Sigma Announces 18-35 f/1.8 DC HSM Art for APS-C

[Albi86]

Surely you know that taking this small sensor, large aperture thing to extremes like this does not result in this Panasonic FZ200 having a lens equivalent of a 600mm f2.8 at full zoom, as Panasonic would like you to think? In terms of framing, yes. In terms of exposure due to ISO tweaking of the sensor, yes. In terms if DoF, no. In terms if light gathering, no! There's no way that a 'slow' Canon 600/4 II on a 1D X as a package gathers less light than that 'f2.8' lens/sensor combo of the Panasonic. Yes, both at f4, 1/1000th of a sec, they'll both have to use an identical ISO to expose the same, but look past using the same rated ISO and guess which combo would work best to get clean images at high shutter speeds in low light?

If you can understand that, then surely you can see the 1.6x crop gives just over a stop less light gathering than a full frame sensor - making f1.8 on crop equal in light capturing terms to an aperture 1.6x smaller on FF - f2.88.

Aperture is just a way to measure the diameter of the iris blade. It's connected to, but it doesn't measure, the real amount of light gathered. This is why T-stops were invented.

[pwp]

Well good! If you shoot APS-C...
Though I may be inspired to get a 7DII.
But is a FF f/1.8 zoom next?

-PW

[LifeAfter]

Imagine what could this lens do with a Metabones speed booster!!!

[CarlTN]

I'm really excited about this lens, and I don't even own a crop body.  I'm excited about the implications for the future.  If this lens can be produced and has good optics (which will be the real issue), it raises so many interesting implications for the future.

A 27-55mm, or even 27-50mm f/1.8 FF lens would be absolutely amazing if it had good optics.  Once the technology is out there, reverse engineering means that this advance will soon be in the hands of other manufacturers.  The very nature of putting out an APS-C only lens means that the price has got to be somewhat reasonable, as there are not (to my knowledge) many APS-C lenses over a thousand dollars US.

+1
My complaint is, it should have simply been a full frame lens.  It still would have been relatively affordable, even if its street price was just under $2k.  Why?  Because if Canon made a full frame lens like this, it would be closer to $3k or even above 3k, and (very likely) not much better, if any.

I can certainly see why it's not full frame.  The smaller pixels and sensor, benefit greatly from a faster lens...because they are starved for light by the nature of their size.

I have been calling for a fast zoom, myself.  Perhaps Tamron will do the full frame version?  I want something bigger and more exotic though, and by the time it ever got designed and hit the market...I will probably be able to afford one: 90-150mm f/0.9 with..."OS"...It's just a fantasy of mine, not realistic that anything like this will ever exist.  Perhaps by 2017 or so, someone will build something...

[rs]

Aperture is just a way to measure the diameter of the iris blade. It's connected to, but it doesn't measure, the real amount of light gathered. This is why T-stops were invented.

T-stops are a measured version of light as opposed to f stops which are theoretical - so they take into account the transmission of light, including effects such as reflections and tinting of glass. However, even T-stops don't take into account the size of the imaging circle or the size of the sensor. That's why a 24-70/2.8 II on FF is more than a worthy rival to this 18-35/1.8 on crop, yet if you mount the same 24-70/2.8 II on crop, it is not.

[Albi86]

Aperture is just a way to measure the diameter of the iris blade. It's connected to, but it doesn't measure, the real amount of light gathered. This is why T-stops were invented.

T-stops are a measured version of light as opposed to f stops which are theoretical - so they take into account the transmission of light, including effects such as reflections and tinting of glass. However, even T-stops don't take into account the size of the imaging circle or the size of the sensor. That's why a 24-70/2.8 II on FF is more than a worthy rival to this 18-35/1.8 on crop, yet if you mount the same 24-70/2.8 II on crop, it is not.

Exactly. F-stops = T-stops in an ideal lens; in practical terms they are a better measure of DoF than they are of light transmission.

I don't agree with the rest though. Or better, I'm not quite sure what you mean.

[indigo9]

Hey guys, read what I wrote. I said f1.8 on crop is brighter than f2.8 on full frame when both have the same shutter speed and ISO, then you start telling me that I'm wrong to say FF is brighter when they're both at f1.8 and the same shutter speed and ISO. I didn't say that, so what gives? I went to great lengths to explain that at the same aperture, shutter speed and ISO, they both expose the same due to the different light gathering of the format being compensated for by the amplifiers being set different. If you can look past that same exposure settings between formats and start to use the higher ISO's with lower noise levels this lower amplification of FF rewards you with, you'll find the true nature of the light gathering of FF lenses on FF sensors.
...

So I was about to chime in that rs is wrong and actually sat down to prove it only to realise that he's right (although not terribly clear).

For the same angle of view:

80mm 1.0x [FF]      @ f/2 aperture = 40mm diam pupil
50mm 1.6x [APSC] @ f/2 aperture = 25mm diam pupil

Given that we have the same angle of view, the amount of light falling on the sensor is purely a function of the pupil diameter. f/2 on full frame is not the same as f/2 on crop.

Exposition
In case you're still not convinced, now assume that you have a FF and APSC sensor using the same production process & the same overall resolution -- they will have the same sensitivity per photon. If you take a shot with the same shutter speed, given that more light falls on the full frame sensor you would expect to use less sensor signal amplification [ie a lower ISO setting].

Alternate exposition
Or looking it at another way, if we keep the sensor signal amplification [ISO setting] constant, more light has fallen on each FF sensor pixel [same resolution remember], so you'll have to reduce the shutter speed of the FF to get the same exposure as APSC. Clearly FF f/2 is not the same as APSC f/2.

[Albi86]

Hey guys, read what I wrote. I said f1.8 on crop is brighter than f2.8 on full frame when both have the same shutter speed and ISO, then you start telling me that I'm wrong to say FF is brighter when they're both at f1.8 and the same shutter speed and ISO. I didn't say that, so what gives? I went to great lengths to explain that at the same aperture, shutter speed and ISO, they both expose the same due to the different light gathering of the format being compensated for by the amplifiers being set different. If you can look past that same exposure settings between formats and start to use the higher ISO's with lower noise levels this lower amplification of FF rewards you with, you'll find the true nature of the light gathering of FF lenses on FF sensors.
...

80mm 1.0x [FF]      @ f/2 aperture = 40mm diam pupil
50mm 1.6x [APSC] @ f/2 aperture = 25mm diam pupil

He is not.

In optics, the f-number (sometimes called focal ratio, f-ratio, f-stop, or relative aperture[1]) of an optical system is the ratio of the lens's focal length to the diameter of the entrance pupil.

You can't compare different focal lenghts. A 50mm is a 50mm both on crop and FF. However, being the former smaller, you crop the edges to a degree which would be equivalent to the angle of view of a 80mm on FF. This is a quick way to grasp the concept, but it's not exact in a way that you can make calculations.

[zim]

and bring on a 24-70 2.8 IS art line for FF I hope it's more 'when' than 'if'

[rs]

Hey guys, read what I wrote. I said f1.8 on crop is brighter than f2.8 on full frame when both have the same shutter speed and ISO, then you start telling me that I'm wrong to say FF is brighter when they're both at f1.8 and the same shutter speed and ISO. I didn't say that, so what gives? I went to great lengths to explain that at the same aperture, shutter speed and ISO, they both expose the same due to the different light gathering of the format being compensated for by the amplifiers being set different. If you can look past that same exposure settings between formats and start to use the higher ISO's with lower noise levels this lower amplification of FF rewards you with, you'll find the true nature of the light gathering of FF lenses on FF sensors.
...

80mm 1.0x [FF]      @ f/2 aperture = 40mm diam pupil
50mm 1.6x [APSC] @ f/2 aperture = 25mm diam pupil

He is not.

In optics, the f-number (sometimes called focal ratio, f-ratio, f-stop, or relative aperture[1]) of an optical system is the ratio of the lens's focal length to the diameter of the entrance pupil.

You can't compare different focal lenghts. A 50mm is a 50mm both on crop and FF. However, being the former smaller, you crop the edges to a degree which would be equivalent to the angle of view of a 80mm on FF. This is a quick way to grasp the concept, but it's not exact in a way that you can make calculations.

Thanks indigo9, nice explanation. Albi86 - a 1.6x crop camera crops the FoV, so to get the same framing, you have to use a different focal length. This 18-35 crop lens is a rival to a 24-70 on FF, not a 16-35. 80mm on FF does frame the same as 50mm on crop - and if they both have the same aperture, the 80mm lens has to have a bigger entrance pupil.

And using indigo9's simple (yet still misunderstood) explanation, to compare the 18-35/1.8 to the 24-70/2.8 on FF we'd need to set them both to a focal length to give an equal field of view. So for the sake of this example, lets use the long end of the Sigma's zoom - set the 18-35/1.8 to 35mm, which is equivalent of 56mm on the 24-70. We get the following:

56mm 1.0x [FF]      @ f/2.8 aperture = 20mm diam pupil
35mm 1.6x [APSC] @ f/1.8 aperture = 19.4mm diam pupil

So, vaguely less light gathering from the new Sigma lens, as well as a vaguely wider DoF.

Neuro - do you want to chime in at this point? People respect your opinion 

And as I said before though, this is a very interesting lens, and its a great option for crop sensor users - while it lets in slightly less light and has a narrower zoom range than the 24-70/2.8, its a great alternative to moving to a more expensive body, and the internal zoom is a great feature for a normal lens.

[Albi86]

...

You are mixing two concepts that are unrelated.

If we talk about framing, then you're right. The smaller effective aperture is the reason why DoF is bigger on crop at a given aperture and angle of view (note that I didn't say focal length).

However, framing has nothing to do with light gathering. f-stops are a function of focal length, not angle of view. A 50mm is a 50mm on every camera. It's the angle of view that changes in relation to sensor size, not the focal length.

Another quote from Wikipedia:

A 100 mm focal length f/4 lens has an entrance pupil diameter of 25 mm. A 200 mm focal length f/4 lens has an entrance pupil diameter of 50 mm. The 200 mm lens's entrance pupil is larger than that of the 100 mm lens, but given the same light transmission efficiency, both will produce the same illuminance at the focal plane when imaging a scene of a given luminance.

[Sella174]

I think the wheels come off because we are using "full-frame" lenses on "crop-frame" cameras, and insist on comparing everything to the "35mm" (aka "full-frame") format.

[Albi86]

I think the wheels come off because we are using "full-frame" lenses on "crop-frame" cameras, and insist on comparing everything to the "35mm" (aka "full-frame") format.

This can be done with no problems. The important thing is not to confuse focal length with angle of view. We can say that a 50mm gives you the same framing ( = angle of view) of a 80mm on FF; however, other parameters related specifically to focal length and not to angle of view are not affected.

[rs]

...

You are mixing two concepts that are unrelated.

If we talk about framing, then you're right. The smaller effective aperture is the reason why DoF is bigger on crop at a given aperture and angle of view (note that I didn't say focal length).

However, framing has nothing to do with light gathering. f-stops are a function of focal length, not angle of view. A 50mm is a 50mm on every camera. It's the angle of view that changes in relation to sensor size, not the focal length.

Another quote from Wikipedia:

A 100 mm focal length f/4 lens has an entrance pupil diameter of 25 mm. A 200 mm focal length f/4 lens has an entrance pupil diameter of 50 mm. The 200 mm lens's entrance pupil is larger than that of the 100 mm lens, but given the same light transmission efficiency, both will produce the same illuminance at the focal plane when imaging a scene of a given luminance.

But to compare a crop lens, crop sensor combo to a FF lens, FF sensor combo, there's no point in comparing two with different framing - otherwise you'd be arguing this 18-35 crop lens is a direct equivalent of a 16-35 FF lens on FF.

To exaggerate, is a 100mm f5.6 large format lens with its huge image circle the same as a 100mm f5.6 lens and its tiny image circle on a compact? Is it wrong to compare lenses which give the same framing? Surely from a photographers point of view, they're two very different lenses?

Wikipedia doesn't take into account imaging circle in that equation you're quoting. The aperture of a lens is a bit like working out the speed of water in a hose pipe. The imaging circle is a bit like the diameter of the pipe. Widen the pipe and keep the speed the same, you get more coming through.

Or think about it like this - imagine a photo taken with a FF lens and a FF sensor. Now you take that same photo and you crop out just the centre 40% - you've taken away 60% of the image - which is also 60% of the light that passed through that FF lens. You're left with only 40% of the light. That's what crop does. You need a faster lens on crop to make it capture the same amount of light in that smaller area.

[Mantanuska]

Given that we have the same angle of view, the amount of light falling on the sensor is purely a function of the pupil diameter. f/2 on full frame is not the same as f/2 on crop.

Exposition
In case you're still not convinced, now assume that you have a FF and APSC sensor using the same production process & the same overall resolution -- they will have the same sensitivity per photon. If you take a shot with the same shutter speed, given that more light falls on the full frame sensor you would expect to use less sensor signal amplification [ie a lower ISO setting].

More light falls on the FF sensor but it is spread out over a bigger area. actual intensity (or in this case it helps to think about it as density) of the light is the same.

The only reason why you are able to use a higher ISO on FF is because the pixels are larger on FF which means better signal to noise.