I'm sorry, my English isn't good enough to explain by first try. I've meant you
may interleave each target patch with the same standard size white patch. This
way your transparency target will contain almost twice more patches. The patch
size is usual.
Due to the interleave there always will be enough contrast between the patches
and/or the spacers. With original target the spacers may become invisible to
the recognition algorithm. But with white patch interleave there always will be
enough contrast, even if the spacers become indistinguishable from original
patch.
Anyway the preliminary actual response profile is highly recommended. With real
profile the strip recognition will know the expected colors better and will
miss less often. суббота, 02 декабря 2017г., 22:27 +03:00 от Roger Breton
graxx@xxxxxxxxxxxx :
Do you mean, your patch size is 50 mm by 50 mm?
Why do you need to use “white neighbors”? To use as Reference White, for
normalization? Just like in monitor profiling?
Best / Roger
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx] ;
On Behalf Of Nikolay Pokhilchenko (Redacted sender "nikolay_po" for DMARC)
Sent: Saturday, December 2, 2017 2:02 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Profiling backlit film (with Monitor?)
Mix normal patches with whites 50/50 with spacers. I had limited success with
such transparent strip mode readings. There will be twice less color patches
but strip reading.
By some spreadsheet calculations you can account white reference for each
patch using it's white neighbors.
May Graeme implement such kind transparency chart reading with per patch white
calibration?
суббота, 02 декабря 2017г., 21:54 +03:00 от Kamil Tresnak
kamil.tresnak@xxxxxxxxx :
Hi Roger,
I have done few profiles with very satisfying results (common large format
print, no hi quality requirements) with CMYK targets around 500 patches. I am
using somewhat bigger patches for easy device (i1pro) positioning.
I never succeed with strip reading, always i have to use patch reading. I am
using chartread -t, exactly as Alexey suggested.
I was experimenting with few LED strips, but some of this having spectrum
with unpleasant peaks. IPad is excellent tip, thank you Alex!
Regards!
Kamil
On Sat, Dec 2, 2017 at 5:35 PM, Roger Breton < graxx@xxxxxxxxxxxx > wrote:
How many patches do you read, Alex? And how big do you make the patch?
Could you write what instruction you type to generate the target?
/ Roger
-----Original Message-----
From: argyllcms-bounce@xxxxxxxxxxxxx [mailto: ;
argyllcms-bounce@xxxxxxxxxxxxx ] On Behalf Of Alexey Gribunin
Sent: Friday, December 1, 2017 12:07 PM
To: argyllcms@xxxxxxxxxxxxx
Subject: [argyllcms] Re: Profiling backlit film (with Monitor?)
Hello Roger!
1. Chartread. I'm measuring with Chartread –t. I prefer it because the i1Pro
calibration is performing relative to lightbox+clear backlit. So the
testform's paper patch (CMYK 0000) is L=100, a=0, b=0, i.e. perfect white
and it's not necessary to recalculate it afterwords.
2. Strip mode vs Patch mode. Theoretically Strip mode is possible but in my
practice it's very unstable because the light intensity of lightbox is not
enough. Maybe it's possible to do with higher intensity lightbox, but it
seems to me that then another problem will arise - i1Pro drift due to
overheating.
So I'm using patch mode. It's slow but it's stable.
3. lightbox. Theoretically an incandescent light source with 80A (or
similar) blue filter would be very good choice. I tried to construct
lightspot for patch measurements with incandescent light source and 80A
filter. CRI of such construction is very good, the spectra is similar to
real D50. But I found that overheating also is an issue. The spectra of such
construction still has quite high amount of IR. So it's better to add an IR
filter also. But I didn't find it and left this idea for simpler light
sources...
In my practice iPad (i.e. LED source) showed himself better than fluorescent
light source. It's more stable and it's also much more usable because I
always have it in my bag :))))
So I would suggest you to try LED first to develop the complete workflow and
then to think about another types of lightboxes.
Best regards,
Alexey Gribunin, UNIT Color Technologies, Moscow, Russia.
On 01/12/2017 19:32, Roger Breton wrote:
All of a sudden, I need to come up with an ICC profile for backlit
media – small world!
Many questions…
First, suppose I use Chartread –e. Can I read in strips with the i1pro?
Or do I need to read in Patch mode?
Second, choice of lightbox. Is it “better” to measure with the media
sitting on a fluorescent light source? Or is it better to measure with
an incandescent light source? Why not LED light source?
I was thinking of using this kind of lightbox:
https://www.ebay.com/itm/1x-JUST-Normlicht-Color-Smart-Light-5000-Tran
sparency-Flat-Viewer-15-x-10/361307134398?hash=item541f9559be:g:LtkAAO
SwNSxVZhQt
I remember reading that the issue of light source “color temperature”
was not important. How is that?
Best / Roger
*From:* argyllcms-bounce@xxxxxxxxxxxxx
[mailto: argyllcms-bounce@xxxxxxxxxxxxx ] *On Behalf Of *Kamil Tresnak
*Sent:* Thursday, November 23, 2017 5:22 PM
*To:* argyllcms@xxxxxxxxxxxxx
*Subject:* [argyllcms] Re: Profiling backlit film with Monitor
Thanks for sharing, Alexey!
Regards,
Kamil Tresnak
In the past I used LCD monitor without matrix (it was damaged, but
lamp itself was good, so I just remove the matrix and used it as a
light table). But the uniformity and stability is not very good, so
it was necessary to calibrate i1Pro every few patches.
Now I'm using iPad (!) and found it much more stable
(auto-brightness=off).
But calibration every 10-15 patches is also highly desirable. It's
necessary to get really repeatable and reliable values.
I'm calibrating device during measurements in two different ways:
1. I'm making ArgyllCMS native calibration every 75 patches.
2. I'm measuring white patch every 15 patches and making linear
spectral correction afterwords in EXCEL.
Best regards,
Alexey Gribunin, UNIT Color Technologies, Moscow, Russia.
