Messages

1

EOS Bodies - For Stills / Re: 5D2 refurb or new 7D as backup to 5D3?

« on: March 23, 2013, 03:46:31 PM »

its a simple fact thats not too hard to grasp that the image quality from canons full frame cameras is significantly better than their crop cameras to date especially at higher iso

If this is such a simple fact, then why is it no one ever posts proof? In the 3.5 years since the 7D was released I have yet to see even one test that showed "significantly better" IQ from the 5D2 over the 7D through ISO 800. Even at 1600 and 3200 I would only call the IQ difference "significant" for large prints. (The 5D2 is usable at 6400 where the 7D really isn't, except for small prints or web shots.)

Lots and lots of posts claiming "it's a fact!" and "everybody knows it!"...but no pictures.

I've had both. I've made plenty of prints, including plenty of 16x20's. At low to mid ISO there's no difference that survives post processing.

Now that we have the FF hyperbole out of the way...for regular professional wedding work I would be torn. Weddings often present terrible light levels where every stop counts. Then again, having one crop and one FF body, and the right lenses, can really expand your range without having to change lenses. (I've shot with crop only and with both.)

I would lean towards one of each...depending on lenses...but I would not fault someone for choosing 2 FF bodies for wedding work. OP is going to have to weigh the non-work advantages (reach; speed) against the work situation (lens collection vs. low light).

2

EOS Bodies - For Stills / Re: Need a treatise on image quality

« on: March 18, 2013, 12:16:44 PM »

I'm in the market for a new body with the criteria being superior low light capabilities, something my 1DsIII is somewhat lacking, and superior IQ.
Dynamic Range and bit depth also rank in the want column as well.  I'm thinking that photomerge will overcome the need for high MP needed for large images.

Noise is NOT primarily determined by pixel size as so many people falsely assume. It is primarily determined by technology and total sensor surface area. When you stitch images what you are basically doing is simulating a larger sensor surface area.

Low ISO noise has been a solved problem for a long time now, and no modern sensor produces noisy low ISO images when properly exposed. People who claim otherwise are zooming to 300% in PS OR using image viewers with horrible scaling algorithms (Apple's Preview is one) OR trying to lift shadows by 3-5 stops. Or all of the above. Suffice it to say, noise is going to be even less of an issue when you stitch multiple frames and thereby simulate a larger sensor.

Dynamic range OTOH is primarily driven by technology and pixel size, and stitching won't help this. Then again, if you're shooting a subject still enough for stitching, you're shooting a subject that's still enough for exposure blending / HDR.

FWIW - I find stitching 3 frames (camera orientation opposite of image orientation) to be relatively easy and practical. Anything more is a chore. Even the lowest Rebel can match a top notch MF film scan with a three frame stitch. It's rare to need or even see prints that big.

I'm thinking that a crop sensor that utilizes the "center sweet spot" of the lens coupled with reduced MP to allow more utilized light would be the answer.

You're not taking technology into account. The current sensors in either format are the lowest noise sensors Canon has produced to date, and noise is being further squashed into oblivion for a given print size by the stitch. Though stitching doesn't help DR, the current sensors are also the highest DR sensors to date, again due to technology.

If DR is truly a concern then 1DX, 5D3, or 6D. If you're doing this for big landscape prints, be aware that the DR gains aren't a great benefit in practice. Generally if your landscape scene exceeds the sensor DR, it REALLY exceeds it, and you need to exposure blend/HDR. In which case any sensor will do.

Don't forget to buy a panoramic head.

You may also want to consider if the shift function on the T/S lenses better serves your needs.

3

EOS Bodies / Re: 7DII and D400 Specs

« on: February 28, 2013, 06:29:50 AM »

What amuses me is people who shoot their mouth off but offer no proof.

I'm sorry...what proof did you offer while chanting "never never never"?

Put up pics you've shot with both lenses, and with a Sony Nex, or whatever you're using...that "proves" it's as sharp as say, a 15mm Zeiss f/2.8 on a D600.

Are we testing specific lens pairs, or crop v FF? Because your 'never never' post implied it could 'never' happen because of something intrinsic to the formats.

As for the 70-200 IS being sharper than the non IS, I have the non-IS, and it's awfully sharp.  I'll put mine up against yours any day of the week young lad.  Whoever's lens loses, has to buy the winner a chess set made of all the Nikon and Canon supertelephoto lenses...oh, and we exchange wives and/or girlfriends...and mistresses, for a month...yours won't be coming home after they've played chess with me!  If I lose...well...I keep my harem chained in an underground lair out in the woods...they aren't really allowed to leave...but I'll send you my neighbor's wife.

See hjulenissen's link. The IS is sharper once you leave center. And...well...I'm not even going to comment on the rest 

4

EOS Bodies / Re: Pixel density, resolution, and diffraction in cameras like the 7D II

« on: February 28, 2013, 06:19:04 AM »

here is the comparison shots i did for another thread
5Dmk3 + 300f4L IS + Canon 2X mk3
vs
EOS-M + 70-200 f2.8L IS II + Canon 2X mk3

While I agree in principle that, out of camera and all other things equal, FF shots are sharper, you've got way too many differences in this test.

But even here, if you sharpen the crop sample they look identical. Sharpening is not an unlimited good, so as long as the gap between component A and B is below a certain amount, post processing can eliminate the gap. This is true FF v. crop at low ISO, but not at high ISO.

I think we can all agree the 70-200 is a sharp lens with lots of resolving power
the 300f4L IS is a much older optical design

For testing purposes you can't make those kinds of assumptions. How do you know that your copy of lens A is sharper then your copy of lens B when used with your copy of teleconverter C at aperture D? Eliminate all relevant differences in the test to isolate the variable you're testing for.

But again I agree with the point. Out of camera, all other things equal, FF is sharper.

5

EOS Bodies / Re: 7DII and D400 Specs

« on: February 27, 2013, 10:23:02 PM »

As I have said before, they may be able to achieve supreme sharpness with $13,000 super telephoto lenses

Oh please. jrista's point about Canon updating their lenses is well taken. But there are plenty of lenses which can produce tack sharp results on crop without a $13k price tag. Canon is updating cheaper lenses as well. As sharp as I thought my 70-200 f/4L was, the 70-200 f/4L IS takes it up a notch, just like jrista's experience with mark II L super telephotos. And it's not $13k.

But when it comes time to shoot a wide angle image with a 24 MP or higher crop sensor, good luck ever getting sharpness on the level that you could from the same pixel count via a full frame sensor and lens, taking in the same angle of view Never going to happen...not ever.

Except that it already happens with Sony/Nikon 24 MP sensors, top notch glass, and a little USM.

We'll never see a wide angle zoom rectilinear lens, that goes to 10mm for a crop sensor camera, that will be sharp to the corners wide open, with no CA, and zero "decentering"...and somehow rival the best 14 or 15 mm wide angle full frame zoom.  Not going to happen.

I'll gladly pit a Tokina 11-16 f/2.8 (good sample) against a Canon 16-35 f/2.8 (good sample).

You keep using that word 'never'. I don't think it means what you think it means 

For people who think crop will 'never' do this or 'never' do that, you better spend some time looking at the world of biology. Because eyeballs, both human and animal, routinely do things that you say much larger crop sensors should 'never' be able to do. If our eyes do it, then it's physically possible, and you're a fool to bet against the march of technology and Moore's Law when it comes to the physically possible.

When DSLRs first hit the scene I heard repeatedly that a DSLR would 'never' out resolve 35mm film (happened at 12-15 MP); 'never' produce large prints that could rival MF film (happened with the 5D2, and then the 16/18 MP generation of crop sensors); and 'never' have DR like neg film (today's FF has more DR then all but a couple emulsions). When the 5D was popular I heard that crop would 'never' out resolve it or have better noise (again, happened with the 16/18 crop generation).

Unless a manufacturing break through renders crop obsolete by making FF just as cheap to build (doubtful, but never say never), we will see crop bodies in the future that out perform today's D800. And we will see even smaller sensors in P&S super zooms that rival today's DSLRs. Not if. When. I guarantee you that designers will exploit every advantage they can get as time goes on, including lenses that change shape, liquid lenses that can alter their characteristics like an LCD can alter its display, super dense sensors, custom in camera processors that put today's best graphics cards to shame, etc, etc. We may even see biologically grown sensors, at which point MF might be as cheap as today's Rebels, shattering another 'never'.

So tell me again what's 'never' going to happen. It amuses me.

6

EOS Bodies / Re: Pixel density, resolution, and diffraction in cameras like the 7D II

« on: February 27, 2013, 10:00:23 PM »

Excellent post. Thank you for digging up and laying out the formulas. I remember where they're at, but I was being too lazy to dig out the book and copy them. You posted them along with a clear explanation.

I would only add that post processing can recover details <MTF50, giving more potential to the 24 MP sensor past it's diffraction "limit". And that diffraction is not the same for all wavelengths, something sensor designers are aware of and will likely exploit in future very high resolution sensors with very high speed in camera processing. At that point you adjust the Bayer pattern to gain detail and process it all down to a file size smaller then the native sensor output, but with more detail then an image from a regular Bayer sensor.

Thanks again for the post!

7

EOS Bodies / Re: 7DII and D400 Specs

« on: February 27, 2013, 06:33:30 PM »

I'm not saying its worse, its just the extra MP don't make any difference to the resolving power once diffraction has set in.

I've pointed out twice now that it actually does. This is not opinion, it's science that was worked out a long time ago. Optical resolution does not work the way the majority of people assume it works, with the weakest link "capping" or "limiting" the entire system to its resolution. The resolution of the total system is always lower than the weakest link. And increasing any component in the chain...not just the weakest...will increase total resolution. Whether the increase is practical and observable is another question entirely, but to wrap your mind around some questions in photography you have to understand how resolution actually works.

Take another example - scan a photo which was a bit blurry -

False analogy. Here you are changing the target resolution rather than the resolution of a component in the optical system.

However, back to the meaning of my original point, do we really need all these MP? Do you need 24MP from your crop camera?

I don't "need" it. But it's not going to hurt anything, and it's a step towards higher resolutions that will show noticeable improvements in the large prints I make. Under the best conditions, it might just improve my prints today.

going from 3 to 6MP is a very real difference. Going from 18 to 36 MP isn't.

Go ahead and compare 36" landscape prints from the 7D (or 5D2/3) and the D800 and say that.

We're at the point of diminishing returns now - especially as lenses and physics are now becoming limiting factors, and virtually no-one needs to print anything that big and that detailed. If you really do, a larger format than APS-C will yield more real life improvements at such high MP counts. Marketing is leading this drive into the unneeded.

As jrista points out, Canon is revamping their entire lens line because they know where this is heading. Moore's Law isn't going to stop because a few people claim they don't "need" higher resolutions. Granted there are ultimate physical limits, but the end of this road is probably 200 MP FF sensors and the equivalent APS-C sensors. The camera may pre process these images to smaller pixel dimensions for better file sizes, but it will use every pixel in doing so, and the resulting output will be stellar, a match for today's MFDBs.

Back when I bought my first DSLR, a 10D, you could have said a larger format would serve me better then incremental DSLR improvements. Except that those incremental improvements added up to a 7D that produces 24" prints to rival anything I've ever made or seen with 645 film. I say keep the improvements coming.

10-22 and 17-55. Admittedly, the 10-22 does have the 17-40 beaten when it comes to detail at larger apertures in the corners, so to call the cheaper L lens comparable is debatable.

The Tokina 11-16 f/2.8 really has the 17-40L beat, and is cheaper.

But the 17-55/24-105 comparison is a good one. The 24-105 when used on FF goes wider, longer, offers more detail, is brighter (f2.8 on crop = f4.5 on FF), and (at least in the UK) cheaper. OK, its vaguely bigger and heavier, but you can't have everything...

f/2.8 != f/4.5 on FF. I cannot shoot a crop body at f/2.8 and a FF body at f/4.5 and hold the same shutter and ISO. I realize what you're getting at (i.e. FF noise or shallow DoF), but it's still not the same. And I would argue the detail claim as well.

That said, they're basically the same price at B&H.

8

EOS Bodies / Re: 7DII and D400 Specs

« on: February 27, 2013, 02:26:31 AM »

jrista - point well proven with some great shots!

9

EOS Bodies / Re: 7DII and D400 Specs

« on: February 27, 2013, 12:06:49 AM »

If you have an aperture so small that both a high MP body and a low MP body have diffraction, the high MP body will have no advantage. It doesn't matter how finely the sensor can resolve the blur projected by the lens, its still just blur.

Once again, resolution does not work like this. First, diffraction is not a hard limit. Second, even at extreme apertures where diffraction approximates a hard limit, a higher resolution sensor will still yield superior results. Why? Because the resolution of a system is not the weakest link in the chain. It is computed from a formula using all the links and is always less than the weakest one. Increasing any link results in a higher final resolution, but the final resolution is always less than the weakest component.

Practically speaking the increase may not matter outside of a lab. And for 18 vs. 24 MP I'm guessing it won't matter much outside of a lab test even at f/2.8. But putting a higher resolution component in the chain will never result in worse performance.

I hate the term DLA because it's inaccurate and it conveys the idea of a "hard limit" that goes down as the sensor resolution goes up. It is not at all consistent with the science of optics. And it plays off another issue in general conversation: the portrayal of resolution as a single number. It's not a single number. It's an MTF curve. Your comment "it's sill just blur" illustrates the problem. The 7D is "diffraction limited" at f/6.9. The way you describe this, f/8 produces "blur." The reality is that detail with X contrast at f/6.9 has some value <X contrast at f/8, and can be restored to X contrast with sharpening. I can make f/6.9, f/8, and f/11 24" prints all day long and you won't be able to tell me which is which.

At some point detail is truly lost, i.e. contrast of 0%. But the point is not immediately past the DLA.

While a 24MP sensor with DLA setting in at f6.0 will allow for some lenses to shine at larger apertures, where will this marketing machine stop? If the next round of crop cameras hit 40mp, and then after that 60, will you still be arguing for it to carry on?

I've seen convincing arguments for 100-200 MP FF sensors, assuming technology allows you to hold the line on noise/DR. Why? Because of another point that's not reflected in simple DLA numbers tossed around on the web: the impact of diffraction is different for different wavelengths of light. And if your sensor design + RAW software takes this into consideration, it can maximize the detail recovered. We will eventually see that point in digital camera design.

There comes a point where making EF-S glass good enough to resolve such detail at the large apertures needed to avoid diffraction becomes unaffordable. We're already at the point where the 17-40L and 24-105L cost less than their EF-S counterparts.

Which counterparts are you thinking of?

But for those than want the extra MP so they can resolve more detail, in all but very select circumstances and with all but the very best glass, they'll be very dissapointed.

I'll agree that 18 vs. 24 MP is not a very big deal and is driven by marketing. But sensor resolution is not limited by diffraction in the way you think it is. And we will see even higher resolutions in the future. If a jump could be made today to 35 or 40 MP while holding the line on noise/DR, it would produce observably better prints.

10

EOS Bodies / Re: 7DII and D400 Specs

« on: February 26, 2013, 03:49:27 AM »

APS-H is a nice middle ground; better IQ than 1.6, better reach than FF.

We're at the point of diminishing returns where it really doesn't matter with FF vs. 1.6x. There's hardly a reach gain today with 1.6x. It's there, but you pretty much have to crop even further and then print 24" or larger. With 1.3x in the newest generation of sensors you would never see the gain vs. FF.

Likewise, there's hardly an IQ gain with FF except at high ISO. So the IQ difference between 1.3x and 1.6x, with the sensor tech the 7D2 will have, would be a scientific footnote, not a human observable quantity.

We're long past the days of the 5D vs. the 20D where IQ and reach differences were immediately obvious. 1.3x loses compatibility with numerous lenses and increases costs for no real advantage. It's not coming back. APS-C gives you lower cost bodies and, in some cases, lower cost glass. FF gives you more high ISO room. That's about it.

11

EOS Bodies / Re: 7DII and D400 Specs

« on: February 26, 2013, 03:42:06 AM »

Ps - I really hope Canon resist the temptation to take their 1.6x crop sensor up to 24mp. It'll suffer from softness due to diffraction from f6.0 onwards - mount an f5.6 lens on there and you've got little in the way of options. Even the legendary 300/2.8 II with a 2x TC III will underperform, and leave you with just one aperture option if you want to attempt to utilise all of those megapixels.

Resolution doesn't work like this. In a format, a 24 MP sensor will always out perform an 18 MP sensor whatever aperture you choose. Now even under the best conditions it won't amount to much and may not be immediately obvious in large prints. But it will always be better by some amount. Diffraction is not a hard limit and is not a reason to cap APS-C sensor resolution or to limit your use of apertures.

The current 7D is "diffraction limited" at f/6.9. I regularly use f/8 and f/11, and after post processing even f/16 is fine. More to the point, the 7D is never "worse" than the 10D, 20D, 40D, etc., not even at f/22 or f/32.

High ISO is another non-reason to cap resolution. Noise for an image is determined by technology and total surface area, not pixel size. At least not until you get into the extremely small pixels on some P&S bodies.

The only possible reason to hold back on increasing the resolution is DR. Pixel size is a factor for DR. But if technology eclipses this, my guess is Canon will match the competition. Too many people will see 18 vs. 24 MP as a negative, and Canon's not going to give up that marketing point.

Given the choice I would take 18 MP if it had noticeably higher DR. But that choice may not be what anyone thinks it is, i.e. the DR gain may be less then the resolution gain. And I doubt Canon will give us the choice. Their competitors have settled on 24 MP. You can guess what they will do.

12

EOS Bodies / Re: Canon EOS 7D Mark II Spec List [CR2]

« on: February 19, 2013, 01:15:02 PM »

Canon cannot put the same AF-system like the 5D Mark III´s in an 7D Mark II for the mentioned price. So don´t expect it.

Canon could put the 1DX's AF system in every body they've got. The AF components are among the cheapest to manufacture. Like everyone else they use AF specs to segment the market, but the 1DX AF is not actually hundreds of dollars more to manufacture then the Rebel's AF.

Whether or not they will put the 5D3's AF in the 7D2 remains to be seen.

13

EOS Bodies - For Stills / Re: Film EOS Body

« on: February 17, 2013, 07:37:45 PM »

Adding another vote for the EOS3. I actually prefer it over the 1-series film bodies due to size/weight. It's an awesome camera, and the used price right now is a steal.

I was never that crazy about eye controlled focus, though it does work with a reduced set of focus points. (It never works well for me with all 45.) But there are a couple other features I wish were on my 7D. Namely the spot metering mode where you can take multiple readings and the camera will compute the exposure, and the additional metering scale to the right of the viewfinder.

14

EOS Bodies / Re: $1200 7D vs. $1800 5D II?

« on: February 14, 2013, 08:40:41 PM »

5D2 if you regularly shoot and make large prints at ISO 1600 and above. OR if you have certain Canon lenses which really work best on FF (T/S; the fast wide primes like the 35 f/1.4).

7D for everything else.

No matter how many times people insist...just insist...there's a HUGE difference when they pixel peep at 300%, they cannot pick out 16x24 or 20x30 prints at ISO 100-800.

And $600 would make a big dent for lenses, either buying one you need out right, or getting you that much closer to one that's more expensive.

15

EOS Bodies - For Stills / Re: Upgrade from 40D

« on: February 08, 2013, 01:50:47 AM »

1) Used 1D Mark III (yeah, submirror AF blah blah)

Sports and wildlife: excellent
Landscape: OK
Low light: OK

2) Used 5D Mark II

Sports and wildlife: OK
Landscape: excellent
Low light: excellent

3) Used 7D

Sports and wildlife: excellent
Landscape: excellent
Low light: OK

Not sure what to say here except that you're probably looking at the 7D or 5D2, and not the 1D3 regardless of the cool factor of having a 1-series camera.

If sports AF/fps are critical, 7D.

If >3200 ISO and/or >11x14 low light prints are critical, 5D2.

Low-mid ISO landscape prints are the same from either after post processing, and you can easily print 16x24 and 20x30.