Do we really benefit from ultra fast lenses?

I read this article

http://www.dxomark.com/index.php/Reviews/F-stop-blues

from DxO and just an honesty disclaymore here but I first saw it on the NEX forum

http://www.dpreview.com/forums/thread/3545800

Basically what DxO argues is that these hyper fast lenses don't really give you the light gathering advantage claimed and manufacturers gain up the ISO to give you a "virtual" ISO advantage. They suggest an F4 as being a good number to rely on.

I do seem to recall some voigtlander users here who showed the shutter speeds Vs. Aperture and it was as expected down to F1.4 but this is just off memory.

If anyone has personal experience in this matter, please chip in for those of us considering buying hyper fast lenses (F1.4 and larger).

Well it was all covered in a thread about an SLR 25 T0.95 lens in this forum.

I have to say a related issue and one that is arguably more interesting, would seem to be the performance of IBIS on Oly's new E-M1. Two reviewers say that it enables them to take sharp images at over 1 second shutter speed. Now these are quite some claims - afterall if it wasnt possible you would look a bit stupid.

It seems a major improvement over the incredibly impressive E-M5 IBIS. I dont recall what claims were made about longer shutter speeds but I dont recall reviewers claiming better than 1/2 a second.

Now if we can hand hold for 1 second you really have to begin to doubt the need for fast glass to be 'fast' at all. Admittedly people also like the DOF control. But to give you and example - Oly is rumored to be bringing out a 7-14 2.8 next year but why? It is not as though 2.8 on a 7-14 will give you shallow DOF and if you can handhold for 1 second (which is mostly likely on a wide-angle) I doubt you would find need for faster than f4.

[No message]

Abrak said:

Abrak wrote:

Well it was all covered in a thread about an SLR 25 T0.95 lens in this forum.

I have to say a related issue and one that is arguably more interesting, would seem to be the performance of IBIS on Oly's new E-M1. Two reviewers say that it enables them to take sharp images at over 1 second shutter speed. Now these are quite some claims - afterall if it wasnt possible you would look a bit stupid.

It seems a major improvement over the incredibly impressive E-M5 IBIS. I dont recall what claims were made about longer shutter speeds but I dont recall reviewers claiming better than 1/2 a second.

Now if we can hand hold for 1 second you really have to begin to doubt the need for fast glass to be 'fast' at all. Admittedly people also like the DOF control. But to give you and example - Oly is rumored to be bringing out a 7-14 2.8 next year but why? It is not as though 2.8 on a 7-14 will give you shallow DOF and if you can handhold for 1 second (which is mostly likely on a wide-angle) I doubt you would find need for faster than f4.

Click to expand...

Can you shoot kids indoors at f4?

I still think fast glass is better but I will have to read that article. I know even with a 20mm f2.8 on my NEX I had a hard time getting good pictures of people indoors without flash. I couldn't get any good pictures with the kit lens and that was at f3.5.

I know with an f1.0 to f1.8 I can get nice, sharp and bright pictures of people indoors.

Shutter speeds that long would only be useful for things that don't move, in which case might as well use a tripod anyway.

Define ultra?

I understand what the article is saying, but I also notice a real difference shooting indoors at 1.8 vs 2.8

mis informed if you take DxO result as a cue cause they do not test lens, they test lens+camera combo and this ISO boost is what the camera did to the exposure , not the lens, and its only there if you use automation. Remember there is always Manual Exposure.

I must though extort DxO for the site really never claim that they were specifically relaying info with the lens and lens only. And its test result on camera body is informative especially with regard to actual ISO. But instead of boosting ISO, I find most of todays camera actually underboost the ISO, and instead rely on the in camera JPEG engine to raise the mid tone and highlight ( which is perhaps understandable ) but its no less frustrating for someone wanting to take control of the exposure for real. One of the thing about ISO is that there is no real absolute ISO, not on a film nor on a sensor. On Film, we have a reference EI ( or so call film speed ) and on a sensor its how a tonal luminesity scale be mapped and recorded ( or not recorded ). Even back in film days its not hard to find film that actually do not really ( strictly technically speaking ) deliver that stated ISO speed.

Whether a Ultra Fast lens can benefit you or not depend on how you employ and how you expose. Its not about the camera. With Film, the film stock pretty much dictate already, but it can be somewhat altered by development. On a digital platform. Fast lens . if used properly is just the same as in old days. You used it to effect an exposure within a reasonable bracket so you can get what you want out of it , either its stopping motion, lowering ISO for less noise, or placing DOF where you wants it. It did not change then and is still the same now.

One of my most frequent use of fast lens is fixing an exposure at a certain aperture / shutter / ISO / WB across a shooting session ( when lighting condition allows that ) so I can get consistent exposure bracket that I know and can plan for , and with a fast lens to allow that ( usually using it at f/2.4 even though the lens in question is in fact a f/1.8 ). So the only thing I am really trying to customize for each shoot is frame and focusing. That leave me with consistent result across the session that I can easily batch develop ( from RAW ) and so.

One of the problem I see with DxO's result and their observation is that they are not very clear as to how they test it. SO is it the camera boost ISO when set to a manual ISO setting or is it doing that on Automation ( say A mode with fixed ISO ) or is it only Auto ISO. How does it fair if its attached to a lens that do not communicate aperture ( adopted Leica M mount lens say. How about effective aperture , is the lens used actually delivering that aperture also ( many lens aperture is a bit mis informing , especially Zoom )

But as a final word, it must be you the photographer to do the judgement, Can you or will you be employing that fast aperture or better corrected ( at stopped down ) performance of a Ultra fast lens and execute an exposure likewise to be able to benefit. Its a question not to the lens, nor to the camera, but to you the user of them.

The "F-stop blues" article doesn't test any mft cameras. I remember seeing a test on those as well but I can't seem to find it. Anyway, the result wasn't the same, in fact the loss of light was much smaller.

Found it! Here: http://www.dpreview.com/forums/post/41615417

Ulric said:

Ulric wrote:

The "F-stop blues" article doesn't test any mft cameras. I remember seeing a test on those as well but I can't seem to find it. Anyway, the result wasn't the same, in fact the loss of light was much smaller.

Found it! Here: http://www.dpreview.com/forums/post/41615417

Click to expand...

Good find. I wonder why it affects M4/3s so little.

Abrak said:

Abrak wrote:

Well it was all covered in a thread about an SLR 25 T0.95 lens in this forum.

I have to say a related issue and one that is arguably more interesting, would seem to be the performance of IBIS on Oly's new E-M1. Two reviewers say that it enables them to take sharp images at over 1 second shutter speed. Now these are quite some claims - afterall if it wasnt possible you would look a bit stupid.

It seems a major improvement over the incredibly impressive E-M5 IBIS. I dont recall what claims were made about longer shutter speeds but I dont recall reviewers claiming better than 1/2 a second.

Now if we can hand hold for 1 second you really have to begin to doubt the need for fast glass to be 'fast' at all. Admittedly people also like the DOF control. But to give you and example - Oly is rumored to be bringing out a 7-14 2.8 next year but why? It is not as though 2.8 on a 7-14 will give you shallow DOF and if you can handhold for 1 second (which is mostly likely on a wide-angle) I doubt you would find need for faster than f4.

Click to expand...

Honestly you look like a shill...

highwave said:

highwave wrote:

I read this article

http://www.dxomark.com/index.php/Reviews/F-stop-blues

from DxO and just an honesty disclaymore here but I first saw it on the NEX forum

http://www.dpreview.com/forums/thread/3545800

Basically what DxO argues is that these hyper fast lenses don't really give you the light gathering advantage claimed and manufacturers gain up the ISO to give you a "virtual" ISO advantage. They suggest an F4 as being a good number to rely on.

Click to expand...

The unwanted effect is a function of photosite size. Not good news for the "NanoPixel" people.

highwave said:

I do seem to recall some voigtlander users here who showed the shutter speeds Vs. Aperture and it was as expected down to F1.4 but this is just off memory.

If anyone has personal experience in this matter, please chip in for those of us considering buying hyper fast lenses (F1.4 and larger).

Click to expand...

Interesting that the article ends with what is a clear typographical error (underlined below):

A possible conclusion of DxO Labs’ measurements is that photographers should consider with caution the maximum f-numbers advertised for lenses. Indeed, depending on the performance of their camera body and sensor, they may not effectively benefit as they expect from such wide apertures.

http://www.dxomark.com/index.php/Reviews/F-stop-blues

semifast said:

semifast wrote:

Shutter speeds that long would only be useful for things that don't move, in which case might as well use a tripod anyway.

Click to expand...

Depends on the user and the situation. Not everyone wants to bring a tripod with them. Some places don't allow them. And sometimes you don't think you'll need a tripod...and then you do.

A good 2-3 years back. At the time, I just dismissed it as the old farts being grumpy, but it turned out they might be onto something. Incidentally, I've seen it postulated elsewhere that this might be why Olympus never produced lenses faster than f2.0 for the Four Thirds mount. I'll leave that debate to Bob and Bustard.

Of course, light gathering isn't the only benefit (supposed as it may be) of a fast lens. Thinner DoF is important to many.

Acrill said:

Acrill wrote:

Define ultra?

I understand what the article is saying, but I also notice a real difference shooting indoors at 1.8 vs 2.8

Click to expand...

For m43 I would say faster than f1.4 but I'm just guestimating things from what I read here and there.

Abrak said:

Abrak wrote:

Well it was all covered in a thread about an SLR 25 T0.95 lens in this forum.

I have to say a related issue and one that is arguably more interesting, would seem to be the performance of IBIS on Oly's new E-M1. Two reviewers say that it enables them to take sharp images at over 1 second shutter speed. Now these are quite some claims - afterall if it wasnt possible you would look a bit stupid.

It seems a major improvement over the incredibly impressive E-M5 IBIS. I dont recall what claims were made about longer shutter speeds but I dont recall reviewers claiming better than 1/2 a second.

Now if we can hand hold for 1 second you really have to begin to doubt the need for fast glass to be 'fast' at all. Admittedly people also like the DOF control. But to give you and example - Oly is rumored to be bringing out a 7-14 2.8 next year but why? It is not as though 2.8 on a 7-14 will give you shallow DOF and if you can handhold for 1 second (which is mostly likely on a wide-angle) I doubt you would find need for faster than f4.

Click to expand...

Agreed,

However, when I'm shooting a scene, I must be able to get everybody and everything to freeze for 1/2 to 1 second. The babies and the birds are especially un-cooperative. I must resort to higher ISO and/or faster lenses. Really, people and things can be so selfish.

I am a mere amateur hobbyist and would call myself a "camera owner" rather than a photographer. I've been pondering this question as well since I have to decide how to spend my lens dollars.

I like taking pictures of people indoors, so I do need at least one fast lens to not only avoid flash use, but also get a shallow DoF for people shots.

However, when shooting outdoors, does one necessarily need a fast lens for low-light? The obvious answer is yes, but what about those shots where I want the whole low-light scene in focus - i.e. with deeper DoF?

I recently took some walkabout evening city shots with the 20mm F1.7 and when checking them later I found that when I wanted to use large aperture to gather as much light as possible while taking handheld shots. either the that nice lamppost in the foreground was in focus or the nice building lit up in the distance, but not both. I was asking myself how to get both and do it without dragging a tripod around.

It may seems like a silly question to all the pros here: For that type of a photography without a tripod, does one need a body with an excellent sensor to handle high ISO and a darn good IBIS/OIS mechanism instead of a fast lens?

Thank you.

I think there's also a consideration of overall system design, mount dimensions, lens corrections, and how big or small the whole thing is to carry/use in real world situations.

Zeiss is coming out with an exceptionally well corrected f/1.4 55mm. It's huge and costly, but the few samples seen to date look good.

When Kodak and Olympus collaborated on the original 4/3rds spec, Kodak had a really great knowledge base of articles online. That's gone, with what's left of the company, but I recall reading a lot about angles of incidence. The mount had a very long ratio of 1.8:1 (registration:image circle diameter) compared to typical 35mm SLRs that had about 1:1 ratios. There was huge importance on keep light rays perpendicular to the sensor surface, perhaps too much. 4/3rds had a nearly 35mm SLR deep mirror box. Approx 39mm vs approx 45mm. Lenses would have to be hugely retrofocal for anything shorter than slight portrait telephotos. It might be that the physical constraints imposed by the mount itself also limited Olympus' desire to spec lenses any faster than f/2, which would, of neccessity, also be huge, see 14-35 and 35-100. As big as 35mm f/2.8 designs of similar reach. One stop faster exposures on a sensor 1/4 the size. So, while very big and fast, still gathering about 1/2 the light.

m4/3 changes the ratio of registration to image circle diagonal to a more "full frame" SLR like 1:1

APSC DSLRs have a 1.6:1 or so ratio, because they're built into the larger mount, and they too are missing, for the most part, any wide fast primes or wide fast ratio zooms, or too many dedicated fast lenses at any focal length. Sigma's 18-35 f/1.8 is an exception, and it too is very large.

In practice, I would think that since the airy disc of a fast aperture lens is smaller, that fast apertures would be of relatively greater benefit on smaller formats* that have both more densely packed sensors, and the optical property of greater depth of field for a given aperture and equal framing (by using wider lenses for the same perspective as a larger format) ie, 25mm f/1.4 on m4/3 vs 50mm f/1.4 on full frame, both take in the same field of view, but the larger format must be stopped down further to get same depth in an identical scene.

If the rule for 35mm was "f/8 and be there", the same rule in m4/3 would be "f/4 and be there".

* fast aperture lenses should still be useful to a point provided two things, that they're well corrected, and built within a sensible mount specification.

eljay0 said:

eljay0 wrote:

It may seems like a silly question to all the pros here: For that type of a photography without a tripod, does one need a body with an excellent sensor to handle high ISO and a darn good IBIS/OIS mechanism instead of a fast lens?

Click to expand...

Depending on your field of view requirements, a wide angle lens might also help things out, since it produces a deeper DoF.

For instance, with a 12mm lens on your m43 camera, set to f/8, focused at something 6 feet away, anything from 2.39' to infinity should be within acceptable focus.

You can play around with the DoF calculations here: http://www.dofmaster.com/dofjs.html

You can use any m43 camera for the calculations, it doesn't have to be the exact model you own.

larsbc said:

larsbc wrote:

larsbc said:

eljay0 wrote:

It may seems like a silly question to all the pros here: For that type of a photography without a tripod, does one need a body with an excellent sensor to handle high ISO and a darn good IBIS/OIS mechanism instead of a fast lens?

Click to expand...

Depending on your field of view requirements, a wide angle lens might also help things out, since it produces a deeper DoF.

For instance, with a 12mm lens on your m43 camera, set to f/8, focused at something 6 feet away, anything from 2.39' to infinity should be within acceptable focus.

You can play around with the DoF calculations here: http://www.dofmaster.com/dofjs.html

You can use any m43 camera for the calculations, it doesn't have to be the exact model you own.

Click to expand...

Thanks for the reply and thanks for the link! That's very helpful. Now, I now what f-stop to use for full scene focus with my 20mm lens. I can see how focus peaking would be very useful.

highwave said:

highwave wrote:

Basically what DxO argues is that these hyper fast lenses don't really give you the light gathering advantage claimed and manufacturers gain up the ISO to give you a "virtual" ISO advantage. They suggest an F4 as being a good number to rely on.

I do seem to recall some voigtlander users here who showed the shutter speeds Vs. Aperture and it was as expected down to F1.4 but this is just off memory.

If anyone has personal experience in this matter, please chip in for those of us considering buying hyper fast lenses (F1.4 and larger).

Click to expand...

Quoting from my own post last month re. the Voigtländer 42.5/0.95 lens:

"I shot a sequence of the same composition on a tripod at ISO 200, primarily to evaluate light transmission at full-stop apertures. Here are the aperture and shutter speed values. (Note that I took the pics at around 8pm, as the sun was setting, so the conditions don't guarantee lab-level accuracy. But I did take 'em all within a one-minute span.)

• "f/0.95 -- 1/250 sec.
• f/1.4 -- 1/160
• f/2.0 -- 1/80
• f/2.8 -- 1/40
• f/4.0 -- 1/20
• f/5.6 -- 1/10
• f/8.0 -- 1/5
• f/11 -- 1/2
• f/16 -- 1 sec."

-Dave-