Working with Commercial Printers

If you plan to send your digital photos to a commercial printer (like Costco, for instance), there are a few things that you need to do:

1) You need to know exactly what printer it uses.  For example, the Costco Photo Center in NE San Jose uses the Noritsu QSS-3411 Photo Printer.

2) You need to know what print resolution it has.  For example, the Noritsu QSS-3411 has a print resolution of 320 pixels per inch.

3) [Ideally] You should have that printer's color profile installed on your computer.  This ensures that the image will look like what you think it should look.

4) You need to send to the printer TIFF files (not JPEG files) processed with Adobe RGB color space and embedded with ICC color profiles.  This will ensure that you exploit all the colors that the printer has to offer, and that you will see what you expect to see.

5) Determine what print size you need to print, and resize your image before you send it to the printer.  For example, your original digital image was taken by a 15-megapixel camera, and you want a 4"x6" from Costco in NE San Jose.  Based on the Noritsu QSS-3411's print resolution of 320 ppi, you need a file that is (4 x 320 = ) 1,280p in width and (6 x 320 = ) 1,920p in length, which is roughly 2.5 megapixels.  So, resize your image to 1280 x 1920.

6) Tell the print operator that he needs to "disable color correction" on your photo.  The best way to do this is to rename the image file to something like "DoNotCorrectColor-XXX.TIFF".  (You can specify "disable color correction" on the order, but many operators will forget to turn it off, and you'll get awful photos.)  Then send the TIFF file (with Adobe RGB and ICC profiles) to the printer.

Question:   But but... butt, my camera outputs JPEG files, not TIFF or Raw files, what can I do?

Answer:  Get another camera.  Just kidding.  You don't have much of a choice now.  Just use JPEG in sRGB color space, and, if possible, embed the ICC color profiles into your JPEG.  All the other steps should be the same.

Graphic Image File Formats

Understanding graphic image file formats is key in how to post-process your digital images.

There are literally hundreds of different graphic image file formats.  Here is a comparison chart that I use, along with a brief description, of the most relevant file formats for photography.

 

 

* Raw:  This file contains the data directly from the camera's image sensor and camera settings, [mostly] unprocessed by the camera's processor, and [mostly] uncompressed.  It uses either the Adobe RGB or the ProPhoto RGB Color Space (depending on the camera manufacturer), with a color depth of 12-14 bits per channel, or up to 42 bits per pixel.  This is not one file format but is a family of about 230 of proprietary (and mostly undocumented) raw formats;  each camera manufacturer has its own raw format;  even each camera model might have its own raw format.  As a camera model is discontinued, the associated raw file format may become unsupported and obsolete.  Some manufacturers may even encrypt this file to prevent third-party access.

 

* DNG:  This file is based on the TIFF file format, and uses a lossless compression.  It's developed by Adobe as an open file format to overcome the limitations of the Raw file formats (ie, proprietary, encrypted, non-standardized, etc.).  They are pushing this file format to be adopted by all camera manufacturers, and to replace the current various Raw formats.

 

* TIFF:  This file uses a [mostly] lossless compression.  It is good for printing because it can use a large Adobe RGB color space (ie, it can exploit the wide range of colors offered by the commercial printers) and because it can embed ICC color profiles.  (Hence, it's "device-"dependent.)  It's not good for display over the Internet because there are a number of its variations (different color depths and compression methods).

 

* BMP:  This file stores a simple description of the image pixels in an uncompressed form.  The drawbacks are that the file size is large, and it's Windows-specific (platform-dependent).

 

* Exif:  This file is basically a JPEG file with TIFF file features.  It contains camera setting data.

 

* JPEG/JFIF:  The JFIF file format uses a lossy JPEG compression on a 24-bit color file, resulting in a small image file size, suitable for display over the Internet.  It's not great for printing because of its lossy compression and its use of the small sRGB color space (ie, its inability to exploit the wide range of colors offered by the commercial printers).

 

* PNG:  This file is a newer version of the GIF file format and has TIFF file features, and it can replace the JPEG file for web applications.  It uses a lossless compression on a 24-bit color file.  The drawbacks are that the file size is larger than JPEG, and some older web browsers cannot read the PNG file.

 

* GIF:  This file uses a lossless compression on a tiny 8-bit color file (that can show only 256 colors), resulting in a tiny file size.  It's good for simple images that do not have a wide range of colors.  It supports animation, making it the preferred choice for animated graphic images to be displayed over the Internet.

 

 

Why is it good to know all of this?

 

* When you buy your camera, if the camera offers Raw format, it should come with a Raw processing software.  Save the software, because this particular camera model's Raw file format may become obsolete one day, and you won't be able to access your Raw image files without this processing software.  (Imagine losing your 35-mm film negatives.)

 

* When you shoot, if the camera allows it, choose the Raw file format and Adobe RGB Color Space.  This way, you have a wider choice and better quality when you post-process your images.

 

* When you post-process your photos, use the TIFF file format in the Adobe RGB Color Space.  Use the 48-bit color depth if you don't mind the heavy toll on disk space.

 

* When you print, send the TIFF files to the printer, with the ICC color profiles embedded in them.

 

* When you send the photos over email or put them up in the Internet, convert the images to JPEG.  This changes the color space to sRGB and 24-bit color depth, resulting in a smaller file size.  If the image does not have a lot of color information in it, you can even convert it to the GIF format to reduce its size even further.

Understanding Color and Color Spaces

Understanding color is the key to having the result that you want, and having your intended audience see what you want them to see.  Color theory can be complex.  Here are some major points.

 

Color is basically the brain's response to and interpretation of a small segment of the electromagnetic spectrum, called the visible light.  Thus, different people see colors differently (just as some people cannot hear certain sounds), and different machines (cameras, monitors, printers, etc.) respond to and "see" colors differently.

 

To standardize all the different "visibilities" of different machines and people, and to make sure that we all "see" the same thing, as intended by the presenter, the industry has developed different "color spaces":

 

CIE 1931 xy chromaticity diagram showing the chromaticities enclosed by
the ProPhoto RGB Color Space (blue outer triangle),
the Adobe RGB Color Space (yellow middle triangle),
and the sRGB Color Space (white inner triangle).
The vertices of the triangles represent the locations of the primary colors.
The cross in the center is the white point.
The areas of the triangle that are outside of the colored area
are colors that cannot be perceived by the human eye.
The color curve shows the color range from blue (380 nm) to red (700 nm).

 

* The L*a*b* Color Space:  This is a large set of the light spectrum that requires 48 bits to represent.  Needless to say, many of the colors in this space are beyond human perception.

 

* The ProPhoto RGB Color Space:  Developed by Kodak, this is a large set of the light spectrum that encompasses about 90% of the colors of the Lab Color Space, and corresponds to 100% of the "likely occurring" real-world colors (even though many are still beyond human perception).  It requires 16-bit per channel to work properly.

 

* The Adobe RGB Color Space:  Developed by Adobe in 1998, this is a set of the visible light spectrum that encompasses about 50% of the colors of the Lab Color Space, and corresponds to what the high-end color printers can reproduce.  It's used by color-aware applications (like Photoshop).

 

* The sRGB Color Space:  Developed by HP and Microsoft in 1996, this is a small set of the visible light spectrum that corresponds to what the CRT color monitors of the mid 1990s can display.  At the time, color monitor technology was poor, so this 8-bit-per-channel color space is small, and it's still in wide use today (because it's the "lowest common denominator"), especially by non-color-aware applications (like web browsers).

 

Why knowing all this is important?

 

* When you take photos, you want to take them in the ProPhoto RGB or Adobe RGB Color Space, if your camera allows you to.  Usually only the digital SLR cameras have that option.

 

* When you want to print your photos, and if you want the print to look the same as what you see on the monitor, you need to work in the Adobe RGB Color Space, work with a color-aware application, [ideally] calibrate your monitor in a dark room, [ideally] know what printer that you will print on, and [ideally] have that printer's profile installed on your machine.  Then tell the lab to "disable the color correction" when printing your photos.

 

* When you want to send your photos over email or to display them in the Internet, you want to export your photos to the sRGB Color Space, then email or display them over the Internet.

Understanding Light

Light makes photography possible.  So, understanding light is key.  Here, we discuss only one light source:  the main light.

When it comes to light, there are 2 properties of the main light that are most relevant to photography:  direction and quality.

1) Direction:  where the light comes from.

* 45-degree:  For people and for scenery, the most common is the 45-degree light (45-deg up and to one side).

* Butterfly:  From the top and slightly forward, most common for the glamour style of portraiture.

* Top:  "The Godfather," heavy and moody.

* Bottom:  "Dracula," dramatic and scary.

* Split:  90-degree to the side, dramatic.

* Back:  Puts the main subject into darkness, if not controlled.

2) Quality:  how "hard" or "soft" the light is.  Technically, it's how large the light source is, and how far it is from the subject.  It's judged by the shadow that it casts:  If the shadow is well-defined, the light is hard.  If the shadow is barely there, or not well defined, then it's soft.  For scenery, generally, you want hard light.  For people, generally you want soft light.  Of course, for every rule, there are exceptions.  For example, in the "Hollywood Golden Age" style of portraiture, only hard lights were used.

Know it, then use it, or control it.  Have your image firmly in your mind before you push the shutter.  Then work everything else to get there.

What camera resolution (how many megapixels) should I buy?

As discussed in the previous posts, most people never use more than 4 megapixels of camera resolution (that equates to a 5"x7" print, that is, if they ever print their images), for most situations.  Therefore, in choosing which camera, other features (and not camera resolution) should be considered, because most cameras nowadays have much larger resolution than needed anyway.

Print Sizes and Camera Resolutions

As a general guide, at 320-ppi printer resolution (as in the case for the Costco Noritsu QSS-3411):

for 4"x6" print:  you need 2.5 megapixels of camera resolution (4 x 320 x 6 x 320 = 2,457,600 pixels)

for 5"x7":  3.6 MP

for 8"x10":  8 MP

for 11"x14":  16 MP

for 16"x20":  33 MP

If you send a file that has a larger camera resolution than needed for the print size that you requested, Costco will simply resize it down before submitting it to the printer, to reduce bandwidth and storage.

If your file has a lower resolution than that is needed for the print (for example, you want to print a 16x20, and you have only a 8-MP file), the printer will interpolate.

Camera Resolution, Print Resolution, Color Depth, File Size

These often confuse people.  It's good to nail down these basic terminology so that we can all speak the same language.

* Camera Resolution:  measured in megapixels, is how many pieces your camera's image sensor divides up the image (into picture elements, or pixels).  This has nothing to do with color depth, even though it affects file size.  For example, the Canon EOS 5D Mark II's image sensor has a physical dimension of 24mm x 36mm;  it divides the image up into 3744 rows and 5616 columns (each of the row/column unit is a pixel).  That's 3744p x 5616p = 21,026,304 pixels = 21 megapixels.

* Color Depth:  measured in bits per pixel or bits-per-channel per pixel, is how each pixel is represented digitally.  For sRGB Color Space, the color depth is 24 bits (per pixel), or 8 bits per channel (Red/Green/Blue) (per pixel).  8 bits per channel is typical for a lot of photo processing, where each pixel is represented digitally as a number from 0 to 255 for each channel (Red, Green, and Blue).  A 0 value means no color for that channel;  a 255 means a pure color, for that channel.  Thus, the total number of colors that can possibly be represented by the 8-bit-per-channel (or 24-bit total) scheme is therefore about 16 million.  Canon DIGIC image processor outputs images that are 14 bits per channel (total 42 bits per pixel).  (More on color spaces later.)

* File Size:  measured in bytes, is the size of the image file on the computer.  This depends on  1) camera resolution, 2) color depth, 3) image file format, and 4) image compression.  For example, for a BMP file format (an uncompressed format), an image of 3744 pixels by 5616 pixels with a color depth of 24 bits (3 bytes), the file size will be 3744p x 5616p x 3bytes = 63,078,912 bytes = 61,600.5 KB = 60.16 MB.  (1 KB = 1024 bytes;  1 MB = 1024 KB.)

* Print Resolution:  measured in pixels per inch, determines how large (the maximum size) your image can be in print, without interpolating.  For example, I use the Costco Photo Center at NE San Jose;  it uses the Noritsu QSS-3411 Photo Printer that has a print resolution of 320 pixels per inch.  If I sent my image file of 3744p x 5616p (full size file) to that printer, the maximum print size that I can get without interpolating would be (3744 / 320 =) 11.7 inches in width and (5616 / 320 =) 17.6 inches in length.  And yes, if you plan to print your photo at a printer, you should know its print resolution.  For most people, the final prints (if they ever print their images) are usually 4" x 6";  therefore, the camera resolution needed is only about 2.5 megapixels (4 x 320 x 6 x 320);  that's what I set my camera to, when I do most of my day-to-day photo-taking.

Which brand of camera is better?

That's a perennial debate.

The short answer:  Canon.

The long answer:  It's all depending on your individual circumstances, as long as it's a Canon -- just kidding.

If you want the most bang for your buck, buy Pentax.

If you want the most pleasing color (overall) for your people photos, buy Olympus.

If you want to waste your money, buy Kodak.

If you want a professional line of cameras and lenses, buy Nikon.

For me, I buy Canon because

* I already have some Canon lenses in the past that I can reuse

* I am more familiar with Canon features and lay-out of buttons

* I like the color rendition of the Canon DIGIC image processor

What lens should be used for what?

There's no hard-and-fast rule.  This is art.  It's all up to your inspiration.

I use this as my guideline:

* For scenery, I use wide-angle lenses that are 1/2x (or shorter) of normal.

* For everyday shooting, I prefer a zoom lens that can go from 1/2 to 2x of normal (that's a 4x zoom, if possible).

* For people, I prefer a tele-photo lens that is 2x of normal.  Longer:  the subject is too far, and I have to yell at them.  Shorter:  their features get distorted.

What is a wide-angle lens? a tele-photo lens?

A wide-angle lens is technically defined as a lens that has a focal length shorter than that of a normal lens.  In a typical case, it's 1/2x.

 

A tele-photo lens is defined as a lens that has a focal length longer than that of a normal lens.  In a typical case, it's 2x.

 

 

Here's a table that I use;  it compares all the Canon and Nikon cameras and their important features:

 

 

 

Why is it good to know?

 

* When you buy, you want to know whether the lens that is packaged in a kit will do what you want it to do.  Retailers often package lenses willy-nilly, so you have to watch out.  You want to make sure that the lens(es) in the kit will cover the zoom range that you need.

 

* When you go on location, you want to bring the fewest number of lenses that will do their job, and you have to match your camera with your lenses.

What is a normal lens?

Knowing this is important when you buy your camera and lenses and when you decide what to take along on your trip.  What you know about a normal lens for a 35-mm film camera might not apply for a modern digital camera.

A "normal" lens is technically defined as a lens that has a focal length (in mm) that is the same as the camera's capturing area's diagonal dimension.  This lens approximates what a human eye sees, in terms of image size and field of view.

For example, for a "35-mm" (a misnomer) film camera, the capturing area (film frame) is 24mm in height and 36mm in width, so the diagonal is 43mm in length.  Therefore, a "normal lens" for a 35-mm camera has 43mm in focal length.

Now, for digital SLR cameras, only a few professional digital cameras have the capturing sensor that is the same size as the "35-mm" film frame ("full-frame" sensor).  These include:

* Canon EOS-1Ds Mark III

* Canon EOS 5D Mark II

* Nikon D3x

* Nikon D3

* Nikon D700

For these cameras, the normal lens is 43mm.  Therefore, the "focal length multiplier" is 1.0.

The rest of the digital SLR cameras have smaller-size sensors, therefore the normal lens for each is different:

Cameras with normal lens of 34mm (focal length multiplier of 1.3):

* Canon EOS-1D Mark IV

* Canon EOS-1D Mark III

Cameras with normal lens of 27mm (focal length multiplier of 1.6):

* Canon EOS 7D

* Canon EOS 50D

* Canon EOS Rebel T1i

* Canon EOS Rebel Xsi

* Canon EOS Rebel Xs

All Nikon digital SLR cameras (except the ones listed above) have a normal lens of 28mm (focal length multiplier of 1.5).

Why knowing all this is important?

* When you buy:  Retailers often package the cameras with all kinds of lenses, some of which are not useful for the consumers.  For example, one offer may be a kit of a digital SLR and a zoom lens of 28-135mm.  Sounds good.  However, when you look at the camera, it's a Canon T1i (for example) which has a focal length multiplier of 1.6, therefore that zoom lens is the equivalence of a 35-mm zoom of 45-216mm, not too bad but it's not for day-to-day shooting because it lacks the wide-angle component.

* When you decide what to take along on your photographic journey:  When I travel, I carry light.  I want one zoom lens that can go from wide-angle to tele-photo.  So if I have to pick one kit, I have to know what camera it is, what the range it needs, and bring with me the proper lens.