[Kunzite]

It's funny how someone could suggest books, yet discount equivalence in the same breath

Because with reading, experimentation and seeing comes understanding, and with understanding you won't have to rely on things like "equivalence".
And because when searching books about equivalence you'd rather find about Stieglitz, than Joseph James

Good question ^^, why is that?

It appears you missed my answer: that assertion is false. A medium format f/5.6 lens is slow.

[ElJamoquio]

This is an EPIC FAIL and a perfect illustration of why the "equivalence" hype is dangerous.

Not hype, wasn't a fail.


So you wouldn't even try before purchasing your TS and experimenting on the small format before. And after that you'd never surpass what the small format TS lens can do, if you somehow had access to a view camera.

Crop capabilities are not included in the "equivalence" theory; so you, as a true believer in "equivalence", are not allowed to crop

Of course cropping is included in equivalence.

By the way, current "full frame" cameras are ranging from 12 to 36MP; none of them can reach the same pixel density as a 24MP APS-C sensor. For most of them, the difference when cropped to APS-C would be easily visible.

'Easily visible'. The sharpness of a 24MP FF cropped to 10MP is perhaps 25% sharpness penalty compared to 36MP cropped to 16MP. That's actually a bit tough to see, it's only half the difference in sharpness between FF and APS-C at the same MP, for reference.

You posted in order to move the direction from my "know your format" to your "know equivalence".

True, but in the context of the thread and your choice of words ('format') I think that's more than fair. If you were talking about your particular camera and lens then you would've gotten your meaning across more quickly if you'd said 'know your camera and lens'.

Since when is computing DOF and estimating shot noise (you're of course incorrect to assume to know SNR by mere multiplication with a "crop factor") is enough in photography?

Absolutely never. But if I can take ten things to remember - and bring it down to 8, I've made quite a bit of improvement.

Since when it is better than seeing with your own eyes?

I can't actually see a picture until I get back to the office.

[jsherman999]

I've used a medium format camera, only once though. If I'm spending your money I'd buy it again, sure. I know it doesn't offer any SNR because of equivalence, though, and I think 36x24 has 'good enough' detail.

This is an EPIC FAIL and a perfect illustration of why the "equivalence" hype is dangerous.

It actually isn't. You can consider what FLs you would want to shoot, find the lenses available for say the 645D/Z that have the equivalent FLs and then convert the f-stops - if you come up with a 'faster' lens for MF, or a very near-hit, then you have a candidate combo for a SNR/noise upgrade by going MF. At that point you'd take the sensor tech into account, and that info is available via other articles and places like DxOMark. You'd be able to tell, without buying anything, if you were likely to see any noise advantage to buying the 645 over your current system, and if there was one almost exactly how much your 'gain' would cost.

Kunzite method: Forget equivalence altogether, just buy and try! Depreciation is a bitch, but maybe there's a liberal return policy on that 645.

.

[Kunzite]

Not hype, wasn't a fail.

Of course it wasn't a fail; it was an EPIC FAIL. Even excluding the typo (which was a bit funny - I assume it was "it doesn't offer any SNR advantage").
You know nothing about SNR until you measure it.

Of course cropping is included in equivalence.

How so? There are so many ways of cropping in image in post, will you compute "equivalence" parameters for all of them? If not, how do you decide for which crops will you do it?

'Easily visible'. The sharpness of a 24MP FF cropped to 10MP is perhaps 25% sharpness penalty compared to 36MP cropped to 16MP. That's actually a bit tough to see, it's only half the difference in sharpness between FF and APS-C at the same MP, for reference.

But what if the "full frame" doesn't have a 36MP sensor, and it has a pretty strong AA filter?

True, but in the context of the thread and your choice of words ('format') I think that's more than fair. If you were talking about your particular camera and lens then you would've gotten your meaning across more quickly if you'd said 'know your camera and lens'.

It's only fair when it goes your way? Hmm...

I can't actually see a picture until I get back to the office.

You're making purchase decisions on a such a short notice, you can't even wait 'till the next day?

---------- Post added 28-07-14 at 09:55 PM ----------

Kunzite method: Forget equivalence altogether, just buy and try! Depreciation is a bitch, but maybe there's a liberal return policy on that 645.

This is an intended misinterpretation, the meaning of what I said is completely the opposite: I'm advocating for a more informed purchase opinion.
For the record, using "equivalence" one would never buy a medium format camera, after all the small format is offering a shallower DOF.

[ElJamoquio]

Of course it wasn't a fail; it was an EPIC FAIL. Even excluding the typo (which was a bit funny - I assume it was "it doesn't offer any SNR advantage").
You know nothing about SNR until you measure it.

I know quite a bit about it actually. I'm betting it's line-on-line with the SNR of the D800, etc at same DOF.

[jsherman999]

You know nothing about SNR until you measure it.

Which DxOmark and places like sensorsgen does for you. But in addition to knowing that, you have to know how much total light will fall on that sensor with the available lenses, compared to what you may be moving from - which equivalence tells you.

Perhaps you missed:

It actually isn't (an epic fail). You can consider what FLs you would want to shoot, find the lenses available for say the 645D/Z that have the equivalent FLs and then convert the f-stops - if you come up with a 'faster' lens for MF, or a very near-hit, then you have a candidate combo for a SNR/noise upgrade by going MF. At that point you'd take the sensor tech into account, and that info is available via other articles and places like DxOMark. You'd be able to tell, without buying anything, if you were likely to see any noise advantage to buying the 645 over your current system, and if there was one almost exactly how much your 'gain' would cost.

Or, you could just buy and try.

.

[ElJamoquio]

You're making purchase decisions on a such a short notice, you can't even wait 'till the next day?

If I want to purchase a Pentax lens or a Pentax camera, I have to purchase it on the internet. There is no try at the shop for me and for the bulk of the users in the United States.

[Kunzite]

I know quite a bit about it actually. I'm betting it's line-on-line with the SNR of the D800, etc at same DOF.

You don't know; you're assuming. But you can't be more wrong - just consider that until recently medium format cameras used only sensors with very different technology, CCDs made by Kodak/TrueSense and Dalsa. They had exquisite low ISO, but some had very poor high ISO. That's common knowledge.

Which DxOmark and places like sensorsgen does for you. But in addition to knowing that, you have to know how much total light will fall on that sensor with the available lenses, compared to what you may be moving from - which equivalence tells you.

Why do I have to know that? It's pointless; I'd rather visually observe the quantity and quality of noise from image samples than staring at a number and try to make sense out of it. Do you know why there are so many review sites around?

I did not miss your idea of buying only from "equivalence" and DXOMark, and trying to figure out cost/performance scores based only on few numbers. I'm just rejecting it as utterly ridiculous.

Or, you could just buy and try.

That's a lie, and an error you pretended not to make - intellectual dishonesty at work. It's not "equivalence" or nothing/trial and error.

If I want to purchase a Pentax lens or a Pentax camera, I have to purchase it on the internet. There is no try at the shop for me and for the bulk of the users in the United States.

And what's the problem?
You have many tools at your disposal to make an informed purchase. You can even have a good idea about a lens rendering characteristics, sharpness, CA etc. by looking at image samples. And you only need one lens to start experimenting - make it a "safe" choice, a normal or something which you know it will work.
Only in "equivalence"-promoting Internet fights, one has to compute precisely AOVs and DOFs.

[ElJamoquio]

You don't know; you're assuming. But you can't be more wrong - just consider that until recently medium format cameras used only sensors with very different technology, CCDs made by Kodak/TrueSense and Dalsa. They had exquisite low ISO, but some had very poor high ISO. That's common knowledge.

I'm referring to the 645Z. I also know the performance of the 645D based on similar CCD chips. Assumption? Sure, the assumption is that the chip hasn't improved fundamentally per unit area when it was made larger. That's a pretty good assumption.

I did not miss your idea of buying only from "equivalence" and DXOMark, and trying to figure out cost/performance scores based only on few numbers. I'm just rejecting it as utterly ridiculous.

That's a pretty silly strawman.

[Kunzite]

And would you even attempt to verify if that assumption holds true? What if the alternatives, let's say Canon and Nikon, are using different sensors, with different noise characteristics? What if such differences can be important?
I don't get why you are insisting with using "equivalence"; you can obtain more pertinent information by direct observation. And you can obtain information outside the restriction of "equivalent" settings.

It's not a strawman, it's what he said:

You can consider what FLs you would want to shoot, find the lenses available for say the 645D/Z that have the equivalent FLs and then convert the f-stops - if you come up with a 'faster' lens for MF, or a very near-hit, then you have a candidate combo for a SNR/noise upgrade by going MF. At that point you'd take the sensor tech into account, and that info is available via other articles and places like DxOMark. You'd be able to tell, without buying anything, if you were likely to see any noise advantage to buying the 645 over your current system, and if there was one almost exactly how much your 'gain' would cost.

[Rondec]

I quite agree Kunzite, know the gear. part of that knowledge is understanding how it performs against other systems and whether a particular shot would start to fall apart on your system. The whole business of books and the net is to reduce experimentation and almost every book on photography invariable gets to crop factor (and by default equivalence) very early experimentation will not change that.
Rondec, I agree with you in full. I just want to query whether the iPhone's camera is inherently bad or is it a matter of being too cropped? . Could it be that the sensor is really not bad but too much algorithms has to be applied image due to it's size and the amount of light it receives? As with Rondec learning the gear seems to be what you are saying but it doesn't change the fact that equivalence is an important part of that knowledge.
I still would like some insight into why MF lens are fast at 5.6 but on a FF it is considered slow. Anyone?

The issues I see with an iphone have to do with the poor lens, lack of dynamic range, poor ergonomics and camera shake due to a combination of light weight/high pixel density on the sensor/bad ergonomics. If you have good light, a scene without too much dynamic range, and manage to keep your hand steady, you can get decent results.

This first photo is an iphone photo, which turned out OK.



Second photo is a K3 photo on a tripod (obviously at a different time of day).

[ElJamoquio]

It's not a strawman, it's what he said:

You said 'your', which reasonable people would think mean 'mine'. So it most certainly was a strawman, as it's something I've never said.

He's not wrong, though, but I don't have time today to debate at you.

[Kunzite]

I said that after quoting jsherman999. You're trying too hard to "prove" me "wrong".

[jsherman999]

Why do I have to know that? It's pointless; I'd rather visually observe the quantity and quality of noise from image samples than staring at a number and try to make sense out of it.

Is there any rational explanation for why you can't do both? (look for sample images after arming yourself with knowledge of equivalence to manage expectations.)

Or check the validity of a "reviewer's" image samples based on what you know about equivalence? After all, we've seen P&S's 'proven' to be as good as aps-c DSLRs based on posted images, and an equivalence-denier might just believe those samples - unless he/she had some basic tools at their disposal that could be used to refute nonsense like that.

.

---------- Post added 07-28-14 at 04:21 PM ----------

...
You have many tools at your disposal to make an informed purchase. You can even have a good idea about a lens rendering characteristics, sharpness, CA etc. by looking at image samples. .

Calm down, stop and think about what you're implying now, in your haste to support your argument. Now - are you proposing that viewing potentially downsampled, PP'd, converted images samples are a better way to determine a lens' sharpness and CA vs. photozone, SLRgear, etc MTF measurements? Keep in mind we're talking about a lens you can't evaluate in person unless you buy. Are you really putting that much weight on someone else's online image samples here?

Even if you do want to weight those images heavily, how would that preclude looking at MTF scores in addition?

Now, In the same vein: how does depending on posted images preclude the use of equivalence as well? What if for example you were considering a portrait combo and the FOV/DOF look you want to consider isn't represented by any online samples you find, wouldn't it be a good idea to (say) figure out what the equivalent f-stop would be for a given AOV and then stop down some lens you do have to that to see what the image DOF would look like on the other format? Why is doing something like that so offensive to you?


.

[Kunzite]

Is there a rational explanation about why you would want to use a more imprecise method, if a better one is so readily available? And if you want to study what's happening in-depth, why working within the constraints of "equivalence"?

You're getting it wrong; the theory is confirmed by practical observations, not the other way around! A theory that doesn't match reality is garbage.
First, you make sure you have a valid practical observation - that you're measuring/observing what you want, that variables are kept under control, that the measurement is repeatable; review sites are helpful in this regard. Using larger samples will reduce observational "noise" (and you might be less constrained by using a precise methodology).
Then, if you wish, you can compare the results with the theory; but that's not necessary, except if you want to confirm the theory. If it doesn't match your beloved "equivalence", it means that "equivalence" failed. Perhaps one of the assumptions needed for it to work is not true in that case.

In some limited circumstances, you might get results from a P&S which are indistinguishable from that of a large sensor camera. This falls under the "know what you're measuring/observing" category. It's a valid result, if and only if that's what you want to know.