RE: Colorsync-users Digest, Vol 7, Issue 117
RE: Colorsync-users Digest, Vol 7, Issue 117
- Subject: RE: Colorsync-users Digest, Vol 7, Issue 117
- From: "Wheeler, Barry" <email@hidden>
- Date: Sun, 2 May 2010 17:24:40 -0400
- Acceptlanguage: en-US
- Thread-topic: Colorsync-users Digest, Vol 7, Issue 117
Getting back to basics on this thread, I'll note that the Library of Congress has moved away from PhaseOne scanning backs to P65 camera backs for most of the high quality imaging we do. Canon cameras are used by the Internet Archive to image books from our general collections. Specialized overhead scanners such as the Digibook and Jumbo Scan are used for maps and books where we want to capture both the content and the appearance of the document with good fidelity. For rare books and other materials where we want the be best resolution, lowest noise, and highest color accuracy we use P65 backs. We still use our one remaining scanning back for high quality images of posters and other large objects. We use a custom Sinar back (developed and operated by a contractor with an extremely skilled staff) in Prints and Photographs.
I've also just recently had a tour of the National Gallery of Art imaging operations. They primarily use multi-shot Sinar backs in their various labs, including one setup for accurate movement of camera and the art so they can capture extremely detailed sections and stitch perhaps a dozen images together to cover a large painting.
In all cases, massive copy stands are used with specialized document cradles and beds. A fully color-managed workflow is required. In addition, a wide variety of camera and lighting techniques have been developed to meet the special needs of individual images. There is still a lot of "art" in imaging science!
-Barry
________________________________________
From: colorsync-users-bounces+bwhe=email@hidden [colorsync-users-bounces+bwhe=email@hidden] On Behalf Of email@hidden [email@hidden]
Sent: Sunday, May 02, 2010 4:44 PM
To: Wheeler, Barry; email@hidden
Subject: Colorsync-users Digest, Vol 7, Issue 117
Send Colorsync-users mailing list submissions to
email@hidden
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.apple.com/mailman/listinfo/colorsync-users
or, via email, send a message with subject or body 'help' to
email@hidden
You can reach the person managing the list at
email@hidden
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Colorsync-users digest..."
Today's Topics:
1. Re: fine art reproduction questions (Andrew Rodney)
2. Re: fine art reproduction questions (Ben Goren)
3. Re: fine art reproduction questions (Mike Strickler)
4. Re: fine art reproduction questions (Ben Goren)
5. Re: fine art reproduction questions (neil snape)
6. Re: fine art reproduction questions (Ben Goren)
7. Re: fine art reproduction questions (Randy Zaucha)
8. Re: fine art reproduction questions (neil snape)
9. Re: fine art reproduction questions (Steve Kornreich)
10. Re: fine art reproduction questions (Ben Goren)
11. Re: fine art reproduction questions (Robin Myers)
12. Fwd: fine art reproduction questions (Mike Strickler)
13. Re: fine art reproduction questions (Ben Goren)
----------------------------------------------------------------------
Message: 1
Date: Sun, 2 May 2010 13:15:35 -0600
From: Andrew Rodney <email@hidden>
Subject: Re: fine art reproduction questions
To: colorsync user list <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
On May 2, 2010, at 10:56 AM, Ben Goren wrote:
> I mean, seriously? In a discussion that includes people posting links to cutting-edge research into spectral imaging at RIT, you post a link to *THAT*? What on Earth are you thinking?
Kind of what I expected from you even though the piece is correct, accurate and backs up what Neil has tried to explain.
Andrew Rodney
http://www.digitaldog.net/
------------------------------
Message: 2
Date: Sun, 2 May 2010 12:16:15 -0700
From: Ben Goren <email@hidden>
Subject: Re: fine art reproduction questions
To: colorsync user list <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
On 2010 May 2, at 10:35 AM, neil snape wrote:
> The reason I said a 50 mm lens is not the right lens to
> use, and yes I know from the beginning you are using a very fine lens for
> copy work is because unless you have absolutely flat surfaces all the
> reflections of the texture will have reflection angles that will be
> detrimental to the copy quality. If you are copying a photographic print or
> very flat art work with careful lighting that is fine.
That may well be why I haven't had any problems with the 50mm CM. I've mostly been shooting watercolors. My mom, predictably my most regular and difficult ``client,'' likes to use cheap paper that curls something fierce, but a couple steel rulers or framing squares along the edges will tame it reasonably well.
As for the lighting, I use four bare strobes (reflector only) placed equally distant about 10' away from the art at a 45o angle. I haven't needed to do the cross-polarization thing yet; I'll burn that bridge when it becomes a problem.
> Yet I am convinced that there are some tricks that can
> give you better results that really do optimise DR with a lower S/N in the
> areas that DSLRs have problems.
See...I haven't found this to be a problem, and I haven't seen any indication that medium format would be fundamentally different.
Unless I'm missing something, even theoretically, the only difference between medium format cameras and high-end DSLRs is the size of the sensor. The sensors themselves are made with the same processes by the same manufacturers. Indeed, it's often the second-tier manufacturers, such as Kodak, making the medium format sensors.
What you get with MF digital over 135 digital is therefore the exact same thing you get with MF film over 135 film: more recording area. And the ``only'' thing that in turn gives you is a smaller capture-to-output enlargement ratio. All of the ``creamy goodness'' you get from MF (and LF, of course), whether film or digital, can directly be attributed to the fact that you're not blowing up the original as much.
(Don't get me worng: the larger formats are indeed greatly superior; I'm just identifying the source of their superiority.)
With that understanding, it should be obvious that, if you keep the levels of magnification the same, the smaller format will have identical quality to a crop from the larger format. That is, an 8" x 10" crop from a 16" x 20" print from 645 will have the same quality as an 8" x 10" print from 135 (within rounding error).
But...if you take two 135 pictures and (skillfully) stitch them together, you've now got a 16" x 20" print that's indistinguishable from what you get with medium format -- precisely because it's the exact same thing. It's a print made with the same magnification (and thus the same levels of visible noise, the same SN ratio, and all the rest).
And, once you start stitching, It's not that much more hassle to (digitally) stitch four exposures than it is two, and get large format quality out of the 135 format DSLR. Or, if you really need it, stitch a few dozen exposures and make 8" x 10" film look like one of those throwaway cardboard film cameras in the impulse aisle in comparison.
(This, of course, assumes lenses of comparable quality, but all the major formats have superlative glass available these days. It *could* be a factor, but it never *should* be a factor.)
> HDR is allows the highest level of
> extraction of the info from the raw processor. It is not limited to only
> high contrast and out of range images.
I still think this is one of those ``in theory, but not in practice'' cases. With the kinds of subject matter we're talking about, the tone curve and color correction that gets applied, and the sensor-to-print magnification involved, I really don't think the human eye is sensitive enough to detect the noise present in a modern DSLR. If you can point to any examples to the contrary, I'd be interested to see them, but I've yet to come across any.
That's not to claim that noise is never a factor, of course; just that, in this particular very limited case, it is, for all practical intents and purposes, invisible.
> Building your own charts could be interesting. If you have the same pigments
> as the art work it'd be optimal.
I originally thought that would be necessary, but my research leads me to think otherwise. Even between hardware-store paints and expensive artist's acrylics, you see the exact same spectral patterns over and over and over again. With enough practice, I'm pretty sure most people could learn to sketch a surprisingly-accurate spectral plot of pretty much any color you'll see.
The paint store favors unsaturated colors whereas the paints artists use also include more saturated, spectrally pure colors. But, aside from the different gamut -- which seems to be as much for fashion reasons as anything else -- there doesn't seem to be anything special about the pigments from any particular source.
(I think I mentioned a couple oddball acrylics I came across that I'll be including in the chart, but I'm pretty sure the local paint store could match them.)
> I am interested in how it goes just the
> same.
So am I!
Sadly, with work schedules and all, it's likely to be another month or two before I'll have time to complete it. But, if it shows promise, I'll post more complete details at that time.
Cheers,
b&
------------------------------
Message: 3
Date: Sun, 2 May 2010 12:28:52 -0700
From: Mike Strickler <email@hidden>
Subject: Re: fine art reproduction questions
To: email@hidden
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii; format=flowed; delsp=yes
Maybe I read too cursorily, but why the concern here with HDR? The
luminance range in studio work, where one can and does control
lighting, after all, does not tax the dynamic range of digital
cameras, which are quite wide compared with, say, transparency film,
which has been happily used in the studio for years. In photographing
paintings the luminance range is also quite short, at most about 50:1
and usually much, much less. We're not talking about the proverbial
redwood forest on a sunny day...
------------------------------
Message: 4
Date: Sun, 2 May 2010 12:29:54 -0700
From: Ben Goren <email@hidden>
Subject: Re: fine art reproduction questions
To: Andrew Rodney <email@hidden>
Cc: colorsync user list <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
On 2010 May 2, at 12:15 PM, Andrew Rodney wrote:
> On May 2, 2010, at 10:56 AM, Ben Goren wrote:
>
>> I mean, seriously? In a discussion that includes people posting links to cutting-edge research into spectral imaging at RIT, you post a link to *THAT*? What on Earth are you thinking?
>
> Kind of what I expected from you even though the piece is correct, accurate and backs up what Neil has tried to explain.
Before we proceed, would you be so kind as to prove your bona fides?
First, kindly explain the applicability of that article to producing colorimetrically-correct reproductions of art. After all, all it is is an informal three-page beginner's-level misleadingly-oversimplified attempt to explain to amateurs why color consistency can be hard to achieve. It perhaps might maybe be useful to get Uncle Bob to stop complaining about why the bridesmaid's dresses look a bit redder than he remembered them, but it's so far removed from the discussion at hand that it ``isn't even worng.'' Frankly, I don't know why the ICC commissioned somebody to author it -- let alone why they accepted it and continue to publish it.
Second, please sketch for us your own workflow for copy work. How do *you* create a colorimetrically-correct reproduction of a piece of art -- or do you just throw up your hands because it's too hard?
Cheers,
b&
------------------------------
Message: 5
Date: Sun, 02 May 2010 21:37:28 +0200
From: neil snape <email@hidden>
Subject: Re: fine art reproduction questions
To: Ben Goren <email@hidden>, colorsync user list
<email@hidden>
Message-ID: <C803A118.47ACE%email@hidden>
Content-Type: text/plain; charset="ISO-8859-1"
on 2/05/10 21:16, Ben Goren wrote :
> Unless I'm missing something, even theoretically, the only difference between
> medium format cameras and high-end DSLRs is the size of the sensor. The
> sensors themselves are made with the same processes by the same manufacturers.
> Indeed, it's often the second-tier manufacturers, such as Kodak, making the
> medium format sensors.
Stitching will have a slight bit of corner fall off, but that you can
correct for so yes stitching is a very good way to come up with a great
resolution for large prints.
There are some pretty major differences though with a MF. There is no anti
alias filter and often the photo-site sizes is bigger than the a DSLR
counterpart, and 16 bit capture makes for highly modifiable files.
I was interested in getting a MF but when testing the Canon 5DII against a
Hasselblad 39 MK II the difference was there but as you agree on print
quality not that noticeable. It's there for pixel peepers are those with
sharp eyes, when you want to look. Where the MF really shines is when you
push curves into the image it stays together far better. Many tests show the
Nikon D3 to be better than Canon from serious tests.
Glad you added what you're copying as water color is a forgiving relatively
flat art work to copy without reflections or need for showing texture. I
sincerely hope you never do have to learn cross polarised light with a
certain direction texture to be picked up. Watercolour is relatively easy to
copy, and correction as well within a normal* workflow.
Sounds to me like you're doing everything needed to produce excellent
results with what you have. The only thing that might be a bit better is use
a more oblique angle for the flash like 30?, you might get a bit more
saturation depending on the type of paper of the painting.
Neil Snape
------------------------------
Message: 6
Date: Sun, 2 May 2010 12:40:30 -0700
From: Ben Goren <email@hidden>
Subject: Re: fine art reproduction questions
To: colorsync user list <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
On 2010 May 2, at 12:28 PM, Mike Strickler wrote:
> Maybe I read too cursorily, but why the concern here with HDR?
If I may play devil's advocate, I *think* the idea is that HDR can be used to reduce shadow noise even when said shadows aren't being pushed or otherwise manipulated.
For example, let's say we're reproducing a gray scale chart. With your normal workflow, ideally, you should be able to put your spectrophotometer on the original and the print and get the same Lab values (within your acceptable margin of error). Further, let's assume that one of the sample patches measures L* = 10, a* = 0, b* = 0.
The camera will, of course, have no trouble capturing that dark-dark gray; it's well within its dynamic range. But, that patch will be noisier than the patch that measures L* = 90, a* = 0, b* = 0.
An HDR exposure, properly done, will produce the same output, but there will be less noise in said dark-dark gray patch.
The first question is: is the noise even visible in the non-HDR version? I sure haven't ever been able to see it.
The second question is: does HDR reduce the noise by an amount sufficient to be worth all the extra headaches? I really doubt it.
Cheers,
b&
------------------------------
Message: 7
Date: Sun, 2 May 2010 12:46:27 -0700 (PDT)
From: Randy Zaucha <email@hidden>
Subject: Re: fine art reproduction questions
To: email@hidden
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
Hi Ben,I was employed by Thom Kinkade for 2 1/2 years to accurately reproduce his paintings for mass distribution. I've seen more cottages and lighthouses than any man should ever see.
We used a Betterlight 100 megapixel back on a 4 x 5 camera outfit. Polarizing filters on camera lense and lighting.
When I first started there, achieving flat lighting was a real pain. Some paintings were 5 feet wide. Later we beta tested an HP system that used luminance mapping to correct the lighting fall off. Now you can get the same tool from Robin Meyer's website. (rmimaging.com) It's called Equalight and it is a very underpriced Photoshop filter. You capture the painting and then slide a white canvas in front of it and capture that. The filter determines the light even/uneveness on the canvas and corrects it in Photoshop. This is extremely important to getting a good reproduction because light fall off make accurate retouching impossible.
I'm not sure what you meant about not being able to use Betterlight due to floors.
The wild card is being able to make a good profile of the camera. HP actually profiled our camera using RGB lights. After exposure I took physical readings with a spectrophotometer from the painting and they were added to the final calculation of the image file. I'm not sure if they ever put that product on the market. Robin?
Once Thom brought back 6 paintings he made on a trip to Israel. He had purchased some paints made locally and they reproduced very differently than what we were used to. That batch took some extra retouching to get a good match. We attributed it to local minerals in the paint.
If you can, get your painter to make a simulation of the Gretag Colorchecker with his paints. Measure it and make a camera profile. That may get you some better accuracy. Be sure to make at least 5 reading per patch and average them.
We were very proud to make extremely accurate reproductions of Kinkade's paintings. The irony was that Thom's older paintings from the 80's were reproduced for offset press. Whoever was doing the color check on press got subjective and told the pressman to go away from the color of the original. They even dot etched the color to the detriment of the color match. Original and reproduction turned out very different and they sell them to this day. The new giclees of his work are extremely accurate reproductions.
Fine art reproduction always involves retouching because the camera will never see color like the eye does. A great camera profile and Equalight 2 can get you about 90% of the way. I set up a local company with those tools and they have youngsters making very accurate painting reproductions on the first print off an HP Z6100 printer.
Monitors...get one with contrast, brightness and individual RGB brightness controls. Then profile it. I use the Spyder 3 Pro at 2.4 Native white point setting.
Randy ZauchaManaged Color
------------------------------
Message: 8
Date: Sun, 02 May 2010 21:56:13 +0200
From: neil snape <email@hidden>
Subject: Re: fine art reproduction questions
To: Mike Strickler <email@hidden>, colorsync user list
<email@hidden>
Message-ID: <C803A57D.47AD4%email@hidden>
Content-Type: text/plain; charset="US-ASCII"
on 2/05/10 21:28, Mike Strickler wrote :
> Maybe I read too cursorily, but why the concern here with HDR? The
> luminance range in studio work, where one can and does control
> lighting, after all, does not tax the dynamic range of digital
> cameras, which are quite wide compared with, say, transparency film,
> which has been happily used in the studio for years. In photographing
> paintings the luminance range is also quite short, at most about 50:1
> and usually much, much less. We're not talking about the proverbial
> redwood forest on a sunny day...
Mike you may be right about a watercolour but the contrast ratio of
oil/acrylic paintings is a lot higher than that. It's not about what the
painting is about but what you are using to capture.
Not sure I follow about the controlled lighting, how does that change the
contrast ratio?
Copy lighting is usually set to 1/10th of an f stop, and visually compared
for shadow density with a pencil for shadow strength and shape. OF course
if you use a Cruse scanner the lights are already done, yet still variable
for texture and reflections one of the great advantages of those scanners.
Film was a lot harder to capture both ends of the extremities, almost always
a lot was left to the scannerist to pull out. Yes digital cameras have a lot
of info in them but the extremities are not always of a usable quality.
Right again on HDR it is theoretically but also mathematically the max you
can use but it is a way to do wonderful thing that Canon DSLRs are not made
to do from a single capture.
We were talking about how to maximise potential with Canon Dslr , not
needed for paintings with a 50:1 contrast ratio!
Neil Snape
------------------------------
Message: 9
Date: Sun, 2 May 2010 13:00:35 -0700
From: Steve Kornreich <email@hidden>
Subject: Re: fine art reproduction questions
To: color <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
What Randy says is the way to go period.
I was doing art repro for almost 10 years in Hawaii.
Always used a better light scan back HID/HMI lights, and invested thousands in both a good rip I used Best Color / EFI and used all of Gretag's color charts.
For me I got my best camera profiles using the Color Checker SG chart and using Monaco Profiler for both camera and output profiles.
Robin Meyeer's is the guru in this topic.
Steven Kornreich
On May 2, 2010, at 12:46 PM, Randy Zaucha wrote:
> Hi Ben,I was employed by Thom Kinkade for 2 1/2 years to accurately reproduce his paintings for mass distribution. I've seen more cottages and lighthouses than any man should ever see.
> We used a Betterlight 100 megapixel back on a 4 x 5 camera outfit. Polarizing filters on camera lense and lighting.
> When I first started there, achieving flat lighting was a real pain. Some paintings were 5 feet wide. Later we beta tested an HP system that used luminance mapping to correct the lighting fall off. Now you can get the same tool from Robin Meyer's website. (rmimaging.com) It's called Equalight and it is a very underpriced Photoshop filter. You capture the painting and then slide a white canvas in front of it and capture that. The filter determines the light even/uneveness on the canvas and corrects it in Photoshop. This is extremely important to getting a good reproduction because light fall off make accurate retouching impossible.
> I'm not sure what you meant about not being able to use Betterlight due to floors.
> The wild card is being able to make a good profile of the camera. HP actually profiled our camera using RGB lights. After exposure I took physical readings with a spectrophotometer from the painting and they were added to the final calculation of the image file. I'm not sure if they ever put that product on the market. Robin?
> Once Thom brought back 6 paintings he made on a trip to Israel. He had purchased some paints made locally and they reproduced very differently than what we were used to. That batch took some extra retouching to get a good match. We attributed it to local minerals in the paint.
> If you can, get your painter to make a simulation of the Gretag Colorchecker with his paints. Measure it and make a camera profile. That may get you some better accuracy. Be sure to make at least 5 reading per patch and average them.
> We were very proud to make extremely accurate reproductions of Kinkade's paintings. The irony was that Thom's older paintings from the 80's were reproduced for offset press. Whoever was doing the color check on press got subjective and told the pressman to go away from the color of the original. They even dot etched the color to the detriment of the color match. Original and reproduction turned out very different and they sell them to this day. The new giclees of his work are extremely accurate reproductions.
> Fine art reproduction always involves retouching because the camera will never see color like the eye does. A great camera profile and Equalight 2 can get you about 90% of the way. I set up a local company with those tools and they have youngsters making very accurate painting reproductions on the first print off an HP Z6100 printer.
> Monitors...get one with contrast, brightness and individual RGB brightness controls. Then profile it. I use the Spyder 3 Pro at 2.4 Native white point setting.
> Randy ZauchaManaged Color
>
>
>
> _______________________________________________
> Do not post admin requests to the list. They will be ignored.
> Colorsync-users mailing list (email@hidden)
> Help/Unsubscribe/Update your Subscription:
>
> This email sent to email@hidden
------------------------------
Message: 10
Date: Sun, 2 May 2010 13:01:41 -0700
From: Ben Goren <email@hidden>
Subject: Re: fine art reproduction questions
To: colorsync user list <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=iso-8859-1
On 2010 May 2, at 12:37 PM, neil snape wrote:
> on 2/05/10 21:16, Ben Goren wrote :
>
>> Unless I'm missing something, even theoretically, the only difference between
>> medium format cameras and high-end DSLRs is the size of the sensor. The
>> sensors themselves are made with the same processes by the same manufacturers.
>> Indeed, it's often the second-tier manufacturers, such as Kodak, making the
>> medium format sensors.
>
> Stitching will have a slight bit of corner fall off, but that you can
> correct for so yes stitching is a very good way to come up with a great
> resolution for large prints.
I agree that peripheral illumination is a problem with stitching -- and even when not stitching, for that matter. I deal with it by avoiding it. If the work is small enough and if I don't need the highest resolution possible, I simply frame it so that the art is well within the sweet spot of the lens.
If I need to stitch, I use lots of overlap -- perhaps even half the frame -- and let Photoshop do its autocorrection magic when creating the panorama. Frankly, it's scary how well it does.
> There are some pretty major differences though with a MF. There is no anti
> alias filter and often the photo-site sizes is bigger than the a DSLR
> counterpart, and 16 bit capture makes for highly modifiable files.
I think I'm gonna stick with the ``theory that doesn't apply to practice in this case'' for all three of those differences.
For really, really critical work, I just shoot close enough or with a long enough lens to get the capture ppi into the 600+ ppi range. At that point, the presence or absence of an AA filter becomes moot. For not-so-critical work, a very little bit of very careful sharpening does the trick quite nicely. (And, to be honest, I'd do the same sharpening for the critical work, even though it's redundant in the print.)
Photosite size is one of those great religious wars, but it's one that I firmly believe is entirely misguided.
Larger photosites have a better SN ratio per site, but they need to be enlarged more in the final print.
For a given format size, assuming you trust the engineers, more megapickles (and thus smaller photosites) are always better. At 100% pixel view, the higher density camera will be noisier, yes. But, at the same size print, the SN ratio will be exactly the same (assuming the same quality of engineering). The kicker, though, is that the higher density camera will have noise that's much finer-grained -- making it much less objectionable -- and it'll capture much more fine detail as well. Personally, I expect the megapickle wars to continue until 135 format DSLRs are in the 200+ megapickle range, since that's about when diffraction starts to be the worst offender at fast apertures.
And I'm all for more bits per pickle...erm...``pixel,'' but the modifications needed for copy work are (or should be) as gentle as they get. Just as I've never actually seen shadow noise in a print when doing copy work, I've never seen posterization, either.
(For landscape or studio work? Oh, yeah. I'd be all over replacing my current gear with MF...if only I could afford blowing more than my remaining mortgage balance on it....)
> The only thing that might be a bit better is use
> a more oblique angle for the flash like 30?, you might get a bit more
> saturation depending on the type of paper of the painting.
Thanks for the tip. Sounds like I need to add more experiments to the list -- curse you!
Cheers,
b&
------------------------------
Message: 11
Date: Sun, 2 May 2010 13:20:43 -0700
From: Robin Myers <email@hidden>
Subject: Re: fine art reproduction questions
To: email@hidden
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
Hello All,
As a fly on the wall this has been an interesting discussion.
Let me clear up one slight thing in Randy's message; EquaLight is not a Photoshop plugin, but a standalone program.
As for the HP spectral system, it is called HP Artist. It is a library of routines that HP has licensed to two vendors; Better Light and Ergosoft. The user selects the camera (only the Better Light and a Nikon are currently supported), lighting, spectral measurements of a white card used for falloff compensation, spectral measurements of the artwork, the white card TIFF image and the artwork TIFF image. After processing the result is a TIFF image with a custom profile. One important note, HP requires a Z3200 printer be attached to the processing computer, no other printers will do (do not ask, HP has made this a rigid requirement). This system was debuted by both Better Light and Ergosoft at Photokina 2008. There are some on this list that can tell you about the results. They were very good without requiring extra processing. If you wanted perfection then a slight bit of tweaking might be needed, but for many the first print was more than good enough.
Robin Myers
rmimaging.com
On May 2, 2010, at 12:46 PM, Randy Zaucha wrote:
> Hi Ben,I was employed by Thom Kinkade for 2 1/2 years to accurately reproduce his paintings for mass distribution. I've seen more cottages and lighthouses than any man should ever see.
> We used a Betterlight 100 megapixel back on a 4 x 5 camera outfit. Polarizing filters on camera lense and lighting.
> When I first started there, achieving flat lighting was a real pain. Some paintings were 5 feet wide. Later we beta tested an HP system that used luminance mapping to correct the lighting fall off. Now you can get the same tool from Robin Meyer's website. (rmimaging.com) It's called Equalight and it is a very underpriced Photoshop filter. You capture the painting and then slide a white canvas in front of it and capture that. The filter determines the light even/uneveness on the canvas and corrects it in Photoshop. This is extremely important to getting a good reproduction because light fall off make accurate retouching impossible.
> I'm not sure what you meant about not being able to use Betterlight due to floors.
> The wild card is being able to make a good profile of the camera. HP actually profiled our camera using RGB lights. After exposure I took physical readings with a spectrophotometer from the painting and they were added to the final calculation of the image file. I'm not sure if they ever put that product on the market. Robin?
> Once Thom brought back 6 paintings he made on a trip to Israel. He had purchased some paints made locally and they reproduced very differently than what we were used to. That batch took some extra retouching to get a good match. We attributed it to local minerals in the paint.
> If you can, get your painter to make a simulation of the Gretag Colorchecker with his paints. Measure it and make a camera profile. That may get you some better accuracy. Be sure to make at least 5 reading per patch and average them.
> We were very proud to make extremely accurate reproductions of Kinkade's paintings. The irony was that Thom's older paintings from the 80's were reproduced for offset press. Whoever was doing the color check on press got subjective and told the pressman to go away from the color of the original. They even dot etched the color to the detriment of the color match. Original and reproduction turned out very different and they sell them to this day. The new giclees of his work are extremely accurate reproductions.
> Fine art reproduction always involves retouching because the camera will never see color like the eye does. A great camera profile and Equalight 2 can get you about 90% of the way. I set up a local company with those tools and they have youngsters making very accurate painting reproductions on the first print off an HP Z6100 printer.
> Monitors...get one with contrast, brightness and individual RGB brightness controls. Then profile it. I use the Spyder 3 Pro at 2.4 Native white point setting.
> Randy ZauchaManaged Color
>
------------------------------
Message: 12
Date: Sun, 2 May 2010 13:24:37 -0700
From: Mike Strickler <email@hidden>
Subject: Fwd: fine art reproduction questions
To: email@hidden
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii; format=flowed; delsp=yes
> What does controlled lighting have to do with it? Well, you can add
> fill, you can modify the source, making it broader, more diffuse.
> This changes the luminance range. As for oil paintings, I've
> photographed many, and the luminance range isn't any greater than a
> glossy photographic print and generally less. If one is shooting 3-
> dimensional objects one can control the lighting ratio and very
> easily keep it within the dynamic range of the CCD, as you did for
> years with film (or still do?). I thought this topic was about
> photographing artwork. If it's about shooting outdoors, where the
> luminance range is in quadruple digits ( a great topic of its own),
> or even location architectural work (another great one), we should
> change the subject line.
>
> On May 2, 2010, at 12:56 PM, neil snape wrote:
>
>> on 2/05/10 21:28, Mike Strickler wrote :
>>
>>> Maybe I read too cursorily, but why the concern here with HDR? The
>>> luminance range in studio work, where one can and does control
>>> lighting, after all, does not tax the dynamic range of digital
>>> cameras, which are quite wide compared with, say, transparency film,
>>> which has been happily used in the studio for years. In
>>> photographing
>>> paintings the luminance range is also quite short, at most about
>>> 50:1
>>> and usually much, much less. We're not talking about the proverbial
>>> redwood forest on a sunny day...
>>
>> Mike you may be right about a watercolour but the contrast ratio of
>> oil/acrylic paintings is a lot higher than that. It's not about
>> what the
>> painting is about but what you are using to capture.
>>
>> Not sure I follow about the controlled lighting, how does that
>> change the
>> contrast ratio?
>> Copy lighting is usually set to 1/10th of an f stop, and visually
>> compared
>> for shadow density with a pencil for shadow strength and shape. OF
>> course
>> if you use a Cruse scanner the lights are already done, yet still
>> variable
>> for texture and reflections one of the great advantages of those
>> scanners.
>>
>>
>> Film was a lot harder to capture both ends of the extremities,
>> almost always
>> a lot was left to the scannerist to pull out. Yes digital cameras
>> have a lot
>> of info in them but the extremities are not always of a usable
>> quality.
>> Right again on HDR it is theoretically but also mathematically the
>> max you
>> can use but it is a way to do wonderful thing that Canon DSLRs are
>> not made
>> to do from a single capture.
>>
>> We were talking about how to maximise potential with Canon Dslr , not
>> needed for paintings with a 50:1 contrast ratio!
>>
>> Neil Snape
>>
>>
>>
>>
>>
>
------------------------------
Message: 13
Date: Sun, 2 May 2010 13:44:07 -0700
From: Ben Goren <email@hidden>
Subject: Re: fine art reproduction questions
To: colorsync user list <email@hidden>
Message-ID: <email@hidden>
Content-Type: text/plain; charset=us-ascii
On 2010 May 2, at 12:46 PM, Randy Zaucha wrote:
> Hi Ben,I was employed by Thom Kinkade for 2 1/2 years to accurately reproduce his paintings for mass distribution.
Sounds like a fun job!
> I've seen more cottages and lighthouses than any man should ever see.
..or maybe not....
> I'm not sure what you meant about not being able to use Betterlight due to floors.
I don't remember typing anything like that. I will state though, that the Betterlight gear is out of my league. I'm afraid budgetary constraints will keep me kluging away with the amateur equipment for...oh, probably the rest of my life, barring any lottery tickets falling into my lap.
> The wild card is being able to make a good profile of the camera.
Of that I am utterly, thoroughly convinced. It's my knowledge and skill in building profiles that has had, far and away, the biggest improvement on my results. I'm sure I could do better work with what I know now using a high-end consumer digicam and open shade than I ever would have been able to when I first started had you turned me loose in your studio without instruction.
> HP actually profiled our camera using RGB lights. After exposure I took physical readings with a spectrophotometer from the painting and they were added to the final calculation of the image file. I'm not sure if they ever put that product on the market. Robin?
I've yet to experiment with this, but I'm pretty sure you could do this with Argyll by adding the various measurements to the appropriate chart reference and measurement files.
> Once Thom brought back 6 paintings he made on a trip to Israel. He had purchased some paints made locally and they reproduced very differently than what we were used to. That batch took some extra retouching to get a good match. We attributed it to local minerals in the paint.
I'll keep that in mind, though I don't expect to run into anything especially exotic any time soon.
> If you can, get your painter to make a simulation of the Gretag Colorchecker with his paints. Measure it and make a camera profile. That may get you some better accuracy. Be sure to make at least 5 reading per patch and average them.
Now *that* is an excellent suggestion. The next time I have somebody dissatisfied with the results -- especially after I've finished making my overly-complex chart -- I'll do just that. Thanks!
> We were very proud to make extremely accurate reproductions of Kinkade's paintings. The irony was that Thom's older paintings from the 80's were reproduced for offset press. Whoever was doing the color check on press got subjective and told the pressman to go away from the color of the original. They even dot etched the color to the detriment of the color match. Original and reproduction turned out very different and they sell them to this day. The new giclees of his work are extremely accurate reproductions.
The fact that the inaccurate old ones still sell very well is good to remind me that I should strive for perfection but not let my shortcomings keep me from sleeping at night. Each job I've done has been better than the previous, but, fortunately, they've all been ``good enough'' so far for the artists' purposes. And dramatically better than what they've been able to do themselves.
> Fine art reproduction always involves retouching because the camera will never see color like the eye does.
I agree...but I'm still at the point where improving the profile has better bang for the buck than retouching. At least, I've revisited some of my mom's works that I shot some time ago that I did extensive retouching on, and I can get better results now from the same files by building better profiles (etc.). And I know a lot of shortcomings in my workflow that I'm working on addressing; when I run out of ways to improve that, I'll start to focus my attention again on retouching.
> A great camera profile and Equalight 2 can get you about 90% of the way. I set up a local company with those tools and they have youngsters making very accurate painting reproductions on the first print off an HP Z6100 printer.
> Monitors...get one with contrast, brightness and individual RGB brightness controls. Then profile it. I use the Spyder 3 Pro at 2.4 Native white point setting.
I'm using an iMac 27", profiled with an i1 Pro. I know there are better options out there, but there are more important things right now for me to be spending my photographic budget on.
The printer is a Canon iPF8100, and I couldn't be happier with it.
Cheers,
b&
------------------------------
_______________________________________________
Colorsync-users mailing list
email@hidden
http://lists.apple.com/mailman/listinfo/colorsync-users
End of Colorsync-users Digest, Vol 7, Issue 117
***********************************************
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden