Jump to content

custom avatars and signatures


Recommended Posts

2nd link -- for Planck's signature

I used this one over the first. The reason being is that I had thought that Wikipedia would use a more recent signature over the one he used in the earlier work, but I was wrong. Fortunately for me, Wikipedia will always want you to save an SVG as a PNG, not to say I am without skill in SVG, but editing it makes it easier. Fireworks opens this with a transparent background, which is preferable over the other version.

As you can probably tell, I had modified the color of the signature. But you may notice it isn't exactly the same color as the other green, being the parts of the face and the German text. This is because i guestimated it using the Hue/Saturation filter.

A very large JPEG (3,590x1,980px). I could obtain the outline by opening it in PS, putting a transparent layer above it, and then "tracing over" the image on this transparent layer, and then reducing it to the appropriate size, and manipulating the fade to white at the bottom. This would be a very tedious, labor intensive process (especially tracing it rather accurately) I wonder(?) how Tripredacus got the image. He said he didn't spend *much time* on his picture, so I'm sure he didn't use this approach.

First, I knew that the image was too large for a signature, so I used an image resize on it. At first I did a whole resize, and then used Transform to my liking. As for getting rid of the white background, no tracing was involved. Since the image was so large, it was likely originally some other format, which seemed to be true. I merely used the magic wand with some varied levels of tolerance to remove the white. And of course, don't forget the "select similar" which doesn't always work but did in this case.

As far as the unused images, this is common in my work on building an image. I will take in multiple different images before determining what i use. In this case, I felt it fine to leave these un-used resources in my post to show you them. It is only because at first I developed the image list before I actually got a look at everything and got an idea.

If the image were traced over at the large size, it wouldn't have to be *too accurate*, since the size reduction would tend to "smooth" things out.

The background was removed prior to shrinking the image, as outlined above.

2) the roman head logo thing, perhaps with a linear gradient on its layer mask to fade its bottom part (although the white overlays from layers 3 to 5 may suffice)

The Gesellschaft used glow at 5 height, 87% opacity, 12 softness, 0 offset. The signature has different settings as it didn't need as much to stand out.

It might look somewhat different than your image, but "Spice is the Variety" of life, or something like that (?).

You can do what you want with the image. That's why I made it. ;)

I just have the tendency to look at how others did it first, then figure how I can do it better. You tend to try to figure out how to do it first, then see how someone else did it afterwards for comparison. Neither approach is wrong, just a difference in approach.

My approach is to do it by accident at first, and not be able to recreate it at a later time if I tried, because I've forgotten how I did it. :D

As noted in previous placed (although I was initially mistaken about how PNGs work) I use Fireworks MX 2004. I have uploaded the source file, which can be opened in (at least) this version of Fireworks and retain all layers. It could possibly be opened in Fireworks 2 (I've had some success at backwards compatibility, but I'm too lazy *myself* to boot up Shiva). :whistle:

http://i579.photobucket.com/albums/ss236/tpx_bst/Designs/Untitled-1.png

Link to comment
Share on other sites


bphlpt

Nah, I just have the tendency to look at how others did it first, then figure how I can do it better. You tend to try to figure out how to do it first, then see how someone else did it afterwards for comparison. Neither approach is wrong, just a difference in approach.

Not only are you never wrong (as I previously indicated), but you are quite the analyst, too !

Is there no end to your talents? (lol)

Tripredacus

I'm dying to redo -- from scratch -- your "powerfully delicate" image just to see what I can come up with.

I'll surely Post the results.

It will be a couple of days before I can continue with it.

(I have some other important business to take care of first.)

2nd link -- for Planck's signature. I used this one over the first.

Thanks for the info.

I will use this one also.

I haven't yet started on removing the background on the signature.

As you can probably tell, I had modified the color of the signature.

For sure !

Changing the color of anything is easy.

What I do is put the new color -- in a big size -- in the layer over what I want I want to change the color of, and then "Alt-Click" exactly in-between these 2 layers, and then link and merge the 2 resulting layers.

A long, long time ago, CoffeeFiend said this technique was called "Clipping Paths".

I merely used the magic wand with some varied levels of tolerance to remove the white.

The magic wand is certainly MAGIC!

I've already removed the background of the large JPEG (at the 3,590x1,980px size) with Mr. Wand, and what pixels He didn't remove, I removed with the Eraser.

I got an absolutely *perfect* job!

Next, I have to shrink it down to the size to fit into the final 380x100 image.

Can you or CoffeeFiend, give me some tips on reducing to maintain a *sharp* image?

What I read on the net a while back said that you shouldn't shrink it *a lot* in 1 step.

It said to shrink it by 50%, again shrink it by 50%, again shrink it by 50%, again shrink it by 50%, etc. until you get close to the final size. Then reduce it to the final size in the final step. (This final reduction very probably won't be by 50%. It might be 70%, 85%, etc.)

So, that's what I've been doing.

This approach seems to work well.

As far as the unused images, this is common in my work on building an image. I will take in multiple different images before determining what i use. In this case, I felt it fine to leave these un-used resources in my post to show you them. It is only because at first I developed the image list before I actually got a look at everything and got an idea.

Yes, I kind of do this also.

When one has *more than enough* images, the *proper ones* seem to coalesce -- as if by magic -- into the final image, or they "synergize" into creating new ideas.

The Gesellschaft used glow at 5 height, 87% opacity, 12 softness, 0 offset. The signature has different settings as it didn't need as much to stand out.

I will start with those settings.

Thanks for that info.

Yes, the signature, by far, uses the darkest green, so it stands out the best with an outer glow around it.

I also think that is the best "treatment" for the signature, since that is kind of the "focal point" of the image.

You can do what you want with the image. That's why I made it.

Many, Many Thanks, Tripredacus.

I'm really looking forward to "constructing" -- (and, hopefully, not "destructing") -- my version.

It's going to be a lot of fun !

My approach is to do it by accident at first, and not be able to recreate it at a later time if I tried, because I've forgotten how I did it.

Boy, I sure know what you are talking about here !

That's why I take it *slow and easy* and LABEL each layer in the .psd file with as *much info* as possible.

Even with this, if I haven't looked at the .psd file for a few weeks, it's often a little difficult to reacquaint myself with what I did.

Honestly, because of this, I often feel like I have pre-Alzheimer's ! (lol)

Many, Many Thanks for Posting this *GEM* ! !

Edited by larryb123456
Link to comment
Share on other sites

The magic wand is certainly MAGIC!

:lol: You should see the selection tools we've got now, like the the quick selection tool and refine edge which is much more magic than the magic wand tool (makes it easy to select things like hair). If you want actual magic, then you should see things like content-aware scaling. Or if you want black magic, check out this

, specifically around the 30 second mark (the image recomposition part). That part didn't make it in time for CS5 though (but the seam carving bits you'll see did)

Can you or CoffeeFiend, give me some tips on reducing to maintain a *sharp* image?

Well, I'm not using such an old version of Photoshop, so I work in completely different ways. To scale something down like that, I typically use smart objects. It gives you the flexibility to scale down, and scale it back up without any loss whatsoever, and it looks great without any fancy tricks. You can literally take something that's like 1000x1000px, scale it down to one single pixel, then back up to 1000x1000 and you end up with the exact same pixels as you had before (it stores how big you want to see it and where instead of destroying the original pixels)

As for Tripredacus, he's using a web designer's tool (Fireworks, not Photoshop) which works in different ways, and also from a different era, so that probably wouldn't really help that much either.

But if you want to do some really fancy sharpening, then there's a LOT of possible tricks one can use. From sharpening in Lab mode (in the luminance channel), sharpening with a high pass filter, you can do selective sharpening with layer masks, there's newer filters that are better than unsharp mask like smart sharpen. Also, using such a filter on a smart object (i.e. using smart filters) lets you edit your sharpening settings later on (you sharpened too much or too little? just tweak the numbers again), there's several 3rd party tools and plugins for sharpening, etc. Also, the settings depend on the source material, what you're doing with it, which medium you're outputting to (i.e. you sharpen more for print than on screen), etc. Also, different programs (like Adobe Camera Raw) don't work quite the same when it comes to sharpening. Also, the amount of noise present in a picture can change your sharpening quite a bit. There's a lot more to discuss about sharpening alone than what one can share in a forum post unfortunately. But no, I don't do that "resize & sharpen a bunch of times" thing.

Link to comment
Share on other sites

Hello, CoffeeFiend:

Thanks for exposing me to the 21st century (my PS 5.0 was from the 20th century).

You should see the selection tools we've got now.

I watched your video 6 or 7 times, asking myself, "Could I achieve the same result with my PS 5.0 work-arounds ?"

My *basic approach* in making a .psd file is dragging a layer that I want to modify into a "New File" (most always with a transparent background), making what changes I want in this New File -- (i.e., resizing, removing background, etc.) -- and then dragging the resultant layer *back into* the file I'm working on.

Consider how I'd approach the case shown in the video, where the sister in the swing -- (at the left) -- is to be moved closer to her brother on the right:

For starters, I'd open the original JPEG, drag it into a New File -- (with a transparent background) -- and then close the original JPEG. (That way, I'd *always* have my original image, unaltered.)

I'd then duplicate this layer in the New File.

On this duplicate layer, I'd *obliterate everything involving the sister*, by mostly using the Rubber Stamp Tool, with the appropriate setting(s) in the Brushes Palette.

The Line Tool is also useful (IMO) in this process, also.

Now, we need to turn attention to removing the background around the sister, swing, and chain holding the swing up.

What I usually do in a case like this is use the Eraser Tool, with Options: "Pencil", "Opacity=100%", and "Fade=0".

I quickly go around the image with the Eraser Tool (but, of course, trying to be as close to the image as possible).

But, of course, after this, the outside of the image still has an "aliased" (i.e., stair-steppy) look.

Next I use a trick I developed myself -- (I honestly don't know if this "trick" is common knowledge or not. It's so simple, it probably is common knowledge.) -- to turn the aliased outer edge into "anti-aliased".

I go around the outer edge -- (a couple of pixels at a time) -- using the Line Tool, with Options: "Normal", "Opacity=100%, Weight=1px, and, of course, with the "Anti-aliased" box checked.

As I go around the outside edge of the image with the Line Tool, I sample the color just inside the image at that point, and use that color in my 2px (or so) anti-aliased straight line.

This technique is very, very fast, and essentially it is based on the Fundamental Principle(s) of Calculus, which state that any curve of any shape can be well approximated by a series of straight-line segments, as long as the length of these segments is short.

To get the *exact* vertical shape of the 2 chains -- (with all the *exact* details) -- I'd probably just crop each chain out -- (as close to the "verticals" as i could get) -- of the image in 2 "New Files" (on transparent layers), and drag the crops back into the original file and position them where they should be.

All of this work sounds "long-winded', but this is the way I'm used to working, and what I've described wouldn't take me very long at all to do.

I typically use smart objects.

But, I don't think *smart* objects would be compatible with *dumb* people ! (lol)

But no, I don't do that "resize & sharpen a bunch of times" thing.

You should.

It would allow you to visit Prehistoric Times, and see the way the Cavemen used to do it.

Many Thanks for this *Instructive Post*, CoffeeFiend.

The video was definitely interesting to watch, but I didn't understand *all* of it, with all the advanced terminology, and especially with the speed at which it went.

However, I did get a strong sense that I could more or less duplicate what was shown using Prehistoric PS 5.0.

Again, Much Appreciated !

P.S.,

Could you please answer this for me ?

It concerns the .psd files I made for bphlpt.

You said that PS 5.0 couldn't even open .psd files made by more-advanced versions.

(I'd guess that's because PS 5.0 would have no way of knowing what the "features" of the more-advanced versions are.)

Finally, my question:

Can much-later versions of PS open PS 5.0 .psd files ?

My guess is that the answer to this question is "Yes".

Kind of like: "Stephen Hawking can understand a 4-year-old -- (when the 4-year-old talks about *anything*) -- but the 4-year-old can't understand Stephen Hawking, especially when he discusses Physics".

Edited by larryb123456
Link to comment
Share on other sites

Could I achieve the same result with my PS 5.0 work-arounds ?

Than the seam carving stuff? No. There is no way to selectively eliminate pixels based on their "frequency" (eliminating rows of the "least important" pixels, that aren't actually in a straight line), while keeping straight lines straight and so on (and simply resizing it will look horrible). I couldn't in a day's work accomplish something as nice as it can do in mere seconds. There were a siggraph presentation that explained the technology behind it if you're curious. But that's just a tiny, minuscule part of the new features anyway ;) We got real 3D editing (like a typical CAD program), vector content, content-aware fill, smart objects and nested smart objects, smart filters, advanced typography, some new and useful/powerful filters like vanishing point and liquify, HDR merge, GPU acceleration, etc. It's from a different century indeed.

You said that PS 5.0 couldn't even open .psd files made by more-advanced versions.

(I'd guess that's because PS 5.0 would have no way of knowing what the "features" of the more-advanced versions are.)

Precisely. The older PS versions can't make sense of the new format, and the new information in it.

Can much-later versions of PS open PS 5.0 .psd files ?

Yes. They open just fine (much like your analogy that followed this).

Link to comment
Share on other sites

As for Tripredacus, he's using a web designer's tool (Fireworks, not Photoshop) which works in different ways, and also from a different era, so that probably wouldn't really help that much either.

I had used Fireworks in CS4 last year and found it had barely any noticeable differences except for the GUI. I'm not even sure why Adobe decided to keep it around. :unsure:

Link to comment
Share on other sites

A while back, in my Post # 136, I discussed my modification of e-t-c's transparent GIF avatar.

The link to his original avatar is:

http://postimage.org/image/4u0c659hd/

The link to my modified version at *that time* is:

http://postimage.org/image/lc3vagpo9/

My *primary objective* at that time was to straighten up the "chopped-off" look of the original, by using *perfect* ovals for the saucer edge and the lip of the cup. I kept the outlines of the modified version as dark and as bold as in the original, to have "harmony" between the "old and the new". I also added some "vapor molecules" above the center of the cup.

e-t-c was very happy with that modified version, so I basically put the image out of my mind (as in "case closed").

However, I saw my modified avatar today on one of e-t-c's Posts, and I saw how I could improve it even more.

In the original and modified version, the saucer and the cup *merge together* in an unsatisfactory way.

So, I corrected that in my second modification to e-t-c's avatar. (I just put a light-colored outline around the cup.)

The link is:

http://postimage.org/image/c6ia5ngbp/

As you can see, I left *everything else* alone (bold/dark outline, etc.).

But, on thinking about it some more, I felt that the bold/dark outline was completely out of place in the picture, since the saucer and cup are very delicate and light-colored images.

So, in my third modified version of e-t-c's avatar, I made a "totally compatible" and "unified" image (IMO) in which I used a thin, light-gray outline around both the cup/saucer unit and the cup sitting on the saucer.

The link for this version is:

http://postimage.org/image/zbsornlgl/

I think this transparent GIF is far superior to the original image and my other modifications.

P.S.

If you look *carefully* at the outer edges of this transparent GIF, you'll notice an aliased (i.e., "stair-steppy") look. All outer edges of an image have to be aliased in order for GIF89a Export to successfully convert it to a transparent GIF. But, in the "grand scheme of things", I don't feel that this look is really objectionable.

Edited by larryb123456
Link to comment
Share on other sites

In Post # 176, e-t-c said he preferred the dark outline around the cup/saucer unit rather than the lighter gray outline in my third modification, shown in my Post #175.

I still felt that the *boldness* of this dark outline, shown in

http://postimage.org/image/c6ia5ngbp

was out of place with the delicacy of the cup and saucer, so I made the outline smaller (i.e., not as thick), as shown in my fourth modification:

http://postimage.org/image/cadxrehjv/

(Since this outline is basically 1 pixel thick, I really can't make it any thinner than this.)

*By far*, this fourth modification is my favorite version.

I really like the way the dark outline makes the cup/saucer "unit" stand out well against the MSFN-colored background, and the way in which the dark outline doesn't overpower the light-gray cup outline as much as before.

I wouldn't have made this image had it not been for e-t-c's comments in Post # 176, so, in that regard, I feel that e-t-c is something of a *co-creator*.

Also, the interesting thing about Art images is:

You might think you have an image *absolutely perfect*.

But, then you make a different version, and it turns out to be *far superior".

Such was the case for me in seeing that modification 4 was far superior to modification 3.

Edited by larryb123456
Link to comment
Share on other sites

So happy you like this one, e-t-c.

It's my fave also.

This little "project" of refining your avatar to get the *best* version really shows the power of collaboration (i.e., the back-and-forth between us).

I'm really glad I re-started on improving your avatar, after so long a time in which I'd considered the avatar finished !

The work was really fun to do.

Edited by larryb123456
Link to comment
Share on other sites

Hello, EVERYONE:

I thought I'd better make this Post, because I **changed (big time!)** a great many of my Posts (previous to this one) which involved making custom signatures and avatars.

I figured that if I didn't give this "HEADS UP", many of you who had already read the old Posts would have missed the *new versions*, because you probably would not have gone back and done some re-reading.

To bring you *completely up to speed* with what it is that I'm talking about, all you need to do is re-read Posts # 137, 140, 144, 147, and 153.

I sure hope that you will enjoy the changes that I made. I also hope that you will enjoy looking at all the pretty (IMO) pictures.

Much appreciated !

Thank you.

Sincerely,

Larry

Edited by larryb123456
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...