DPI And ICC Profile

I use DPP for my post processing and my question refers to "Convert And Save".
In this window, what should the DPI be set at. It defaults to 350. Does this have to do with the printers printing resolution? Also, there is a checkbox labeled "Embed ICC Profile". Can some one tell me what this means? Thanks

Well, first of all, the "DPI" (meaning dots per inch) does only pertaing to when you are preparing an image to print and, in fact, is a term that gets confused because a more accurate term for digital imaging is "PPI" meaning pixels per inch, which is in today's technology actually describes what we are doing -- redering an image according to the resolution of pixels per printed inch. In fact, digital printers print using a whole lot of dots and use a "dithering" process to bled things together, and they take the image you pass to them and in the driver software optimize it to what they can effectivly use for the print.

For uses such as the Web or personal sharing/displaying, the figure means nothing -- it's just stored in the file Exif. When you upload an image to be viewed on the Web, you typically use a size in pixels that will suit a "normal" display -- POTN, for example, has a size limit of 1024 pixels at the widest dimension, and Web displays simply show these images at the resolution of the monitor you are viewing it on, which is typically around 90-100 ppi, so the "DPI"/PPI of the file is ignored.

As to imbedding the icc profile, I'm not positive as to whether Canon is referring to the color space or not, as in sRGB or aRGB (or another)...if so, then that would be to give proper information to color-managed software about how to display colors, but I'm not sure about that and really should not matter to color-managed software because the image should be converted to a color space and good software should be able to "read" that. So, maybe someone who understands what Canon actually means can fill in here.

OK, I just checked the User Manual for DPP and it is referring to the color space, but then goes on to list several color spaces and says that if you leave the box unchecked the profile is still imbedded. But it doesn't list sRGB or aRGB, which is curious.

Now, sRGB is the "generic" color space and so if you have DPP set to work in sRGB you are safe. What I'm not sure of is if you are set to aRGB and then convert without embedding the profile and try to open that in a color-managed app or browser. I've never messed with that.

I guess to be save check the box, especially if you are working in anything but sRGB. I myself haven't sed DPP for "fancy" color-managed stuff (I use Lightroom for my "serious" processing) so can't shed any more light on the subject!

James Emory wrote in post #11456745
I use DPP for my post processing and my question refers to "Convert And Save".
In this window, what should the DPI be set at. It defaults to 350. Does this have to do with the printers printing resolution?

Yes. That label should actually be PPI. I set mine at 300 for compatibility with my HP printer (Canon too). I would imagine an Epson user may want 360.

I never use the Convert and Save function, preferring to send any printable image directly to Photoshop for final editing before saving as a TIF (or JPG).

Another confusing setting. You want to keep that checked as it tags your converted image with whatever default RGB color space you selected in the Preferences (sRGB, Adobe RGB, etc.). Tools > Preferences > Color Management.

HOSTED PHOTO
please log in to view hosted photos in full size.

tonylong wrote in post #11457123
Well, first of all, the "DPI" (meaning dots per inch) does only pertaing to when you are preparing an image to print and, in fact, is a term that gets confused because a more accurate term for digital imaging is "PPI" meaning pixels per inch, which is in today's technology actually describes what we are doing -- redering an image according to the resolution of pixels per printed inch. In fact, digital printers print using a whole lot of dots and use a "dithering" process to bled things together, and they take the image you pass to them and in the driver software optimize it to what they can effectivly use for the print.

For uses such as the Web or personal sharing/displaying, the figure means nothing -- it's just stored in the file Exif. When you upload an image to be viewed on the Web, you typically use a size in pixels that will suit a "normal" display -- POTN, for example, has a size limit of 1024 pixels at the widest dimension, and Web displays simply show these images at the resolution of the monitor you are viewing it on, which is typically around 90-100 ppi, so the "DPI"/PPI of the file is ignored.

As to imbedding the icc profile, I'm not positive as to whether Canon is referring to the color space or not, as in sRGB or aRGB (or another)...if so, then that would be to give proper information to color-managed software about how to display colors, but I'm not sure about that and really should not matter to color-managed software because the image should be converted to a color space and good software should be able to "read" that. So, maybe someone who understands what Canon actually means can fill in here.

Thank you, Tony, much appreciated.

tonylong wrote in post #11457196
OK, I just checked the User Manual for DPP and it is referring to the color space, but then goes on to list several color spaces and says that if you leave the box unchecked the profile is still imbedded. But it doesn't list sRGB or aRGB, which is curious.

Now, sRGB is the "generic" color space and so if you have DPP set to work in sRGB you are safe. What I'm not sure of is if you are set to aRGB and then convert without embedding the profile and try to open that in a color-managed app or browser. I've never messed with that.

I guess to be save check the box, especially if you are working in anything but sRGB. I myself haven't sed DPP for "fancy" color-managed stuff (I use Lightroom for my "serious" processing) so can't shed any more light on the subject!

I like to shoot in RAW and then convert to 16 bit TIFF files. Seems to look a little better than JPEG or my eyes are playing tricks on me.

ChasP505 wrote in post #11457805
Yes. That label should actually be PPI. I set mine at 300 for compatibility with my HP printer (Canon too). I would imagine an Epson user may want 360.

I never use the Convert and Save function, preferring to send any printable image directly to Photoshop for final editing before saving as a TIF (or JPG).



Another confusing setting. You want to keep that checked as it tags your converted image with whatever default RGB color space you selected in the Preferences (sRGB, Adobe RGB, etc.). Tools > Preferences > Color Management.

Hosted photo: posted by ChasP505 in
./showthread.php?p=114​57805&i=i91727391
forum: RAW, Post Processing & Printing

Chas, what do all these settings mean, I do have the MP990. I have none selected. I don't get what these MP990 numbers related to. Nevermind, I don't know how to do what you did as to the copy screen and get it into this post. I did the copy screen OK and put it into a Word document but when I went to upload it, I got an error saying invalid file. Can you help me out here if you have time?
I wanted to show you what I was talking about here.

James- Just ask a specific question about a specific setting that you're not sure of.

And I use screenprint32 for screen captures.

****If you are talking about all the possible Printer Profile selections for your MP990... Unless you print from DPP, you should leave that as "none". (I have a profile selected because I was showing someone else the setting options, but I never print from DPP either.

James Emory wrote in post #11457932
I like to shoot in RAW and then convert to 16 bit TIFF files. Seems to look a little better than JPEG or my eyes are playing tricks on me.

Well, first of all the embedded profile has nothing to do with tiff or jpeg -- it is the color space your Raw converter uses to render colors and "builds in" to your image. Checking the "Embe ICC Profile" ensures that color-managed apps will "understand" what profile was used and will be able to interpret the color values properly.

Now, tiff and jpeg are file formats, jpeg undergoes "lossy compression", tiff not (although some software offers a "lossless compression". Jpegs are 8 bits per channel only, tiff gives you the choice of 8 or 16 bits.

But, a high quality jpeg compared to the same image converted to a tiff should really match, since all the processing has been done prior to conversion. But, many of us do use tiff when we have to go into Photoshop to do further editing to keep the maximum ability to edit without doing damge, and to avoid the one step of extra jpeg compression. As to why you might be seeing a difference between your tiff and your jpeg, hard to say. Play around!

James Emory wrote in post #11458028
Chas, what do all these settings mean, I do have the MP990. I have none selected. I don't get what these MP990 numbers related to. Nevermind, I don't know how to do what you did as to the copy screen and get it into this post. I did the copy screen OK and put it into a Word document but when I went to upload it, I got an error saying invalid file. Can you help me out here if you have time?
I wanted to show you what I was talking about here.

To get a screen shot you want to copy it into a new document in an image editor, like Photoshop or Paint. Have the document sized to a reasonable display size. You can also crop the screen shot. Then save the finished product as a jpeg so that you can upload it to here. for a screen shot you should be able to easily have a file size that is quite a bit smaller than the 300Kb max here, so you can use the Attach tool in the Advanced toolbar of this window to upload it from your computer, then to attach it to the post.

tonylong wrote in post #11458182
Well, first of all the embedded profile has nothing to do with tiff or jpeg -- it is the color space your Raw converter uses to render colors and "builds in" to your image. Checking the "Embe ICC Profile" ensures that color-managed apps will "understand" what profile was used and will be able to interpret the color values properly.

Now, tiff and jpeg are file formats, jpeg undergoes "lossy compression", tiff not (although some software offers a "lossless compression". Jpegs are 8 bits per channel only, tiff gives you the choice of 8 or 16 bits.

But, a high quality jpeg compared to the same image converted to a tiff should really match, since all the processing has been done prior to conversion. But, many of us do use tiff when we have to go into Photoshop to do further editing to keep the maximum ability to edit without doing damge, and to avoid the one step of extra jpeg compression. As to why you might be seeing a difference between your tiff and your jpeg, hard to say. Play around!



To get a screen shot you want to copy it into a new document in an image editor, like Photoshop or Paint. Have the document sized to a reasonable display size. You can also crop the screen shot. Then save the finished product as a jpeg so that you can upload it to here. for a screen shot you should be able to easily have a file size that is quite a bit smaller than the 300Kb max here, so you can use the Attach tool in the Advanced toolbar of this window to upload it from your computer, then to attach it to the post.

Thanks Tony, I'll try it out.

James Emory wrote in post #11458688
Thanks Tony, I'll try it out.

I have Paint, and I don't see anywhere where I can open a new document on Paint. Thanks anyway, I'll look in help and see if anything is there.

James Emory wrote in post #11458736
I have Paint, and I don't see anywhere where I can open a new document on Paint. Thanks anyway, I'll look in help and see if anything is there.

If I remember correctly (my God, I haven't seen Paint in 15 years) you can paste into Paint. If not, the freeware Irfanview not only can be pasted into but also has a very nice screen capture tool.

James Emory wrote in post #11458736
I have Paint, and I don't see anywhere where I can open a new document on Paint. Thanks anyway, I'll look in help and see if anything is there.

tzalman wrote in post #11458908
If I remember correctly (my God, I haven't seen Paint in 15 years) you can paste into Paint. If not, the freeware Irfanview not only can be pasted into but also has a very nice screen capture tool.

Heh! It's funny because it's been a long time for me as well. Once not too long ago I tried it out on my XP machine so I could help people who don't have Photoshop get screen shots for the Raw Conversion thread, but that was it...

Until just recently -- I've been working on a little Win7 laptop without Photoshop and a thread was active that prompted me to do a screenshot of my DPP Raw histogram for the 1D3 and I figured "why not" and so opened Paint and, oh my, there are no menus with the old-style easy to find stuff! I had to poke around on the toolbar -- the first tool on the left opens a list that includes New I believe, open the other lists and you will see things as well.

So, here's a quick approach:

First to do the PrintScn, make it easy on yourself: make sure the actual dialog you want is active by clicking it then you want to just copy the active window. On a "normal" keyboard that is Alt-PrtScn but on my laptop you use the fn key with those two. Now you have something in your clipboard.

Now, open Paint. You should see a small blank document open.

In the left side of the tool bar you will see the little tool button I was referring to but don't click that for this -- look below it and you should see Paste in bold (active) letters. Click that and click Paste and your screenshot should appear as the now active document.

Note that initially your whole shot/document will be surrounded with a selected/resize "box". At this point you can do as you wish -- to save the entire image just click that top tool button and the familiar File menu should open with Save/Save As. jpeg is in the list.

Before you save, though, consider image size. For POTN you just want to make sure your pixel dimensions are in line, 1024 at the largest, before you save something so you would look a bit on the right of the toolbar past the Select tools and see a list with Crop, Resize and Rotate -- click Resize. It will likely defaut to doing a "Percentage" resize, so click Pixels and it will show the actual dimensions and allow you to make changes on the image size.

A couple other good things to try are cropping and compositing "pieces" of one or more screenshots.

To crop, first Select All (Ctl-A) to activate the crop tool. Then you can just draw a crop rectangle. Unfortunately I can't find an "OK" function to just make everything else go away like you have in Photoshop, so I just did a Copy (Ctl-C) then, back to a Ctl-A to select the whole image then Delete. You can then Ctl-V Paste the crop in and then you can if you wish resize the canvas to fit it. To do this don't have the image selected or you will resize the image itself. If you click outside the image the selection box should disappear and the canvas should show little control/drag points so you can drag the canvas closer to the borders of the crop.

Or, you can keep the canvas larger and redo another screen shot. Just be careful working with a second shot overlaying the first one for stuff like cropping -- make sure you have the second one select, crop, copy, then again delete but only the top image so the back one will be preserved. Then you can past the new crop in and move it as you wish.

There! Believe it or not I was figuring all this out as I was typing it, so I've been tutoring myself!

tonylong wrote in post #11460511
Heh! It's funny because it's been a long time for me as well. Once not too long ago I tried it out on my XP machine so I could help people who don't have Photoshop get screen shots for the Raw Conversion thread, but that was it...

Until just recently -- I've been working on a little Win7 laptop without Photoshop and a thread was active that prompted me to do a screenshot of my DPP Raw histogram for the 1D3 and I figured "why not" and so opened Paint and, oh my, there are no menus with the old-style easy to find stuff! I had to poke around on the toolbar -- the first tool on the left opens a list that includes New I believe, open the other lists and you will see things as well.

So, here's a quick approach:

First to do the PrintScn, make it easy on yourself: make sure the actual dialog you want is active by clicking it then you want to just copy the active window. On a "normal" keyboard that is Alt-PrtScn but on my laptop you use the fn key with those two. Now you have something in your clipboard.

Now, open Paint. You should see a small blank document open.

In the left side of the tool bar you will see the little tool button I was referring to but don't click that for this -- look below it and you should see Paste in bold (active) letters. Click that and click Paste and your screenshot should appear as the now active document.

Note that initially your whole shot/documant will be surrounded with a selected/resize "box". At this point you can do as you wish -- to save the entire image just click that top tool button and the familiar File menu should open with Save/Save As. jpeg is in the list.

For POTN you just want to make sure your pixel dimensions are in line, 1024 at the largest, before you save something so you would look a bit to the right of the toolbar past the Select tools and see a list with Crop, Resize and Rotate -- click Resize. It will likely defaut to doing a "Percentage" resize, so click Pixels and it will show the actual dimensions and allow you to make changes on the image size.

A couple other good things to try are cropping and compositing "pieces" of one or more screenshots.

To crop, first Select All (Ctl-A) to activate the crop tool. Then you can just draw a crop rectangle. Unfortunately I can't find an "OK" function to just make everything else go away like you have in Photoshop, so I just did a Copy (Ctl-C) then, back to a Ctl-A to select the whole image then Delete. You can then Ctl-V Paste the crop in and then you can if you with resize the canvas to fit it. To do this don't have the image selected or you will resize the image itself. If you click outside the image the selection box should disappear and the canvas should show little control/drag points so you can drag the canvas closer to the borders of the crop.

Or, you can keep the canvas larger and redo another screen shot. Just be careful working with a second shot overlaying the first one for stuff like cropping -- make sure you have the second one select, crop, copy, then again delete but only the top image so the back one will be preserved. Then you can past the new crop in and move it as you wish.

There! Believe it or not I was figuring all this out as I was typing it, so I've been tutoring myself!

Thanks for taking the time to explain, Tony. Much appreciated. I've copied and printed. Will try it out.