About 99.9% of the time, I am perfectly able to work within the dynamic range of my Canon cameras, whether by single capture in natural light, by adding lights or reflectors, or by exposure blending for static scenes. That doesn't challenge the validity of the call for greater DR. I completely agree with these two posts (my emphases):
Hrow wrote:
This is an absolutely crazy position to take and you may want to unbox yourself from the corner you have walked into. In your own pictures there are multiple examples where better shadow detail would be in order. Does that make you a bad photographer? Does it make you a lazy photographer for not carting around a ton of lights to properly light some of those classic cars? Wouldn't you rather have the inherent capability of being able to pull the shadows rather than shoot a car show with an assistant and multiple lights to get exactly the same effect? Would the car show organizers even let you in with multiple light stands and softboxes that might blow over and damage some very expensive cars? Do you want to spend the hundreds or perhaps thousands of extra dollars for a portable lighting kit to realize the same benefit of shadow detail that Nikon gives you in a camera that's already $500 cheaper?
Nikon and Sony can already do this so it isn't like anyone here is asking for the impossible or the unreasonable. They are asking that a beneficial, existing technology be available from a camera manufacturer with whom they have invested a lot of money. Seems perfectly reasonable to me. What seems unreasonable is carrying lights and reflectors around a car show (referencing your images) or backpacking them into the wilderness to be prepared for the off chance to shoot a happened upon mushroom, nursery log, or waterfall.
artd wrote:
This is exactly the kind of rationalization I don't understand. You are comparing a solo photographer working to get a single photograph to having the resources of an entire film crew?
So if you have a high contrast scene and you want to retain highlights and not have blocked up shadows, you have the following choices:
1. Set up extra lighting gear to balance the lighting of the scene to fall within your sensor's dynamic range.
2. Use no extra gear, but take additional exposures and use exposure blending in post production.
3. Use no extra gear, no extra exposures, take one shot exposed for the highlights and lift the shadows in post.
Isn't it obvious which option is the simplest and most efficient?
Options 1 and 2 are not even feasible a lot of times. If you are working in a studio, or if you have an entire production crew backing you up, sure, bring in all the lighting gear that you want. If you are shooting a static scene on a tripod, blending exposures works great (I do it a lot). But a lot of times we don't have such luxuries.
So again, it seems we are spending a lot of time constructing rationalizations instead of just accepting the facts. Canon's sensors don't have the same latitude as the competition. If you do option 3, your result will be worse if you are using a 5Dx sensor vs an EXMOR sensor. It doesn't matter how you spin it, that's just the reality. If your shooting and processing style never bumps up against the limitations of the Canon sensor, then that's perfectly fine. But some do. ...Show more →
morganb4 wrote:
So on the 5D3, at what ISO does the shadow banding become less obvious?
I don't know, but the dynamic range would probably become even with the D800 once you get to ISO1600, maybe a touch below. Wonder what is taking DxO so long.
digitalbug30d wrote:
just like when one sees things that arnt there...LSD,Psychotic break....
;-)
digitalbug30d wrote:
you can keep using your camera as a lab test tool fine by me..
double ;-)
skibum5 wrote:
maybe your only experience is with studio and set shots, natural world photography, random on the spot photography, is a different game
That's where I shoot most often - though not quite all the time. But I don't suppose it would matter if I pointed out that I don't have a problem makingphotographs of these subjects either with this technology, would it? ;-)
To be clear, I've often written that I'm happy to see continuously improving photographic technology.
Arun Gupta wrote:
I still can't find the 5D3 in www.amazon.fr (the French Amazon site). I'm told DXOMark is French.
Maybe Canon is making sure France is the last to get it, at least until all early 5D3 sales around the rest of the world are complete fearing how it will look compared to D800 sensor. (it does seem like everyone other than Switzerland and France have it now right?)
Have to say although I'm someone that has stated I haven't seen much issue with banding on my 5D II in shadows (noise is another matter) seeing what the Sony sensors can now do would be a godsend. HDR or even exposure fusion is not always viable for a multitude of reasons and being able to do the crazy pushing of shadows would open up a lot of possibilities. You can work around these things, but it's not always convenient and sometimes you don't even bother taking the shot. One example I can think of (and I'll post an example tonight) is shooting in caves. You have a lot of dark unlit areas full of beautiful formations, but you can't reveal the detail as you have had to expose for the one lit formation.
Just for reference...I find that sometimes my underexposures are the result of misinterpreting the actual exposure accuracy as visualized on the LCD output of my camera. For instance, some shots I expected to be about 2/3 stop underexposed based on my 5D LCD info actually turned out to be almost 2 stops underexposed, and they had to be brightened accordingly...other shots look to have some pure whites in them now but looked correctly exposed according to the LCD...before any real processing has taken place.
morganb4 wrote:
pwned
ok ok...well just a little but this doesnt mean I am going to jump ship to Nikon...yeah that bar example does impress the hell out of me...
gdanmitchell wrote:
To be clear, I've often written that I'm happy to see continuously improving photographic technology.
You have jamed most banding & shadow noise reports since the day 5d2 was born. What kind of photographic improvement you are talking about here...from you or Canon Neither of you done any good in term of 2003 IQ
could it be Nikons software is better at this than Canons,no-one has chime in about which software does a better job I would like to see comparisons of ACR,DPP,Lightroom,Photoshop with the same image pull shadow detail,and see which one does it without producing banding and the threshold thereof.
digitalbug30d wrote:
could it be Nikons software is better at this than Canons,no-one has chime in about which software does a better job I would like to see comparisons of ACR,DPP,Lightroom,Photoshop with the same image pull shadow detail,and see which one does it without producing banding and the threshold thereof.
does that make sense?
Google search 5d2 low iso banding...we have tons of these reports already archived on the web.
jeremy_clay wrote:
The banding is not software related unfortunately.
there is a chance software MIGHT be able to fix the 5D3 banding, there is a chance the type left in the 5D3 might be fairly consistent shot to shot (that said it won't reduce amp noise so you still won't be able to get anything like D800 DR but you would be able to fully use every little bit the 5D3 has, still over 2 stop less usable although it might close things a good half a stop) unlike with the 5D2 where significant parts of the banding vary too much frame to frame to remove well in software.
if i get one i will check it out and see if some pre-processing software can write out a new RAW with most banding removed or not
digitalbug30d wrote:
could it be Nikons software is better at this than Canons,no-one has chime in about which software does a better job I would like to see comparisons of ACR,DPP,Lightroom,Photoshop with the same image pull shadow detail,and see which one does it without producing banding and the threshold thereof.
Canon is going to need to either invest more energy into sensor development, license Sony sensors, or be content to lag behind in sensor technology.
speedmaster20d wrote:
A few points regarding FPN for those who are interested.
Canon use Analog CDS in their sensor architecture. In this scheme the reset signal is read first and stored in the form of charge in the CDS unit capacitors. Then the actual read signal is transferred to CDS. The differential charge remains in the capacitors until it is transferred to the ADC chip. The entire sensor is read row by row, one row at a time. In this method the FPN resulting from variation in pixel amplifiers are cancelled, however the variation in the CDS units themselves cannot be cancelled. This also requires high-clock ADC operation because the entire row of pixels is quantized at once. High frequency noise further degrades the signal stored in CDS units. As a result fixed patten noise from CDS will remain. If the signal level is too low (low pixel amplification) it will fall below the noise floor of the CDS circuit, therefore the post ADC data will show extreme banding and poor quality. This is why you cannot achieve results identical to ISO400 by underexposing your photo at ISO100 and then digitally scaling the data. The reason is analog CDS.
Sony on the other hand, have developed a clever technique that eliminates the use of analog CDS circuit. In this method signal is digitized right at the column readout line. The CDS operation is performed digitally by flipping the direction of ripple counters in the ADCs. Since every column has its own ADC, operation is parallel and does not require a high clock ADC. This eliminates high frequency noise. Since there is no analog CDS circuit there is variation (banding) that results from it and it is possible to read very small signals even without pixel amplification. This is called Column-parallel architecture
This method has been fully patented by Sony and exclusive to their image sensor technology. So until Canon comes up with a more clever technique without violating Sony's patents there is not much they can do to reduce FPN or read noise except for using higher precision fabrication and tighter yield criteria which increases cost significantly.
It is easy to sit at home and complain but these issues are quite difficult for the electrical engineers to solve. ...Show more →
;-)
Neither of you done any good in term of 2003 IQ