Fun Arias rant on APS-C vs. FF

[sdsr]

Is this more than a disagreement about what "very little" means (or, put differently, about whether the differences you see matter)? dtaylor doesn't say they look the same, after all....

He implied the difference is not noticeable to which we responded (based on actually having owned a 7D and 5D) that we disagree. That is all we are stating. Neither person is wrong as we are simply stating opinions. What others choose to do with that info is their business.

Note this is based on the 7D. Modern day APS-C cameras with Sonikon sensors is a different debate. Perhaps in those cases the difference is less. I'm sure someone with both could comment.

Well, I have both (Canon 5DIII, 6D & SL1, and Sony a7r & a6000) and can certainly see differences between FF and APSC (and between the 5DIII and 6D, for that matter), both within and between camps. I've not done anything resembling scientific testing, but I get the impression that the differences are similar (though at least for low ISO noise the Sony camp wins because their APS-C and FF sensors both have less noise at low ISOs and better DR).

I've never used a 7D, though that doesn't really affect the point I was trying to make with my comment above, which is merely that whether something is noticeable, whether a difference is negligible, depends on who's looking, how and why, so that some of this evident disagreement may be in some sense spurious. You could sit two people, A and B, side-by-side in front of the same photograph and A could say to B "they look the same to me", B could respond "but what about (say) this patch of noise here?" and A could in turn respond "Oh, OK, I see that now that you point it out, but it's not the sort of thing I notice because for me it's an unimportant element in how a photo looks". (I've actually done something like this myself.) But if A were instead to respond to B "no, I don't see it at all," that could mean that A or B needs new glasses....

 

[dtaylor]

First of all, read my post first- I said both 5D (FF) and 50D (APS-C) that I have used appeared less noisy than my 7D. So it has nothing to do with confirmation bias.

First of all, read my post first - I did not say your opinion was from confirmation bias. My suggested theory is that you don't know how to size/compare properly. (And having three cameras in the mix would not preclude confirmation bias.)

Was my 7D out of spec? I think not- I have heard of similar noise issues from many people.

Noise issues which don't show up in scientific tests? I guess Canon sent the special noise free 7D's to the reviewers.

Secondly, your first guess is wrong- I don't pixel peep when looking for noise. When I say, ISO xxx was unacceptable it means I looked at it full screen on my 24" monitor and it looked bad. That's a pretty low bar, wouldn't you say?

As a testing methodology yes, that is very low bar. Your software, scaling algorithm, monitor, monitor settings, all affect the final result. For example: at the right scaling in Apple Preview on a MacBook (non-Retina) you can make a FF 5D3 or D600 look noisy at base ISO because the scaling mechanism screws up.

Finally, please don't go mentioning "scientifically reproducible tests" without citing them.

Pretty much every review on Earth noted that the 7D was cleaner then the 50D, and at 3200 the 5D has visibly more chroma noise (though I think it's visibly even and tests a bit better on luminance). Go review the Dave Box tests at Imaging Resource.

 

[dtaylor]

Yeah it's perplexing. The 50D may be as noisy as the 7D, but it lacks the waxy AA filter that just smears details on the 7D.

LOL! Google "Comparometer." Pick the 50D and 7D. Click the ISO 100 test scene with the fabric and the bottles, etc. Pay careful attention to which one looks "waxy."

In fairness this is about the JPEG engines. The 7D has the better JPEG engine, and both cameras are capable of higher IQ in ACR. But it's the 50D that was "waxy" in JPEG. The AA filter isn't an issue with either body.

It's the combination of inconsistent focus (burst mode in AI Servo and/or AF single point even on static subjects),

http://www.prophotohome.com/news/2010/03/autofocus-torture-test-updated-canon-1d-mkiv-nikon-d3s-added/?utm_source=MailingList&utm_medium=email&utm_campaign=pph%20newsletter%20autofocus%20mkiv%20robg

If you experienced poor AF you should have returned the body for repair work/calibration.

drab colors

It's digital. You decide the colors.

 

[sagittariansrock]

First of all, read my post first- I said both 5D (FF) and 50D (APS-C) that I have used appeared less noisy than my 7D. So it has nothing to do with confirmation bias.

First of all, read my post first - I did not say your opinion was from confirmation bias. My suggested theory is that you don't know how to size/compare properly. (And having three cameras in the mix would not preclude confirmation bias.)

Was my 7D out of spec? I think not- I have heard of similar noise issues from many people.

Noise issues which don't show up in scientific tests? I guess Canon sent the special noise free 7D's to the reviewers.

Secondly, your first guess is wrong- I don't pixel peep when looking for noise. When I say, ISO xxx was unacceptable it means I looked at it full screen on my 24" monitor and it looked bad. That's a pretty low bar, wouldn't you say?

As a testing methodology yes, that is very low bar. Your software, scaling algorithm, monitor, monitor settings, all affect the final result. For example: at the right scaling in Apple Preview on a MacBook (non-Retina) you can make a FF 5D3 or D600 look noisy at base ISO because the scaling mechanism screws up.

Finally, please don't go mentioning "scientifically reproducible tests" without citing them.

Pretty much every review on Earth noted that the 7D was cleaner then the 50D, and at 3200 the 5D has visibly more chroma noise (though I think it's visibly even and tests a bit better on luminance). Go review the Dave Box tests at Imaging Resource.

You mentioned that I shall attribute the noisier image to 7D. How does that relate to my inability to size images? Of course, it will be affected by my alleged inability to compare images, but then that is just your guess, isn't it? Do note that the 50D was mentioned anecdotally, I did say that I have NOT done ANY comparisons. So yes, there could be bias, and my statement implicitly expressed that possibility. Although it is cognitive bias, not confirmation bias.

A low bar for performance is not the same as low bar for technique.
I am not a professional pixel-peeper or someone with lots of free time in hand. I don't objectively test cameras or formats so all the parameters you mentioned are moot. My finding was, at the same monitor settings and when displayed full screen, the 7D images looked noisy. Yes, I know it isn't the same ratio of enlargement. Who cares? All that matters to me is whether or not an image looks good at the size I will use it. If the one from 5Dc looks better, I will keep that camera. And I did.

So what are you trying to prove through your offensive statements, unsupported information and uneducated "guesses"? If I read some 10 reviews saying 7D is supposed to be better, the images that looked noisy before will now appear clean?

Put less stock in theories and calculations, and rely on your eyes. If you feel 7D is better, good for you. Don't come trying to tell me what I should feel.

 

[MichaelHodges]

In fairness this is about the JPEG engines. The 7D has the better JPEG engine, and both cameras are capable of higher IQ in ACR. But it's the 50D that was "waxy" in JPEG. The AA filter isn't an issue with either body.

I don't shoot in jpeg, so I really can't comment in this context.

If you experienced poor AF you should have returned the body for repair work/calibration.

I did. Five times.

It's digital. You decide the colors.

Feel free to enjoy a Rebel XTi with the 18-55 non-IS wide open. After all, you choose the colors.

 

[Zv]

Is this more than a disagreement about what "very little" means (or, put differently, about whether the differences you see matter)? dtaylor doesn't say they look the same, after all....

He implied the difference is not noticeable to which we responded (based on actually having owned a 7D and 5D) that we disagree. That is all we are stating. Neither person is wrong as we are simply stating opinions. What others choose to do with that info is their business.

Note this is based on the 7D. Modern day APS-C cameras with Sonikon sensors is a different debate. Perhaps in those cases the difference is less. I'm sure someone with both could comment.

Well, I have both (Canon 5DIII, 6D & SL1, and Sony a7r & a6000) and can certainly see differences between FF and APSC (and between the 5DIII and 6D, for that matter), both within and between camps. I've not done anything resembling scientific testing, but I get the impression that the differences are similar (though at least for low ISO noise the Sony camp wins because their APS-C and FF sensors both have less noise at low ISOs and better DR).

I've never used a 7D, though that doesn't really affect the point I was trying to make with my comment above, which is merely that whether something is noticeable, whether a difference is negligible, depends on who's looking, how and why, so that some of this evident disagreement may be in some sense spurious. You could sit two people, A and B, side-by-side in front of the same photograph and A could say to B "they look the same to me", B could respond "but what about (say) this patch of noise here?" and A could in turn respond "Oh, OK, I see that now that you point it out, but it's not the sort of thing I notice because for me it's an unimportant element in how a photo looks". (I've actually done something like this myself.) But if A were instead to respond to B "no, I don't see it at all," that could mean that A or B needs new glasses....

Yes OK that makes sense. I see what you're getting at. Two people can see the same thing differently.

 

[Skirball]

Feel free to enjoy a Rebel XTi with the 18-55 non-IS wide open. After all, you choose the colors.

What does non-IS wide open have to do with colors?

 

[dtaylor]

You mentioned that I shall attribute the noisier image to 7D. How does that relate to my inability to size images?

Most people drop images in PS and zoom to 100% forgetting that higher MP images are magnified more at "100%".

Of course, it will be affected by my alleged inability to compare images, but then that is just your guess, isn't it?

You are correct that I am guessing as to why you believe the 7D is noisier when reproducible tests say otherwise.

I don't objectively test cameras or formats so all the parameters you mentioned are moot.

No, your opinion is by your own admission moot.

It simply amazes me how people will stubbornly cling to opinions even when they fully admit their opinions are not based in any way on anything objective, reproducible, or reasonable. No wonder Zack went on his rant. I'm behind him 100%.

Yes, I know it isn't the same ratio of enlargement. Who cares?

Now I have no need to guess.

So what are you trying to prove through your offensive statements,

Pointing out the truth is offensive? But it's not offensive at all when people appeal to hyperbole to prop up clearly false statements?

and uneducated "guesses"?

My guess was spot on as you just admitted.

Put less stock in theories and calculations, and rely on your eyes.

Silly statement considering I'm telling you to rely on your eyes with a proper comparison.

Human nature just amazes me sometimes...

 

[dtaylor]

I wonder if anyone can tell me which of these came from which camera? Or brand? Or even format? All scaled so that on a 96ppi monitor it's like looking at sections from a 36" print.

http://imgur.com/B2DUlmJ

NEG-LI-GI-BLE

 

[sagittariansrock]

You mentioned that I shall attribute the noisier image to 7D. How does that relate to my inability to size images?

Most people drop images in PS and zoom to 100% forgetting that higher MP images are magnified more at "100%".

Of course, it will be affected by my alleged inability to compare images, but then that is just your guess, isn't it?

You are correct that I am guessing as to why you believe the 7D is noisier when reproducible tests say otherwise.

I don't objectively test cameras or formats so all the parameters you mentioned are moot.

No, your opinion is by your own admission moot.

It simply amazes me how people will stubbornly cling to opinions even when they fully admit their opinions are not based in any way on anything objective, reproducible, or reasonable. No wonder Zack went on his rant. I'm behind him 100%.

Yes, I know it isn't the same ratio of enlargement. Who cares?

Now I have no need to guess.

So what are you trying to prove through your offensive statements,

Pointing out the truth is offensive? But it's not offensive at all when people appeal to hyperbole to prop up clearly false statements?

and uneducated "guesses"?

My guess was spot on as you just admitted.

Put less stock in theories and calculations, and rely on your eyes.

Silly statement considering I'm telling you to rely on your eyes with a proper comparison.

Human nature just amazes me sometimes...

As unfocused stated in a different context- you let your keyboard run off before your brain and now you have to keep it up... (paraphrasing)

1. I did not drop images into PS and zoom to 100%. As I clearly stated before, I set the image full screen on a 24" monitor to see how it looks. Being higher resolution, the 7D gains an advantage in terms of pixel downsampling, but being a smaller sensor, it has the disadvantage of enlargement. Regardless, if one is trying to gauge whether his photos look good and presentable, all that matters is looking at them at the same size that your viewers will be looking. For me, that is either a 24" monitor or an 8x10 print. For someone comparing a 7D and a 5Dc, comparing them with similar enlargement of pixel makes sense. But I never owned both cameras at the same time. Which brings me to my second point.

2. My comments about the 7D noise is not an opinion, it is an observation. I owned the 7D and loved it except when I needed to use it at high ISOs. I could not use it at ISO 3200 and above, ISO 1600 was best avoided and ISO 800 was borderline. Not an opinion, an observation at the extremely low performance bar that full screen images from 7D showed unacceptable noise. I have frequently lamented that fact, and still I stuck with the camera for all the other attributes and the great and inexpensive EF-S lenses. So I have no idea why you are blowing this whole issue up as appealing to hyperbole to prop up clearly false statements. What would be my reason for doing that? Being able to keep the 7D would have saved me a lot of money. In fact, I tried to wait for 7DII but it took too long and I just went for the 5DIII. The difference was immediately apparent.
So, why the comment about 7D vs 5Dc? I own the 5Dc now, and had considered selling it after a small job, but I realized that I was able to use images at ISO 1600 without any problem. So, given my subjective scenario, saying 'my 7D was worse than my 5Dc" makes sense.

3. Your guess was I don't know how to compare images. Well, as a scientist, I have published my work that included various images of neurons, that needed to be counted and examined for shape and form. These were fluorescent microscopy images taken using various microscopes and magnifications. I am fortunate that the expert reviewers that accepted my work for publication didn't share your view!

Now, less cross-examine your rants:

4. You still haven't cited any "reproducible" tests or shown any of your own images. You really expect someone to comment on those low-res images without any ISO information? If you're so convinced, why not post some of your OWN 7D images at ISO 1600 without any de-noising. Let everyone see for themselves?

5. You stated one guess at the beginning at the post (PS>100%, etc) and when I said I used a different way to compare, you concluded that supports your guess.

6. Zack Arias is essentially speaking about APS-C sensors that perform better, and not having used Fuji I am willing to take his word for it. I am not saying APS-C sensors cannot be good. In fact, I mentioned he is partially correct in my first post. The reason I got irritated by the video is the weird argument that FF vs APS-C is negligible because they are both so much smaller than MF etc. So you wouldn't care if your arm broke vs it got twisted just because they are so much better than your arm getting ripped off? That's terrible logic. And he is annoying and obnoxious. But he feels like a sweetheart compared to you. I made a subjective comment, stating my observation with my own camera. And you keep trying to prove that I have an 'opinion' against the 7D based on 'what? certainly couldn't be the reviewers who unanimously praise it's high ISO performance'. It is not the words you say that are offensive (well they are, but I don't care), it is the nature of your argument- how you are making a normal, civil discussion into a fight.

7. I am surprised human nature sometimes amazes you. It should always amaze you, considering how foreign it should feel...

Say what you want to say to this... I am not going to waste any more of my time responding to your posts.

 

[DRR]

I wonder if anyone can tell me which of these came from which camera? Or brand? Or even format? All scaled so that on a 96ppi monitor it's like looking at sections from a 36" print.

http://imgur.com/B2DUlmJ

NEG-LI-GI-BLE

I think you have missed the point entirely.

In a studio with tripod and lights at ISO 100 and f/11, yes the difference is negligible. In fact, photographers have put an iPhone up against a Hasselblad with "negligible" difference.

http://www.photigy.com/iphone-vs-hasselblad-still-thinking-buying-new-camera/

Now I am certainly not saying that FF is "better" than APS-C, or iPhone, or pinhole. It's a tool to give you a certain result. APS-C has its advantages over FF, it's up to the photographer to choose the best tool for the job.

But to say broadly, that the difference between FF and APS-C is negligible, is ignorant. It's not negligible. High ISO performance is significantly better with FF. That alone is enough to make me want FF. It's not up to Arias or you or anyone else to judge the needs of others and tell me where my priorities should lay.

 

[Keith_Reeder]

High ISO performance is significantly better with FF.

Nope, not even close to true.

It's precisely this sort of unconditional sweeping statement - which has been demonstrated more times than tongue can tell to be nonsense - that sticks in some folks' craw.

Fact: my 7D and especially my 70D, are night-and-day better at low light/high ISO than, say, the old 5D - so right there is a perfect example of the lie that the statement "High ISO performance is significantly better with FF" is.

Other things being equal a FF sensor should beat a crop sensor of the same technological maturity, but that's an utterly different statement to the "FF is always better" meme that gets interminably churned out.

Let's see a 5D do better than this: 10,000 ISO with my 70D. Straight out of Capture One 7 Pro, default NR and no PP NR.

And check the Exif in the image - there's hardly any light on that scene...

Here's a 100% crop from the same camera -"only" 6400 ISO this time, but again, straight out of Capture One, at default NR.

Each easily beating what the older FF cameras (including the older 1Ds pro bodies) can do.

Of particular significance is that neither the 7D nor (especially) the 70D throw out the high ISO banding/pattern noise that bedevilled the 5D/5D Mk II - so right there, you've got a significant high ISO performance improvement.

 

[DRR]

Fair enough. Full frame of the same generation, has significantly better ISO performance than APS-C of the same generation.

For the record I've been touting "best tool for the job" since day one, so I am not lying (e.g. trying to deceive) as you suggest, but thank you for helping me clarify. I had thought this was a pretty basic assumption ("of the same generation") but a conclusion can always be nitpicked to inaccuracy. I'll be more careful in the future when posting here.

 

[MichaelHodges]

It's precisely this sort of unconditional sweeping statement - which has been demonstrated more times than tongue can tell to be nonsense - that sticks in some folks' craw.

A bit dramatic, isn't it?

Also, comparing a first generation FF to newer crops doesn't make much sense. Of course technology is going to improve. But it also improves on FF, too, not just crop. Why not compare the 70D to the 6D? I have both sitting right here. There's no comparison.

The poster was correct when he said that crops don't compare to FF in terms of noise and low light handling. Anyone can create an outlier buy comparing extremes or cherry picking, or ignoring parallel sensor improvements in both formats over time.

Let's see a 5D do better than this: 10,000 ISO with my 70D. Straight out of Capture One 7 Pro, default NR and no PP NR.

And check the Exif in the image - there's hardly any light on that scene...

Plastic knobs should be smooth, not rough and grainy. I'd say the 70D didn't get the job done. And I think using a crop at ISO 10,000 is masochistic. Sure, you can take pictures that way, but why?

Of particular significance is that neither the 7D nor (especially) the 70D throw out the high ISO banding/pattern noise that bedevilled the 5D/5D Mk II - so right there, you've got a significant high ISO performance improvement.

I owned the 7D and the 5D II, side by side. There is no comparison on lowlight performance. And, my 7D did have banding issues. The 70D is much better than the 7D (which I now own).

I don't know. Some of these arguments here feel like people defending what they bought, rather than any objective discussion. This feels like one of those.

 

[Skirball]

High ISO performance is significantly better with FF.

Nope, not even close to true.

It's precisely this sort of unconditional sweeping statement - which has been demonstrated more times than tongue can tell to be nonsense - that sticks in some folks' craw.

Fact: my 7D and especially my 70D, are night-and-day better at low light/high ISO than, say, the old 5D - so right there is a perfect example of the lie that the statement "High ISO performance is significantly better with FF" is.

Other things being equal a FF sensor should beat a crop sensor of the same technological maturity, but that's an utterly different statement to the "FF is always better" meme that gets interminably churned out.

Let's see a 5D do better than this: 10,000 ISO with my 70D. Straight out of Capture One 7 Pro, default NR and no PP NR.

And check the Exif in the image - there's hardly any light on that scene...

Here's a 100% crop from the same camera -"only" 6400 ISO this time, but again, straight out of Capture One, at default NR.

Each easily beating what the older FF cameras (including the older 1Ds pro bodies) can do.

Of particular significance is that neither the 7D nor (especially) the 70D throw out the high ISO banding/pattern noise that bedevilled the 5D/5D Mk II - so right there, you've got a significant high ISO performance improvement.

I'm amazed you decided to take that much time typing an extensive "response" comparing 10 year old FF technology to a modern APS-C.

 

[c.d.embrey]

I don't know. Some of these arguments here feel like people defending what they bought, rather than any objective discussion. This feels like one of those.

People base their response on their self esteem. If you own a (pick one) camera, you have to defend that brand/format. If you don't do this you (and other people) may think that you bought the wrong camera! And being wrong about your camera choice lowers your self esteem, and your perceived standing in your community.

 

[dtaylor]

As unfocused stated in a different context- you let your keyboard run off before your brain and now you have to keep it up... (paraphrasing)

You can always tell that someone is about to provide a rational and logical proof of their statements when they open with personal insults ;D

1. I did not drop images into PS and zoom to 100%. As I clearly stated before, I set the image full screen on a 24" monitor to see how it looks. Being higher resolution, the 7D gains an advantage in terms of pixel downsampling,

That depends entirely on the program used to view the samples and the scaling to screen resolution. Some popular viewers will chew up files, Apple's Preview being an infamous one. In print I would agree.

2. My comments about the 7D noise is not an opinion, it is an observation.

To others it is an opinion until you provide examples.

What would be my reason for doing that?

Oh, I believe that when you looked at your screen you saw noisy images. It just doesn't mean what you think it means.

3. Your guess was I don't know how to compare images. Well, as a scientist...blah blah blah...I am fortunate that the expert reviewers that accepted my work for publication didn't share your view!

One has zero to do with the other. That said, I have an honest question: are you this defensive when someone critiques your scientific work?

Now, less cross-examine your rants:

Oh boy here we go! ;D

4. You still haven't cited any "reproducible" tests or shown any of your own images.

Just head on over to DPReview or Imaging Resource.

You really expect someone to comment on those low-res images without any ISO information?

Those are 50% views of one of the Imaging Resource studio test scenes. I remember which cameras I picked...but now I forget the order...hmmm...guess that was Arias' point ;D

If you're so convinced, why not post some of your OWN 7D images at ISO 1600 without any de-noising.

And listen to all the whining that I didn't do X or Y or Z right and that must be the reason it looks good but FF SURELY would be better if I had just shot it at the same time???

Nah...just go look at professionally produced and unbiased results. IR's studio scene can vary in lighting, but their Dave Box scene is strictly controlled.

And he is annoying and obnoxious. But he feels like a sweetheart compared to you.

Let's get down to it, shall we? You said the 7D was noisier then the 5Dc and 50D. I said professional tests showed otherwise. You got hurt that I contradicted you. The rest of this is fluff.

So...is no one ever supposed to contradict you online? Point you to evidence that shows something other then your opinion? Is this how you practice science?

I made a subjective comment, stating my observation with my own camera. And you keep trying to prove that I have an 'opinion' against the 7D based on 'what?

Not at all. Just pointing out your opinion is contradicted by hard evidence and is likely due to some other factor.

how you are making a normal, civil discussion into a fight.

First you say that. Then you say...

7. I am surprised human nature sometimes amazes you. It should always amaze you, considering how foreign it should feel...

Who is starting a fight? And who, btw, is blatantly violating forum rules? Did I personally attack you or insult you once? You've done it three times now in this one response.

If you can't handle the possibility of disagreement then maybe you shouldn't offer your opinion