gogopix
Graf
On the DPreviw forum there was a thread by Andrej Kolev who is trying to make a DNG raw converter.He may have broken the code on why the M8 has only 10MB files and the DMR 20MB.
He says the M8 files have only 8 bits per channel rather than 16, but that they are non linear mapping to 14 bit luminance range per color RGB. That means in fact M8 has less information than DMR - since the DMR is linear and 16 bit.
When you map 8 bits non-linear to 14-bits linear you can recover dynamic range but you do lose detail,since you dont know all the gradations in between. I wonder why Leica would do this.
It could be he is wrong, and the lossless compression just codes redundancy and thus all detail can be recovered (just mapping 8 bit to 14 bits will NOT do it.
There are a lot of situations where non linear mappings lose very little since the eye (as well as ear) are logrithmic devices, that is, it can take a factor of 10 i nsound to have a perceived 'doubling' of sound (that is why 10-20 db is about 'twice' as loud perceived, but sound energy is actually 10 times!
maybe Leica tested and found a clever mapping that preserves detail, but going from 20MB per image to 10 MB for same number of pixels is half the infrormation. Whether it makes any difference only time will tell. If the coding uses redundacy(like patches of same color) and just compresse more cleverly, well then nothing will be lost.
But since we have SO much time on our hands I thought it would be interesting to speculate on something besides lenses and Seals
Ok, Ok many will notice same topic a few fora. I want to see where the technical guys are!
So who can address this? (BTW images may not REALLY be affected)
Victor
PS Pre-order FIRMLY staying in place
He says the M8 files have only 8 bits per channel rather than 16, but that they are non linear mapping to 14 bit luminance range per color RGB. That means in fact M8 has less information than DMR - since the DMR is linear and 16 bit.
When you map 8 bits non-linear to 14-bits linear you can recover dynamic range but you do lose detail,since you dont know all the gradations in between. I wonder why Leica would do this.
It could be he is wrong, and the lossless compression just codes redundancy and thus all detail can be recovered (just mapping 8 bit to 14 bits will NOT do it.
There are a lot of situations where non linear mappings lose very little since the eye (as well as ear) are logrithmic devices, that is, it can take a factor of 10 i nsound to have a perceived 'doubling' of sound (that is why 10-20 db is about 'twice' as loud perceived, but sound energy is actually 10 times!
maybe Leica tested and found a clever mapping that preserves detail, but going from 20MB per image to 10 MB for same number of pixels is half the infrormation. Whether it makes any difference only time will tell. If the coding uses redundacy(like patches of same color) and just compresse more cleverly, well then nothing will be lost.
But since we have SO much time on our hands I thought it would be interesting to speculate on something besides lenses and Seals
Ok, Ok many will notice same topic a few fora. I want to see where the technical guys are!
So who can address this? (BTW images may not REALLY be affected)
Victor
PS Pre-order FIRMLY staying in place
Last edited:
gogopix
Graf
No replies?
does every one realize that this is fundamental?
I have no fear that Leica is throwing away data, but come on, should be at least ONE technical Math guy here!
does every one realize that this is fundamental?
I have no fear that Leica is throwing away data, but come on, should be at least ONE technical Math guy here!
AndyPiper
Established
Leica Users Forum went all over this - it is USUAL for 10-12 Mpixel cameras to produce RAW files of around 10 Mbytes. The exceptions tend to get slammed for their "outrageous" file sizes (Sony R1, DMR, both at around 20Mb).
A RAW file is essentially a 16-bit grayscale TIF file - one 16-bit brightness value from each pixel (plus some administrative overhead). There are several ways to losslessly compress/decompress that 2-bytes-per-pixel data to get a 10-Mbyte package. And then decompress the compressed file to get 20Mbytes of RAW data and demosaic THAT into a 30 or 60 Mbyte (8 or 16bit) full-color RGB image.
But like you, I'll have to wait for a math guy to detail all of them.
A RAW file is essentially a 16-bit grayscale TIF file - one 16-bit brightness value from each pixel (plus some administrative overhead). There are several ways to losslessly compress/decompress that 2-bytes-per-pixel data to get a 10-Mbyte package. And then decompress the compressed file to get 20Mbytes of RAW data and demosaic THAT into a 30 or 60 Mbyte (8 or 16bit) full-color RGB image.
But like you, I'll have to wait for a math guy to detail all of them.
jlw
Rangefinder camera pedant
gogopix said:
That does seem to be the simplest likely explanation...
ClayH
Diana camera, coffee
This particular 'the sky is falling!!' discussion was beat to death on the leica-camera forum last night. Bottom line is that unless you happen to be the engineer who created the chip and firmware, most comments are just ridiculous speculation at this point. The RAW file images that are beginning to leak out seem to be pretty good. I don't think Leica is going to turn out a $5000 camera-phone.
gogopix
Graf
The sky is falling??
I'm ALWAYS the last to hear these things!
But have you seen the cosmic ray thread? LOL
I'm ALWAYS the last to hear these things!
But have you seen the cosmic ray thread? LOL
greggebhardt
Well-known
gogopix said:The sky is falling??
I'm ALWAYS the last to hear these things!
But have you seen the cosmic ray thread? LOL
It has been discussed, the Sky is Falling, but it is OK!
Oh Two
Established
Decompression?
Decompression?
Decompression is for divers. Expansion is for.......?
Decompression?
Decompression is for divers. Expansion is for.......?