Join ePHOTOzine, the friendliest photography community.
Upload photos, chat with photographers, win prizes and much more for free!
Scanning high contrast transparencies - Peter Bargh looks at a scanning technique to help you get better results from your high contrast transparencies.
If you are one of the many photographers who's making the transition from film to digital you will no doubt have a collection of negatives or transparencies to convert to digital. This is done using a film scanner and with the variety of high specification models available today can be done very well.
Scanning negatives is relatively easy because they usually have a lower contrast range than transparencies, but you may find scans from your transparencies lack the tonal range with either detail missing in the highlights or shadows, or both. This is just like the days when you used to get prints made. They never looked as good as those from negatives, yet when projected they leaped off the screen.
There is a solution that requires an fair amount of extra work at the computer but will help deliver much better results.
Here's a shot taken while on holiday in Greece. While it not be a typical subject to shoot it does display the usual level of contrast that you'd get when shooting transparency film in bright conditions. I exposed making sure the highlights wouldn't be overexposed. This has resulted in strong deep shadows. When projected the power of the lamp lifts the detail in the shadows lift while maintaining the highlights. When commercially printed the shadows are black and the highlights white - both with no detail. If you scan the transparency using the scanner's default or auto settings you get a similar disappointing result.
To get around this problem I can scan twice. Once for the shadows and once for the highlights and then merge the two in Photoshop. Here's how:
1 I'm using the Konica Minolta Dimage Scan Elite 5400 with Konica Minolta's Dimage Scan 1:1, but similar features can be found on most film scanner software. If the software you are using is too basic, try using VueScan -it supports over 400 scanners.
Most scanner software will create a thumbnail view of the strip of negatives or mounted transparencies that you insert in the scanners carrier. Pick the shot you want to scan from the thumbnails and make a preview scan. This is a larger version that's easier to refer to when making adjustments to tone or colour in the driver software. The first screenshot shows what will happen when you scan using the default settings.
2 If you go into the advanced option found under the Image correction tab you'll find a range of things you can adjust such as sharpness, tone and colour. The first option is the one we will use - Tone Curves and histogram - which brings up a dialogue box with various controls. This is like the Curves and Levels adjustments that you may be familiar with in Photoshop. The top one is Curves and by clicking on the graph and dragging up or down you can increase and decrease tone or contrast, depending where you click from and pull to.
For this technique I'm going to go for the simpler option of levels using the histogram in the lower half of the dialogue box. For those of you that are unfamiliar with this, the histogram is a graph of the tones in a photo. It ranges from black at the left to white at the right. Everything in between covers the full tonal range of the photo from extreme shadows to highlight. The peaks indicate that there are more of that particular tone in the photo. So in this example it's easy to see that most of the detail is recorded in the mid tone to shadow area - tones such as the deep blue sky take up the section about one third into the graph from the left.
There are three triangular sliders below the Histogram that can be adjusted to change how the tones are distributed. The black one represents shadows, grey mid-tones and white is highlights. If we drag the black slider across to the right the darkest point of the photo will change from the extreme shadows to the dark grey area. All detail left of this will then become black so you increase the number of shadow tones anything left of the black marker will become black. This is known as clipping and can remove necessary tonal information. Similarly moving the white slider to the left clips the highlights. Dragging the grey slider adjusts where in the photo the mid tones should be and it's this one we will use to create two different scans.
|3 For the first scan I'll improve the shadow detail by dragging the mid tone (grey) triangle to the left slowly. Notice the whole photo in the preview starts to become lighter as the mid tones are being adjusted to what was originally the darker regions of the photo. You'll also notice the curve above has changed shape to an arc. Adjust so that the areas in shadow have enough detail without being too light and over exposed. Then close the tone curve window and click the full scan option to digitise the photo.|
|4 Keep your scanning software open and go back to the preview. This time we will adjust the histogram so that the highlights are exposed correctly. In this photo the two areas where the lost detail is most concerning is in the centre of the two octopus. By dragging the mid-tone slider to the right you will see this detail improves but the shadows and most of the mid tones become black. This will not matter as the only areas we will salvage in the next step are the highlights. When you have a preview with good detail in the highlights click to make the full scan. make sure you keep all the size and colour settings the same for both scans. You should only change the histogram.|
|5 Open both of the scans in Photoshop. Copy the one with just the highlights, scanned correctly in step 4, and paste it on top of the version with the correct shadows, scanned in step 3. Do this using Select>All (Ctrl+A), Select>Copy (Ctrl+C) and Select>Paste (Ctrl+V). Alternatively click anywhere in the photo of the highlight version and, while holding the mouse down, drag the photo to the shadow version to quickly copy. Tip: If you hold down the Shift key while dragging the new layer will centre itself rather than falling offset where you drop it. The photo will now look dark because the new layer hides the layer behind it.|
|6 Create a Layer Mask by going to Layer>Add Layer Mask and select Hide All. The layer icon will now have a black rectangle next to the layer thumbnail. This signifies that the whole of the top layer has been masked so the bottom layer is showing through. If you had selected Reveal All you would see a white square and all of the above layer would be visible still.|
7 Parts of the mask can now be removed so that the top layer appears over the bottom layer. This is done using the Paint brush. If you select a large brush with 0% hardness to give a soft edge you can paint using white to reveal the detail. Set the mode to normal and Opacity to about 15%. This allows you to build up the detail. I've illustrated this by highlighting the area I've done inside the red border. Each time you go over the same area you reveal another 15% of the layer behind the mask. If you make a mistake you can paint the mask back using black or use the eraser which will already be set to black.
Tip: Use the X key to toggle between foreground and background colour or the B (brush) and E (eraser) keys to toggle between tools.
8 Work your way around all the highlight areas painting the mask with white to reveal the darker highlights through the masked layer. If you make a mistake press Ctrl+Z to undo. The illustration to the right shows before and after painting with the mask.
|9 When you've finished, flatten the layers - Layer>Flatten Image.|