[Rondec]

As Class A implies, If you want truly equivalent images (same FOV and DOF and shutter speed,) then there's no advantage in low-light/SNR for larger sensors of the same generation, because to get those equivalent images you need to stop down the FF camera about 1.3 stops to get the same DOF for that FOV. If you're shooting in low-light when you do this, you'll need to boost ISO an equivalent amount to maintain the same shutter speed... thus introducing more noise.

In a practical sense, the larger sensor does give a 'low light' advantage because you're probably rarely interested in forcing the image to have the exact same DOF as the aps-c shot would - you're either willing to accept the more shallow DOF - or in many cases, even welcome it.

In other words, we use equivalence to describe the relationship between formats - this doesn't mean that we're constantly striving for equivalent images. This gives FF a very real practical low-light advantage in real-world shooting, and it's why DxOMark scores it like that.

Also, you have to consider the lenses available to both formats. A 50mm f/1.8 on FF is capable of delivering images that would require about a 33mm f/1.2 lens on aps-c, and one of those doesn't exist (yet). Your f/2.8 zooms on FF are the equivalent of f/1.8 zooms on aps-c. Etc...

Sigma has shaken this up now with that excellent 1.8 zoom, but I hear rumours that they're coming out with an f/2 constant 24-70 for FF, and the line moves again

.

I could be wrong, but I think Sigma is using "speed boosting" technology to design these faster zooms and not sure that would be usable on full frame.

[jsherman999]

I could be wrong, but I think Sigma is using "speed boosting" technology to design these faster zooms and not sure that would be usable on full frame.

I don't think anyone knows anything yet, but this is the rumor that's making the rounds.

[Rondec]

I don't think anyone knows anything yet, but this is the rumor that's making the rounds.

I did a web search and found that, but doesn't seem to have anything backing it up. Seems like it would be expensive and brutally big (the f1.8 zoom is big, but pretty reasonably priced, considering). But certainly Sigma seems to be coming up with a lot of pretty sweet lenses that give Canon/Nikon a good run for their money.

[RonHendriks1966]

I don't think anyone knows anything yet, but this is the rumor that's making the rounds.

That was a photoshop fake.

[falconeye]

Where the 50 1.2 would be absolute killer is at f2.0 or f2.8 on a FF format. The lens would be nice and sharp in the in-focused areas and would still retain the really narrow DOF.

Zeiss is doing a new 1.4/55 which s tack sharp across the field fully opened. No need to stop down. But it is medium format quality at medium format price ($3000).

[falconeye]

That was a photoshop fake.

The photo may have been shopped. But it is relatively straightforward to scale all measures of a 18-35/1.8 by a factor 1.5 and derive a 27-53/1.8 for a full frame image circle and same LW/PH resolution capabilities. Of course, the larger elements will be more expensive to manufacture (after all, the lens would collect more light). But they would allow for slightly larger production tolerances too.

Of course, in real life, one would use the 27-53/1.8 design as a starting point only to cover the 24-70 range. And a few modifications are in order to possibly utilize the shorter registration distance (after scaling) and to arrive at an even higher LW/PH figure. F2 seems like a reasonable compromise here.

So, even if the rumor is 100% fake, I would still assume that this lens or a very similiar one is in the makes.

[falconeye]

The biggest reasons (in order IMO...but I am right) for FF vs. APS:

1) ISO advantage from more signal
2) Better DR related to signal
3) Easier with the wides in lens design especially
4) More legacy backwards compatibility
5) Shallower DOF control
6) Likely greater resolution

Cons:

1) Much more expensive camera body
2) Larger, heavier camera body
3) Less telecentricity for longer FL's
4) Much bigger, long glass
5) Web imaging lessens visible difference via-a-vis APS
6) Newer sensors really require new glass

Of all factors compared, I found that focus accuracy improves the most when reducing crop (assuming PDAF and expressed as the power in a crop^power law). Still true for manual focus accuracy using an OVF. Other effects (from your list) are less impacted. The second most impacted factor is feasible lens resolution. So overall, practically achievable resolution is my #1 differentiator. The bigger glass argument assumes a choice of non-equivalent lenses (that's fine, but should be mentioned, since it is optional). Same for the bigger body (here, the choice is more on the vendor's side though). A small FF body with small lenses would combine advantages from both lists.

DR can be made equal if APSC sensors would be made with deeper walls compensating for the smaller surface. But it requires new developments like Canon's dual pixel AF to overcome the APSC AF problems at higher resolutions.

---

Last edited by falconeye; 08-05-2013 at 05:10 PM.

[Pål Jensen]

But how much dynamic range and/or resolution do you need? At a certain point the law of diminishing returns set in.
As for fast lenses; most photographers go through that phase (I've been there with 50/1.2, 85/1.4, 135/1.8 and 300/2.8) and then move on and take a more pragmatic view of lenses.

[jsherman999]

But how much dynamic range and/or resolution do you need?

You don't always need it - but to have it available is very nice at times.



.

As for fast lenses; most photographers go through that phase (I've been there with 50/1.2, 85/1.4, 135/1.8 and 300/2.8) and then move on and take a more pragmatic view of lenses.

Can your 'pragmatism' magically bring more light to the scene? Fast lenses add to your system's power and capability, they are not something you graduate from when you (or your sensor) get 'better'.

.

[Pål Jensen]

Can your 'pragmatism' magically bring more light to the scene? Fast lenses add to your system's power and capability, they are not something you graduate from when you (or your sensor) get 'better'.

.

People rarely buy what they need but what they want; thats were pragmatism comes in. I've shot lots low light work but came to the conclusion that ultrafast lenses weren't that useful as you need a some DOF in most images anyway - something that also help for sharpnes and focus accuracy. Todays sensor technology have made fast lenses far less needed. Not that long ago 200ISO were max for publishable results. 200ISO F:4 (with an 1.8 lens) below....

[Class A]

In other words, we use equivalence to describe the relationship between formats - this doesn't mean that we're constantly striving for equivalent images. This gives FF a very real practical low-light advantage in real-world shooting, and it's why DxOMark scores it like that.

I agree with a lot of what you've said, but not with the above.

The equivalency argument is not based on the idea that you'd always want to take equivalent images.

It is used to make fair comparisons.

The "real practical low-light advantage" of FF, does not stem from the sensor size. It comes from the faster lenses. That's why I don't entirely agree with the DxOMark approach of granting better low-light scores to FF sensors. They should be comparing sensors, not lenses.

[Pål Jensen]

I agree with a lot of what you've said, but not with the above.

The equivalency argument is not based on the idea that you'd always want to take equivalent images.

It is used to make fair comparisons.
.

If it was the idea of fair comparison we would in the film days compared the Pentax 67 with 3200ISO film with a Pentax MX with 50ISO (or something like that). No one did. Fair comparison is to use the same film cause thats the only way to spot the real differences between the formats - and thats after all what we're after - we don't want them to be equal. Thats why we never heard about equivalency in the film days.
Remember, you MUST take exposure into your equivalency calculation, if not, you have no image and then your results is not relevant for photography. This correct approach will show that there can never be true equivalency between formats and that whatever equivalency you choose will be a subjective decision that will (in real life) change from image to image. This also provide the best reason, apart from differences in image quality, to choose a format over the other.

[jsherman999]

I agree with a lot of what you've said, but not with the above.

The equivalency argument is not based on the idea that you'd always want to take equivalent images.

It is used to make fair comparisons.

Yes, I know, which is what I said. The reason I fleshed it out explicitly is that so many people read a discussion on equivalency and then ask a variation of the question, "but why do you need to try to take equivalent images all the time?" **(see edit) which indicates that they may not even understand why equivalency is being discussed, much less the particular tenets of equivalency.

The "real practical low-light advantage" of FF, does not stem from the sensor size. It comes from the faster lenses. That's why I don't entirely agree with the DxOMark approach of granting better low-light scores to FF sensors. They should be comparing sensors, not lenses.

Stating it like this ^^ isn't ideal, IMO, because... I don't know the best way to put it besides "it doesn't map to real life options". It doesn't really come from the faster lenses, because the lenses we're talking about are usually the same lenses - or at least lenses with the same maximum apertures. DXOMark gets this, and their scores tend to map directly with the user experience - and thus are fairly useful.

EDIT ** Just like this:

we don't want them to be equal. Thats why we never heard about equivalency in the film days.

.

[Digitalis]

Zeiss is doing a new 1.4/55 which s tack sharp across the field fully opened. No need to stop down. But it is medium format quality at medium format price ($3000).

that lens is actually a medium format lens - it is important to note it isn't a planar design like most 50mm lenses Zeiss have designed for 35mm format, it is a distagon design which is more commonly used with 50mm lenses on Medium format, much like the Zeiss CFI 50mm f/4 Distagon for the hasselblad V system.

[falconeye]

that lens is actually a medium format lens - it is important to note it isn't a planar design like most 50mm lenses Zeiss have designed for 35mm format, it is a distagon design which is more commonly used with 50mm lenses on Medium format, much like the Zeiss CFI 50mm f/4 Distagon for the hasselblad V system.

Oh, interesting.
OTOH, it is F/1.4, an unheard of aperture for MF. So, it probably is a mix of both worlds ...