A700: 12 or 14 bits?

I'm looking at the Pop Photo review of the A700 and in the Vital Statistics section it says:

Imaging: 12.2MP effective CMOS sensor captures images at 4272x2848 pixels with 14 bits/color in RAW mode

I thought the A700 was 12 bits. Is this a typo? I know the D300 and the 40D are 14 bits, and I was surprised the A700 was only 12 bits when it was announced, given that Sony manufactures sensors for Nikon. I looked at the A700 manual but it doesn't indicate the bit depth...

Here is a link to the article:

http://www.popphoto.com/cameras/4724/camera-test-sony-alpha-700.html

Cheers, Keith
http://www.kotay.net/keith/photo/photo.shtml

I know I saw it somewhere that my Jpeg shots are 8 bit and my RAW shots are 14 bit with the A-700 . I think it was in the IDC Sony converter where you get to choose the end Bit depth .

I will look .

UnderDriven said:

I'm looking at the Pop Photo review of the A700 and in the Vital
Statistics section it says:

Imaging: 12.2MP effective CMOS sensor captures images at 4272x2848
pixels with 14 bits/color in RAW mode

I thought the A700 was 12 bits. Is this a typo? I know the D300 and
the 40D are 14 bits, and I was surprised the A700 was only 12 bits
when it was announced, given that Sony manufactures sensors for
Nikon. I looked at the A700 manual but it doesn't indicate the bit
depth...

Here is a link to the article:

http://www.popphoto.com/cameras/4724/camera-test-sony-alpha-700.html

Cheers, Keith
http://www.kotay.net/keith/photo/photo.shtml

Click to expand...

12 bits (just one of many typos)
Nikon is 12/14 it scales back when shooting high FPS.

I have actually seen some 40D users asking for 12 bit switch because "14 bits just makes my files bigger."

I have yet to see the value of 14 bits past marketing since the color gamut of the sensor is already larger than any we using in image work, larger than what the human eye can see by many times, etc. Larger than any output device.

2^14 is 16384 shades of each primary color. The Human eye can do about 500 shades.

14 bit color is 4,398,046,511,104 colors the human eye can do about 16,000,000. Some women with an extra cone type can see 100,000,000 colors.

We do need more than 8 bit for processing overhead to protect colors.. but I am with the "it only makes my files larger" crowd..

------------
Ken - KM 5D (A700 Joy)
http://www.cascadephotoworks.com

Ken_5D said:

12 bits (just one of many typos)
Nikon is 12/14 it scales back when shooting high FPS.

Click to expand...

Hi Ken. I thought it was 12...

Ken_5D said:

I have yet to see the value of 14 bits past marketing since the color
gamut of the sensor is already larger than any we using in image
work, larger than what the human eye can see by many times, etc.
Larger than any output device.

Click to expand...

I think the idea is to better represent subtle variations in color. With extra bits you could enlarge the gamut itself (meaning the area of representable color in the gamut graphs), or you can keep the area the same and provide a finer representation of the colors inside the gamut area (which is what they are trying to do I think)...

In any case, I shoot RAW exclusively on my R1 and output from Bibble in 16-bit TIFF, so I'm used to big files. I figure a couple of extra bits couldn't hurt...

Cheers, Keith
http://www.kotay.net/keith/photo/photo.shtml

Just wait, the bit depth will become a marketing race like MP. It's already started.

I've seen it in audio where they moved from 16bit to 24bit and now some 32 and 48 bit. And that's just recording, greater bit depth is used for processing. And the sampling rate is moving through 96khz headed for near 200khz. And audio files already made photography disk eating seem like child's play.

Want to guess how soon it's 32bit or even 64bit in photography?

Walt

12 or 14 bit is not that important. More important is the distribution of tonal range. I can confirm that the A700 performs extremely well. When I am working with astrophoto tonal range is essensial, and the A700 delivers much smoother gradients than the KMD7 and A100. Also, the noise level and smoothness of dark areas is much, much better than anything I have seen from my previous Sony and KM digital cameras.

This 30 minute exposure (5 x 5 minutes) of the Pleiads shows very smooth dust and gas areas around the bright stars. ISO 1600, 600 mm focal length.

You are waking up the old astronomy geek... who dreamed of taking shots like that years ago.

What are you using...
What is the best budget set up so I can start trying to do shots like that...

I see lots of low cost goto tracking scopes under $1000.

used to be quite adept at finding stuff with setting circles and a star chart. Are any of these drives going to track accurately with the A700 on them?

------------
Ken - KM 5D (A700 Joy)
http://www.cascadephotoworks.com

Nordstjernen said:

12 or 14 bit is not that important. More important is the
distribution of tonal range. I can confirm that the A700 performs
extremely well. When I am working with astrophoto tonal range is
essensial, and the A700 delivers much smoother gradients than the
KMD7 and A100.

Click to expand...

Hi Nordstjernen. As I said in another post, smoother gradients is the purpose for having more bits. If you feel that the A700 has better gradients, perhaps it is due to the extra bits?

Nordstjernen said:

This 30 minute exposure (5 x 5 minutes) of the Pleiads shows very
smooth dust and gas areas around the bright stars. ISO 1600, 600 mm
focal length.

Click to expand...

Great shot! I assume you shoot RAW--am I correct?

Cheers, Keith
http://www.kotay.net/keith/photo/photo.shtml

I don't think 12 or 14 bit is the big question, but how the bit depth is treated by the camera and software.

I do shoot cRAW + jpg. For my astro pictures and most other work I use the cRAW data.

I don't know if these "cheap" telescopes do track well enough for the A700 and prime focus photography. Since I got my Losmandy G-11 equatorial mount and Takahashi FS-102 telescope I have been very satisfied with it. So actually I don't know todays marked. Try an astro forum!?!

And NO! I will NOT stop that!!! Please go to this link to see some more of my A700 astrophotos:
http://www.dyxum.com/dforum/forum_posts.asp?TID=23019

Hi guys,

Many of you will know that 12bit/14bit is done at the ADC(Analog Digital Conversion), which means each pixel of CCD/CMOS sensor data was analog. Through the ADC it will change to a discrete value. This is going to be boring so I'll make it short and clear.

"Quantization noise" is a model of quantization error introduced by quantization in the analog-to-digital conversion (ADC) process in telecommunication systems and signal processing
(omited ...)
16-bit ADC has a maximum signal-to-noise ration of 6.0206 · 16=96.33 dB

• from wikipedia ( http://en.wikipedia.org/wiki/Quantization_error )

N-bit ADC SNR will be 6.02N dB

You can see 16-bit ADC SnR ratio is 96.33 dB.
14-bit ADC SnR ratio will be 84.29 dB
12-bit ADC SnR ratio will be 72.25 dB

As you see, more bits will result less noise.

What do you think about this?