Re: Myth busted: Canon's 14-bit snake oil
All this 14 bit discussion is really funny! If I was a Canon marketing manager I'd have gone all the way and put in all Canon cameras, in big shiny letters, 32 BIT SUPER EXTRA HIGH QUALITY RAW IMAGE FORMAT! (who cares?, the "real" quality is 10 bits anyway, so I can put any number just to attract customers! [:D])
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Jon Ruyle
if the user is patient, it can be virtually eliminated it in software- improving DR by several stops.
Well, it's only the late-stage read noise component that is improved here (e.g. noise in the row-level components, secondary gain stage, ADC, etc.). For Canon, that's where most of the read noise is, so we could get 2-3 stops improvement at ISO 100 with multiple readouts. It's similar to the benefit we get with image stacking in astrophotography (but not as huge). For Nikon the benefit is smaller because their read noise is more largely composed of early stage read noise (e.g. rts, trapped carriers, etc.).
Quote:
Originally Posted by Jon Ruyle
I mean, why stop at 4 reads? Why not do 100? Or as many as it takes?
It would be nice if Canon let us control options like that.
Quote:
Originally Posted by Jon Ruyle
Metadata ISO and auto ISO are a horse of one color.
You're quite right.
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
<div>
Quote:
Originally Posted by Alan
The bigger question is: when is the technology going to be available to freeze my head?[img]/emoticons/emotion-2.gif[/img]
<div>[:D]</div>
</div>
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
<div>
Quote:
Originally Posted by Carlos Lindado
All this 14 bit discussion is really funny! If I was a Canon marketing manager I'd have gone all the way and put in all Canon cameras, in big shiny letters, 32 BIT SUPER EXTRA HIGH QUALITY RAW IMAGE FORMAT! (who cares?, the "real" quality is 10 bits anyway, so I can put any number just to attract customers![img]/emoticons/emotion-2.gif[/img])
I know, me too. If the customers ask why our new 32-bit files are the same size as 10-bit files, we can tell them it's an advanced new lossless compression technology called "debunking".
</div>
Re: Myth busted: Canon's 14-bit snake oil
After reading all of this, i feel bad for scooby, he lost his job on this one :/ poor scooby-doo, Dang it Daniel why you make him lose his job. ^^
I wonder if canon is like "oh snaps we been busted" pack up and run away, or "man now we have to do some work around here"
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Steven23
I wonder if canon is like "oh snaps we been busted" pack up and run away, or "man now we have to do some work around here"
How do we make them, and everybody else, notice this discussion here?
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Colin500
How do we make them, and everybody else, notice this discussion here?
<div style="clear: both;"]</div>
We don't. Daniel's tried, and all he gets is form letters in return.
Canon is in the business to make money and they, like any other company, gauge their business with marketing studies. They're not going to listen to a tiny minority, when they can sell bazillions of cameras now, knowing that they can improve them incrementally, and have the same people that bought the previous versions, go out and buy the new ones. It's just the way the business world works.
That's why these discussions don't do much, other than let people banter/rant about what could be, or should be.
It's a Mobius strip conversation, which has no end.
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Alan
That's why these discussions don't do much, other than let people banter/rant about what could be, or should be.
Not much?!? Some of us live to rant [:)]
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Jon Ruyle
Not much?!? Some of us live to rant [img]/emoticons/emotion-1.gif[/img]
And some of us live to find those people and goad them into ranting. (You just love seeing me foam at the mouth, don't you, Jon?) [;)]
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Daniel Browning
And some of us live to find those people and goad them into ranting
[8o|]
Re: Myth busted: Canon's 14-bit snake oil
Quote:
Originally Posted by Jon Ruyle
Quote:
Originally Posted by Daniel Browning
And some of us live to find those people and goad them into ranting
[img]/emoticons/emotion-39.gif[/img]
<div style="clear: both;"]</div>
I love a good rant, too. I doubt, though, that Daniel has foamed at the mouth. [:D]
Re: Myth busted: Canon's 14-bit snake oil
I forgot to respond to this earlier:
Quote:
Originally Posted by Jon Ruyle
I looked at DXOMark. They give the D700 a dynamic range score of 12.2 stops and the 5DII a score of 11.9 stops. Doesn't this mean that the D700 has only 0.3 stops more dynamic range? What am I missing? (One way to reconcile this would be that the canon exposes for more highlight headroom than the Nikon, but I don't think that is true).
I meant to say D7000, not D700. The D7000 is where Nikon really made leaps and bounds in dynamic range. As for the 5D2's 11.9 stops, that's because DxOMark only measures random noise. If they were to measure non-random noise such as "banding" (temporal FPN), they would show a much lower number -- I'd say less than 10 stops.