Jump to content

12, 14, 16 bit DNG Files


Timmyjoe

Recommended Posts

Advertisement (gone after registration)

I am totally confused. Been using my M8.2 for about six months now, just loving it, and always shoot RAW/DNG files. I could have sworn the bit rate for the files was something above 8 bit, but when I import the files into Photoshop, they all read as 8 bit files. Am I missing something?

 

Can't seem to find the information anywhere. Again, I could have sworn the Leica digital M cameras shot in a higher color bit rate than 8. Am I wrong?

 

Thanks for any and all info.

Link to post
Share on other sites

The M8 digitizes 14 significant bits per pixel that are mapped to 8 bits (by mutiplying the values by 4 and taking the square root) to save space. The raw converter expands those values to 14 bits again. These won’t be the original 14 bit values, though, as only 256 different 14 bit values can be expressed in 8 bits, but finer distinctions made possible by storing the full 14 bits would mostly be noise anyway.

 

(This is the long version; the short version would be: Don’t worry, everything’s fine.)

Link to post
Share on other sites

Yes- but he is asking about the camera. I think the confusion is arising from not differentiating between color depth of the camera and the ( unrelated ) bit rate in postprocessing. However, basically we agree, just have the raw processor expand to 16 bits files and all is fine. although I tend to suffer from paranoia and expand DMR and uncompressed M9 files to 32 bits...:(

Link to post
Share on other sites

Advertisement (gone after registration)

Jaap, the OP is saying that the files open up in PS as 8 bit files. That's a function of the raw processor.

Actually I was wondering what he meant. Since he talked about the M8 resolving more than 8 bits, I assumed he was referring to the number of bits per pixel in the original raw file. Obviously the colour depth of an image in Photoshop has nothing to do with the number of bits in the raw file.

Link to post
Share on other sites

Thanks for the responses. I am trying to understand how to get the most color depth from my M8.2.

 

So Michael, what you are saying is that the M8.2 records in 14 bit color, but then compresses that 14 bit color down to 8 bit, thereby eliminating the extra color information?

 

And what raw converter expands them back out to 14 bit (though it doesn't sound like that is going to help me with getting better color depth)? I usually dump the RAW/DNG files directly into Aperture 2, then edit them in Photoshop CS3.

 

Is there anyway to get real 14 bit color depth from the M8.2?

 

Thanks for any and all info.

Link to post
Share on other sites

So Michael, what you are saying is that the M8.2 records in 14 bit color, but then compresses that 14 bit color down to 8 bit, thereby eliminating the extra color information?

Strictly speaking it isn’t colour data that is eliminated since the raw file only records luminance data. Each pixel is sensitive to either red, green, or blue light; the amount of light in the other two channels has to be interpolated from the data of neighbouring pixels sensitive to these colours.

 

And what raw converter expands them back out to 14 bit (though it doesn't sound like that is going to help me with getting better color depth)?

Every raw converter compatible with the M8 will do that. The 8 bit data ist first expanded to 14 bits; the actual internal calculations may then be performed with 16 or 32 bits, or using floating point arithmetic, depending on the converter used.

 

The colour depth of the eventual image depends on the raw conversion settings; you can usually choose between 8 and 16 bits per channel (i.e. 24 or 48 bits for the RGB data). This has nothing to do with the number of bits per pixel in the raw file.

 

Is there anyway to get real 14 bit color depth from the M8.2?

If the question is: Can one make the M8.2 store uncompressed 14 bit data? then no, it isn’t supported. Only the M9 offers that option. But then it is not like you are missing much. As mentioned above, the difference between uncompressed and compressed data is mostly noise anyway.

Link to post
Share on other sites

  • 2 weeks later...

Tim, I believe Michael wrote an exhaustive article for LFI on the topic of the M8's luminance-information-compression-algorithm, some years back. I don't remember the exact issue, but you should be able to find it via the LFI store. It's a pretty enlightening read, and once you understand (as far as us mere mortals can understand such high-tech chit-chat) how it works and what it does, you understand why more bits wouldn't mean more quality for the M8's files.

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...