Create an Awesome 3D Anaglyph Poster
In this tutorial, we'll take inspiration from these movies of yesteryear and recreate a retro B-movie styled poster. I'll walk you through entire creation process, step-by-step using photography and 3D renders. Then we'll take a jump into the third-dimension — by adding effects only visible when viewed through anaglyph 3D glasses. So hold onto your seats, because stunning sights really will leap at you in 3D: just as the poster says!
This tutorial requires the following resources:
And you'll need a pair of 3D glasses which you'll find supplied with 3D comics and DVD's, or available online.
Before the advent of computers, the movie industry relied on stop-frame or stop-motion animation to bring their monsters to life; special effect artists such as Ray Harryhausen used this technique on classics such as "One Million Years BC" and "The Valley of Gwangi". The 1950s also witnessed a 3D movie explosion — and Hollywood's recent 3D revival means moviegoers of today can experience the next generation of 3D technology.
What is a Stereoscopic 3D Image?
Anaglyph images are used to provide a stereoscopic 3D effect, when viewed with 2 color glasses (each lens a chromatically opposite color, usually red and cyan). Images are made up of two color layers, superimposed, but offset with respect to each other to produce a depth effect. Usually the main subject is in the centre, while the foreground and background are shifted laterally in opposite directions. The picture contains two differently filtered colored images, one for each eye. When viewed through the "color coded" "anaglyph glasses", they reveal an integrated stereoscopic image. The visual cortex of the brain fuses this into perception of a three dimensional scene or composition.
Create a new A4 portrait document (21cm x 29.7cm) at 300dpi in RGB Mode and choose White from the Background Contents drop-down menu.
Ensure you've got Snap and Snap To > All selected under View. Now pull in some guides, snapping them to all four edges of the canvas.
We'll need a generous amount of bleed for offsetting the 3D anaglyph effect; so go to Image > Canvas Size (Option + Command + C), then add 2cm to both the Width and Height fields (23cm x 31.7cm). Next, highlight the central Anchor point reference and set the Canvas extension color to White.
This is a great technique to quickly add both bleed and guides; it's also cuts out the calculation work of manually placing guides via the View menu.
Drag the street scene image in as a new layer, hit Command + T to Transform and position as shown. Don't worry about filling the bottom of the canvas — if you refer to the final design, you'll see it's covered by the orange bar. Now label the layer "City."
Over the next few steps, we'll be dropping in a more dramatic sky, so we need to isolate the buildings. First, switch to your channels palette to determine which holds the most contrast — in this instance its the Blue channel. Duplicate it by dragging its thumbnail into the Create New Channel icon. Next, hit Command + L to access the Levels dialogue box and use the settings shown.
Now use a large black, hard-edged brush to fill in the buildings and a white brush on the sky. Some areas will be impossible to mask, such as the gantry above the right tower. Remove by plotting a path with your Pen Tool (P) using the Paths option.
Next, generate a path-based selection and fill with white. You can also use the Dodge and Burn Tools (0) — set to Highlights and Shadows respectively to achieve a clean mask.
For complex areas, it's best to toggle the visibility of the top RGB composite channel as you work. Now draw further paths, make selections and fill with white to delete more unnecessary detail on the channel.
When your channel mask is complete Command-click its thumbnail to generate a selection (by default, this will select the white areas) and hit Delete. You can now delete your extra channel.
Drop the "Sky.jpg" in as a new layer beneath it, Transform and name it "Clouds." Finally, target your "City" layer and go to Layer > Matting > Defringe by 2 px — this will remove any light-colored pixels which may have been included in the original selection.
The channel mask has made a neat job of removing the original sky, but there's still some odd color halos remaining on some of the edges. Go to Image > Adjustments > Replace Color, then select the unwanted colors with the eyedropper and adjust the Replacement and Fuzziness sliders accordingly. You may need a few passes using this technique to remove all the unwanted colors.
You can also use the Sponge Tool (0) set to Desaturate on any remaining edge halos.
Add a new layer above the "City" and use the Clone Stamp Tool (S) set to Current & Below to remove the figures from the scene. When you're done hit Command + E to Merge Down.
Next, we'll add the 3D creature render to the scene; open "Monster.jpg" from the "source" folder, then add a new channel. Open "Monster_mask.jpg" also from the "source" folder and Select All > Copy. Now target the new channel in the "Monster.jpg" and Paste.
Generate a channel-based selection, target the composite channel and Copy to the clipboard.
Paste as a new layer in your working file at the top of the layer stack and label it "Monster." Defringe by 2 px, then Transform and position as shown.
To allow space for the base graphics, enable your guides by pressing Command + Semicolon, then Command-click all your layer thumbnails and Shift-drag them to the top of the canvas.
Next, we'll create the effect of falling building debris as our beast rampages the streets. Shift-drag "Destruction_1.psd" from the "source" folder in as a new layer, name it "Explosion 1" and position below the "Monster" layer. Duplicate it, add a slight Motion Blur and rename it "Explosion blur 1." Now set its Blending Mode to Soft Light.
Now for some falling debris in the foreground; Shift-drag "Destruction_2.psd" also from the "source" folder as a new layer, name it "Explosion 2" and position above the "Monster" layer. Duplicate, add a small amount of Motion Blur and rename it "Explosion blur 2." This layer can now be set to Lighten.
To integrate the beast into the scene we need to give him a shadow; add a new layer beneath the "Monster" and label it "Monster shadow." Command-click the "Monster" layer thumbnail to make a selection and fill with black on the "Monster shadow" layer. Hit Command + T, then right-click and select Flip Vertical.
Now Shift-drag the layer content down and drastically squash it. Transform again, this time selecting Flip Horizontal. Transform a final time using the Skew option and pull the bottom middle handle to the right. This may take a few attempts to get it looking right — it's also helpful to use the shadows on the "City" layer as a guide.
Use your artistic licence to fill any areas with black using a small, hard-edged brush. Also, feel free to remove areas with the Eraser Tool (E) using a small, hard-edged brush.
Set the "Monster shadow" to Multiply and reduce its Opacity to 58%. Next, give it a Gaussian Blur, again gauge the amount of blur you use to the shadows on the "City" layer — I used a small amount of 1.5 px.
Reopen the "Sky.jpg" and hit Q on your keyboard to enter Quick Mask Mode. Now select a large, soft-edged brush to paint over the central area. Exit Quick Mask Mode by pressing Q again and Copy the active selection to the clipboard.
Paste your selection as a new layer below the "Monster shadow" and name it "Smoke." Now set it to Screen and lower the Opacity to 45%. This layer now blends in nicely and has the effect of smaller particles of debris dust. Next, highlight all your layer thumbnails and drag them into a group folder called "MAIN SCENE."
We now need to add a group of people fleeing from the monster. Open the first runners image. Now clip the first runner from the background with the Pen Tool. This stage is quite involved if you're not familiar with how the Pen Tool operates — but it's worth taking the time and effort here.
You can fine-tune your path at any time by holding the Command key to access the Direct Selection Tool to adjust individual direction/anchor points as required. Also, remember to select the Subtract From Path Area to plot the inner sub-path on the figure's arm. When you're done save the path by double-clicking its thumbnail.
Next, click the Create New Path icon and clip the second figure to keep the paths independent. Now repeat this step for the third figure.
Use your paths to Copy > Paste the figures as new layers within a group folder called "SMALL FIGURES" at the top of the stack. Transform and position them in the foreground, then Flip Horizontal so the light matches the main image. Now add some more figures from the second runners image to the right and distance them slightly as shown.
Apply the same shadow techniques as detailed in steps 14 — 16 to the first figure, but leave its Opacity at 100% for the moment.
Continue adding shadows for the remaining figures — remember to use the existing shadows as reference, so they look convincing.
Now adjust the Opacity of each shadow layer to suit — I set the nearest to 55% and the furthest to 40%, then marginally tweaked the ones in between.
Now's the time to tweak elements and apply adjustments. First, reposition your "Clouds" layer by moving it down to the bottom of the canvas. Next, press Option while selecting Levels from the Create New Fill or Adjustment layer drop-down menu. Check Use Previous Layer to Create Clipping Mask in the next window and hit OK. Now enter the settings shown below.
Use the same technique to clip a Hue/Saturation adjustment layer to the "Monster" layer. First, change the Master setting, then alter the Reds and Magentas from the Edit drop-down menu as shown.
Clip further Levels adjustment layers to each of your running figures. Adjust each one accordingly so figures within the creature's shadow are much darker — this is best achieved by pulling the white Output slider to the left.
Add a normal Hue/Saturation adjustment at the top of the stack to affect all layers.
Next, add a another normal Levels adjustment at the top, again affecting all layers.
For this step, we'll use an advanced blending technique on the "Smoke" layer. Double-click the thumbnail to access its Layer Style, then Option-drag the left half of the whitepoint Blend If slider to 153.
Finally, feel free to duplicate any of the running figures and Transform/Flip Horizontally as I've done.
Many low budget B-movie posters used artistic impressions to draw in the audience, these scenes were more-often-than-not misleading and usually superior to the actual film's special effects!
A quick way to reproduce a painted look is by using the Median Filter. Start by adding an empty layer at the top of the stack, then go to Image > Apply Image and hit OK in the next window. You'll now have a merged composite layer which you can label "Merged." Next, use the fly-out menu at the top of the layers palette to select Convert to Smart Object. Now go to Filter > Noise > Median and enter 5 px.
Since we've used a Smart Filter, it can be adjusted by double-clicking its name below the layer thumbnail.
Set the Rectangle Tool (U) to Shape layers and pick # f49f22 for its fill. Now add a bottom block as an uppermost layer as shown.
Now to add our hero and heroine; Open "Models.jpg" from the "source" folder. I chose this image because both models are improvising. I also like how the role reversal injects a little humor.
Next we'll isolate the figures using a channel mask as we did with the city at the start of the tutorial. Duplicate the blue channel and carry out a Levels adjustment as shown.
Now use the Burn Tool set to Highlights around the edge of the figures as indicated.
Set the Dodge Tool to Shadows and lighten the background gray around the edge of the figures.
Finally, clean up the channel mask using hard-edged black and white brushes.
Generate a selection from the channel, then hit Shift + Command + I to inverse it. Now select Modify > Contract by 1 px. Target the top composite channel, then switch to your layers palette and hit Command + J to float the selection as a new layer. Next, defringe by 3 px, then sandwich a neutral gray layer (I used #3b3b3b) to check the figure edges.
As you can see their hair needs some attention which we'll fix after this step. First, clip a Levels adjustment layer to the models and use the settings below.
Now set the Burn Tool to Midtones and use a soft-edged brush at 50% Opacity to darken the edges of the models' hair.
Target your Levels adjustment layer and apply Merge Down. Drag the layer into your working file at the top of the stack and name it "Main Figures." Now convert it to a Smart Object and Transform/position bottom-right as shown.
Clip an Exposure adjustment layer to the "Main Figures" to boost the contrast.
Create a new folder labelled "MAIN TITLE" above "Exposure 1." Add your main title within the folder using the Comic Script font at around 85 pt. Adjust its tracking to -60, then carefully reduce the kerning between some of the character pairs, such as the "O" and "R."
For a more detailed explanation of tracking and kerning have a look here.
Once you're happy with your spacing, go to Layer > Type > Convert to Shape. Now hit Command + T, then right-click to select Perspective. Now pull the right, top or bottom control handle to distort the text. Remember, once your type has been committed to a Shape Layer you cannot edit the text.
Duplicate the Shape Layer and position beneath the original, then double-click its layer thumbnail and change the fill to black. This layer can then be nudged down and right using the arrows on your keyboard to create a shadow. Now select both layers and rotate/resize as shown.
Add a mask to the "MAIN TITLE" folder, then Command-click your "Monster" layer from within the "MAIN SCENE" folder to generate a selection. Now use a black, medium-sized, hard-edge brush to reveal the monster's jaw.
Add another folder labelled "3D-DIMENSION" and type the text as shown using the Headline One font. Adjust the tracking and kerning, then convert to a Shape layer. Transform using the Distort option, pulling the top-right control point down, then drag the middle right control point to the left to squash it.
Add a shadow below it as you did for the main heading. Duplicate the layer again and position at the top. Now change its fill to #f49f22, then target the Vector Mask thumbnail. Now grab the Path Selection Tool (A) and click-delete all the type paths excluding the 3-D letters.
Add another folder and name it "SMALLER TITLES." Set the remaining text on seperate layers using #f5da64 and white with the Headline One font.
Now let's add the famous Psdtuts+ logo! Go to File > Place, then navigate to the "PSDTUTS_logo.ai" from within the "souce folder." The logo will now import as a Smart Object layer - which you can position bottom-right.
As were recreating a 1950's style poster, let's age it slightly by reducing the brilliance of the whites. A non-destructive way of achieving this is to add a #fff200 Color Fill layer at the top of the stack. Now change its Blending Mode to Multiply and reduce its Opacity to 15%.
OK, that's our poster finalized — now we'll move onto creating the 3D layers. Apply an uppermost merged layer (using the same method as step 26) and name it "Green 3D." Duplicate this layer and rename it "Red 3D." It's important that these two layers stack as shown for the 3D effect to work. We'll be offsetting the "Red 3D" layer next as well, so its a good idea to enable your Guides.
Target the "Red 3D" layer, then access Levels. Use the drop-down menu to select the Green channel and change the Output whitepoint to zero. Don't hit OK just yet though…
Now select the Blue channel and set its Output whitepoint to zero and click OK.
The "Red 3D" layer can now be set to a Blending Mode of Screen to make the "Green 3D" layer visible.
Next, target the "Green 3D" layer, access Levels again and set the Red channel's Output whitepoint to zero. I've disabled the visibility of the "Red 3D" layer so you can check your image matches the screenshot.
Now for fun part! Ensure the Move Tool (V) is highlighted and the "Red 3D" layer is targeted. Nudge the layer content by tapping the left keyboard arrow ten times to the left. The result looks pretty dreadful right? Now pop your 3D glasses on and you'll notice a slight 3D effect, but nothing too brilliant. This is because the 3D layers apply to the entire image — to make it pop, we need to contrast selective areas using a mask.
Grab the Crop Tool (C), and snap to your guides and accept the Crop prompt — this will delete the right-hand mis-registration from the previous step.
Drop the "Green 3D" and "Red 3D" layers into a new folder called "3D" — maintaining their stacking order. Now target the folder and go to Layer > Layer Mask > Hide All. Set your Foreground to black, then generate a selection from the "Monster" layer and hit Delete to fill the active selection with white.
Ensure the Move Tool is not highlighted in the toolbar and nudge the selection ten times to the left, then fill with white again. Pop your 3D glasses on again to check the result — the monster should now jump off your screen, but hold tight, we're not finished yet!
We'll now modify the mask further, so for clarity, I'll only show the mask (you can view your mask in full-screen mode by Option-clicking its layer thumbnail). Now Shift-drag a black to Transparent Linear Gradient from the bottom to add a sense of distance and scale to the creature.
Generate a layer based selection from your black "3-Dimension" layer and fill with white, nudge the same number of steps and fill with white again. This process needs to be repeated using selections from your white "3-Dimension" and orange "3-D" type layers. Now repeat with selections from both your shadow and white "TERRORSAUR!" type layers.
Use the same fill and nudge technique with selections from all your "Explosion" layers including the blurred duplicates.
Now we'll reverse the process and fill other areas with black. Make a selection from your "AMAZING! EXCITING! SPECTACULAR!" type layer and go to Select > Modify > Expand by 4 px (this is to avoid any unwanted white edge interference), then hit Shift-delete to fill with black. Now nudge and fill with black once again.
Keep in mind that white will reveal 3D areas and black will hide them — if you're unsure of how your mask is progressing, Option-click its thumbnail and use your 3D glasses.
Now fill a selection from your "Main figures" layer with black to delete any stray debris. There's no need to nudge and fill again — because all we're doing is deleting overlapping white areas.
Now fill selections with black from all your layers within the "SMALL FIGURES" folder, apart from their shadows. The mask is almost done, all that's left is just to tidy up unnecessary white specs as indicated using a black, hard-edged brush.
Now put your 3D glasses on, sit back and enjoy! For optimum viewing, select Full Screen Mode and hit Tab to hide your palettes. Also, stand up and step away from your screen, then move your head from side-to-side for a real sense of 3D!
I hope you've enjoyed this tutorial and discovered some new techniques along the way. These 3D techniques can be applied to any image, but work best if they have a dramatic, or exaggerated perspective. I'd like to thank Photographer John Wood for supplying the model image, the models themselves and the make up artists for all their hard work.