Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
04-21-2007, 09:46 AM   #1
Inactive Account




Join Date: Sep 2006
Location: Lone Tree, Colorado, USA
Posts: 77
A "Bit" Of A Primer, Please

I have zero formal computer education, but have used various models successfully for 10+ years. My question:

Am I saving files in 4096 colors?

Situation:

K100D, RAW files only.
Process PEF files in Pentax Photolab v3.00. Save files as 16 bit TIFF files.

This whole bit thing thoroughly confuses me. Are the PEF files 16 bit? I know that the JPEGs are 8 bit, but after that I lose it in comprehension of what the H I'm doing.

Would someone please run through this bit concept as it relates to files? I'm particularly interested if I'm wasting my time with the 16 bit TIFF files.

The bottom line- am I indeed saving my files at 4096 colors?

Thanks for any enlightenment provided.

04-21-2007, 10:11 AM   #2
Inactive Account




Join Date: Oct 2006
Location: Boise, Idaho
Photos: Gallery | Albums
Posts: 1,414
The number of bits tells you how many times you multiply 2 times itself. In other words, 24 bits (8 times three for RGB) is 2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2x2 = 16,777,216 colors.
04-21-2007, 10:12 AM   #3
Veteran Member
volosong's Avatar

Join Date: Sep 2006
Location: Antelope Valley, SoCal
Photos: Gallery
Posts: 663
16-bit is way better!

Someone will have to correct me, but from what I understand, when talking about 8-bit and 16-bit images, it is being in the context of each individual color channel. There are three channels in the computer world...red, green, and blue. Each color channel, if it is 8-bit data, consists of one possible values between 0 and 255 (256 possible color values). A density number (DN) of zero will be black while a DN of 255 will be white. Combine the three color channels, 8-bit each, and you have what is sometimes referred to as a 24-bit image.

This is what computer monitors (and televisions and any other number of display devices) show. If you have ever played around with the color display on your computer, you might have noticed that one mode is 16 million colors. This is the mode where you display 24-bit (8-bits per channel) data. Depends on the memory of your video card. Probably not an issue these days as memory is relatively inexpensive.

The advantage of 16-bit data (16-bits per color channel) is that the dynamic range is much greater. 256 possible values for byte (8-bit) (per channel) data and ~64,000 possible data values for 16-bit (per channel) data. As you might surmise, you don't lose data when in the 16-bit mode.

(But...isn't Pentax RAW data only 12-bit (per channel) data?)

The disadvantage of 16-bit data is that it cannot be directly displayed, on a computer monitor or printed out on a hard-copy print. It need be reduced to 8-bit per channel data. In your post processing, the conversion from 16-bit to 8-bit data would be your last step before printing hard copy.

Not sure where you are coming up with 4096 data values. Perhaps that is what the old GIF format was. It was byte data (8-bits per pixel) divided up between the three color channels. If you were saving an image as byte data, you would see the artifacts in relatively flat-field portions of your image. For example, you will see lines on a wall behind your subject created by the distance falloff from a distant lamp.

Not sure if any of this make any sense to you. Bottom line...working in 16-bits is way better as you have more latitude and control over what happens. As your last step, save as a 8-bit image.
04-21-2007, 03:39 PM   #4
Inactive Account




Join Date: Sep 2006
Location: Lone Tree, Colorado, USA
Posts: 77
Original Poster
QuoteOriginally posted by volosong Quote
16-bit is way better!

The disadvantage of 16-bit data is that it cannot be directly displayed, on a computer monitor or printed out on a hard-copy print. It need be reduced to 8-bit per channel data. In your post processing, the conversion from 16-bit to 8-bit data would be your last step before printing hard copy.

Not sure where you are coming up with 4096 data values.

Not sure if any of this make any sense to you. Bottom line...working in 16-bits is way better as you have more latitude and control over what happens. As your last step, save as a 8-bit image.
Steven-

Where I get lost is is that the Pentax Photolab software allows me to save in 16 bit TIFF format....and I have been. I've been successfully printing in that mode since I started with my K100D. As a point to note, the Photolab software gives the option to save TIFFs in either 8 bit or 16 bit.

Now, I'm not sure where I came up with 4096 colors either.

I'm still sort of lost in the wilderness when it comes to the detail that my prints have.......according to the math, saving a 16 bit TIFF file yields 65,536 colors.

The question of the day still is: Am I also printing at that same high quality?

04-21-2007, 04:00 PM   #5
Inactive Account




Join Date: Sep 2006
Location: Göteborg, Sweden
Posts: 810
The raw files are 12-bit things, converted into 16-bit files for your PP happenings. Having that amount of steps means you can edit all day long, save losslessly in TIF whenever you fell it calls for it, and end up in nice pictures with no jumps in the color tones.

I assume you convert the raw files into a color space as well as to a bit depth. I use aRGB as that is what my printer handles. My monitor is somewhere between sRGB and aRGB but as I do my prints myself I see no problems using aRGB. Some prefer ProPhoto but that is very wide and I'll never see all the colors, not on screen, not when printed, so i prefer to have the somewhat smaller aRGB space to work in thus minimizing the risk of cause banding or odd color phenomenas.

So, in the end you have an aRGB 16-bit file. Now when it is time to print it your printer driver will handle the numbers and make them into something your printer can handle. The driver will make for all possible colors that your printer's inkset can provide and yes, I would say that is it more than 8-bits of gradations possible with any decent printer.

For publishing on the Net it is a somewhat other story where you can convert your picture to 16-bit sRGB for your PP just to have full control of the colors until you do the final conversion to an 8-bit sRGB JPG file.

hth, somewhat,
04-21-2007, 04:59 PM   #6
Inactive Account




Join Date: Oct 2006
Location: Boise, Idaho
Photos: Gallery | Albums
Posts: 1,414
I am not quite sure where you get 65,536 for a 16 bit tiff file. 8 bit and 16 bit are per color so you multiply them by three to get the total bit depth. That means that you are getting 24 bit or 48 bit color. 24 bit color is 16.7 million colors (multiply 2 by itself 24 times). 48 bits is 281 trillion colors.

Last edited by davemdsn; 04-21-2007 at 10:35 PM.
04-22-2007, 07:22 AM   #7
Senior Member
jslifoaw's Avatar

Join Date: Apr 2007
Location: Toronto/Victoria
Posts: 262
4096 may have come from the RAW files being 12-bit (2^12= 4096). Which means that the RAW files can record one of 4096 values of brightness for that pixel which, using most digital cameras, will be a measurement of the brightness of red, green, OR blue at that point. In the RAW file, half of them will be green, and a quarter will be each of red and blue.

Each pixel only records the value for one channel of colour, not all three, so your application takes each colour at a time, and then fills in the blanks between pixels of the same colour which are one of the other two colours, and then combines all three for the full colour image.

In 8-bit mode, those "blanks" can be one of 256 values. In 16-bit, 65536. So for a particular pixel in 8-bit mode, it could be one of around 256 different values in 16-bit mode because there is that much more precision which is rounded in 8-bit mode. This applies while you apply filters like contrast, etc. because every time it computes in maximum precision what the new red, green, and blue values should be, it has to round each to whatever is closest in the mode you use.

It's like if I report a financial statement in thousands of dollars, one million is 1000. However, $1000499 would be reported as 1000 as would $999500, and those differ by a measureable amount. And then if you compute other ratios based on the number 1000, you will propogate the error, which is the equivalent of posterization in Photoshop.
04-22-2007, 06:48 PM   #8
Inactive Account




Join Date: Sep 2006
Location: Lone Tree, Colorado, USA
Posts: 77
Original Poster
QuoteOriginally posted by Jonas B Quote
The raw files are 12-bit things, converted into 16-bit files for your PP happenings. Having that amount of steps means you can edit all day long, save losslessly in TIF whenever you fell it calls for it, and end up in nice pictures with no jumps in the color tones.

I assume you convert the raw files into a color space as well as to a bit depth. I use aRGB as that is what my printer handles. My monitor is somewhere between sRGB and aRGB but as I do my prints myself I see no problems using aRGB. Some prefer ProPhoto but that is very wide and I'll never see all the colors, not on screen, not when printed, so i prefer to have the somewhat smaller aRGB space to work in thus minimizing the risk of cause banding or odd color phenomenas.

So, in the end you have an aRGB 16-bit file. Now when it is time to print it your printer driver will handle the numbers and make them into something your printer can handle. The driver will make for all possible colors that your printer's inkset can provide and yes, I would say that is it more than 8-bits of gradations possible with any decent printer.

For publishing on the Net it is a somewhat other story where you can convert your picture to 16-bit sRGB for your PP just to have full control of the colors until you do the final conversion to an 8-bit sRGB JPG file.

hth, somewhat,
Jonas-

Yes, I use aRBG for printing at home. Judging by the IQ of some of my images on that other web site, yours is a good reminder to utilize sRBG for the web. I had better start that.

In the final analysis, it appears that I read you as saying that something higher than 8 bit color gradiation is possible with a reasonable printer when printing from a 16 bit TIF file. It seems therefore that shooting in RAW and saving in 16 bit TIF vice 8 bit JPG does have a positive effect.

My thanks to all in this discussion.....my understanding has been elevated somewhat.

Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
bit, camera, colors, dslr, files, pef, photography, tiff
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
K-7 Movie "Sound" Primer dopeytree Video and Pentax HDSLRs 58 04-06-2010 11:00 AM
Sports "Highside Exit" took 1st Place in DPReview "Missed It by THAT much, Part 1" Challenge MRRiley Post Your Photos! 27 02-21-2010 08:26 PM
K-x battery issue "primer"? switters Pentax DSLR Discussion 13 12-03-2009 12:24 PM
"Hunger for a DA*50-135?" or "The DA*50-135 as a bird lens!" or "Iron age birds?" Douglas_of_Sweden Post Your Photos! 4 08-13-2008 06:09 AM
Just a little bit of "potty" humor. :) heatherslightbox General Talk 5 06-09-2008 04:05 AM



All times are GMT -7. The time now is 11:05 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top