Showing posts with label make over. Show all posts
Showing posts with label make over. Show all posts

Even Adobe Photoshop CS5 Can't Make This Look Better

I've got to admit it, the garage got away from me over the winter.  Some things are there for good reason.  At least it was a good reason four months ago.  In the lower left is the handle of the snow blower.  For a while it seemed like it was being used every other day for some serious snow removal.  Ten inches, twelve inches, fifteen inches.  It was a tough winter.  The reason for today's image is a promise to the wife to straighten up the garage.  I wanted a frame of reference for a starting point so, if it doesn't meet with her complete approval, at least I could show the improvement.  While I was taking a couple of shots I figured it was easy enough to fire away with the camera set on auto bracketing and see what a fringe HDR image might look like.  It turned out to be a nine shot bracket.  All nine shots were not used to make the HDR conversion.  The lightest and darkest didn't have enough information to worry about, so they were dropped.  The "proper" exposure, one two stops under and one three stops over were selected for the HDR.  I really wasn't going for any sort of exaggerated reality.  I wasn't going for any sort of reality at all.  I figured on maxing out most of the sliders in Adobe Photoshop CS5's HDR Pro.  Then the colors were pumped up even more.  The result is that I can identify every culprit that needs attention and form a plan to work toward.  For that it is a successful image.  For art?  I don't think so.  To find out just what was done, hit the "read more".

Talk about double processed, today's image was probably triple processed at least.  The first was a trip through HDR Pro.  The Detail slider was pretty much maxed out.  The Strength and Radius sliders were brought up, with the Strength slider pushed more than the Radius.  Vibrance was cranked up to about 80% and the Saturation thrown toward its upper limit.  It was then opened into CS5.
Once in CS5 the image took a trip through CS5's HDR Toning.  Detail was shoved up again and the roles reversed for the Strength and Radius.  This time the Radius was pushed harder.  More Vibrance, more Saturation brought the image to a pretty bright state.  The Exposure and Gamma were brought down to arrive at a better tonal range (like it made any difference). 
The third bit of processing was going through individual (Red, Yellow, Green, Cyan, Blue, and Cyan) Hue Saturation Adjustment Layers where each color was brought up to just below the point of going neon. 
There is a lot of false colors in today's image, but I really don't have a problem with that.  The back and left side wall is actually painted white.  Plastic bags became blue.  The cabinets look like they are made of much nicer wood than the y really are.  The floor is a cement floor but appears to possibly be a dirt floor.  All in all, it was a fun image to throw caution to the wind and move sliders all over the place.  I made have to do an insert showing the cleaned garage.  We'll see.

Pull My Strings, I'm Your Puppet (Warp)

Ever wonder what a straight, staid businessman who rides a straight, staid motorcycle does on the weekends?  Well, based on today’s image, he trades in the staid motorcycle and goes cruising on his “chopper”.  Instead of sitting upright, he leans into the wind, puts on flashier colors and lets his wild side show through.  Actually, if you take a close look at the smaller image you’ll find it’s the same image.  Today’s image started life as that thumbnail on the right.  I’ve never taken a shot at using Adobe Photoshop CS5’s Puppet Warp applet, until today.  I saw the clips about what it could do in the teaser videos before CS5 was released and thought “that’s kinda neat”, but never came across a reason to do something serious with it.  I guess the holiday break was a good time to tax my brain and look at a couple features in Photoshop that aren’t exactly mainstream.  Between Wednesday’s messing with the Liquefy Filter applet and today’s playing with its Puppet Warp I’ve been able to add a couple more arrows to the quiver.  Today’s image is about Puppet Warp, but there’s several other techniques that were used to get the end result.  To find the list (and short explanations) of what was used, hit the “read more”.

The biggie was, obviously the Puppet Warp (Edit/Puppet Warp) applet. The motorcycle was stretched out into a hog or chopper or a custom bike. Puppet Warp overlays an image with a mesh of triangles and allows each intersection to be moved independently. That doesn’t mean moving one node won’t affect the ones around it. As an example, the front forks got to be a little squirrely, bending in directions that would really weaken the bike and undoubtedly caused a structural failure. To “fix” that little problem a copy of the front wheel and a portion of the fork were selected (square Marquee Tool (M)), copied (Ctrl C) and Pasted (Ctrl V) onto its own Layer. Once there, the Free Transform (Ctrl T) applet was used to Scale, Rotate and align the wheel and fork to match up with the reasonably straight portion that resulted from the Puppet Warp operation. The rider’s legs were brought forward to more of a chopper rider’s position rather than the straight up and down angle in the original. The rider’s upper body was also bent forward to give a more aggressive pose.


An Alpha Chanel Mask was created to pull the bike and rider out of the city streets and permit replacing the background. I’ve been asked several times “why Alpha Channels rather than just a selection or erasing the unwanted areas”. That answer is easy. Once a selection is used and the worker moves on to something else, that Selection is gone. Once something is erased (yuck) its unrecoverable. A Selection (however you get it), once saved as an Alpha Channel, can to refined, adjusted back and forth, using over and over again on other Layers, uncoupled from what’s being Masked and either the image or the Mask moved independent of each other to do things like reposition clouds within a scene and anything else you can think of to do with a Mask. An Alpha Channel Mask is just so much more flexible than a straight Selection or erasing anything. Biggest tip of 2010, never use the Eraser Tool.


The background came from a trip to Lime Rock Park racetrack. It had a little red Porsche 914 zipping along, but also had about the right amount of (in camera) motion blur to make it right. As brought in, the bike looked like Bikezilla. The background had to be enlarged using the Free Transform (Ctrl T) applet. The Porsche sort of showed through around the bike, so some Clone Stamp (S) work was need to get rid of the car.
 A little shadow was put under the bike and rider and they were both “colored” to make them a little more interesting. A soft Brush (B) and colors chosen from the Swatch Panel were “painted” on the various bits and pieces needing a splash of color. Each color was applied on its own Layer and the Bland Mode changed to Color to place the colors over the detail below.


With the faux colors applied, Hue/Saturation Adjustment Layers were applied for each color (in this case: Red, Yellow, Green, Cyan, Blue, and Magenta). Extraneous color splashes were Masked out on the Adjustment Layer’s built in Mask.


Save it, close it, go to it in Adobe Photoshop Lightroom 3, put a vignette on it and reduce any noise and we were done. Hope you get a kick out of it.

And Now, For Something Totally Different...

If you’re a frequent reader of The Kayview Gallery you know that most of the images have the color saturation maxed out.  Big, bright, bold colors that make use of Adobe Photoshop CS5’s Hue/Saturation Adjustment Layers in a reasonably unique way  At least I haven’t done the way we do it being used by others.  Making a separate Adjustment Layer for each of the possible colors.  (Red, Yellow, Green, Cyan, Blue , and Magenta.)  The reason for using the ALs in this manner is not so each color can be adjusted independently, but to have a mask available for each color.  One part of an image might need a high amount of blue saturation, while other parts go neon with the same amount.  Having a mask specifically for the blue adjustment gives the ability to take away the neon blue areas.  If the blues in other areas need a different amount of saturation additional blue Adjustment Layers can be added.  This gives better control over the results.  Better than using shades of gray on a single mask.  Several images here on the gallery have had multiple Hue/Saturation Adjustment layers for one or more colors.  Today’s image takes the control of the colors through the use of the Hue/Saturation Adjustment Layers in the opposite direction.  Today’s image bounced back and forth between Adobe Photoshop CS5 and Adobe Photoshop Lightroom 3 several times.  Some things are better done in CS5 and others in LR3.  To find out what was done where, hit the “read more”.
First an admission, this is a remake of a very old image. So old that the Meta Data doesn’t even record the date it was originally taken. One of the things I did when I was at Intel was to introduce new digital things to the people in the retail stores. Our team would have all the latest gadgets to take around and amaze the sales people with “the next big thing”. We talked about USB, DVDs, processors and all manner of things as they were coming out on computers. Our team was given some of the first retail level digital cameras to play with in the stores. The first one we had was from HP and was 640 x 480 resolution. Less than one Megapixel images. Today’s image probably comes from the first digital camera I bought for my personal use. It was a three Megapixel Fuji and it was a thousand dollar camera. Today’s image might go back to the early of mid 1990s. Today’s image is a modification of the original alteration of the image as it came out of the camera. Even back then I was altering the images.

What’s different in today’s edition of the image? The bottom quarter was much darker than what’s seen today. The Adjustment Brush in LR3 was used to make targeted adjustments of the water and the rocks. The water got slightly more Brightness than the rocks. The major trees were darkened by reducing the Brightness along their trunks. The rocks along the stream at about the midpoint on the left side got a little bump in brightness, as did the sun lit area on the opposite shore.

One of the major and first adjustments was to remove as much pixelization as possible. Going to the Effects Panel and bringing the Noise Reduction sliders up to the merged the wider surfaces, reducing the appearance of noise. Was it was was actually pixelization due to the low resolution.

The sky was brought down just a touch. Anything to really bring the sky to a deeper blue resulted in red outlines defining each cloud.

An enlargement of the original hangs in our bedroom. It just might be replaced by today’s remake.

Some People's Free Advice Actually Is Worth More Than You Paid

I don’t know of any photographer who doesn’t think he/she’s done the absolute best treatment of an image that can be done. Every image is a ten. Last night I got a little comeuppance. I showed a photographer a print of one of my images (click this link to see the original). He said he loved the image, but the sky was kind of dull. He did say he’d rate it a nine, but nine ain’t ten. Since the key to the image was a mask of the trees and the fine detail of the branches I told him he could pick a sky that would be better and I’d have a go at it. He said he’d like to see bigger, fluffier clouds and not have the image look so sullen. The original intent was the sullen, slightly dark look of a cove with a storm approaching. Bigger, fluffier clouds would change the tenor of the image from foreboding to inviting. Not a bad change. The shot could go either way. Today’s image is the result of about fifteen minutes of tweaking. The hardest part of the transformation was the sharpening. I was attacking it from the wrong position. I figured I’d be quick and clever about the sharpening and it wound up that I was neither. It had me stumped for several minutes. Since I had the mask saved as an Alpha Channel I could call on it whenever I needed to mask the trees against anything that came up. Because Smart Objects come with their own Layer Mask I figured I’d take advantage of that to skip one step and sharpen the Smart Object after applying a couple of Filters (filters on Smart Objects become Smart Filters). Before converting the Layer to a Smart Object I selected the Alpha Channel Mask (CTRL click on the Alpha Channel icon). I put a slight blur on the upside down copy of the sky and added an Ocean Ripple filter. That’s where the problems started. To find out what the solution to the problems was, hit the “read more”.



With the Smart Object Mask, as soon as I did a typical High Pass Filter sharpening everything went dark. First guess was that the mask needed to be inverted. That just muddied up the clouds rather than darkening the trees. I went back and forth several times, trying different combinations with no luck. The purpose was that I didn’t want to sharpen the clouds. Sharpened clouds tend to look a little phony to me. After messing with it a time I figured I go back to the “old fashioned” two step process and put the Mask on the Layer itself. Duh! That worked fine. Trying to shortcut the process didn’t speed things up, but it did provide a valuable lesson. If you use the mask that comes with a Smart Object you’re applying the Mask to the Smart Filters. If you use a normal Layer mask, you’re applying the mask to the entire image. It’s always the simple things that can trip you up.


Speaking of Smart Filters, two were used on today’s version of the image. They both come in on the same layer, so talking about the layer should probably comes first. It’s pretty obvious, comparing the “old version” and the “new version” that the cloud Layer was imported. In order to give a more realistic impression a copy of the clouds was made, Free Transform (CTRL T) was begun and Flip Vertical was selected. Since it had the mask that went along with the cloud Layer it was unlinked from the Layer itself. The Layer was then slid down to serve as a reflection in the water. Well, reflection can’t be as sharp (in water) as what they’re reflecting, so a Blur Filter was used and blurred to taste. Flat water with the older, dull sky worked. Flat water with the new and improved (?) sky doesn’t. So, I put in an Ocean Ripple Filter. In the Filter Gallery that comes up with that and many other filters the Size and Magnitude were adjusted to give just a hint of movement.

Is the “New Version” better than the “Old Version”, or just different? Let me know what you think.

Deconstructing An Image

How an image is presented is a big deal when it comes to a recording of a scene or art. As has been talked about on the blog before, intent is the key to figuring out how an image should be displayed. In today’s image we see three different representations of what could be done to distinguish the shot. On the far left we have pretty much the straight shot. The only things that have been done are sharpening and applying a vignette. Would the image be anything special if left natural? No! It’s a shot of a flower with the seeds in focus and the back pedal going a little soft. Not a big deal. The center shows the image as a graphic, a line drawing type of treatment. This can be used to study an image, to pick it apart to determine what’s important and what can be diminished. There’s an application for this type of image. It could be used by a botanist to describe the markings on the pedals or study the shape of the seeds. If many flowers were treated in this manner, someone with the proper background might be able to discern problems in one batch over another. There could be a problem at the farm or nursery a specific number of specimens came from. It could help the grower track down what is going on at his/her facility that causes illnesses in the flowers. The center portion of the image could be used for analytical purposes. The right hand section of the image is another type of graphic representation of the original image. The might be used as a study by a muralist. Murals aren’t typically painted as continuous tone objects. Murals are meant to be viewed at a distance, especially those found outside, spanning multiple stories of a building. For the painter to even attempt continuous tones would make him/her crazy. Once an artist breaks down the image to a manageable number of colors, the job of creating the larger than life image becomes relatively easy. There are several ways to break down images for specific purposes. Being able to do the break down for clients is an art. Understanding the needs determines the amount of “art” that needs to be included in an image. To review the two techniques (center and right) used to produce today’s overall image, hit the “read more”.



The easiest method (that I know of) to get the center image would be to use the Filters/Stylize/ Find Edges filter, convert the result to B&W and use a Levels Adjustment Layer to bring the whites up and the blacks down. This control is the key to this type of structure.


The right portion of the shot can be accomplished in several ways. The easiest is a simple posterization (Filter/Artistic/Poster Edges). Pick a low number of posterization steps. Once you get to eight or above the crispness of the colors starts to go more toward continuous tone. Standing with four, five or six steps is good for “most” images.

The key is to work with the client to determine their needs. What they will doing with the alteration of the straight shot is key to what “effect” to give.

Makeover Wednesday - Adding Pop With Spot Color

Today we're revisiting an image we discussed back on July 1st. At that time the post talked about the impossible Depth of Field in the shot. What we're doing today is responding to a question asked about creating Pop in a shot by using spot color. Unfortunately, the question came without an accompanying image to work on. I went through some of the things we'd already looked at and decided the pocket knife in today's image could be made the center piece for using spot color. An additional reason for playing with a spot color situation was an offer for a tutorial and a Photoshop Action being sold online. The write up for the Action said it would take the complexity out of doing spot color. My question was "what complexity?". Lately I've seen things for sale as Actions or Plug-ins for PS that are as close to scams as I've ever seen. One was an "Action" being sold to give a "Rule of Thirds" grid on your image to check placement of points of interest. What's so hard that someone would spend money on that? Set your grid spacing to 33% (yep, you can call out percentage for grid spacing) and then use Ctrl comma to toggle your "Rule of Thirds" grid on and off. I can't see charging someone money for a setting. There's actually a secondary issue we'll tackle before we ever get to trying to make the knife sing. The plastic trim pieces on a knife like the one shown typically has a smooth finish. In the original image it appears to have a texture of some sort. We'll take less than one minute and "fix" that inconsistency.   Hit the "read more" to fiinsh today's makeover.

Here's the quick fix to remove the false texture from the knife handle. On your base image, go to the Channels Panel, click on each channel (Red, Green and Blue) to determine where the noise (that's what the "texture" is after all) resides. It might be spread over all three channels or, as in the case of today's image, it may be in only one channel. The noise in the knife handle is only in the red channel. Select the Red Channel and then Filters/Blur/Surface Blur. Use the Radius (low value) and Threshold (typically a little higher value) sliders to reduce (eliminate) the noise. The advantage of using the Surface Blur filter is that it does very little to any edges or complex details. It works only on "surfaces" with some area to them. Repeat the process in each channel having a noise issue. HINT: Use the preview that come with the Surface Blur dialog box extensively to check on how much noise is being removed. Any more than the minimum required is unnecessary.)

Now, on to the focus of today's post, Spot Color. Make a copy of the base layer. Use the B&W Adjustment Layer to convert the entire image to B&W. (If you have older versions of Photoshop, use whatever you normally use to convert an image to B&W. Desaturate (bad idea), Channel Mixer (better), or Hue/Saturation (still better), or whatever you typically do to get to a B&W image. Then, just use the Brush tool with the foreground color set to black to paint back the handle to its original color. Done.

That was the shortest paragraph of today's entire discussion, and offered several options on how to get to the B&W portion of the image. Someone wants to "sell" an Action to accomplish that? Be careful of what's available to "easy the workload" in Photoshop. Most of what I've seen are things that Photoshop already does with a pretty bow put on them.