lcms-user — For users of the little cms library

Re: [Lcms-user] Mixing Lab color

[Armindo Da S. <tec...@wa...>]

What is the name of the profile I should use for CMYK?

  ----- Original Message -----=20
  From: Mart=ED Maria=20
  To: Armindo Da Silva ; Lcm...@li...=20
  Sent: Friday, January 25, 2002 7:33 PM
  Subject: Re: [Lcms-user] Mixing Lab color


  Hi,

  I'm not 100% sure, but I belive you cannot do that.

  Lab works as light, that is, additive. You add several=20
  lights and got a resulting one. No problem since there=20
  is no additional information needed.

  However, substractive colorspaces DOES need some=20
  additional information, since they are also reflective.=20

  You need to know illuminant (the source of light the inks
  are blocking off) and the media white of substrate (the=20
  chromaticity of paper) the gamut (the whitest you can get,=20
  that is no more that the substrate, and the darkest, the point
  at inks blocking all but ambience light) etc etc.

  So, a Lab color minus another lab color can result in a lot of=20
  colors, on depending on a lot of things. It *is* device dependent
  at all :-)

  In the other hand, assuming you have a 100% difusse perfect=20
  reflective substrate and same illuminant as Lab white point,=20
  Lab is an euclidean  space, and as a such should have=20
  substraction defined. Do you have tried a plain vector substraction?

  Please note that real world is far away of this latter. If you want =
this for
  anything remotely related with printing, your best move will be to =
convert
  to CMYK and operate in a real-word

  >One solution would be to convert all lab color to CMYK :
  >
  >and doing c:=3D (x1*c1 + x2*c2 + x3*c3) / (x1+x2+x3)
  >                    m:=3D(x1*m1 + x2*m2 +x3* m3) / 3 (x1+x2+x3)
  >                    y ....
  >                    k ....
  > ...
  > and then converting to Lab.

  Aha! This is what I was suggesting on first e-mail. I guess this is =
the best=20
  way at all. It handles ALL aspects I was talking off.

  > the only problem is that I will lose precision (double conversion)

  No too much, if you use 16 bits, that is enough. At least this is
  the maximum precission you can get from Photoshop, to say an
  example.

  Regards,
  Mart=ED.

    ----- Original Message -----=20
    From: Armindo Da Silva=20
    To: Lcm...@li...=20
    Sent: Friday, January 25, 2002 9:27 AM
    Subject: Re: [Lcms-user] Mixing Lab color


    No it's not really what I want to do :

    let's say Col 1 is (in Lab)  80 /10/35
                    Col 2 is (in Lab)  30 /40/65=20
                    Col 3 is (in Lab)  10 /10/-15=20

    I have to mix 10% of Col1 with 40% of col 2 and 90% of col 3
    So what is the formula to have Lab of the resulting color

    But as I said it's should work on soustractive maneer, because 0% =
mean no color and 100% mean full color

Re: [Lcms-user] Mixing Lab color

[<ma...@li...>]

Hi,

I'm not 100% sure, but I belive you cannot do that.

Lab works as light, that is, additive. You add several=20
lights and got a resulting one. No problem since there=20
is no additional information needed.

However, substractive colorspaces DOES need some=20
additional information, since they are also reflective.=20

You need to know illuminant (the source of light the inks
are blocking off) and the media white of substrate (the=20
chromaticity of paper) the gamut (the whitest you can get,=20
that is no more that the substrate, and the darkest, the point
at inks blocking all but ambience light) etc etc.

So, a Lab color minus another lab color can result in a lot of=20
colors, on depending on a lot of things. It *is* device dependent
at all :-)

In the other hand, assuming you have a 100% difusse perfect=20
reflective substrate and same illuminant as Lab white point,=20
Lab is an euclidean  space, and as a such should have=20
substraction defined. Do you have tried a plain vector substraction?

Please note that real world is far away of this latter. If you want this =
for
anything remotely related with printing, your best move will be to =
convert
to CMYK and operate in a real-word

>One solution would be to convert all lab color to CMYK :
>
>and doing c:=3D (x1*c1 + x2*c2 + x3*c3) / (x1+x2+x3)
>                    m:=3D(x1*m1 + x2*m2 +x3* m3) / 3 (x1+x2+x3)
>                    y ....
>                    k ....
> ...
> and then converting to Lab.

Aha! This is what I was suggesting on first e-mail. I guess this is the =
best=20
way at all. It handles ALL aspects I was talking off.

> the only problem is that I will lose precision (double conversion)

No too much, if you use 16 bits, that is enough. At least this is
the maximum precission you can get from Photoshop, to say an
example.

Regards,
Mart=ED.

  ----- Original Message -----=20
  From: Armindo Da Silva=20
  To: Lcm...@li...=20
  Sent: Friday, January 25, 2002 9:27 AM
  Subject: Re: [Lcms-user] Mixing Lab color


  No it's not really what I want to do :

  let's say Col 1 is (in Lab)  80 /10/35
                  Col 2 is (in Lab)  30 /40/65=20
                  Col 3 is (in Lab)  10 /10/-15=20

  I have to mix 10% of Col1 with 40% of col 2 and 90% of col 3
  So what is the formula to have Lab of the resulting color

  But as I said it's should work on soustractive maneer, because 0% mean =
no color and 100% mean full color

Re: [Lcms-user] Mixing Lab color

[Armindo Da S. <tec...@wa...>]

One solution would be to convert all lab color to CMYK :

and doing c:=3D (x1*c1 + x2*c2 + x3*c3) / (x1+x2+x3)
                    m:=3D(x1*m1 + x2*m2 +x3* m3) / 3 (x1+x2+x3)
                    y ....
                    k ....

if I 'm not wrong.

and then converting to Lab.

the only problem is that I will lose precision (double conversion)

also, should I use CMY or CMYK?

thanks

Armindo

Re: [Lcms-user] Mixing Lab color

[Armindo Da S. <tec...@wa...>]

No it's not really what I want to do :

let's say Col 1 is (in Lab)  80 /10/35
                Col 2 is (in Lab)  30 /40/65=20
                Col 3 is (in Lab)  10 /10/-15=20

I have to mix 10% of Col1 with 40% of col 2 and 90% of col 3
So what is the formula to have Lab of the resulting color

But as I said it's should work on soustractive maneer, because 0% mean =
no color and 100% mean full color

Re: [Lcms-user] Mixing Lab color

[<ma...@li...>]

Hi Armindo,

If I recall correctly, Lab does work additively.=20

If you want to add ink to see what is resulting color, you need to work=20
in CMYK. Add desired amout of ink by adding CMYK components
and then convert to RGB  or Lab by using reverse  transform.=20
I.e, CMYK->RGB or CMYK->Lab.

Example, to add some ink to a Lab color, convert color to CMYK,=20
add desired ink and back to lab. The impact of ink on color is=20
device dependent.

Is this what you want to do?

Regards,
Marti.


----- Original Message -----=20
  From: Armindo Da Silva=20
  To: Lcm...@li...=20
  Sent: Friday, January 25, 2002 6:54 AM
  Subject: [Lcms-user] Mixing Lab color


  Hello,

  I would like to mix some Lab colors for example 10% Col1 + 20% col2 =
+90% col 3

  what would be the result value?

  note that this is for ink (0% =3D white) 100% =3D pure color, so it's =
soustractive mixing not additive

  thanks

  Armindo

[Lcms-user] Re: LCMS and TIFF images

[<ma...@li...>]

Hi,

> As I want a quick solution, I was thinking about using Perl to access
> the TIFF files and use LCMS for the conversion tables.
>
> Now before I start, I wanted to ask whether someone else is doing the
> same thing, or already has done it.

It is quite easy to do this conversion, since lcms does include
a utility to apply profiles on TIFF. It also handles embedded profiles.
 If you downloaded ver 1.08, check the tifficc sample. ('make tifficc'
will build it) You will need libtiff to build the program, but is likely
you have libtiff already installed.

For the GIMP, there is also a plugin for color management
using lcms, http://www.freecolormanagement.com

Hope this helps
Martí Maria
The little cms project
http://www.littlecms.com
ma...@li...

----- Original Message -----
From: "Ulrich Windl" <Ulr...@rz...>
To: <in...@li...>
Sent: Friday, January 25, 2002 7:34 AM
Subject: Q: LCMS and TIFF images


> Hello,
>
> I have a scanner that can make TIFF images with embedded profiles (e.g.
> sRGB). Using LCMS I'd like to do color space conversion (e.g. to
> monitor profile). I know LCMS library can provide a mapping table.
>
> As I want a quick solution, I was thinking about using Perl to access
> the TIFF files and use LCMS for the conversion tables.
>
> Now before I start, I wanted to ask whether someone else is doing the
> same thing, or already has done it.
>
> Most preferred solution would be an extension for GIMP 1.2 (Linux).
>
> Regards,
> Ulrich
>
>
>

[Lcms-user] Mixing Lab color

[Armindo Da S. <tec...@wa...>]

Hello,

I would like to mix some Lab colors for example 10% Col1 + 20% col2 +90% =
col 3

what would be the result value?

note that this is for ink (0% =3D white) 100% =3D pure color, so it's =
soustractive mixing not additive

thanks

Armindo

Re: [Lcms-user] Profile/conversion problem

[<ma...@li...>]

Hi,

>thanks for the fast answer. Today I found the time to check it with some
>other images. It works almost perfect! There is almost no visible
>difference. If you use Photoshop to get the color values you find a small
>difference like 1% in Cyan, 2% in Magenta, 0% in yellow
>So I think nobody will notice it if the file is printed.

Thanks to you for let me know.

I wonder about these differences. In fact I wonder why ALL cmm
does give slightly different results, ICM, ColorSync, Adobe ACE. :-?

Regards,
Marti.


----- Original Message -----
From: "Gunnar Lieb" <ki...@f1...>
To: <Lcm...@li...>
Sent: Thursday, January 24, 2002 3:52 PM
Subject: Re: [Lcms-user] Profile/conversion problem


Hi Martí,

thanks for the fast answer. Today I found the time to check it with some
other images. It works almost perfect! There is almost no visible
difference. If you use Photoshop to get the color values you find a small
difference like 1% in Cyan, 2% in Magenta, 0% in yellow
So I think nobody will notice it if the file is printed.

Thanks again,

Gunnar

am 21.01.2002 23:08 Uhr schrieb Martí Maria unter ma...@li...:

> Hi,
>
>>> I'm not able to get similar values by doing it in LCMS. The olny profile I
>>> can use is Tiflab8space.icm. If I don't specify the profile and it uses the
>
> Thanks for the images.
>
> Sigh. :-( Ok. It happens that Photoshop does _not_ follow the TIFF 6.0 spec
> and stores Lab using D50 instead of D65. As a result, Lab images saved in
> Photoshop does behave quite different, specially on blues.
>
> This "bug" pops up quite frequently, so, since it seems Photoshop is doing the
> "true" standard instead of the spec, I have put a profile for this case in the
> site:
>
> http://www.littlecms.com/tiff8adobe.zip
>
> In the zip  is also the code to generate the profile, if anyone is interested
> on how
> to generate weird profiles using lcms. This profile is smaller that the
> anterior
> Tifflab8spac.icm and not subjected to any Linotype copyright anymore.
> I will check it carefully this week, but for now seems to work quite well on
> your
> test images.
>
> The CMYK bitmap does give pretty same values that Photoshop. One important
> thing: your Lab TIFF does have a embedded profile. This is wrong because Lab
> does not need any profiles and worst, the profile is operating in RGB, and
> tagged
> as Lab. If you are using TIFFICC to do checking, assure to use -n to ignore
> embedded profile.
>
>
> Best Regards,
> Martí Maria
> The little cms project
> http://www.littlecms.com
> ma...@li...
>
>
> ----- Original Message -----
> From: "Martí Maria" <ma...@li...>
> To: "Gunnar Lieb" <ki...@f1...>; <Lcm...@li...>
> Sent: Monday, January 21, 2002 9:41 AM
> Subject: Re: [Lcms-user] Profile/conversion problem
>
>
>> Hi,
>>
>>> I'm trying to reproduce a color conversion from LAB->CMYK. The original
>>> conversion is made by Heidelberg Linocolor.
>>>
>>> I'm not able to get similar values by doing it in LCMS. The olny profile I
>>> can use is Tiflab8space.icm. If I don't specify the profile and it uses the
>>
>> Are you decoding a TIFF file?. Tiflab8space.icm is only valid on 8-bit TIFF.
>> This profile does hold Lab AND a special decoding  for 8 bit TIFF.
>> If your Lab source is any other, INCLUDING 16 BIT TIFF, try
>> lcmslabi.icm. Programatically, you can use the new built-in Lab profile:
>>
>> cmsCreateLabProfile(NULL)
>>
>> That  is present on ver 1.08. I would recommend this latter, since it
>> takes 0 bytes on disk. If you are using tifficc, lcmslabi.icm will do the
>> job. I have done a lot of testing on Lab->CMYK and works quite well.
>>
>> About the embedded profile... if the file is storing data on Lab space, there
>> is no need of embedded profile, data is already device independent.
>>
>> Regards,
>> Martí Maria
>> The little cms project
>> http://www.littlecms.com
>> ma...@li...
>>
>>
>> ----- Original Message -----
>> From: "Gunnar Lieb" <ki...@f1...>
>> To: <Lcm...@li...>
>> Sent: Sunday, January 20, 2002 2:13 PM
>> Subject: [Lcms-user] Profile/conversion problem
>>
>>
>>> Hi,
>>>
>>> I'm trying to reproduce a color conversion from LAB->CMYK. The original
>>> conversion is made by Heidelberg Linocolor.
>>>
>>> I'm not able to get similar values by doing it in LCMS. The olny profile I
>>> can use is Tiflab8space.icm. If I don't specify the profile and it uses the
>>> embeded I get strange colors. It looks like solarized!. If I use the
>>> mentioned LAB profile there a big differences in the color. I.e. I get in
>>> - Linocolor conversion a blue with c96 m72 y0 k0
>>> - LCMS conversion with c95 m66 y13 k5
>>>
>>> I tried the t option with all values, but the conversion remained different.
>>>
>>> So with does the embedded profile produces such a strange color?
>>> Are there any other options to look for?
>>>
>>> Thanks,
>>>
>>> Gunnar
>>>
>>>
>>> _______________________________________________
>>> Lcms-user mailing list
>>> Lcm...@li...
>>> https://lists.sourceforge.net/lists/listinfo/lcms-user
>>>
>>>
>>
>
>


_______________________________________________
Lcms-user mailing list
Lcm...@li...
https://lists.sourceforge.net/lists/listinfo/lcms-user

Re: [Lcms-user] Profile/conversion problem

[Gunnar L. <ki...@f1...>]

Hi Mart=ED,

thanks for the fast answer. Today I found the time to check it with some
other images. It works almost perfect! There is almost no visible
difference. If you use Photoshop to get the color values you find a small
difference like 1% in Cyan, 2% in Magenta, 0% in yellow
So I think nobody will notice it if the file is printed.

Thanks again,

Gunnar

am 21.01.2002 23:08 Uhr schrieb Mart=ED Maria unter ma...@li...:

> Hi,
>=20
>>> I'm not able to get similar values by doing it in LCMS. The olny profil=
e I
>>> can use is Tiflab8space.icm. If I don't specify the profile and it uses=
 the
>=20
> Thanks for the images.
>=20
> Sigh. :-( Ok. It happens that Photoshop does _not_ follow the TIFF 6.0 sp=
ec
> and stores Lab using D50 instead of D65. As a result, Lab images saved in
> Photoshop does behave quite different, specially on blues.
>=20
> This "bug" pops up quite frequently, so, since it seems Photoshop is doin=
g the
> "true" standard instead of the spec, I have put a profile for this case i=
n the
> site:
>=20
> http://www.littlecms.com/tiff8adobe.zip
>=20
> In the zip  is also the code to generate the profile, if anyone is intere=
sted
> on how
> to generate weird profiles using lcms. This profile is smaller that the
> anterior
> Tifflab8spac.icm and not subjected to any Linotype copyright anymore.
> I will check it carefully this week, but for now seems to work quite well=
 on
> your
> test images.
>=20
> The CMYK bitmap does give pretty same values that Photoshop. One importan=
t
> thing: your Lab TIFF does have a embedded profile. This is wrong because =
Lab
> does not need any profiles and worst, the profile is operating in RGB, an=
d
> tagged
> as Lab. If you are using TIFFICC to do checking, assure to use -n to igno=
re
> embedded profile.
>=20
>=20
> Best Regards,
> Mart=ED Maria
> The little cms project
> http://www.littlecms.com
> ma...@li...
>=20
>=20
> ----- Original Message -----
> From: "Mart=ED Maria" <ma...@li...>
> To: "Gunnar Lieb" <ki...@f1...>; <Lcm...@li...>
> Sent: Monday, January 21, 2002 9:41 AM
> Subject: Re: [Lcms-user] Profile/conversion problem
>=20
>=20
>> Hi,
>>=20
>>> I'm trying to reproduce a color conversion from LAB->CMYK. The original
>>> conversion is made by Heidelberg Linocolor.
>>>=20
>>> I'm not able to get similar values by doing it in LCMS. The olny profil=
e I
>>> can use is Tiflab8space.icm. If I don't specify the profile and it uses=
 the
>>=20
>> Are you decoding a TIFF file?. Tiflab8space.icm is only valid on 8-bit T=
IFF.
>> This profile does hold Lab AND a special decoding  for 8 bit TIFF.
>> If your Lab source is any other, INCLUDING 16 BIT TIFF, try
>> lcmslabi.icm. Programatically, you can use the new built-in Lab profile:
>>=20
>> cmsCreateLabProfile(NULL)
>>=20
>> That  is present on ver 1.08. I would recommend this latter, since it
>> takes 0 bytes on disk. If you are using tifficc, lcmslabi.icm will do th=
e
>> job. I have done a lot of testing on Lab->CMYK and works quite well.
>>=20
>> About the embedded profile... if the file is storing data on Lab space, =
there
>> is no need of embedded profile, data is already device independent.
>>=20
>> Regards,
>> Mart=ED Maria
>> The little cms project
>> http://www.littlecms.com
>> ma...@li...
>>=20
>>=20
>> ----- Original Message -----
>> From: "Gunnar Lieb" <ki...@f1...>
>> To: <Lcm...@li...>
>> Sent: Sunday, January 20, 2002 2:13 PM
>> Subject: [Lcms-user] Profile/conversion problem
>>=20
>>=20
>>> Hi,
>>>=20
>>> I'm trying to reproduce a color conversion from LAB->CMYK. The original
>>> conversion is made by Heidelberg Linocolor.
>>>=20
>>> I'm not able to get similar values by doing it in LCMS. The olny profil=
e I
>>> can use is Tiflab8space.icm. If I don't specify the profile and it uses=
 the
>>> embeded I get strange colors. It looks like solarized!. If I use the
>>> mentioned LAB profile there a big differences in the color. I.e. I get =
in
>>> - Linocolor conversion a blue with c96 m72 y0 k0
>>> - LCMS conversion with c95 m66 y13 k5
>>>=20
>>> I tried the t option with all values, but the conversion remained diffe=
rent.
>>>=20
>>> So with does the embedded profile produces such a strange color?
>>> Are there any other options to look for?
>>>=20
>>> Thanks,
>>>=20
>>> Gunnar
>>>=20
>>>=20
>>> _______________________________________________
>>> Lcms-user mailing list
>>> Lcm...@li...
>>> https://lists.sourceforge.net/lists/listinfo/lcms-user
>>>=20
>>>=20
>>=20
>=20
>=20

Re: [Lcms-user] Profile/conversion problem

[<ma...@li...>]

Hi,

> > I'm not able to get similar values by doing it in LCMS. The olny profile I
> > can use is Tiflab8space.icm. If I don't specify the profile and it uses the

Thanks for the images.

Sigh. :-( Ok. It happens that Photoshop does _not_ follow the TIFF 6.0 spec
and stores Lab using D50 instead of D65. As a result, Lab images saved in
Photoshop does behave quite different, specially on blues.

This "bug" pops up quite frequently, so, since it seems Photoshop is doing the
"true" standard instead of the spec, I have put a profile for this case in the site:

http://www.littlecms.com/tiff8adobe.zip

In the zip  is also the code to generate the profile, if anyone is interested on how
to generate weird profiles using lcms. This profile is smaller that the anterior
Tifflab8spac.icm and not subjected to any Linotype copyright anymore.
I will check it carefully this week, but for now seems to work quite well on your
test images.

The CMYK bitmap does give pretty same values that Photoshop. One important
thing: your Lab TIFF does have a embedded profile. This is wrong because Lab
does not need any profiles and worst, the profile is operating in RGB, and tagged
as Lab. If you are using TIFFICC to do checking, assure to use -n to ignore
embedded profile.


Best Regards,
Martí Maria
The little cms project
http://www.littlecms.com
ma...@li...


----- Original Message -----
From: "Martí Maria" <ma...@li...>
To: "Gunnar Lieb" <ki...@f1...>; <Lcm...@li...>
Sent: Monday, January 21, 2002 9:41 AM
Subject: Re: [Lcms-user] Profile/conversion problem


> Hi,
>
> > I'm trying to reproduce a color conversion from LAB->CMYK. The original
> > conversion is made by Heidelberg Linocolor.
> >
> > I'm not able to get similar values by doing it in LCMS. The olny profile I
> > can use is Tiflab8space.icm. If I don't specify the profile and it uses the
>
> Are you decoding a TIFF file?. Tiflab8space.icm is only valid on 8-bit TIFF.
> This profile does hold Lab AND a special decoding  for 8 bit TIFF.
> If your Lab source is any other, INCLUDING 16 BIT TIFF, try
> lcmslabi.icm. Programatically, you can use the new built-in Lab profile:
>
> cmsCreateLabProfile(NULL)
>
> That  is present on ver 1.08. I would recommend this latter, since it
> takes 0 bytes on disk. If you are using tifficc, lcmslabi.icm will do the
> job. I have done a lot of testing on Lab->CMYK and works quite well.
>
> About the embedded profile... if the file is storing data on Lab space, there
> is no need of embedded profile, data is already device independent.
>
> Regards,
> Martí Maria
> The little cms project
> http://www.littlecms.com
> ma...@li...
>
>
> ----- Original Message -----
> From: "Gunnar Lieb" <ki...@f1...>
> To: <Lcm...@li...>
> Sent: Sunday, January 20, 2002 2:13 PM
> Subject: [Lcms-user] Profile/conversion problem
>
>
> > Hi,
> >
> > I'm trying to reproduce a color conversion from LAB->CMYK. The original
> > conversion is made by Heidelberg Linocolor.
> >
> > I'm not able to get similar values by doing it in LCMS. The olny profile I
> > can use is Tiflab8space.icm. If I don't specify the profile and it uses the
> > embeded I get strange colors. It looks like solarized!. If I use the
> > mentioned LAB profile there a big differences in the color. I.e. I get in
> > - Linocolor conversion a blue with c96 m72 y0 k0
> > - LCMS conversion with c95 m66 y13 k5
> >
> > I tried the t option with all values, but the conversion remained different.
> >
> > So with does the embedded profile produces such a strange color?
> > Are there any other options to look for?
> >
> > Thanks,
> >
> > Gunnar
> >
> >
> > _______________________________________________
> > Lcms-user mailing list
> > Lcm...@li...
> > https://lists.sourceforge.net/lists/listinfo/lcms-user
> >
> >
>

Re: [Lcms-user] Profile/conversion problem

[Gunnar L. <ki...@f1...>]

Hi Mart=ED,

thanks for the answer.
The image is an 8-Bit TIFF. I tried also the other profile you recommended
but it also converts it to this wired colors???

I just tried to convert it in Photoshop using the profile just to be sure
that Linocolor doesn't do any extra conversion. The image in Photoshop look=
s
like the one from Linocolor.

I have put two sample image and the profile on the FTP Server. It would be
great if you find the time to check it. Please let me know.
The files are:

LAB image:  http://199.108.103.177/staff/gunnar/lcms/44414LAB_small.tif
CMYK image: http://199.108.103.177/staff/gunnar/lcms/44414CMYK_small.tif
(Converted using Linocolor)
Profile:=20
http://199.108.103.177/staff/gunnar/lcms/Offset_Euro_pos_U340_K95.icm

Thanks,

Gunnar



am 21.01.2002 14:01 Uhr schrieb Mart=ED Maria unter ma...@li...:

> Hi,
>=20
>> I'm trying to reproduce a color conversion from LAB->CMYK. The original
>> conversion is made by Heidelberg Linocolor.
>>=20
>> I'm not able to get similar values by doing it in LCMS. The olny profile=
 I
>> can use is Tiflab8space.icm. If I don't specify the profile and it uses =
the
>=20
> Are you decoding a TIFF file?. Tiflab8space.icm is only valid on 8-bit TI=
FF.
> This profile does hold Lab AND a special decoding  for 8 bit TIFF.
> If your Lab source is any other, INCLUDING 16 BIT TIFF, try
> lcmslabi.icm. Programatically, you can use the new built-in Lab profile:
>=20
> cmsCreateLabProfile(NULL)
>=20
> That  is present on ver 1.08. I would recommend this latter, since it
> takes 0 bytes on disk. If you are using tifficc, lcmslabi.icm will do the
> job. I have done a lot of testing on Lab->CMYK and works quite well.
>=20
> About the embedded profile... if the file is storing data on Lab space, t=
here
> is no need of embedded profile, data is already device independent.
>=20
> Regards,
> Mart=ED Maria
> The little cms project
> http://www.littlecms.com
> ma...@li...
>=20
>=20
> ----- Original Message -----
> From: "Gunnar Lieb" <ki...@f1...>
> To: <Lcm...@li...>
> Sent: Sunday, January 20, 2002 2:13 PM
> Subject: [Lcms-user] Profile/conversion problem
>=20
>=20
>> Hi,
>>=20
>> I'm trying to reproduce a color conversion from LAB->CMYK. The original
>> conversion is made by Heidelberg Linocolor.
>>=20
>> I'm not able to get similar values by doing it in LCMS. The olny profile=
 I
>> can use is Tiflab8space.icm. If I don't specify the profile and it uses =
the
>> embeded I get strange colors. It looks like solarized!. If I use the
>> mentioned LAB profile there a big differences in the color. I.e. I get i=
n
>> - Linocolor conversion a blue with c96 m72 y0 k0
>> - LCMS conversion with c95 m66 y13 k5
>>=20
>> I tried the t option with all values, but the conversion remained differ=
ent.
>>=20
>> So with does the embedded profile produces such a strange color?
>> Are there any other options to look for?
>>=20
>> Thanks,
>>=20
>> Gunnar
>>=20
>>=20
>> _______________________________________________
>> Lcms-user mailing list
>> Lcm...@li...
>> https://lists.sourceforge.net/lists/listinfo/lcms-user
>>=20
>>=20
>=20
>=20
> _______________________________________________
> Lcms-user mailing list
> Lcm...@li...
> https://lists.sourceforge.net/lists/listinfo/lcms-user

Re: [Lcms-user] Profile/conversion problem

[<ma...@li...>]

Hi,

> I'm trying to reproduce a color conversion from LAB->CMYK. The original
> conversion is made by Heidelberg Linocolor.
>
> I'm not able to get similar values by doing it in LCMS. The olny profile I
> can use is Tiflab8space.icm. If I don't specify the profile and it uses the

Are you decoding a TIFF file?. Tiflab8space.icm is only valid on 8-bit TIFF.
This profile does hold Lab AND a special decoding  for 8 bit TIFF.
If your Lab source is any other, INCLUDING 16 BIT TIFF, try
lcmslabi.icm. Programatically, you can use the new built-in Lab profile:

cmsCreateLabProfile(NULL)

That  is present on ver 1.08. I would recommend this latter, since it
takes 0 bytes on disk. If you are using tifficc, lcmslabi.icm will do the
job. I have done a lot of testing on Lab->CMYK and works quite well.

About the embedded profile... if the file is storing data on Lab space, there
is no need of embedded profile, data is already device independent.

Regards,
Martí Maria
The little cms project
http://www.littlecms.com
ma...@li...


----- Original Message -----
From: "Gunnar Lieb" <ki...@f1...>
To: <Lcm...@li...>
Sent: Sunday, January 20, 2002 2:13 PM
Subject: [Lcms-user] Profile/conversion problem


> Hi,
>
> I'm trying to reproduce a color conversion from LAB->CMYK. The original
> conversion is made by Heidelberg Linocolor.
>
> I'm not able to get similar values by doing it in LCMS. The olny profile I
> can use is Tiflab8space.icm. If I don't specify the profile and it uses the
> embeded I get strange colors. It looks like solarized!. If I use the
> mentioned LAB profile there a big differences in the color. I.e. I get in
> - Linocolor conversion a blue with c96 m72 y0 k0
> - LCMS conversion with c95 m66 y13 k5
>
> I tried the t option with all values, but the conversion remained different.
>
> So with does the embedded profile produces such a strange color?
> Are there any other options to look for?
>
> Thanks,
>
> Gunnar
>
>
> _______________________________________________
> Lcms-user mailing list
> Lcm...@li...
> https://lists.sourceforge.net/lists/listinfo/lcms-user
>
>

[Lcms-user] Profile/conversion problem

[Gunnar L. <ki...@f1...>]

Hi,

I'm trying to reproduce a color conversion from LAB->CMYK. The original
conversion is made by Heidelberg Linocolor.

I'm not able to get similar values by doing it in LCMS. The olny profile I
can use is Tiflab8space.icm. If I don't specify the profile and it uses the
embeded I get strange colors. It looks like solarized!. If I use the
mentioned LAB profile there a big differences in the color. I.e. I get in
- Linocolor conversion a blue with c96 m72 y0 k0
- LCMS conversion with c95 m66 y13 k5

I tried the t option with all values, but the conversion remained different.

So with does the embedded profile produces such a strange color?
Are there any other options to look for?

Thanks,

Gunnar

Re: [Lcms-user] order problem?

[<ma...@li...>]

Hi,
 
> wich dwflags do you suggest me to get best color acuracy when converting
> from Lab to RGB

If you are going to translate a few values, cmsFLAGS_NOTPRECALC, since
it returns almost immediatly. Else it assumes you will translate a big amoult of
data and does compute a lookup table. Building such table can take 1-2 seconds, 
but then cmsDoTransform() goes a lot faster. There is an additional 
cmsFLAGS_HIGHRESPRECALC for building huge lookup tables, intended 
only for situations where gamma of both spaces are quite different, like sRGB 
and XYZ, for example. Is not really needed, but included for completness sake.

Any flags should give same values, but the most accurate would be 
cmsFLAGS_NOTPRECALC, at expense of speed. Also it takes no 
additional memory.

Regards,
Marti.

Re: [Lcms-user] order problem?

[Armindo Da S. <tec...@wa...>]

Marti,

wich dwflags do you suggest me to get best color acuracy when converting
from Lab to RGB
for the moment I use :

PrinterTransform := cmsCreateTransform(hLabProfile,
                TYPE_Lab_8,
                hPPaperProfile,
                TYPE_BGR_8,
                RenderIntent,0);

thanks

Armindo

Re: [Lcms-user] order problem?

[<ma...@li...>]

Hi,

I am attaching a sample on how to do whole stuff on Delphi.

Compile by DCC32 -CC testfrm.pas (it is a console program)

The parameters are as all other functions... I do use C convention 
of destination first, then source. strcpy() for example works in such
way. As you noted, the mistake is in  cmsFloat2LabEncoded() that should read:

PROCEDURE cmsFloat2LabEncoded(wLab: Pointer; Lab: LPcmsCIELab); StdCall;

Sorry again. I didn't catch the bug because it compiles and works ok
no matter the wrong parameters; Delphi doesn't check the pointer type!

Another question... the apparent loss of precission, or that's same, different 
values, are due to the quantization induced by 8 bit. Using 16 bit on RGB
the sample gets pretty same values after rounding.

 
Regards,
Marti.

[Lcms-user] order problem

[Armindo Da S. <tec...@wa...>]

I think it is=20
PROCEDURE cmsFloat2LabEncoded( wLab: Pointer;Lab: LPcmsCIELab); StdCall;

[Lcms-user] order problem?

[Armindo Da S. <tec...@wa...>]

Marti,

Are you sure the parameters are in the good orders in these functions?

PROCEDURE cmsLabEncoded2Float( Lab: LPcmsCIELab; wLab: Pointer); =
StdCall;
PROCEDURE cmsFloat2LabEncoded( Lab: LPcmsCIELab; wLab: Pointer); =
StdCall;

Armindo

Re: [Lcms-user] Re: Conversion problems

[Armindo Da S. <tec...@wa...>]

Marti,

I may do something wrong using the utils functions :

var EncodedRGB: ARRAY[0..2] OF Word;
        EncodedLab: ARRAY[0..2] OF Word;
        Lab: cmsCIELab;

now to convert from Lab16 to RGB16 I do

        Lab.L:=StrtoFloat(edit1.text);
        Lab.a:=StrtoFloat(edit2.text);
        Lab.b:=StrtoFloat(edit3.text);

        cmsFloat2LabEncoded(@Lab,@EncodedLab);


        cmsDoTransform(hLAB2RGBTransform, @EncodedLab, @EncodedRGB, 1);


        edit4.Text:=inttostr(EncodedRGB[0]);
        edit5.Text:=inttostr(EncodedRGB[1]);
        edit6.Text:=inttostr(EncodedRGB[2])

       but I always get 8,8,8 as rgb result

       What am I doing wrong?

Re: [Lcms-user] Re: Conversion problems

[Armindo Da S. <tec...@wa...>]

Thanks Marti,

I will make some test and will let you know.

Armindo

[Lcms-user] Re: Conversion problems

[<ma...@li...>]

Hi Armindo, happy new year.

I'm CC: the response to mailing list, perhaps anyone else is interested.

> I have tryed a old example you gave me 2 year ago in delphi and if I put
> 100,100,100 in Lab and then convert it to RGB I have more than 255 in R.
>

This is because Lab= (100, 100, 100)  is out of gamut of RGB space. Is normal.
Lab=100, 0, 0 is a total highlight. On depending on profile it could be that the
extrapolation of a=100, b=100 give even more red.

> second problem if I convert from a lab value to RGB then back to Lab I don't
> have initial value do you know why?

Depends on profile. But most don't result on this change. I tried on the
icctrans sample program and it gives me same result if rounded properly...
But as said, this depends on profile. If you are using the built-in profiles,
cmsCreate_sRGBProfile() and cmsCreateLabProfile() the results are
reversible.

> how to use in delphi these utils function
> void cmsLabEncoded2Float(LPcmsCIELab Lab, const WORD wLab[3]);
> void cmsFloat2LabEncoded(WORD wLab[3], const LPcmsCIELab Lab);

There is a mistake in lcmsdll.pas, where parameters are swapped: It should be

PROCEDURE cmsLabEncoded2Float( Lab: LPcmsCIELab; wLab: Pointer); StdCall;
PROCEDURE cmsFloat2LabEncoded( Lab: LPcmsCIELab; wLab: Pointer); StdCall;

Sorry for this, I'm updating sources as well.

These are utility functions to simplify conversion form/to encoded form. You need to
use a variable of type 'cmsCIELab'... example;

var
  Lab: cmsCIELab;
  LabEncoded: ARRAY[0..2] OF Word;

       ....
        cmsDoTransform(@sRGB, @LabEncoded, 1)
        cmsLabEncoded2Float(@Lab, @LabEncoded)

And then you can access Lab.L, Lab.a and Lab.b in floating point and properly decoded.

Regards,
Marti.


----- Original Message -----
From: "Armindo Da Silva" <tec...@wa...>
To: "Martí Maria" <ma...@li...>
Sent: Tuesday, January 15, 2002 7:45 AM
Subject: Conversion problems


> Hola Marti,
>
> First of all Happy new year!
>
> I have tryed a old example you gave me 2 year ago in delphi and if I put
> 100,100,100 in Lab and then convert it to RGB I have more than 255 in R.
>
> second problem if I convert from a lab value to RGB then back to Lab I don't
> have initial value do you know why?
>
> how to use in delphi these utils function
> void cmsLabEncoded2Float(LPcmsCIELab Lab, const WORD wLab[3]);
> void cmsFloat2LabEncoded(WORD wLab[3], const LPcmsCIELab Lab);
>
> thanks
>
> Armindo
>

[Lcms-user] Re: Questions on CMS

[<ma...@li...>]

Hi,

>>Proofing is a bit more complex since it involves two intents.
>I agree and is a stumbling block for the end user. Unfortunately, many of 
>the terms used for CMS can obscure the meaning of things for end users.  

Yes. But has some logic indeed.
For example, I want to proof how a photo will look on a given printer. 
The "main" transform would be perceptual, since what I will use to 
print is perceptual to maintain contrast. Then, I could choose to 
view the proof  perceptually, that is, matching the _appearance_ of 
the printed sheet, or absolute, that is the rendered  colors on sheet 
ready to side-by-side matching. 

Most users are unaware the adaptation stuff done by the our eye/brain, 
so they choose "absolute colorimetric"... Only to next  complain about 
"yellowish" result. 

Absolute proof does completly depends of the illuminant used, and 
since ICC system is based on D50, proof is viewed as done on a D50 
box. 

In the other hand, perceptual will us give a overall proof on how 
will look the result, no matter wich illuminant and with the only
 requeriment of fully adapted observer. A nice feature IMHO. 
But it cannot match side-by side unless the room illuminant were D50.
Not very frequent after all. 

So, what to choose as default?
I belive each one has its virtues/defects, and the best move would be
to let enduser to select between them, perhaps giving perceptual as 
a reasonable default (Windows ICM does use abs. colorimetric by 
default)


>My only comment is I have one client set up like this and 95% of their 
>images have a workflow like this:
>
>Scanned RGB tiffs - no auto level adjustments in the scan> PS 6> Source=scanner 
>profile > Destination= CMYK Press Profile. Working space is Bruce RGB or Adobe 
>RGB.  

Very typical workflow. I assume workspace to be efectively null, 
since PS6 does use workspace only to modify image. This can be done 
by lcms by just ONE transform, scanned RGB tiff -> CMYK Press profile. 
You can use TIFFICC utility to do some testing. TIFFICC can do
the conversion even more accurately, by using 16-bit TIFF. 
If you want to do softproofing to your monitor, you will need another 
transform, form CMYK output space to monitor RGB. Here applies the
intent stuff I was commenting before.


>At least my clients have, uncommonly I find, : 1) Accurate monitor 
>profiles refreshed every couple of months. ( I tape the settings over, 
>once they are set properly.) 2) A fairly accurate press profile with 
>matching inks and paper. 3) Fairly consistent soft proofs from machine to 
>machine.   

These are the three big issues on the stuff, I agree. Specially the
monitor, that IMHO has a vital role.

> Testing by me shows relative colorimetric gives us very similar results. 

As it should. 

Perceptual = Relative colorimetric + appearance corrections 

In your case all would be same except on gamut boundaries. But there 
are gamut mapping algorimts that does move all colors, in this case 
intents are different. Although this is really a rare case.


>Can you tell me which program can or by default put in the preview tag? 
>Is there a way in little CMS to do this programmatically?

As said I think a aproximation could be done by two transforms, and 
indeed I would avoid the preview tag. Not because lcms being buggy,
that is not, just because most profile builders does stored bogus
content in preview tags!! surprising but true. Chris Cox did notify 
me that Adobe itself is ignoring such tag for this reason.



>I am playing with the monitor profiler.  Works well - no crashes.  I'll 
>have some questions about that later on. I'll have a chance to test the 
>scanner profiler soon. 

Glad to know :-)

Best regards,
Marti.
 

[Lcms-user] Re: Measurement Tool crashing

[<ma...@li...>]

Hi,

>You might want to add some Corel target reference files to your package as
> well. I don't know the copyright implications but you'll decide...

Many thanks!

After taking a close look, I realize these the profiler does
not accept CORELTAR.REF just because it misses the
required field "NUMBER_OF_SETS". The other target
reference, "coreltgt.ref" is accepted without complains.

If you can edit "CORELTAR.REF" manually, just add:

NUMBER_OF_SETS 168

The target is then accepted. Just place the files in a
directory and select this directory pressing the
"Select target vendor and type". I do use the "targets"
folder as root of target types, but any directory could work.

I agree the profiler should give a clean explanation on what
is happening instead of hunging up. I will fix this soon.

Then, it only remains to build suitable templates for measurement
tool ... but for this I do need Corel target geometry. Anyway,
I will do some searching on this stuff.

It would be nice to add Corel & GretagMacbeth  support in next
release... but I'm still little worried about copyright stuff.

Thanks again,
Marti.


----- Original Message -----
From: "Gabor DEAK JAHN" <dj...@tr...>
To: "Martí Maria" <ma...@li...>
Sent: Wednesday, January 09, 2002 11:23 AM
Subject: Re: Measurement Tool crashing


At 1/4/2002 03:52 PM, you wrote:

Martí,

You might want to add some Corel target reference files to your package as
well. I don't know the copyright implications but you'll decide...

These targets came bundled with various Corel software packages. There must
be other versions as well, with different PROD_DATE values. As far as I
know, they are basically standard IT8.7/2 targets with the addition of the
two pictures on the right.



--------------------------------------------------------------------------------



Regards,

Gabor DEAK JAHN

-------------------------------------------------------------------
Gabor DEAK JAHN -- Budapest, Hungary.
WWW: www.tramontana.co.hu
E-mail: dj...@tr...

[Lcms-user] Re: MNPROF

[<ma...@li...>]

Hi Ian,

> I am not understanding Mnprof very well.=20

 Well, MNPROF was a first try of a simple, easy to use monitor
 profiler. It generates very small (per file size) profiles.
 These profiles are usefull for embedding, and good as as a=20
 aproximation of monitor colorspace.=20

 Then, lately, we have released a more elaborated pack,=20
 including scanner and monitor profiler. Although the=20
 quantity of controls does make the new profilers a bit more
 complex, the quality of obtained profiles are a lot better.
 =20
 You can find new profiler pack at

    http://www.littlecms.com/profilers.htm

 Anyway, old MNPROF still is capable of giving some aid.

>=20
> 1. Is the IT8 test picture an alternative to using the wizard?
>=20

  No. The IT8 is just a sample. MNPROF can be used to check the
  results of profile. All you need is a JPEG or TIFF with embedded=20
  profile. Then, you can load the image, run the wizard to calibrate=20
  the monitor, and check the results on the image. This of course=20
  only works with images having colorimetric information (embedded
  profile). The IT8 is just a sample of such images. To check the=20
  difference between with/without color management, just click the=20
  small icm icon on left of status bar.=20


> 2. When I calibrate the gamma, it goes right down to 1.3 in order to
> make the central square merge into the outer one. Is something wrong?

  Yes. The gamma method I was using is not very good. This has been=20
  fixed in new profiler. Anyway, if you know the gamma of your monitor,=20
  set the adjust to the numerical value, no matter what the checker=20
  patterns says. But such low value is not normal. Could be because=20
  you have already modified hardware gamma by adobe gamma or other=20
  means.

>=20
> 3. When I attempt the final adjustment, there is an error message =
#4096
> "input profile operating on wrong colour space". The programme stops

  Could be a problem of old MNPROF. As said, we are dropping this old
  profiler to the new one.=20

>=20
> 4. When a profile is saved, one is invited to make notes (name of
> monitor etc. I could not find these notes when I looked at the saved
> profile with Wordpad.
>=20

  ??? This sounds weird. A way to check this strings is to right click
  on the profile withing explorer, go to "Properties". There is=20
  a tab wich give info on profiler. Here your comments should appear.=20


> 5. Photoshop and Paintshop Pro each offer a monitor gamma. Is this an
> alternative to something like MNPROF? It has been pointed out that =
such
> a gamma will be valid only for the programme producing it. In other
> words, a monitor calibrated with the Adobe gamma will look wrong with
> Paintshop Pro. Why are these gamma so exclusive? Does Adobe assume =
that
> one will never want to use a programme other than theirs?

  Adobe gamma does two things. One, sets the gamma ramp of the monitor,=20
  so, it is actually modifying hardware gamma. Two, does generate a=20
  profile for  this modified gamma. In such way, after adobe gamma runs, =

  any profile taken before is invalidated, since gamma is now changed.=20
  Both mnprof and new profiler are simply making profiles of existing=20
  adjusts, and does not change anything. MNPROF does generate same kind
  of profiles that adobe gamma, the new profiler does generate profiles=20
  more sophisticated, taking into account viewing condition as well.=20
  I did check adobe gamma, and it can load profiles produced by=20
  MNPROF and some (not all) of the profiles generated by new profiler.
  Windows ICM (so, PSP) does accept all profiles.

  =20
=20
> 6. What exactly is a generic profile? Is it something like sRGB Colour
> Space? Is it an ICC or ICM profile associated with a particular make =
of
> monitor (like those listed in Microsoft Windows)? Is it a profile =
other
> than one worked out on one's own computer on one's own monitor using a
> colour management programme - like MNPROF?

 A generic profile is a profile for a family of devices instead of a=20
 particular one. For example, if you have a Philips 17B, the vendor does =

 give a generic profile for ALL Philips 17B. But this is not to be =
accurate=20
 since there are important variations even within same brand model.=20
 Color management needs a different profile for *each* Philips 17B to =
work
 properly.

=20
> 7. Is it true that each and every monitor manufacturer produces an ICC
> or ICM profile and will publicize it if one keeps on asking? I feel
> convinced that I will never get anything out of Fujitsu-Siemens and
> suspect that it is partly for this situation that programmes like =
MNPROF
> are produced.

 That are generic profiles. And anyway, if Fujitsu-Siemens does give=20
 such generic profile, it would be less accurate that any one produced=20
 by any profiler. Why? just because each monitor behaves differently,=20
 and the gamma and white point does depend of factors like hours of use, =

 ambient light, etc.

 You are goint to get best results by profiling YOUR monitor, not by =
using
 the generic profile. There is no magic or device hardware adjust on =
profiles,
 just a set of measurements about gamma, white poin, primaries, etc.


>=20
> 8. My monitor (an LCD Fujitsu-Siemens) uses a different language from
> yours. for example, the default brightness and contrast settings are =
128
> and go to a maximum of 255.

  Ouch! All profilers does work on CRT monitors, and are not well suited =

  for LCD. This is true on our free profilers and in a commercial ones =
from
  other companies too.  LCD does not exhibit "normal" gamma, but a =
s-shaped=20
  curves hard to reproduce. Also, primaries are quite different. There =
is a=20
  lack   of support for these devices primarily because the hard =
dependecy of=20
  viewing angle.=20
 =20
  Anyway, I have been able to profile such devices with a colorimeter=20
  (X-Rite DTP92) and the results of new profiler are reasonably.  So, =
despite=20
  you will not get perfect color match, it should be a lot better that =
nothing.

  Regards,
  Mart=ED Maria
The little cms project
http://www.littlecms.com
ma...@li...

=20
----- Original Message -----=20
From: "Ian Ross" <ia...@kr...>
To: <ma...@li...>
Sent: Saturday, January 05, 2002 2:37 PM
Subject: MNPROF


> Hello Marti,
>=20
> Many thanks for letting me contact you again about colour management.
>=20
> I am not understanding Mnprof very well.=20
>=20
> 1. Is the IT8 test picture an alternative to using the wizard?
>=20
> 2. When I calibrate the gamma, it goes right down to 1.3 in order to
> make the central square merge into the outer one. Is something wrong?
>=20
> 3. When I attempt the final adjustment, there is an error message =
#4096
> "input profile operating on wrong colour space". The programme stops
>=20
> 4. When a profile is saved, one is invited to make notes (name of
> monitor etc. I could not find these notes when I looked at the saved
> profile with Wordpad.
>=20
> 5. Photoshop and Paintshop Pro each offer a monitor gamma. Is this an
> alternative to something like MNPROF? It has been pointed out that =
such
> a gamma will be valid only for the programme producing it. In other
> words, a monitor calibrated with the Adobe gamma will look wrong with
> Paintshop Pro. Why are these gamma so exclusive? Does Adobe assume =
that
> one will never want to use a programme other than theirs?
>=20
> 6. What exactly is a generic profile? Is it something like sRGB Colour
> Space? Is it an ICC or ICM profile associated with a particular make =
of
> monitor (like those listed in Microsoft Windows)? Is it a profile =
other
> than one worked out on one's own computer on one's own monitor using a
> colour management programme - like MNPROF?
>=20
> 7. Is it true that each and every monitor manufacturer produces an ICC
> or ICM profile and will publicize it if one keeps on asking? I feel
> convinced that I will never get anything out of Fujitsu-Siemens and
> suspect that it is partly for this situation that programmes like =
MNPROF
> are produced.
>=20
> 8. My monitor (an LCD Fujitsu-Siemens) uses a different language from
> yours. for example, the default brightness and contrast settings are =
128
> and go to a maximum of 255.
>=20
> Sorry to be a nuisance and for the level of ignorance shown. I had
> thought of joining your mailing list but doubt whether I am qualified. =
=20
>=20
> Many thanks in advance and Happy New Tear.
> --=20
> Ian Ross
>=20
>=20
>=20
> --=20
> Ian Ross
>=20
>=20

[Lcms-user] Re: Littlecms troubles...

[<ma...@li...>]

Hi,

>I was hoping that you could help:
>
>I spent a lot of time, short of thoroughly looking through your code,
>trying to figure out packing order of RGB images:
>
>TYPE_RGB_8  suppose to specify data that is R,G,B,R,G,B,....
>arranged as oppose to TYPE_RGB_8_PLANAR RRR...,GGG...,BBB...
>Right?
>
>Well, I definitely pass the data as RRR...,GGG...,BBB... and
>yet with TYPE_RGB_8_PLANAR my image comes out all scrambled
>while with TYPE_RGB_8 it is looking fine.  Is there a bug in
>release version 1.08?

It happens that planar layout was only supported on input. I'm
attaching a modified cmpack.c that does support output too.
Simply replace the old cmspack.c with this one and the problem
should be gone.

Regards,
Martí Maria
The little cms project
http://www.littlecms.com
ma...@li...



----- Original Message -----
From: <wi...@ma...>
To: <ma...@li...>
Sent: Wednesday, January 02, 2002 7:32 PM
Subject: Littlecms troubles...


Hello Marti,

I was hoping that you could help:

I spent a lot of time, short of thoroughly looking through your code,
trying to figure out packing order of RGB images:

TYPE_RGB_8  suppose to specify data that is R,G,B,R,G,B,....
arranged as oppose to TYPE_RGB_8_PLANAR RRR...,GGG...,BBB...
Right?

Well, I definitely pass the data as RRR...,GGG...,BBB... and
yet with TYPE_RGB_8_PLANAR my image comes out all scrambled
while with TYPE_RGB_8 it is looking fine.  Is there a bug in
release version 1.08?

Your help is greatly appreciated.  I definitely did not make a mistake
in passing the data or in specifying the transform.  I used the
same input and output data arrangement.

What's going on?

Have you seen this before?

THank you,

Witek



--------------------------------------------------------------------
mail2web - Check your email from the web at
http://mail2web.com/ .