I’ll repeat:
" Ahhh, OK.
I got the original heron image, and the reason you’re having problems is that it’s a CMYK JPG.
Convert to RGB, then resave for web. "
Comcast’s dumb-ass servers probably choke on CMYK JPEGs.
Alright, let the learning begin. The results on the black background look very good. BTW, thanks for finding the heron RGB problem. It wasn’t an issue on the gifs because before PS would let me erase the background on the logo, I needed to convert it from indexed to RGB. I got in the habit of doing the same thing with the heron before starting to work on it. Oh well….
To start, the black background on the site I’ve been using for the postings is there just for testing these images. Since they were leaving a white halo, black seemed to be the best. I haven’t decided yet what the production site will have as a background so part of the reason for learning all this is so I can make the adjustments as needed in the future.
Let me go through the process I can piece together based on all the posts in the thread. For this summary I’ll use "logo" to mean the brush lettering image and "heron" for the heron image. BTW, these are both images we had a graphic artist do for us and the intent is to use them on the web and in printed output. We already have them on our existing site but as jpg files so they have a white rectangular background. I’d prefer to have control over the background.
Keep in mind as we go through this that I’m fairly new to image processing (can take a digital photo and play around with it, but not this kind of detail). I even have to look up words like antialiasing as part of my homework.
-Preparing the JPEG file: Both images start out as jpg files. The logo’s mode is indexed while the heron is CMYK. Since PS won’t allow background removal in indexed and CMYK is for print not the web the first step will be to resave the files (at least the versions destined for the web) in RGB. Right? Is this also the time to make the white selective color image adjustment posted earlier?
-Save as the new jpg: I’m also assuming that the save is set in a way for optimal quality so I set the image quality setting to 12. There are other options in the save as dialog for jpg. Mine has the default of format option to baseline (standardized). Is that right? Anything else I should learn about at this stage?
-Save as a gif: Is this the right point to create a high quality gif? Like with the digital photos I mess with sometimes (there are photos on the final site) I tend to save "masters" as I go along adjusting the images. I’m assuming that is the same technique for this gif file. In other words, the file would be saved with total emphasis on quality not size for now. I’m assuming the matte is applied when I do the final save for web(?).
-Remove the background: After the save the technique I used to remove the background was to magic wand in the background area (outside the image content) and then use the background erase tool (set to a size that covers the whole image) touching in the background area. I then inverse the selection and bg erase the areas inside the "e"’s.
-Correcting the color: I don’t think I will mess with this but I’m assuming that there may be subtle color differences resulting from the mode change(?) What’s the approach here (assuming I have the step described right)?
Am I close?
Tom
If you had a graphic artist do them for youand you plan on using them for web as well as printthey exist somewhere as vector files, right?
Having the vector originals could render most of the process you’re hashing through virtually unneeded.
Oh, and just FYI…
I used the Mac-only plugin "Peel Off White" from <
http://www.bergdesign.com/plugins/>
The heron was just a wee bit more hinky, because after the RGB conversion "Peel-Off White" left some of the color translucent.
If the graphic artist didn’t create them originally as vector files, s/he should redo them as such.
I’ll contact the graphic artist. She did the work a few years ago so hopefully she’s still around.
The great news about Peel off White is that I have PS Elements on my Mac (hopefully it will work with Elements). Full PS on my PC laptop.
I’m going to go ahead and change the web site background to white now to see your images against that color. Did you use both POW and the matte feature when the image was saved for web?
Tom
I matted using black, so the GIFs I created won’t work against a white BG.
Our graphic artist is still alive and well. She created the originals in Photoshop so they aren’t vector. She will either send the original psd files or will do the background erase for us.
Tom
Either way, Thom…
Pay attention to that technique I posted about Selective Color in post # 7.
It’s good for cleaning up dirt in the white areas of images with JPEG compression and in the white background of scanned drawings.
Thanks again
To make sure I got the most from all this it seems that there were four topics of use: one was that some of the halo was produced when the jpg compression happened (avoidance: save max quality and never do more than one compress), the second was the freebee Mac plugin used to clean up the edges, the third was the selective color cleanup (which was done on the image before the first save), and finally the matte work done when the image was finally saved for web.
Assuming that’s correct I got the guidance I needed here. What I will need to do is to understand how each is effecting the halo since ultimately I will want the image to be background neutral as much as possible. For the lettering (except for the flames) this will simply be a matter of using the color in the lettering for all the matting, etc. For the heron, things will get a bit more complicated and I’ll probably have to pick a background before doing the matte work.
Thanks,
Great help.
Tom
I started the learning by reading about the selective color adjustment. I understand, sort of, what’s going on. In the adjustment posted in the thread, all CMYK was removed from White. I hear the words, but….
First, the dumb question: does anything in the image need to have been selected before using the command or, it there is no selection will the command apply to the whole image? I’m assuming that I can have it either way (i.e. if no selection the adjustment is done on the whole image, if I have done some selection I get that area matted out of the adjustment.
Now to what’s actually happening. I think the basic question I have is how are the colors being recognized? The colors include RGB and CMYK plus white and neutral. On the logo white, red, yellow and magenta were "found". Very, very cool what this function allows you to do, btw. I’m assuming it’s here that color correction can be done.
Back to the question… How does it know what a magenta is (for example)? I understand quite a bit about the hex and RGB representations of color. Does it have to to with the hex for the color? For example, is a red anything with a #FFxxxx pattern?
Finally, to the adjustments made for the white. Obviously the adjustment was to remove all variations from pure white from the image. But why? I’m assuming it has to do with something else being done later (perhaps the background erase?).
One of the cool things about the tool was that I was able to select White and then run all the sliders to the right (black). Here you can see the halo quite clearly. I’m assuming that halo are pixels that for some reason or another PS is not seeing as white. ?????
I’m spinning a bit here…
Tom
