why should I have to? I've read the reasonings that you call "market speak". I had the need, which is why I looked into this at all. And so I bought the product.

And guess what....it works as advertised.

So why should I be bothered by you trying to say it won't work when it does?

Other things give good results as well and are far cheaper. The pringles lids, coffee filters etc. I've even showed my tests with those.

David Barker91965 said:

I'll have to fall back on examples - what if the subject is
principally orange? How can the expodisc fail to report to the
camera that the scene is (on average) rather orange? And when it
does, the CWB will adjust away from orange, giving the net result
that the "corrected" images will be too cold?

Click to expand...

I'm sure over time I'll discover limits to the usefulness of the ExpoDisc. For now, it's helped me do exactly what it said it would -- get spot on color right out of the camera -- instead of having to try and correct the colors in post processing.

It is very easy to use. As someone who shoots more than 10,000 picts a year -- I'm ecstatic to FINALLY be getting good color in all the lighting situations I shoot in -- right out of the camera.

It appears that there are those who have no problem accepting AWB or who do not find it a PIA to do color correction in post processing.

Hooray for them. But that's not me.

I shoot a lot of indoor shots and now I can finally just bring the camera home and post the shots on the web. The only post processing required is Noise Ninja, which works in batch mode so I just set it and forget it and come back later to good picts.

For the relative few that I actually care to print, those I'll spend the necessary time to get "just right" in post processing.

David Barker91965 said:

Btw, please don't confuse the issue with the snow example, which
has nothing to do with WB. It's not helping!

Click to expand...

You want to help? Then why don't you put your time and money where your mouth is. Where are your test photos? Where is your solution? Why don't you put up or shut up.

Lee

As I said, it's at the extremes that we need most help.

Having said that, this infuriating camera struggles in what should be the least extreme circumstances, particularly indoors. I really don't understand why the AWB makes such a hash of the job under tungsten light.

I acknowledge the excellent results given by ExpoDisc, Pringles, coffee filters etc. I want to understand why they work so well. It seems to be more or less by coincidence that the 180 degree hemisphere of light sampled is often closely indicative of the colour balance of the scene, but if so why can't Canon sample that light? (This is a rhetorical question).

DB

pbleic said:

David Barker said:

Take Doug's example of a person standing in front of a red
dirigible hangar as an example, and you'll quickly see what I mean
(I hope). The camera will "see" a diffuse reddish scene, the CWB
will produce a complementary adjustment, and hey presto! - the
final image looks dreadful. The red hangar will be grey (or would
be if the camera could adjust that far) and the person will look
ill.

I know this is an extreme example, but surely it's in the tough
situations that WB is difficult to get right?

DB

pbleic said:

Read what their website has to say:

How does the ExpoDisc read reflective light?

The prismatic front of the ExpoDisc collects light from 180 degrees

• it is not looking only at what is directly in front of the

camera. In fact, to get a reading influenced by a dominant color
central subject, you would have to be very close in order to reduce
the impact of skylight and other ambient light spilling in from the
sides to contribute to the WB reading. When you step away from the
central subject, the influence of skylight and ambient light will
become more apparent.

The many faceted planes in the front lens provide the most
efficient means (better than a dome) of collecting multiple samples
of all light sources. The prismatic pattern of the lenticular front
element transmits an increasingly larger percentage of ambient
light as the source swings more off axis, yielding an accurate
average of all illumination for perfect WB.

The near-perfect neutral balance (+ - 3%) of the ExpoDisc provides
digital programs the best means for determining the amount of
correction required in any lighting. You will download near-perfect
WB.

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Click to expand...

--
DB

Click to expand...

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

Click to expand...

--
DB

David Barker91965 said:

Yeah - I've read that lot as well!! I know HOW it reads the
reflected light. What I want to know is why the $^@!\ that sample
of light is relevant to the setting of an accurate CWB, except by
coincidence, more or less.

Click to expand...

What does it matter? If it works it works. How hard is it to point towards the sun or the lights in order to take an incident shot versed shooting at the scene? It's not hard at all.

But it mixed light, you'll have to decide which of the incicent lights to select or try to get the correct blance between them -- or you just shoot the scene.

Now if you have mixed lighting in front of a giant red hanger, then I guess you are screwed. You'll have to use a pringles lid for that shot.

Lee

A very reasonable post. You are right, the pringles lid/coffee filters get folks a good deal of the benefits of the expodisc.

Whether it's worth the $100 to get even better is certainly up to the individual.

Lee

DWBell said:

Another fascinating post, thanks to you.

I think I like many have found such a huge leap in quality from AWB
(or one of the presets) to the Pringles lid (or coffee filters etc)

• that for the outlay involved (ie a couple of cents and a slight

heartburn), we're simply enamoured with the reults! The next step
probably seems to most (me included at this stage in my
ability/requirements) to be a huge outlay ($100) for very small
increase in quality.

I'm sure at some stage we may also feel the need for perfect WB in
all conditions where nearly close won't do - it's just a shame
people can't accept that and (sadly like so many other topics on
this forum) will fight and miss-inform to suit their particular
corner.

But anyway, until we reach that stage, I would appreciate if you
could keep firing the info our way, so that those of us that do try
to advance our own knowledge and ability may learn from your
learning.

--
Kind Regards
DWBell

Click to expand...

LeeBase said:

snip] ... I bought the product.
And guess what....it works as advertised.

So why should I be bothered by you trying to say it won't work when
it does?

Click to expand...

You've a short memory - I agree that it works, and I'm keen to explore "why?" and "how?" I accept that you're not and I won't challenge you any more on this.

LeeBase said:

Other things give good results as well and are far cheaper. The
pringles lids, coffee filters etc. I've even showed my tests with
those.

Click to expand...

You have done a wonderful job, and helped many people increase their understanding of CWB. Judging from Paul's phone conversation you should be on commission!

LeeBase said:

I'm sure over time I'll discover limits to the usefulness of the
ExpoDisc. For now, it's helped me do exactly what it said it would
-- get spot on color right out of the camera -- instead of having
to try and correct the colors in post processing.

It is very easy to use. As someone who shoots more than 10,000
picts a year -- I'm ecstatic to FINALLY be getting good color in
all the lighting situations I shoot in -- right out of the camera.

Click to expand...

Message received.

LeeBase said:

LeeBase said:

Btw, please don't confuse the issue with the snow example, which
has nothing to do with WB. It's not helping!

Click to expand...

You want to help? Then why don't you put your time and money where
your mouth is. Where are your test photos? Where is your
solution? Why don't you put up or shut up.

Click to expand...

I'm sorry, I seem to have touched a nerve. I thought we could debate these things here on the forum, but maybe we should all shut up and save Phil the cost of any more servers. Seriously, I'll post some examples if you like, but I'm not sure there's much life left in this discussion.

--
DB

DWBell said:

Another fascinating post, thanks to you.

I think I like many have found such a huge leap in quality from AWB
(or one of the presets) to the Pringles lid (or coffee filters etc)

• that for the outlay involved (ie a couple of cents and a slight

heartburn), we're simply enamoured with the reults! The next step
probably seems to most (me included at this stage in my
ability/requirements) to be a huge outlay ($100) for very small
increase in quality.

I'm sure at some stage we may also feel the need for perfect WB in
all conditions where nearly close won't do - it's just a shame
people can't accept that and (sadly like so many other topics on
this forum) will fight and miss-inform to suit their particular
corner.

But anyway, until we reach that stage, I would appreciate if you
could keep firing the info our way, so that those of us that do try
to advance our own knowledge and ability may learn from your
learning.

--
Kind Regards
DWBell

Click to expand...

--
Paul

------------------------------------------------
Pbase supporter
Photographs at: http://www.pbase.com/pbleic
--------------------------------------------------
Copyright 2003, 2004 All rights reserved.

LeeBase said:

What does it matter? If it works it works.

Click to expand...

Just simple human curiosity. It's how we discovered about electricity, and fire, and beer, gravity, the rainbow, DNA and interesting stuff like that.

--
DB