Bit Depth

in Academic
By qadirsh/Dr Qadir (1,195) Send mail to this user on November 12, 2007 4:01:29 AM CST

Introduction
Bit Depth
    (A) Quality of Image
    (B) Art of Photography
Conclusion

Quality of a photograph basically depends on the intensity of bit depth

Introduction

Quality of photograph is the main theme in the art of photography. It is always a question as to how this objective can be achieved. It has definitely multi-facial attribution related to different mechanism and approaches yet the bit depth that we would discuss here is one of the dominating factors in delivering out put. None of the factors can be seen in a photograph but the bit depth can.

Bit Depth

It is really an irony of fate with the art of photography that in most sites and forums this art, under its own caption, is replaced by a dominating debate on the type of camera thus becomes a commercial for different brands of camera in the outset. Camera is a tool not the tale. Nevertheless the man behind the camera and the skill that he/she creates the image with are basically ignored. No doubt that a good tool is of paramount importance in delivering a good result yet the person behind it is even more important. When we look at Mona Lisa the tool and material used are never perceived rather the skill and virtue of Leonardo himself, who created that masterpiece, do. When it comes to the camera, I understand, any camera that may have the mechanism to deliver more bit depth is the better.

(A) Quality of Image

The quality of image amongst others, I believe, depends much more upon the bit depth of the image which comes from a camera in-put that supports more than 8 bits (1 byte) in single channel and more than 24 bits (3 bytes) in three channels of RGB. Even if we calculate 256 shade of different colours in three channels, it equals (256x256x256=16 777216 ) or say 16.77 million colours as normally we hear. In resolution we normally get bigger dimension not depth. Commonly we take the resolution as depth not dimension. Is it not so? Which camera can deliver more than 8 bits in single channel and more than 24 bits in 3 channels (RGB)? This is the question that deserves appropriate answer and detailed discussion. The technology and material used in manufacturing such a camera makes it much expensive as well. Commercially this aspect of the camera does not come under discussion while other facilities do. It would be nice if some one add more information on it as the best tool.

(B) Art of Photography

Nevertheless the art of photography, in my opinion, should be discussed in the perspective of art itself not the camera. The camera should be placed in its own forum especially not under the title of art of photography which discusses the following aspects:

1. Rule of third
2. Exposure
3. Focus
4. Depth of Field (DOF)
5. Lighting / White balance
6. Fore, Middle, and Backgrounds
7. Cropping/Framing
8. Color / Tonal Range
9. Leading lines
10. Dark vs. Light areas
11. Balance
12. Diagonals, S-Curves, etc
13. Expression

Hope it may make a sense. I would like to add a portrait which was taken with a Rollie in B/W which demonstrates the quality, though the question of bit depth which is especially related to digital technology was not looming those days. I added colour in post processing and changed the background alone. In this exposure the skill of the artist in terms of modeling light maneuvering, focusing, sharpness, expression and balance etc and the quality of the photo as a result of a good tool is distinctive.

Conclusion

The idea that I want to put forth is that we should keep both the camera and the art of photography in two separate section and do not intermix them all together, thus, do not cause aberration and save time for those who are after some thing in the article and ultimately find nothing related that he/she is after.

Read 33,375 times

Reply 

From mattman944/Matt (2,538) Send mail to this user on November 15, 2007 3:39:39 AM CST

The bit depth of the FINAL IMAGE only needs to be as good as the human eye. Studies have shown that the human eye can perceive 10 million colors. So, 8-bits per color (16 million) is sufficient for a final image. Standard JPEGs are 8-bits per color. Most computer video cards are 8-bits per color. Cards that have a maximum specified bit depth of 32-bits usually have 8-bits per color. The other 8-bits are used for other purposes or are unused.

Here is a test you can do yourself. Open a new Photoshop image with a size of about 800 x 600. Set your background color to (R=0,G=0,B=200). Fill the image with this color. Use the rectangular marquee tool to select a rectangle in the middle of the screen about 1/3 to 1/2 the size of the image. Set the foreground color to (R=0,G=0, B=210). Fill the selected area with this color. Press Ctrl-H to hide the “marching ants”. You should be able to easily see the center rectangle when you use the colors values specified. Now, modify the colors until you find the minimum difference that your eye can detect. Always keep two of the colors at the same value for the background and foreground and vary the third color (blue in this example). For most background colors, the minimum difference I can detect is 2 or 3. There were a few background colors were I could detect a difference of 1. Remember, these are near ideal conditions. Finding color limitations due to bit-depth in a real image is much harder. Try using a background color of white (255,255,255). If you can see a (255,255,251) foreground rectangle on a (255,255,255) background your monitor is adjusted well. Try using a background color of black (0,0,0). You will probably find a much larger difference is required to see the rectangle. If you can see a (0,0,4) foreground rectangle on a (0,0,0) background your monitor is adjusted extremely well.

So, why all the fuss about Raw images (typically 12-bits per color)? The most important advantage is that you can use the extra data to bring out more detail in the shadow areas of your image during post-processing. You do this by adjusting the contrast, curves, levels, etc. The details on how to do this are beyond the scope of this reply. There are also other less important advantages: To prevent the comb effect (missing brightness levels) in your histograms when you increase contrast. To avoid the losses associated with JPEG compression. This non-lossy advantage is overrated by many people. At high quality settings, JPEG compression is very good. Try this test: take a Raw image and put in on one layer and the same image saved as a JPEG (quality=9) on a second layer. Then change the layering from normal to difference. The screen will be black. Use the eye-dropper to look for the maximum difference. As we saw in the previous test, a difference of a few is only perceptible in ideal conditions.

Let me end by saying that Raw is an advanced tool. I only recommend it be used by advanced photographers. Beginners will not have the skill necessary to take advantage of Raw. Using Raw will not make you a better photographer any more than using a pro-level camera will. For most photographers, your money and time will be better spent on books and by shooting more (especially if you can find a better photographer to shoot with).

Reply 

From jondee/John (1,044) Send mail to this user on November 25, 2007 5:01:10 AM CST

I have to disagree about RAW being an "advanced tool." It is as near to a negative as a digital image system can provide. Just as a film print is a subset of the information on the negative, a jpeg is a subset of the data recorded by the camera. Your remarks about the sensitivity of the human eye are spot on, but, a greater bit depth offers a greater range of data from which an image can be abstracted. Also, for beginners, RAW offers a means of learning to see what they did wrong. Shooting RAW+jpeg also offers a very good tool for learning the camera, if RAW is available. Jpegs from the camera are sometimes excellent, almost always adequate, but occasionally they are failures. With the RAW and exif data to work from, it is possible to understand where the mistakes are made. I think your last point is also very good. Reading and practice are the best means of improving photographic skills. However, even cheap DSLRs come with RAW processing software (well, Nikon charges extra) and it is a fine means of expanding the learning arena.

Reply 

From mattman944/Matt (2,538) Send mail to this user on January 13, 2008 4:19:00 AM CST

I have found that beginners do not have the skill level to successfully use RAW. They typically cannot get the image as good as the JPEG would have been. A beginner can improve their photography more by reading and shooting, than by learning to use RAW successfully. Whey they are past the beginner stage, shooting RAW+JPEG is an excellent idea for learning to use RAW.

Reply 

From bsswartz/Bailey (185) Send mail to this user on December 12, 2007 4:57:24 PM CST

RAW allows you to change the color temperature and such after exposure also, which is very nice if you are shooting in dynamic conditions. Using a program like Photoshop Lightroom, raw formats are NOT an advanced tool... they are presented just like jpegs, but you can do 10x more with them.

Reply 

From peter13/Peter (31) Send mail to this user on November 21, 2007 1:20:56 PM CST

What kind of article is that?

Reply 

From johnfalky/John (60) Send mail to this user on December 28, 2007 9:34:28 PM CST

The writer made a good attempt, but I find it technically incorrect. The Quality of an image is first determined by a person's skill at using whatever piece of equipment he or she is using. The technical hoopla in which many photographers bury themselves can become a distraction.

Reply 

From alfredos/Frank (937) Send mail to this user on April 29, 2008 10:35:35 AM CDT

I think the Bit Depth is one of the most important parts of digital imaging. Since i was working in a laboratory for digital image processing i understood how important is the bit depth. Sometimes it was very difficult to "put" the important part of an image into the range of an 8 bit grayscale line-camera. There were just 256 levels because this small 8-bit format (2^8) from the darkest part to the lightest part of the image. Sometimes it was too less to get enough diffences out of the image for the analization-program. Some levels more often would have been an atvantage to see differences for example in a nearly white part of the surface. A grayscale camera (every digital camera is a grayscale camera in truth) with 1024 levels (10 bit) would have shown more details between of levels 230 to 240 and not only 10 steps. The same to normal photography. If I have a shot which has most dark parts and most bright parts it would be very difficult for the camera software, to project the differences into an image of 256 levels. You can imagine, that there must be something lost, perhaps the white clouds on a blue sky havent a smooth blending from white to light-blue. Instead there are some steps between blue and white and this makes it impossible for you, to post process the image because the information is lost. Thanks god the camera developer made cameras with raw format and were break-through the 8-bit-format. Imagine, a 12-bit-sensor can theoretical capture 4096 steps (levels) of gray per channel, so if we considder one channel like blue, there would be a big range of blue shades between black and white. If you think about, you would think that the big range of 4096 steps would be poor in truth and you think your eye could catch more than just 4096 steps of one color? May be it is possible and therefore we got the 48-bit depth cameras which has 16-bit per color. This means 65536 steps for each color RGB. Isn't it marvelous? It means that i have a lot of shades of blue in light-blue sky, it means i am possible to change something into a small range of shades... if I can process the 48-bit-format. After all image-processing i have to (out-) put it into a small 24-bit jpg image to show it here... :-)

Reply 

Return to articles