CMYK color space vs sRGB

[zayca]

Hi,

Why I get so much havier files when exporting pdf with cmyk (90mb) vs srgb(30mb) ?

[zayca]

Settings are the same. I am just testing it. So does it seem normal to have 3 x havier file in cmyk then srgb? (due to 4th channel? but the space is limited...) I just can't make the file to look the same in srgb( I did it in cmyk space, but when I set the color space on export as srgb, some colors are off)

[v_kyr]

15 minutes ago, zayca said:

I did it in cmyk space, but when I set the color space on export as srgb, some colors are off

There are certain CMYK gammut colors which aren't inside the sRGB gammut and have to be mapped to nearest other colors.

[zayca]

It's portfolio. Many images, almost no text. Rather no vectors. 

[zayca]

3 hours ago, BofG said:

It's not so straightforward with images - there can be colour profile interactions between the document and the images that can inflate the size, if the profiles have to be included separately for example vs the whole document being in device CMYK.

I'm not expert enough to fully explain it, I can point you in a general direction though - look into your image profiles and whether you are converting them on export.

I don't understand to be honest. I guess that if I link  images with sRGB profile to document, they are converted to export color space (CMYK) during the process. All images should be converted for export. 

 

I set document space to CMYK, I import some drawings, images, I add some colors etc. It look ok in Publisher. Now (due to size of the file) I would like to export it as sRGB BUT I WOULD LIKE all media to look the same as edited in cmyk. 

[lacerto]

On 12/14/2021 at 12:33 AM, zayca said:

All images should be converted for export.

 

On 12/14/2021 at 12:33 AM, zayca said:

I would like to export it as sRGB

There is a kind of a conflict between these two goals. But if your placed images are in RGB in a CMYK mode document, then their visual outlook will be adjusted to simulate the target color space. You can keep much of that outlook in a PDF export if you convert the native colors (color of shapes, text, etc.) to CMYK, and keep the export file size compact, if you do not converl the image color space when exporting (meaning that the placed RGB images will not be converted). This happens if you export using e.g. PDF/X-3 or PDF/X-4 presets, or the default "PDF (press ready)" profile (the export settings have "Convert image color spaces" option unchecked). 

The reason your CMYK export (e.g. using PDF/X-1a, which forces images to be converted to CMYK) produces bigger file size might be related, as mentioned, to producing one channel more, or using a lower export DPI (e.g., if using "PDF (digital - small size)" setting when exporting to sRGB. In certain situations Publisher might also upscale low-resolution files to document DPI, but to be able to tell the exact cause, an example file would be needed of both kinds of exports.

[zayca]

 

This is document created in cmyk. When I set color as sRGB in document setup it looks like this on the right (more saturated). Is there any way to make it look the same as cmyk but in srgb color space? (less saturated)

 

 

[v_kyr]

sRGB:

• tulpen_srgb.afphoto

[lacerto]

13 hours ago, zayca said:

This is document created in cmyk. When I set color as sRGB in document setup it looks like this on the right (more saturated). Is there any way to make it look the same as cmyk but in srgb color space? (less saturated)

As long as you have the placed images as image layers, switching the document color mode will not convert the colors, so if you have an sRGB image in CMYK color mode, its visual outlook will be adjusted to simulate the document CMYK color target. When you change it back to sRGB mode, it will once again be shown in full sRGB and typically more saturated than in CMYK mode. If you want an RGB image to look more or less the same as it looks in CMYK mode, rasterize the image while you are in CMYK mode. Now the image pixels are in CMYK color mode (which is a destructive change). Now if you switch the document to RGB color mode, the image pixels will be converted to sRGB, and it would look pretty much the same in both color modes. But each time you switch the color mode, the image pixels would actually get converted, which is not recommended.

Top half: rasterized in CMYK mode, converted to RGB by switching the document color mode. Bottom half: the original RGB image in sRGB color space.

[zayca]

12 hours ago, Lagarto said:

As long as you have the placed images as image layers, switching the document color mode will not convert the colors, so if you have an sRGB image in CMYK color mode, its visual outlook will be adjusted to simulate the document CMYK color target. When you change it back to sRGB mode, it will once again be shown in full sRGB and typically more saturated than in CMYK mode. If you want an RGB image to look more or less the same as it looks in CMYK mode, rasterize the image while you are in CMYK mode. Now the image pixels are in CMYK color mode (which is a destrucitve change). Now if you switch the document to RGB color mode, the image pixels will be converted to sRGB, and it would look pretty much the same in both color modes. But each time you switch the color mode, the image pixels would actually get converted, which is not recommended.

Top half: rasterized in CMYK mode, converted to RGB by switching the document color mode. Gottom half: the original RGB image in sRGB color space.

Tahnk You so much. Well, it helped however I don't understand why exactly.

I took one flower. Each is made by separate layers (color, outline, other tweaks). I rasterize it. ( why doas it exectly mean? i was already an pixel art not vector). Then I set whole document as sRGB. That flower now looks good (same desaturated color as in CMYK). So the result is ok. But I lost all editing options as leyers are now rastrized and merged.

[lacerto]

15 hours ago, zayca said:

I rasterize it. ( why doas it exectly mean? i was already an pixel art not vector).

If you have a raster file linked or embedded in a document, it retains it original color space as long as it is an "image" layer. When you rasterize it, it becomes a "pixel" layer and subsequently all changes would be applied directly on pixel level rather than on meta data. Rasterizing is a destructive operation so as mention, you lose editing options. In Publisher, you can, however, turn a pixel layer again back to an image layer and regain benefits of image layers (you can, however, apply adjustments also on pixel layers).

As an alternative to rasterization, you could have an RGB color mode Publisher document and just use a Soft proof adjustment on the RGB layers (groups) which you wish to see displayed in specific CMYK target color space. This would allow you to keep objects and raster image layers in RGB color space yet see them previewed as they were in CMYK, and allow you also produce RGB based (leaner) PDFs.

softproofrgb.pdf

But at production time, you'd need to turn off the softproof adjustments and just produce the print file, either one actually converted to CMYK (resulting in larger file size),or one where placed raster images are retained in RGB color space, but profiles included to tell how they should be rasterized when printed.