PhilDWedding wrote:
I think this is unlikely if I'm honest. I'm expecting the 1dx to be a "little bit" better than the d3s, according to Canons own view of the two stop increase compared to the 1d4.
That would make it 1 stop better than the 5d2 and about 1-2 stops better than the 5dc.
ISO3200 -> ISO25,600 is 3 stops and I doubt that... but we'll see I suppose.
yeah sadly he is dreaming if you he thinks the 1D X will have 25,600 looking like 3200 on the 5D2
3 real stops is, for one, better than the 2 stops they mention! and even for 2 stops they say it's a combo of better real performance and better jpg NR so....
alundeb wrote:
1 Stop better than the 5D2 at ISO 3200 unprocessed is realistic, but at the maximum I would expect, and still very difficult to get to. It will mean a sensor effiiciency of 66%, The D3s gets to 57% (sensorgen.info)
However, the read noise signature of the 5DII is not pretty, showing in as banding and blotching at 12800 and above. The potential for multiple stop improvements lies in at what extreme ISO you can get an image that doesn't look like crap. That may well be 2-3 stops better than the 5DII.
Lastly, we have been disappointed by Canon marketing claims about noise performance before..... ...Show more →
yeah they certainly could get a good 3 stops better true not banded DR out of it than the 5D2 if they really did fix it up, but they are still using the same A/D non column type
Hard for me to get excited about this, certainly not to the point of replacing my 1Ds MarkIII. FPS is nice for the sports, wildlife photographers, unimportant for the landscape and architecture I do. More ISO unimportant for me, I never go above 800 anyhow and almost never go even that high. Don't know enough about DIGIC 5 to know how significant it will be. All in all there are some nice refinements but not the blockbuster I was expecting four years after the 1DsIII was introduced and the price seems really high for what it is. The only thing that might tempt me would be if after a year or so of experience there's a general consensus that IQ is noticeably better than the 1DsIII.
Seems kind of funny or ironic or weird or something that Canon originally was the big mpx company, Nikon was the less mpx, better noise company. Then Nikon joined Canon with more mpx at the high end. Now Canon joins former Nikon with fewer mpx.
DanielPaul wrote:
If the 5D-follow-up is the new MP king (meaning 36MP+), it will not be the pure IQ king. More MP != more IQ. It would win in things like resolving power, ability to crop, and all the things that are guaranteed with more MP. But with that would be more noise, artifacts, and all the things that come with making pixels smaller.
but since when is it fair to compare cams at the pixel level?
you need to do normalized comparisons
also before you think higher MP is so utterly ruinous look at crops taken from same absolute area of 5D2 and 7D sensor and the 7D gives a touch better SNR than 5D2! In fact if you shoot a far off bird with both and then downsample the 7D bird to the 5D2 bird pixel size you will get the bird from the 7D cam to have slightly better SNR and LESS artifacting since the demoasic errors have been averaged away by the downsample!
Ralph Conway wrote:
:-) To be honest: I was talking about native ISO. That is, what the sensor does. 5D II does not support ISO 12.800. That is just calculated and imho a marketing gag (or: better a bad image than none).
I guess, Canon would not have set D1Xs native ISO to 51.200 if one can not use the images done.
That is the really improvement that blows away all competitors. Next to AF and 12 FPS.
Letīs wait for the pictures.
Ralph
native iso is just marketing gag too
heck nikon had an ISO6400 cam almost a decade ago
on a 5D2 the ISO1600 is the last true ISO where the extra amping does any good, beyond that it just clips off more highlights each stop up while not controling noise any better, same for most canon models, although the 7D and probably 1D4 do make ISO3200 a touch better than ISO1600
alundeb wrote:
There is no reason to continue the resolution debate, since you have obviously never experienced how people react to images with detail level at 200MP equivalent.
Yes I did. And it was amazing. What now?
Your comments about image noise reveal that you know little about it. There is plenty of literature. This is a bit technical, but covers the main aspects:
Thank you. As far I can remember I did nothing mention anything about image noise. I was talking abou how great a comparable 5D MK II ISO 3.200 noise level would be to me if comparable to ISO 51K of 1DX.
alundeb wrote:
I will join your euphoria in my own way, I am happy that not all photographers use high resolution as a tool to produce stunning images, which means we can differentiate our aesthetic expressions and all be happy
Most of the time I use high resolution as a tool to produce mediocre images. When is Canon going to introduce a stunning image feature?
After working with the d7000 and the 7d I find your findings to be wrong. Also the buffer on the d7000 is crap.
I highly doubt you'll see your desired bodies from canon ( the ones you list). How many product lines would that be?
Rebel, xxd, 5 series, 3 series, 1 series, 7 series. skibum5 wrote:
and yet nikon keeps putting better and better AF into all their models, i see say the D7000 even tracks better than the 7D
come on, 5D needs serious best effort AF
anyway:
1DX 18MP, super speed 12-14fps, FF, giant brick
5D3 34MP, 6.3fps, prior 1 series level AF at the worst, FF, compact with optional grip
or maybe the latter split to
5D3 38MP4-6fps, prior 1 series level AF, FF, video with a bit of line skipping but a little better than on current models
3D 26-28MP, 6.3fps, prior 1 series AF at worst, FF, flawlessly sampled video
5 and 3 series small bodies with optional grips
the truly serious sports shooters need 8fps min so they have to get 1DX so 1Dx is protected for it's main target market (although the way canon talks they make it sound like they want the 1Ds crowd to all get 1DX instead of D800/5D3/3D but I don't see how that will hold up) ...Show more →
alundeb wrote:
There is no reason to continue the resolution debate, since you have obviously never experienced how people react to images with detail level at 200MP equivalent.
How do people react? I do not believe I have seen any kind of 200MP image from a camera. What should I expect?
alundeb wrote:
I will join your euphoria in my own way, I am happy that not all photographers use high resolution as a tool to produce stunning images, which means we can differentiate our aesthetic expressions and all be happy
But I don't understand how you expect a sensor to collect 240% of the light falling onto it, that is what is necessary to get 3 stops ISO improvement with 5dII 3200 as a reference
Uh, what?
Canon didn't claim 3 stops.
If they did though, it could be achieved by raising the quantum efficiency 50% (from 40% to 60%, which is a bit above where the D3s is) and then dropping the read and amp noise by 75%. Now, a 3 stop improvement in a single sensor generation is insane, but it's not as if it's a complete absurdity. 2 stops is entirely possible, in some ways I'd expect no less.
The buffer is horrible, yes. But it's not as horrible as initially thought. you get a second and a half of shooting in RAW/RAW+JPEG Fine, versus just over a second on a 60D and 2 seconds on a 7D. And it clears faster than the 60D with UHS-I cards. In other words, it's actually better than its direct competition and competitive with the class above.
It does track well. I shoot wakeboarding and capoeira quite often, and my in focus rate is nice and high, often higher than a couple of the 7D shooters at the lake (though that's obviously anecdotal). It tracked as well as the D300s I tried, though initial lock on was definitely slower and it has more of a tendency to focus on the background until you get used to the larger sensors (the sensors at each point are bigger than the D300 equivalent).
Whiteknight over.
I agree with you on the product lines though - I expect there'll be 6-7 DSLRs in the lineup, though the bottom two may get eaten by mirrorless cameras sooner rather than later.
RobertLynn wrote:
After working with the d7000 and the 7d I find your findings to be wrong. Also the buffer on the d7000 is crap.
I highly doubt you'll see your desired bodies from canon ( the ones you list). How many product lines would that be?
Rebel, xxd, 5 series, 3 series, 1 series, 7 series.
tonywong88 wrote:
Most of the time I use high resolution as a tool to produce mediocre images. When is Canon going to introduce a stunning image feature?
If they did though, it could be achieved by raising the quantum efficiency 50% (from 40% to 60%, which is a bit above where the D3s is) and then dropping the read and amp noise by 75%. Now, a 3 stop improvement in a single sensor generation is insane, but it's not as if it's a complete absurdity. 2 stops is entirely possible, in some ways I'd expect no less.
My response was directly to Ralph, I know Canon didn't claim 3.
You are wrong, raising QE 50% only improves noise by approximately 1/2 stop. You need a doubling (100% increase) to improve one stop.
And again, reducing read noise by 75% does not improve midtone SNR at ISO 3200. It does help at ISO 51200 though.
