collection index pageculture index pagefungi index pageother info index pagedownloads index pageHome page

CREATION AND PROCESSING OF DIGITAL PHOTOGRAPHS


 

Hardware and Software:
(organized around a single workstation
on a central table in our lab)


click for larger photo


click for larger photo

1. A relatively fast PC computer (Pentium III or better, > 512 MB of RAM, at least one GHz clock speed) to easily process data using image-editing software.

2. Many of the older photos present on this website were taken with an analog single-chip video camera -- high quality digital cameras for microscope applications were prohibitively expensive at that time. For a time, we upgraded to a 3-CCD (a chip to handle each of the three input colors: red, green and blue) analog video camera (Sony DXC-970 "PowerHAD"). Because these are video cameras, a "frame-grabber" card had to be installed in the computer to freeze a single frame of the live image appearing on the monitor. We used the FlashBus made by Integral Technologies, Inc.

3. Digital cameras have come down in price and are more affordable (now in the $6000-7000 range), we use a Nikon DXM1200. A newer model now is available from Nikon, the DXM1200F (found under the "microscope" section of Nikon's home page, which captures images up to a resolution of 3840 x 3072 dpi.

4. Adaptors which allow the video camera to be attached to either a compound microscope (to capture photos of mounted spores/mycorrhizae, etc.) or a stereomicroscope (to capture images of intact spores, roots, mycelium, etc.).

5. A high-quality stereomicroscope. We use Nikon SMZ-U for routine photos because it uses the same adapter as the compound microscope and the camera can be readily interchanged.

6. A high-quality compound microscope, preferably equipped with differential contrast optics to better resolve thin colorless layers and other difficult-to-see subcellular structures. We use a Nikon E600 Eclipse Microscope with DIC prisms matched to each objective (10, 20, 40, 100X).

7. A dye-sublimation printer with 4-color ribbon and black & white ribbon, capable of printing from standard size photograph paper (8.5 x 11 inches the U.S.). We use a Tektronix Phaser 450 Printer because of "colormatch" software that makes it easier to match CYMK output to paper with RGB output on the computer monitor. The nicest feature of a dye-sublimation printer is that high quality photographs can be obtained with a minimum resolution of 150 dpi. This is important because video cameras output a maximum image size of 640 x 480 pixels so that the largest quality print size is 10.8 x 8 cm or 4.3 x 3.2 inches). However, these dimensions are more than enough for voucher and manuscript photos -- the targetted output of this system. With the higher resolutions obtained from the digital camera (300 dpi or higher), then any color printer suffices. Epson makes inkjet printers with permanent inks that don't fade as fast as those of other brands and don't bleed when exposed to moisture.

8. Image-editing software. Many different applications are available, but we use the workhorse that does everything (and does it extremely well): Adobe Photoshop. It takes a while to learn most operations in this application, but the return far exceeds the time investment.


Steps in creating a digital image using a digital camera:

1. Load application that accompanies the digital camera. For the Nikon DXM1200 digital camera, this application is called "ACT-1". A window appears that contains the "live" image from a microscope (see far left photo below).

2. Arrange the specimen in the window to clearly show all the features desired, adjusting the lighting to provide the best contrast.

3. "Grab" the image using menu or button commands in the window and save the image as a TIF file (NOT jpeg, because jpeg uses a compression algorithm and some pixel data are lost).

4. Once saved, the image then is opened in Photoshop and cropped to a size that best frames it. The image is edited to match color under the microscope and the background darkened to optimize contrast. To play it safe, I always create a duplicate image (go to "image" in the menu bar, and then "duplicate in the menu -- see center photo below). Some applications will automatically add a scale bar (after calibration). Alternatively, a structure in the photo can be selected and size (in µm) measured. It is best to find an object that measures in an increment that can be somewhat standardized across photos: we use measures of 5, 10, 20, 50 µm. Going back to Photoshop, a separate layer is created and a bar drawn over the image of the object in the layer below (see far right photo below). Bar length is typed above the bar (it will appear on a third layer). Move the bar and text to an unobtrusive part of the photo. SAVE this edited image in Photoshop's native PSD format (thereby retaining all the layers, so they can be edited or deleted later if necessary). Flatten the image and save it again as a JPG file. You now have three files: (i) the original TIF image to archive, (ii) the layered file containing the maximum information that also is editable individually and can be altered at any time to change resolution (e.g. highest for printed photograph, 72 dpi for presentation on computer monitors), and (iii) a compressed file of the edited image whose resolution also can be changed as needed. It is this multiple filing system which provides maximum flexibility for any future use, whether anticipated or not.

click on photos for larger images

5. For photos archived as information only (not for general viewing), notes or observations can be included as separate layers on the PSD file. When the file eventually "goes public", these layers are deleted and the file resaved in another format (thus preserving the original).

6. All photos are easily organized in folders organized hierarchically, after which they can be backed up or archived on compact disc.


Resolution issues:

1. Web-based images should not exceed 72 dots-per-inch (dpi). At higher resolution, the image will appear larger on a monitor because it has a fixed number of pixels. Larger monitors have more pixels and can be set at higher resolutions (1024 x 768, 1260 x 1600, etc.), and so these are ideal for viewing high resolution images. When working with images, we use a 21-inch monitor.

2. Images for Microsoft Powerpoint presentations should not exceed 100 dpi. Extra pixels above this resolution are not used, even though the entire file is loaded.

3. Images for printing should have a minimum resolution of 150 dpi for dye-sublimation printers and 300 dpi for many other color printers.

4. There are times when resolution may need to be increased to obtain a good print image (usually when resolution was low when first shooting the photo with an analog camera or from a low-end scanner). To preserve fine detail (such as inner wall structure of spores, etc.), then the image should not be resampled (pixels added) when increasing resolution. To do this, uncheck the "resample image" option. The result will be a smaller printed image (the same number of pixels are present, but with more per inch with no change in file size (see example below). The size of the image on the monitor will not change, since no pixels were added or subtracted.

5. Resolution of images from a digital camera usually is greater than that which is optimal for monitor-based displays (web pages, image catalogs, etc.). Here, the image is resampled ("resample image" is checked) when the resolution is changed from 150 to 72 in the example below. Notice that the file size is greatly reduced (3.75Mb to 885 Kb), but the image size remains unchanged. It is amazing how much detail is preserved when so much pixel data is discarded. To experiment, duplicate the original image, perform these modifications on the copy, and compare quality with the original.

Data on the original image

Increasing resolution (no resampling of pixels)

Decreasing resolution (resampling of pixels)