• Log in with Facebook Log in with Twitter
Photo

Sigma 18-35mm f/1.8


  • Please log in to reply
52 replies to this topic

#1 felix

felix

    Advanced Member

  • Members
  • PipPipPip
  • 153 posts
  • LocationGermany/Switzerland

Posted 18 April 2013 - 07:16 AM

Hi everyone! 

I didn't think I'd ever see an f/1.8 zoom, but here it is: http://www.sigma-glo...5_18/index.html

 

Sounds pretty interesting!  ;)



#2 mst

mst

    Advanced Member

  • Moderators
  • 1,820 posts
  • LocationWesterwald, Germany

Posted 18 April 2013 - 08:12 AM

Wow, interesting indeed!

-- Markus
Editor (Nikon, Leica, Samsung reviews)
photozone.de

#3 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 3,737 posts
  • LocationThe Netherlands

Posted 18 April 2013 - 09:49 AM

Pretty nice going into FF territory, but that weight makes it a bit curious on APS-C... 810 grams for a ~28-55mm f2.8 lens... The Canon EF 24-70mm f2.8 L USM II weights 805 grams and the Tamron 24-70mm f2.8 VC USD weights 825 grams.

That is one hefty APS-C standard zoom.



#4 PuxaVida

PuxaVida

    Advanced Member

  • Members
  • PipPipPip
  • 757 posts
  • LocationIstanbul

Posted 18 April 2013 - 10:18 AM

Cool design... I think the zoom range is also well chosen. One can pair it with various alternatives from UWAs. And for the short tele range, I guess we'll be seeing another "new standart" from Sigma soon  :)... A 35-120mm f/2.8 would be nice for instance...

 

And as BC mentioned, I agree this one's quite heavy compared to the 24-70mm f/2.8 FFs... Maybe because of the build material and 17 glass elements.

 

Serkan



#5 popo

popo

    Advanced Member

  • Members
  • PipPipPip
  • 1,305 posts
  • LocationUK

Posted 18 April 2013 - 12:51 PM

Very interesting. This specific range is not a major interest area for me, but I'm hopeful that is an indication they will try longer focal lengths in future. Having said that, the tradeoffs are tricky. This could replace budget primes, but I can't see this lens being in any way cheap! Still could be a poor mans alternative to a high end prime set.

Actually, this lens would be far more interesting if they were to attempt this on APS-C mirrorless systems. The reduced need for retrofocus should hopefully simplify the design somewhat, reducing size and weight, and cost too. Imagine pairing this with a future X-Pro2!

dA Canon 7D2, 7D, 5D2, 600D, 450D, 300D IR modified, 1D, EF-S 10-18, 15-85, EF 35/2, 85/1.8, 135/2, 70-300L, 100-400L, MP-E65, Zeiss 2/50, Sigma 150 macro, 120-300/2.8, Samyang 8mm fisheye, Olympus E-P1, Panasonic 20/1.7, Sony HX9V, Fuji X100.


#6 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 3,737 posts
  • LocationThe Netherlands

Posted 18 April 2013 - 02:19 PM

Cool design... I think the zoom range is also well chosen. One can pair it with various alternatives from UWAs. And for the short tele range, I guess we'll be seeing another "new standart" from Sigma soon  :)... A 35-120mm f/2.8 would be nice for instance...

That then would have to be a 35-120mm f1.8, of course ;)

Sigma already has a 50-150mm f2.8 EX DC OS HSM, which will pair it nicely. Only this APS-C lens rivals the 70-200mm f2.8 full frames almost in size and weight... But it does have very sharp optics.

 

And as BC mentioned, I agree this one's quite heavy compared to the 24-70mm f/2.8 FFs... Maybe because of the build material and 17 glass elements.

 

Serkan

Well, obviously it also is heavy because of its f1.8-ness... f2.8 standard zooms for APS-C weight between 500-650 grams already.



#7 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 1,233 posts
  • LocationSwitzerland

Posted 18 April 2013 - 08:24 PM

Brightcolours, I've no idea how you calculate a f/2.8 - it has f/1.8 and you can't just multiply the aperture by the crop factor - that would only be fair, if the lens would have to bring enough light to a double as large sensor.

 

It has an aperture which is hard (if ever) to get for a range between 18 and 24mm - except some pricey 24/1.4 FF lenses. I can see that as a standard zoom for D7100. Given what the 35/1.4 performs, Sigma is really reaching for a high place in lens heaven.  :rolleyes:



#8 mst

mst

    Advanced Member

  • Moderators
  • 1,820 posts
  • LocationWesterwald, Germany

Posted 18 April 2013 - 10:38 PM

Brightcolours, I've no idea how you calculate a f/2.8 - it has f/1.8 and you can't just multiply the aperture by the crop factor - that would only be fair, if the lens would have to bring enough light to a double as large sensor.


It's absolutely fair if you want to compare lenses with equivalent DOF. In this regard a f/1.8 lens on APS-C is the equivalent of a f/2.8 lens on full frame. Which also shows: there is actually no size advantage if you take as much as possible into account.

-- Markus
Editor (Nikon, Leica, Samsung reviews)
photozone.de

#9 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 3,737 posts
  • LocationThe Netherlands

Posted 19 April 2013 - 07:11 AM

Brightcolours, I've no idea how you calculate a f/2.8 - it has f/1.8 and you can't just multiply the aperture by the crop factor - that would only be fair, if the lens would have to bring enough light to a double as large sensor.

That is just the nature of things. To get an equivalent lens, you multiply the focal length to get similar field of view. You multiply the f-value to get a similar aperture size and similar DOF. 

 

You can just multiply the focal length by crop factor, and similarly, you can just multiply the f-value by crop factor.

 

Simple example:

Suppose we have a 85mm lens on an 1.6x crop camera (compared to 135 format "full frame), with an focal length/aperture ratio of f/1.8.

To get a similar FOV we multiply by crop factor: 85 x 1.6 = 136mm.

To get a similar aperture we multiply the f-value by crop factor: f1.8 x 1.6 = f2.88, so lets say about f2.8.

Lets check if that is correct. 

Size of the aperture of the 85mm f1.8 lens:

85mm / 1.8 = 47mm aperture.

Size of the aperture of a 135mm f2.8 lens: 

135mm / 2.8 = 48mm aperture.

 

The 85mm f1.8 lens on 1.6x crop and the 135mm f2.8 lens are equivalent. They offer each respective sensor the same FOV angle, they offer a similar aperture size for the light to pass through, they will offer a similar DOF.

 

So this Sigma lens offers for APS-C what the f2.8 standard zooms offer on FF.

 

It has an aperture which is hard (if ever) to get for a range between 18 and 24mm - except some pricey 24/1.4 FF lenses. I can see that as a standard zoom for D7100. Given what the 35/1.4 performs, Sigma is really reaching for a high place in lens heaven.  :rolleyes:



#10 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 1,233 posts
  • LocationSwitzerland

Posted 19 April 2013 - 09:51 PM

Brightcolours, a f/2.8 is slower than f/1.8 no matter which sensor size.

 

That was the reason I wrote "you can't just multiply the aperture by the crop factor" - only if it comes to DOF, but not to a gain of 1.5 f-stops faster shutter speeds, not to a gain of a relatively narrow DOF in wide-angle (for APS-C). To compare that shutter speed, it has to be a 28-55/1.8 for FF! And that sounds much less cheap than a 28-55/2.8 - each manufacturer can do that.

 

I admit, 1.5 f-stops are easily done by contemporary VR/OS devices, one can handhold up to 4 f-stops - but one doesn't get short shutter speeds by VR//OS.

 

It looks long and thin in the pictures, but given the filter size 72mm, it's a huge piece of gear.



#11 popo

popo

    Advanced Member

  • Members
  • PipPipPip
  • 1,305 posts
  • LocationUK

Posted 19 April 2013 - 10:07 PM

If we are talking about strict equivalence, as best we can, then you DO have to multiply the f number.

Effective focal length, aperture, and ISO need to be multiplied. Shutter speed has to remain unchanged. Reducing the aperture means it would get darker, but you offset that by increasing the ISO too. Noise would be about the same since you have a bigger sensor.

Of course, if you don't care about strict equivalence, you can ignore that. But it doesn't stop it from being the case.

dA Canon 7D2, 7D, 5D2, 600D, 450D, 300D IR modified, 1D, EF-S 10-18, 15-85, EF 35/2, 85/1.8, 135/2, 70-300L, 100-400L, MP-E65, Zeiss 2/50, Sigma 150 macro, 120-300/2.8, Samyang 8mm fisheye, Olympus E-P1, Panasonic 20/1.7, Sony HX9V, Fuji X100.


#12 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 1,233 posts
  • LocationSwitzerland

Posted 19 April 2013 - 10:29 PM

What kind of equivalence? These number acrobatics are just ridiculous - for what reason has the ISO be multiplied? Does the sensor gets less sensitive or more just because of the sensor size?

 

You're saying

 

a 50mm becomes a 75mm (not true)

f/1.8 becomes f/2.8

ISO 200 becomes 300

 

Left numbers FF, right numbers APS-C

 

That doesn't make any sense

 

If you go on with that you end up with 500mm, f/18 and 2000 ISO for a normal point and shoot - what kind of equivalence that should be, I just fail to understand.   :blink:



#13 popo

popo

    Advanced Member

  • Members
  • PipPipPip
  • 1,305 posts
  • LocationUK

Posted 19 April 2013 - 11:22 PM

If you have two cameras of different sensor size and stood in the same spot, and wanted to get the same photo, you need to do what I said. If you don't care about making it match, you can do whatever you like.

One step at a time, you multiply the crop factor to the focal length so you get the same field of view.
Bit if you then set the same f number, you will get different depth of field. So you need to multiply that by the crop factor too.
But a higher f number would be darker right? So we have shutter speed and ISO left. We can't use a different shutter otherwise we'd have a different amount of motion blur. So we set a higher ISO.
In this case, you have look at the sensor in area terms. The bigger sensor has a bigger light collection area, offsetting what would have been higher noise.

As for a compact camera being equivalent to high f numbers and ISO on a DSLR, that's exactly the case. Point and shoots generally have massive depth of field, and have to apply a lot of noise reduction to get a nice looking image out.

dA Canon 7D2, 7D, 5D2, 600D, 450D, 300D IR modified, 1D, EF-S 10-18, 15-85, EF 35/2, 85/1.8, 135/2, 70-300L, 100-400L, MP-E65, Zeiss 2/50, Sigma 150 macro, 120-300/2.8, Samyang 8mm fisheye, Olympus E-P1, Panasonic 20/1.7, Sony HX9V, Fuji X100.


#14 popo

popo

    Advanced Member

  • Members
  • PipPipPip
  • 1,305 posts
  • LocationUK

Posted 20 April 2013 - 08:23 AM

From another forum:

First pictures with the lens http://lcap.tistory....5mm-f18-Preview

And on full frame http://lcap.tistory....with-5D-Mark-II

Looks good but confirms I'm NOT buying this lens due to the backwards zoom ring.

dA Canon 7D2, 7D, 5D2, 600D, 450D, 300D IR modified, 1D, EF-S 10-18, 15-85, EF 35/2, 85/1.8, 135/2, 70-300L, 100-400L, MP-E65, Zeiss 2/50, Sigma 150 macro, 120-300/2.8, Samyang 8mm fisheye, Olympus E-P1, Panasonic 20/1.7, Sony HX9V, Fuji X100.


#15 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 3,737 posts
  • LocationThe Netherlands

Posted 20 April 2013 - 08:28 AM

Brightcolours, a f/2.8 is slower than f/1.8 no matter which sensor size.

f2.8 is only slower than f1.8 on the SAME sensor. Or film. The sensor/film defines the speed. The f-value only describes the hole size. If you change the sensor or film, the speed will change. 

That was the reason I wrote "you can't just multiply the aperture by the crop factor" - only if it comes to DOF, but not to a gain of 1.5 f-stops faster shutter speeds, not to a gain of a relatively narrow DOF in wide-angle (for APS-C). To compare that shutter speed, it has to be a 28-55/1.8 for FF!

No, you are wrong there. if you set the cameras to an equivalent ISO value, the exposure times will also be the same. If you think of ISO still as "film sensitivity", you must think and read a bit more on the subject. ISO says nothing about sensitivity in the digital age.

And that sounds much less cheap than a 28-55/2.8 - each manufacturer can do that.

 

I admit, 1.5 f-stops are easily done by contemporary VR/OS devices, one can handhold up to 4 f-stops - but one doesn't get short shutter speeds by VR//OS.

Just change the ISO setting to an equivalent ISO setting, If you use ISO 200 on APS-C, use the equivalent ISO setting of 200 x 1.5 x 1.5 (square) = ISO 450. If you have a sensors with the same amount of sensels (pixels if you want), at ISO 200 on 1.5x crop and on FF with ISO 450 each sensel will register the same amount of light during the same exposure time.

It looks long and thin in the pictures, but given the filter size 72mm, it's a huge piece of gear.



#16 Brightcolours

Brightcolours

    Advanced Member

  • Members
  • PipPipPip
  • 3,737 posts
  • LocationThe Netherlands

Posted 20 April 2013 - 09:23 AM

What kind of equivalence? These number acrobatics are just ridiculous - for what reason has the ISO be multiplied? Does the sensor gets less sensitive or more just because of the sensor size?

The sensor gets MORE sensitive. If you have more surface, the sensor will catch more light. If you have bigger pixels, the pixels will catch more light. Hence: the FF sensor is MORE sensitive by definition. 

You're saying

 

a 50mm becomes a 75mm (not true)

No one says that 50 mm "becomes" 75mm. It is like this (for 1.5x crop compared to 135 format):

To get a similar field of view, you have to use an EQUIVALENT focal length (one that gives you the same image). So if you use 50mm on 1.5x crop format, you must use 75mm on 135 ("FF") format.

f/1.8 becomes f/2.8

Again the same: to get a similar DOF, you have to use an EQUIVALENT f-value, which will give the same aperture size. So if you use f1.8 on 1.5x crop, use f2.8 on FF for a similar DOF.

ISO 200 becomes 300

Wrong. The same applies again as above. To catch the same amount of light to form the image, and/or to take the image in a similar exposure duration, you will have to use an EQUIVALENT ISO setting. So if you use ISO 200 on 1.5x crop, you have to use ISO 450 on 135 format, for a similar amount of light forming the image and a similar exposure time.

Left numbers FF, right numbers APS-C

 

That doesn't make any sense

It does make sense in every way you look at it. But you have to not make thought process errors, like that f-values are linked to film sensitivity.

If you go on with that you end up with 500mm, f/18 and 2000 ISO for a normal point and shoot - what kind of equivalence that should be, I just fail to understand.   :blink:

I think you are confusing yourself here... "normal point and shoot" (I am guessing you mean compact digital cameras with a small sensor, here....) use lenses with super small focal lengths. Of course, if you take a compact digital ultra zoom model which reaches a 500mm FF EQUIVALENT, then yes, 500mm for FF for that 500mm equivalent setting on the compact digital. 

 

Lets take a "typical" compact digital as example, a Canon PowerShot S100. It has a 7.44 x 5.58mm sensor size. This S100 has a 5.2-26mm f2-5.9 lens.

And lets compare it to a "typical" 135 format FF DSLR, a Nikon D3s. It has a 36 x 23.9mm sensor size. Which lens would be equivalent on the D3s to the lens of the S100?

 

36 / 7.44 = ~4.8x crop. I take the width and not the diagonal, to get the same FOV when using the camera in horizontal orientation.

 

To get the same FOV, the D3s will need a 5.2x4.8= 25mm to 26x4.8= 124mm lens.

So, basically a Nikon 24-120mm lens.

To get the same DOF (same size aperture), we will need to use 2x4.8= f9.6 to 5.9x4.8= f28.3.

 

Of course, we can not get a lens with such small max. aperture on FF, so the lens will always have more capabilities than that of the compact digital camera (which of course is the reason both type of cameras exist). 

 

So, an equivalent lens to the S100's 5.2-26mm f2-5.9 on 135 format will be a 24-120mm f10-f32 lens, basically. One will have to settle for the Nikon 24-120mm f4 VR or its older sibling and stop it down accordingly.

 

To match the amount of light forming the image, and keep the same exposure time (both have nothing to do with the equivalence of the lens, we are talking about the equivalence of the sensor now), we have to set an equivalent ISO setting.

Suppose we use ISO 200 on the digital compact. We then have to use 200 x 4.8 x 4.8 = ~ ISO 4600 in the D3s.

This then will result in a similar exposure times, when we set the lenses at equivalent settings.



#17 vodanh1982

vodanh1982

    Advanced Member

  • Members
  • PipPipPip
  • 39 posts

Posted 20 April 2013 - 09:51 AM

What kind of equivalence? These number acrobatics are just ridiculous - for what reason has the ISO be multiplied? Does the sensor gets less sensitive or more just because of the sensor size?

 

You're saying

 

a 50mm becomes a 75mm (not true)

f/1.8 becomes f/2.8

ISO 200 becomes 300

 

Left numbers FF, right numbers APS-C

 

That doesn't make any sense

 

If you go on with that you end up with 500mm, f/18 and 2000 ISO for a normal point and shoot - what kind of equivalence that should be, I just fail to understand.   :blink:

Let's compare 50 f/1.8 vs 75 f/2.7

From http://en.wikipedia.org/wiki/F-number

 

D (crop)=f/N=(50/1.8) = 27.7

D (FF) = (75/2.7) = 27.7

 

So two lenses have the same "opening"

 

Because the FF is larger the light intensive is lower. At the same shutter speed and ISO the FF image is underexposed. To make 2 images at the same exposure, we need to increase ISO of FF. Now the FF has higher ISO than the crop.

The question: will the FF at higher ISO has more noise than the crop at lower ISO?

There are two factors that affect the noise

-The sensor technology.

-The total amount of light get to the sensor.

Assume two sensors are made with the same technology. Because of the same "opening", two sensors receive the same amount of light with the same shutter speed. In conclusion, two sensors has the same noise even at different ISO.

In reality, the result maybe a little bit off because the crop and the FF are not manufactured using the same technique, micro-lenses, spacing, light fall off sensor (non-rounded sensors), etc

 

Other factor that is not equivalent is the resolution. Lens is limited by their own resolution in term of LP/mm. Assume the same sharpness and even sharpness distribution for the whole area of the lens, the FF will have more resolution because it is bigger. The total resolution = lens resolution x picture height.



#18 Sylvain

Sylvain

    Advanced Member

  • Members
  • PipPipPip
  • 556 posts
  • LocationBrussels, Belgium

Posted 20 April 2013 - 01:15 PM

A possible game changer, go Sigma, go.



#19 Klaus

Klaus

    Chief Editor

  • Moderators
  • 4,458 posts
  • LocationSydney, Australia

Posted 20 April 2013 - 11:22 PM

A possible game changer, go Sigma, go.


Is it ? Wirh this focal length range ?

Fast standard zoom lenses are the most overrated lenses in general.
Not overly fast compared to primes, overpriced, heavy.
  • dave9t5 and x-vision like this
Chief Editor
photozone.de

#20 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 1,233 posts
  • LocationSwitzerland

Posted 21 April 2013 - 08:18 PM

With all this math mambo-jambo you, popo and brightcolours, just seem to forget:

There's not such a lens for APS-C like there are some very pricey ones for FF - and to use them, you just need to get an also pricey FF-body. Which, following Klaus' post are overrated and overpriced anyway, so all the people buying these are just wasting money? ;) :huh:

I don't get the point trying to calculate equivalence for FF, this is not a question for an APS-C shooter how "lame" this lens is compared to what is available for FX. At the moment it's just outstanding for APS-C and if it can be used at open aperture, it still gives 1.5 f-stops more than the usual 17-50/2.8 "fast standard zooms" which I don't find that overpriced (except a Nikon 17-55/2.8, ok, point taken).

And DoF? At 18mm and open aperture it's 0.86m, not breathtakingly narrow. f/2.8 (with APS-C and same focal length) would give 1.47m.

One can ask if a Sigma 17-70/2.8-4 OS couldn't be a bit more useful, because of the OS which compensates more than the 1.7 missing f-stops (I was interpolating until 35mm) - but there's no other way to get a narrow DoF and short shutter speeds.

And, by the way, it's not correct to calculate those ISO differences like you did: the borders between the sensels keep the same size, more or less, so in reality the sensitive area decreases more when the sensors get smaller with the same pixelcount.

"The sensor gets MORE sensitive. If you have more surface, the sensor will catch more light. If you have bigger pixels, the pixels will catch more light. Hence: the FF sensor is MORE sensitive by definition."

So a film 810 inch and 100 ISO is per your definition much more sensitive than a film 135 with 100 ISO?

Strange - my old spotmeter didn't list film formats as parameter. Must've been a cheap one... :rolleyes:

It will capture more photons, in absolute numbers - but not more on the same area than the smaller frame will catch, too.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users



© by photozone.de