form wrote:
I think high ISO is pretty good. There don't seem to be too many lighting situations that are that interesting that justifies/needs greater than a very good ISO25600 with f/1.4 primes. Although we could go farther to get faster shutter speeds, it seems very adequate for most purposes.
To me, low ISO dynamic range seems more important because it's something that anyone can derive benefit from if they ever work in high contrast environments, e.g. landscapes or anything outdoors. It's not really an extreme or rare situation, it's mostly like having the ability to retain daytime skies and people in the shadows in a single image, or retaining some detail in subjects that would be either completely blown-out or completely unrecoverable from the darkness....Show more →
In that case, I'd definitely switch to Nikon.. Canon won't ever release a camera capable of low ISO dr. It's pointless waiting.
There's no guarantee they won't come up with good low ISO dynamic range sensors in the next few years, I just feel it isn't all that likely for the next generation.
I intend to keep a set of Canon equipment at any time, even when I use Nikon gear. I make changes gradually, not suddenly.
form wrote:
There's no guarantee they won't come up with good low ISO dynamic range sensors in the next few years, I just feel it isn't all that likely for the next generation.
I intend to keep a set of Canon equipment at any time, even when I use Nikon gear. I make changes gradually, not suddenly.
I agree. Very very unlikely. But I'd bite the bullet and just switch if I was you. Seize the day!
form wrote:
I think high ISO is pretty good. There don't seem to be too many lighting situations that are that interesting that justifies/needs greater than a very good ISO25600 with f/1.4 primes. Although we could go farther to get faster shutter speeds, it seems very adequate for most purposes.
To me, low ISO dynamic range seems more important because it's something that anyone can derive benefit from if they ever work in high contrast environments, e.g. landscapes or anything outdoors. It's not really an extreme or rare situation, it's mostly like having the ability to retain daytime skies and people in the shadows in a single image, or retaining some detail in subjects that would be either completely blown-out or completely unrecoverable from the darkness....Show more →
High ISO isn't just for interesting lighting conditions. Sometimes it's for making a photo despite the lighting conditions.
These problems of high ISO and low ISO have existed since the invention of photography. Photographers have dealt with them in many ways, from using flash and other added lighting in dark settings, to using reflectors and fill flash in bright contrasty conditions. Of course, not every situation has a practical or easy solution. Likely there will never be a sensor that covers every situation perfectly.
I'm all for improvements in every area, from lenses to flashes to autofocus, etc. Dynamic range is somewhere on the list, but not particularly prominent and not something I would switch systems over. I have thousands of photos from various recent Nikons and there is absolutely nothing in their dynamic range that would make me want to switch. Differences in color stand out for me much more than any differences in dynamic range. Whether they are good or bad photos depends on too many other factors.
Having learned photography in the film era and having used the crummy first generation of digital, I feel that current sensors from Canon and everybody else are pretty darned good. Sure there are differences, but for me not compelling ones.
I think Canon really did not have much to offer in APS-C technology at this moment in time that is why they delayed till the last moment before Photokina and released a new version of the 7D with improved AF and some other B&W. If there was new sensor tech on offer it would have been highlighted and promoted much more aggressively. 7d's were selling below a $1000 but now atleast they can try to command a more decent price at this prosumer level.
Fred Miranda wrote:
An open source tool would be great!
This goes in line with the early samples we are seeing. The new camera' sensor is improved from its predecessor and just as good as the 70D.
If you want to poke about the masked areas yourself and such, do a search for a program called "iris" and download it. It lets you examine the true RAW file, no conversion and you can find the blackpoints, whitepoint, StdDev, set a narrow range around the blackpoint to examine shadow banding characteristics, etc.
form wrote:
There's no guarantee they won't come up with good low ISO dynamic range sensors in the next few years, I just feel it isn't all that likely for the next generation.
The 5D3 doesn't have the low ISO DR you want. The D800/D810 has amp glow at higher ISO that you don't like. The D750 doesn't have 1/8000 shutter or a PC sync port.
Form, at some point you are going to have to choose the lesser of evils and move forward. There has never been one piece of gear that I have owned where I didn't say I wish it had that, or I wish it didn't do that. Unless you find a company to custom make your dream body you are going to have to accept less then perfection.
dgdg wrote:
I guess Joey we've categorized everyone now?
1. Complainers
2. Complainers about the complainers
3. Coveters silently complaining
4. Complainers about those so perplexed that no equipment will ever make them happy (a subset of #2 I suppose but worth noting separately)
5. Those happy taking photos with their current gear, have fun talking about it with others on the forum, ready to purchase/use new gear from time to time as desired for continued enjoyment.
David
I'm glad we have category 5! For what it's worth, I'm excited about the 7DII, and my preorder is in. Canon focused on the things I needed with APS-C: fps and high ISO performance. Perhaps the 5D4 will bring bring higher DR for low ISO landscape shots.
Here's the bottom line in the next 8 weeks: If you want a crop camera action body from Canon, you can get a 7Dm2 which has noticably better IQ than the original 7D even if it's only as good as the 70D sensor. If you want one from Nikon, you can get a D300s with 12mpx which has more noise and less DR at any ISO over 200 compared to the 70D sensor. Oh, and you can get tele lenses which weigh about 2 lbs more.
I never truly believed Canon would drop it's very best new fabbed sensor in a sport body. I guess we'll find out soon enough (if Psychic1 is right) whether they have made any improvements.
form wrote:
Right now the least of the evils for me is the D750. It's also the least expensive of the evils.
Only gets you about 9.5 mpx on the duck as compared to the 7D2.
I don't know... When it comes to my wildlife body, I'm not sweating low ISO DR that much. Like I said either in this thread or earlier, I'm usually shooting wildlife at an ISO where the Exmor DR advantage is mostly gone.
mabidally wrote:
I think Canon really did not have much to offer in APS-C technology at this moment in time that is why they delayed till the last moment before Photokina and released a new version of the 7D with improved AF and some other B&W. If there was new sensor tech on offer it would have been highlighted and promoted much more aggressively. 7d's were selling below a $1000 but now atleast they can try to command a more decent price at this prosumer level.
I guess you can see that glass as half empty. I see that the 7D2 has Canon's best APS-C camera ever, with a long list of improvements over its already rather good predecessor. Unless you have some first-hand information from Canon headquarters, I don't see what info supports your belief that they "delayed till the last moment" for anything related to the 7D2 or Photokina.
1. Canon's latest sensor tech is shown in lower end cameras below the 5d3, 1dx, etc. That is their latest mainstreamed tech, what's in the 7d II.
2. The latest sensor tech does not significantly improve dynamic range at low ISOs.
3. The primary improvement to the latest sensor tech is a duel pixel focus system enhancer for live view.
From this I draw the (IMO plausible) impression that they are more interested in enhancing live view focus than with other things. From that I can get the idea that low ISO dynamic range is not something they are as worried about as other things (like live view focus). From their statements of solid sales in these camera sectors, I can get the idea that they feel generally comfortable with how their products are selling in the market. If the products are marketing well, they will probably be more inclined to stay the course. That course seems to be paying little attention to low ISO dynamic range.
Stay the course, it's working financially. That matters more than delivering what a niche few "complainers" voice their desires to have, while the vast majority continue to denounce the complainers while silently some of them wish for the exact same improvements (and will laud Canon for bringing those improvements home if it ever happens).
Don't be so pessimistic!. Canon holds several patents regarding on-chip ADC on each row/column of pixels. And surely they indeed know this lack is their weak point, and are silently listening their clients.
A few days ago I found an interesting document. It seems that there exist so called well capacity adjusting schemes to improve dynamic range, which may cause or not a worse image quality depending on their implementation. This document studies a couple of them:
Fortunately, it seems that not a single DSLR maker is currently using the well capacity response adjustment (something I have personally think of in the past, without knowing its name and its drawbacks). It is better to have a moderate DR than a huge DR with poor SNR. The other scheme studied in the document would be too difficult to achieve (e.g. one ADC per pixel).
Meanwhile we hope that Canon will simply follow the Sony route, without complicating our image quality tests with more variables than that of the DR...
skibum5 wrote:
If you want to poke about the masked areas yourself and such, do a search for a program called "iris" and download it. It lets you examine the true RAW file, no conversion and you can find the blackpoints, whitepoint, StdDev, set a narrow range around the blackpoint to examine shadow banding characteristics, etc.
Thanks for the reference. I think that I have heard of it in the past, but did never take a look (maybe because it seemed to be a windows-only app, and I usually reject such apps at first). I also have heard of it recently in a ML forum thread. Googling there are several Iris out there, it seems that it could be just this one, am I right?:
I just have written a small app to read the PGM files generated by dcraw -E or -D and perform simple analysis on them (stdev, substraction, etc). The idea would be to use also the metatada provided by dcraw/exiftool to perform automatically basic tasks without switches or user intervention, whenever possible (e.g. add the proper intelligence to find the usable masked pixels) or at least allow automating bulk data processing over a set of test shots (even if requiring switches). Of course, command-line driven only...
cgarcia wrote:
Canon holds several patents regarding on-chip ADC on each row/column of pixels.
This is very good news!
I was afraid that patent restrictions would prevent them from moving in that direction. I am sure that we will see new full frames in the next 6-9 months. They were rumored for February/March time frame.
In any case, Canon never announces 2 major DSLR body upgrades at the same time. Thx!
I was afraid that patent restrictions would prevent them from moving in that direction. I am sure that we will see new full frames in the next 6-9 months. They were rumored for February/March time frame.
In any case, Canon never announces 2 major DSLR body upgrades at the same time. Thx!
Last year the SL1 and T5i were rolled out simultaneously.
cgarcia wrote:
Don't be so pessimistic!. Canon holds several patents regarding on-chip ADC on each row/column of pixels. And surely they indeed know this lack is their weak point, and are silently listening their clients.
I am curious if Canon is struggling with the implementation of both DPAF and on-chip ADC on their sensors.
cgarcia wrote:
The other scheme studied in the document would be too difficult to achieve (e.g. one ADC per pixel).
Should the fabrication processes used to produce camera sensors catch up with what is being used to produce CPUs today, an ADC per pixel (or Bayer quad, or...) would likely become possible.
Even Sony's process is around 10 years behind commercial fabs, which are themselves years behind the best, i.e. Intel, Samsung, TI, and a few others that I'm forgetting.
Thanks for doing the tests very interesting
I'm hoping that iso 1600 on the new camera is as good as 800 on the old 7D , it looks like it will be close
I'm normally at iso 400 or 800 and for my subjects get great IQ
sometimes I want to go higher but don't as my 7D isn't quite as good at 1600 and above
At the moment I open up the apeture and focus stack but it looks like new camera will make my life easier
