[Class A]

During any other shooting situation, as in 99% of the average person's shooting, there is no difference.

Not true.

There is always a difference -- even when the same DOF is used -- due to the different enlargement factors of the two formats. This difference impacts on AF accuracy, lens faults, and acuity.

If you look at the subset of FF advantages I originally posted again, you'll see that they include quite a bit more than just "one stop when you are shooting wide open".

[Rondec]

Not true.

There is always a difference -- even when the same DOF is used -- due to the different enlargement factors of the two formats. This difference impacts on AF accuracy, lens faults, and acuity.

If you look at the subset of FF advantages I originally posted again, you'll see that they include quite a bit more than just "one stop when you are shooting wide open".

Of course there are differences. The biggest difference tends to be with regard to printing size. But how big do you have to print before you actually see a difference between a D600 and a K3?

I think we need to be honest. A difference can be statistically significant and still not real world significant. If folks are shooting jpegs, doing minimal post processing, and mainly posting to web/printing 12 inches or less on a side, they are quite unlikely to see a difference between APS-C and full frame, except for a little bit more shallow depth of field.

Even the shallow depth of field thing is a little bit questionable. I would counter to all of this that if people shooting APS-C want to get more shallow depth of field, they just use a longer lens. But equivalence people get angry at that possibility: "That's not right," they argue. "You have to use an equivalent lens! Find a 35mm f0.9 lens for your APS-C camera." But I don't care about equivalence in that situation, I just care about getting the shot that I want and if shooting with an FA 77 gives me the shot I want, why would I stick a 30mm lens on my camera just to try get some kind of full frame equivalence?

I have said this multiple times in this thread, but people need to ask themselves what they are frustrated with their current gear. There are certain limitations that going full frame will address -- printing bigger and getting a little shallower depth of field are probably the biggest -- but simply going full frame will not fix bad photos.

[Class A]

why would I stick a 30mm lens on my camera just to try get some kind of full frame equivalence?

You wouldn't use the 30mm to get some kind of full frame equivalence.

You would use it because you want the perspective, i.e., the distortion of angles that a 30mm provides compared to a 77mm, say.

Of course you can throw out the background more out of focus by using a longer focal length (with the respective speed), or just by getting closer, but both these options change the design of the image. If you want a particular angle distortion and a particular framing, there is no option but to use the respective focal length. On FF, you then have more options to use shallow DOF, if you want it and are prepared to pay the respective price for it.

There are certain limitations that going full frame will address -- printing bigger and getting a little shallower depth of field are probably the biggest -- but simply going full frame will not fix bad photos.

You don't necessarily need to print bigger to see the advantages. Even on regular output sizes, I find that FF has a better contrast between sharp subject and OOF background. The difference may not be night and day; it can be subtle, just like it is subtle between a great lens and an OK lens.

Going FF will not fix bad photos, for sure. Using an FA 31/1.8 does not fix bad photos either. But it is a really, really nice lens that often rewards me with a special feel, even if it is just subtle. Does my Sigma 28/1.8 frustrate me? No, on the contrary, it is remarkably good and it could be argued that no one needs more IQ and bokeh than this very fine lens delivers. But still, if I don't need to get close, I prefer the FA 31/1.8 without hesitation.

It will be the same with the FF camera. My K-5 II would be more than competent for almost all applications, but the FF model will provide that bit of extra which means a lot to some.

[BrianR]

If someone wants to deviate from the simple practice of comparing images which are identical in all image-relevant parameters for the purposes of discussing differences between sensor formats, they should have a better argument than just "I can get the framing right by changing the subject distance and I don't care about the perspective change and its impact on DOF because I make the arbitrary decision to use the same lens on both formats for the same purposes". Continuing to use the same lens for the same purpose (e.g., portraits) despite a change in sensor size does not make sense from a photography aesthetics viewpoint and while it is something that someone could choose to do regardless, it just makes a comparison between sensor formats incredibly harder because one compares apples to oranges.

The "Available Lens" argument is a very strong one in practice- if I'm going to use the lenses I have available, how will they differ on FF? The 12 (or 13?) year digital aps-c compromise in Pentax land makes this a very practical and relevant question- there's literally tons of FF glass already being used on aps-c bodies by people itching to move to FF..

For me a 1.5x change in sensor size is not so huge that a lens will necessarily need to be assigned a new role on a new body. My aps-c prime bag holds 14mm, 28mm, 50mm, and a 100mm, I don't miss in-betweeners. Ymmv depending on how precise you are about perspective, available working space relative to your subjects, and what sort of compromises you're happy to live with.

I'm not deciding what an "apples to apples" comparison is.

I admit that I'm a proponent of "apples to apples" comparisons, in particular when people performing "apples to oranges" comparisons arrive at incorrect conclusions and then teach their incorrect findings to others.

Can't argue with that, but it doesn't make a well-formed apples to oranges comparison irrelevant if you've got oranges in your hand and are considering buying some apples.

I'm not sure about the "9 times out of 10" but in principle I agree.

8 out of 7 of the statistics I quote are made up nonsense, but 6 out of 7 times I do have something resembling a point. Maybe 5 out of 8 times.... 2 out of 9?

BTW, the utility of the "equivalence" approach is that it takes out all variability of the parameters by enforcing an apples to apples comparison. There is no end to the scenarios one could construct to argue the case for one particular property holding for some format in specific circumstances. A pretty fruitless endeavour, AFAIC. Hence, it is very useful to provide a level playing field and then observe which differences are present, if any. That's what falconeye did and if only more read his articles and understood them, a lot of the fruitless discussions could be avoided.

I'm not at all opposed to "equivalence", I think it's grand. It answers some meaningful questions, and anyone doing the reading will pick up tools to answer other questions as well. It's just not the case that the level playing field is where everyone plays or wants to play, and that needs to be respected.

---------- Post added 02-25-15 at 08:09 AM ----------

Not true.

I think Norm's point is that for the average person, 17 out 18 of their photos will be identical in the eyes of the average viewer whether they used FF or aps-c to take the photo. Given the generally small sized web outputs and overall awesomeness of aps-c these days, I'd say this is true.

There is always a difference -- even when the same DOF is used -- due to the different enlargement factors of the two formats. This difference impacts on AF accuracy, lens faults, and acuity.

I honestly don't believe you can stress the 'different enlargement factor' enough. It just explains so much.

[Rondec]

You wouldn't use the 30mm to get some kind of full frame equivalence.

You would use it because you want the perspective, i.e., the distortion of angles that a 30mm provides compared to a 77mm, say.

Of course you can throw out the background more out of focus by using a longer focal length (with the respective speed), or just by getting closer, but both these options change the design of the image. If you want a particular angle distortion and a particular framing, there is no option but to use the respective focal length. On FF, you then have more options to use shallow DOF, if you want it and are prepared to pay the respective price for it.


You don't necessarily need to print bigger to see the advantages. Even on regular output sizes, I find that FF has a better contrast between sharp subject and OOF background. The difference may not be night and day; it can be subtle, just like it is subtle between a great lens and an OK lens.

Going FF will not fix bad photos, for sure. Using an FA 31/1.8 does not fix bad photos either. But it is a really, really nice lens that often rewards me with a special feel, even if it is just subtle. Does my Sigma 28/1.8 frustrate me? No, on the contrary, it is remarkably good and it could be argued that no one needs more IQ and bokeh than this very fine lens delivers. But still, if I don't need to get close, I prefer the FA 31/1.8 without hesitation.

It will be the same with the FF camera. My K-5 II would be more than competent for almost all applications, but the FF model will provide that bit of extra which means a lot to some.

My experience is that when shooting portraits, the main reason that you use a wider angle lens is when you don't have enough shooting distance or, when you want to include more environmental features. The idea that you would shoot with a 30mm on full frame for portraiture for the "perspective shift" is probably not accurate. Yes, it will look different from an 85 mm lens shot, but I don't think the difference will be a beneficial one, unless you are shooting a different kind of photo (whole body).

As to the whole full frame is "better," it probably depends on the application and the skill of the photographer and quality of the glass used. Sure, the size of the sensor plays a factor, but there is a lot more to it than that.

[Tony Belding]

I found this video to be the best explanation of the difference between full-frame versus crop sensors:

You Tube

You Tube

And our word of the day is "negligible".

[BrianR]

And our word of the day is "negligible".

I thought it was 'Fuji'?

[normhead]

Thanks! The other day I saw brand new K-30+18-55 WR kits on sale for just $349! Is that a great deal, or what? I already have a K-3, but I almost bought one anyway, just because! (I bought a used HD DA 55-300 WR for $269 instead)


I sold my K-5 to help pay for a K-3, but sometimes I still wish I had one of the 16mp Pentax DSLR's. I really liked the K-5, but the K-3's improved AF is noticeable. Still, it seems like the K-5's high ISO IQ is a bit cleaner. I am nobody's knowledgeable pixel-peeper, or in any way eminently qualified to back up this perception. It just seems to be so when I scroll though photos by both cameras in Lightroom. Have you ever wanted to "move up" to a K-3?

On a more relevant note, a 33mm f/0.8 lens would be pretty cool! Wouldn't it?

Thanks,


Dave the Perennial Newbie

I have a K-3 and a K-5 and the K-5 only gets used for nostalgic purposes. The AF is so good on the K-3, when you go back to a K-5 there are noticeable issues. In terms of IQ I'm not sure there is much difference. You trade bit of DR for bit of resolution.

[Class A]

And our word of the day is "negligible".

The word of the day could -- as well as "Fuji" for which this is an advert -- also be "manipulation".

When he says his APS-C sample is "roughly" APS-C sized then he is right about "roughly" because the FF negative should be 2.25 times as large. The size difference does not look that big, does it? It looks more like a factor of 1.8.

Why make the difference visually smaller than it is?

Here is a proper comparison:


The difference in field of view is certainly not negligible and the size difference is big enough to be noticeable w.r.t. IQ.

Why argue that bigger is better and then state that 2.25 times bigger is "not better enough"?

He doesn't even scratch the surface of falconeye's treatment.

I have to fully agree with Zack Arias on one point, though. He is no Joe McNally.

[Tony Belding]

The word of the day could -- as well as "Fuji" for which this is an advert -- also be "manipulation".

What? He picked up a Fuji and mentioned it by name, therefore logically his entire presentation is somehow invalid? Is that what you are implying? Is that what passes for logic nowadays?

I'll admit, I have lusted after the Fuji X-T1 myself. It actually resembles those great old 35mm film cameras in size, appearance, and (most importantly) controls. However, that's just an emotional response on my part. After a while I realized it would be a lot cheaper to indulge that impulse by getting some film and actually shooting those great old 35mm film cameras.

However, when he picks up the Fuji in one hand and the Nikon in the other, it does illustrate the difference in portability. That's more a result of the Fuji being mirrorless than due to it having a smaller sensor, though.

When he says his APS-C sample is "roughly" APS-C sized then he is right about "roughly" because the FF negative should be 2.25 times as large.

1.5. The FF negative is 1.5X the size of APS-C. That's why we keep talking about a 1.5X "crop factor".

Why argue that bigger is better and then state that 2.25 times bigger is "not better enough"?

Because it's not 2.25X. It's 1.5X. And when dealing with human perceptions (in many areas, not only visually), I've found a good rule of thumb is that you have to double the magnitude of a stimulus for an improvement to be noticeable.

I actually do a lot of M4/3 shooting here. It is actually just about half the size of FF, and I reckon (other things being equal) the difference in performance between M4/3 and FF should be just about noticeable.

[BrianR]

1.5. The FF negative is 1.5X the size of APS-C. That's why we keep talking about a 1.5X "crop factor".

Linear Dimension vs Area. Class A is using 'bigger' to refer to the area, which differ by a factor of 2.25.

Either way, the 'aps-c negative' used in the video is too large relative to the FF one, measuring pixels in a screen shot I get a crop factor of about 1.35 between the two (corresponding 1.83X more area). This is visually misleading and of dubious integrity coming from someone trying to sell an aps-c camera.

[jsherman999]

What? He picked up a Fuji and mentioned it by name, therefore logically his entire presentation is somehow invalid? Is that what you are implying? Is that what passes for logic nowadays?

He's funded by Fuji, Fuji pays for his travel, etc. That video and it's author have been discussed on dpreview. Class A's statements are correct re the rest of it as well.

.

[cfraz]

Linear Dimension vs Area. Class A is using 'bigger' to refer to the area, which differ by a factor of 2.25.

Either way, the 'aps-c negative' used in the video is too large relative to the FF one, measuring pixels in a screen shot I get a crop factor of about 1.35 between the two (corresponding 1.83X more area). This is visually misleading and of dubious integrity coming from someone trying to sell an aps-c camera.

Maybe Fuji measures length the same way they measure ISO.

[Tony Belding]

He's funded by Fuji, Fuji pays for his travel, etc. That video and it's author have been discussed on dpreview. Class A's statements are correct re the rest of it as well.

OK, let me rephrase that...

So he's funded by Fuji, and Fuji pays for his travel, etc. Therefore his entire presentation is invalid? Is that really your argument; is that what passes for logic nowadays? If he'd had a Pentax K-01 in his hand instead of the Fuji, would the conclusion somehow be different?

As for the sensor sizes... You can look them up on Wikipedia. In fact, you can see them right here: Full-frame digital SLR - Wikipedia, the free encyclopedia

Here's the deal with APS-C and FF: The technical advantage of FF, the specification advantage of FF is, for the most part... Negligible.

That doesn't mean FF is worthless. That doesn't mean we (and Pentax) shouldn't go there. K-mount was designed for 135 format. We're already basically toting around FF-sized DSLRs already. If you're toting a FF-sized camera, you might as well have a FF sensor. We've got lenses for it already. Third parties are cranking out lenses for it. The only reason APS-C exists is because it's cheaper, but the cost of big sensors is coming down.

FF is a prestige item right now. I'd like to imagine someday it won't be like that, and APS-C DSLRs might just go away. (That's assuming all DSLRs don't go away, but that's yet another subject.)

However... If you want a bigger sensor that delivers a truly useful and satisfying upgrade in performance from APS-C, you should be looking at medium format.

[jsherman999]

OK, let me rephrase that...

So he's funded by Fuji, and Fuji pays for his travel, etc. Therefore his entire presentation is invalid? Is that really your argument;

The presentation isn't automatically invalid because of his sponsorship ties.

You would look for factual problems in the presentation, and then see if those problems helped build a case (or strong implication) to buy a camera just like the Fuji he was holding.

Lo and behold, there are factual problems in support of a Fuji-Friendly purchase decision therein.

Here's the deal with APS-C and FF: The technical advantage of FF, the specification advantage of FF is, for the most part... Negligible.

I disagree. I've shot around 90,000 frames on aps-c and 60,000 on FF now, and the difference is not 'negligible'. (I guess it completely depends on how you personally define negligible.)

However... If you want a bigger sensor that delivers a truly useful and satisfying upgrade in performance from APS-C, you should be looking at medium format.

The price is the concern there. For around $700 more than the K3 you can get a FF camera. Entry-level medium format is about $7000 more. The lenses are also extremely expensive... and not especially fast, if you'e looking for a low-light advantage.