For those of you that think the 7D is good with noise...

idsurfer wrote in post #12609927
I'm not going to read this entire post (so I prob shouldn't even chime in) but I am going to have to say my 60D is pretty poor above 400 ISO. The idea of shooting at high (800, 1600) is absolutely ridiculous if you have any desire for clean images. I'm kinda new and have never used a FF, but after an outing with friends in a low light bar/restaurant two nights ago I won't be leaving the house without a flash again.

Are you exposing to the right at the higher ISOs? That may make a big difference in how much noise you get.

ETTR is not always possible and rarely useable under conditions where noise is going to be an issue.

So, what do I do ? I'm shooting a concert ISO 3200, F2.8 and 1/125s to get a reasonably accurate exposure. Tell me how I can use ETTR ?

Shoot at F2...

If you can show me how to do that with a zoom, I'd be greatful

Let's not forget, whatever a 7D can do with noise a 5d2 can do it better.

bohdank wrote in post #12610974
If you can show me how to do that with a zoom, I'd be greatful

Let's not forget, whatever a 7D can do with noise a 5d2 can do it better.

Use a 1.2 prime

bohdank wrote in post #12610974
Let's not forget, whatever a 7D can do with noise a 5d2 can do it better.

Except make lots of it...

rhys216 wrote in post #12609035
The fact is glass is fast becoming the bottleneck in resolution as sensor density increases,

We're not even close yet. I think you are coming from a viewpoint where you think that getting a white pixel next to a white pixel is a good or desirable thing. You get more and more resolution, when you get a finer sampling rate, even if getting 100% contrast or even 50% contrast between neighboring pixels becomes impossible.

That statement is so unwieldy that I will wait until you rephrase it, to reply.

With a very sharp lens, the AA filter of the 7D, and it's large pixels, are the main source of blur (blur due to pixel coarseness, of coarse, may be pixelated and not look blurred at all, but the basic idea that the location of photon strikes is obscured more by bigger pixels still holds true).

There is no brick wall limit at work here, except one: large pixels have a hard limit on resolution. Above a certain resolution, nothing can be resolved, at any contrast level.

rhys216 wrote in post #12609114
^^^
Rubbish, AA filters are needed (although my personally preference is the weaker the better) because moire can be a real problem, hence why most camera manufactures use them.

And if most people don't want/need this much data, why is Canon trying to pack soo many pixels into a 1.6 crop?

AA filters are not needed if the lens softness and diffraction, combined with high pixel density, provide enough blur. With much higher pixel densities than we have now, we can drop the AA filter with no penalties. We could drop it now if our lenses started at f/64.

TI didn't say that most people don't want current pixel densities (though many actually don't). What I said was that sufficient pixel density to fully resolve good lenses; IOW, much higher than we have now, would be a burden to people's current workflows. A compromise solution would be to have super high densities, but output smaller files, including reduced-size RAWs.

iso 5000 f2.8

John Sheehy wrote in post #12611624
We're not even close yet. I think you are coming from a viewpoint where you think that getting a white pixel next to a white pixel is a good or desirable thing. You get more and more resolution, when you get a finer sampling rate, even if getting 100% contrast or even 50% contrast between neighboring pixels becomes impossible.

"As an aside, I think this is probably going to put the last nail in the Megapixel war’s coffin. The 4/3 companies have already said 12 Mpix is as far as they intend to go. I suspect the full-frame manufacturers are going to call a halt at 30 Mpix or so, just because there’s no sense in it: they’re already out-resolving the quality control of their best lenses."

http://www.canonrumors​.com …-is-soft-and-other-myths/

John Sheehy wrote in post #12611670
AA filters are not needed if the lens softness and diffraction, combined with high pixel density, provide enough blur. With much higher pixel densities than we have now, we can drop the AA filter with no penalties. We could drop it now if our lenses started at f/64.

While I agree, with these two points in principal, I seriously doubt we will be seeing camera's from Canon/Nikon without AA filters any time soon, unless they release a camera purely for stills, because moire is a serious issue for video...

rhys216 wrote in post #12611852
"As an aside, I think this is probably going to put the last nail in the Megapixel war’s coffin. The 4/3 companies have already said 12 Mpix is as far as they intend to go. I suspect the full-frame manufacturers are going to call a halt at 30 Mpix or so, just because there’s no sense in it: they’re already out-resolving the quality control of their best lenses."

http://www.canonrumors​.com …-is-soft-and-other-myths/

And right above that he said this:

"To summarize, these points seem pretty obvious to me. I’m not certain they’re all completely correct". In other words, it is another internet opinion.

rhys216 wrote in post #12611852
"As an aside, I think this is probably going to put the last nail in the Megapixel war’s coffin. The 4/3 companies have already said 12 Mpix is as far as they intend to go. I suspect the full-frame manufacturers are going to call a halt at 30 Mpix or so, just because there’s no sense in it: they’re already out-resolving the quality control of their best lenses."

http://www.canonrumors​.com …-is-soft-and-other-myths/

And yet, the lenses on point'n'shoot cameras make possible useful pixel densities far in excess of what we get with even the highest resolution crop DSLRs, and the pixel densities on those crop DSLRs already yield resolutions far in excess of 30 megapixels (in particular, the 7D's resolution scaled up to full frame would get you 46 megapixels).

Either lens resolving power winds up scaling roughly with target sensor size (so that the smaller the sensor size, the better the resolving power limit), or it doesn't. If the former, then no format will be better than any other (within some limits) for resolving detail. If the latter, then DSLR lenses have a long way to go before they hit any particular limits.

The evidence thus far suggests the latter, for the resolving power of lenses has been improving over time as technology improves (the Sigma 17-50 f/2.8 OS, for instance, has resolving power capability that was reserved for primes previously).

rhys216 wrote in post #12611852
"As an aside, I think this is probably going to put the last nail in the Megapixel war’s coffin. The 4/3 companies have already said 12 Mpix is as far as they intend to go. I suspect the full-frame manufacturers are going to call a halt at 30 Mpix or so, just because there’s no sense in it: they’re already out-resolving the quality control of their best lenses."

http://www.canonrumors​.com …-is-soft-and-other-myths/

The problem is that people define lenses as "soft" or "sharp". It's not that simple. Just because a lens shows softness on an image from an 8mp image does not mean that the image it projects onto a 30mp sensor will not have more data to work with. And it's generally meaningful data (i.e. provide detail/information about the scene).

I doubt I'm the only one who has experienced this. You have an old DSLR (my first was a 300D), a lens shows itself to be a bit soft, yet even so when you put it on a newer DSLR you can still recognise more detail in the scene from the same lens.

I love more pixels. GIVE ME MORE! So long as the noise per sensor area does not get worse (and I'm certainly always happy for the SNR to get better!) then I welcome it.

As for the comments about higher densities but smaller output files, we already have lower resolution RAW files from the recent models. I can't say I've used them so I'm not sure if they're junk or something, but the idea has already been implemented.

micallef1990 wrote in post #12611709
iso 5000 f2.8

What's the rest of the EXIF info? SS?