Noct lenses and ultra fast lenses - meh

MunchmaQuchi

I don't get too excited about fast lenses (< 1.4) and Nocts and here's why.

1. They don't resolve very well wide open - even the best ones.
2. Modern digital ISO can address low light situations better than a fast lens
3. Size, weight, and cost (and this is from someone that owns all 3 Otii)
4. Depth of field nonsense - play around with DOF calculator and you'll see - studio and field work will even drive this point home further. Lastly, razor thin DOF is unnatural to the depth we typically see.
5. Poor transitions into OOF.
6. LOCA and SLOCA when wide open and typically for the next 3-4 full stops.

Each one of these is reason enough but in reality, all of them occur at the same time when trying to use an ultra fast lens wide open or within a few stops of WO.

There is a joy in holding a Zeiss 1.2 or Leica 0.95 in your hand and looking through the barrel of the lens but all of that ends with the frustration of trying to get any decent photos out of these lenses. At the end of the day, it's just for bragging rights. Most sports cars capable of going 200mph have average speeds of 25mph over the life of the car. Just like a Prius.

NSXTypeR

MunchmaQuchi said:

Ist post only because I lost my old ID. Nikon user (and famous at that) - I wrote the Sony A7r vs Nikon 810 blog that had over 8,000 readers. In that blog I documented over and over how the D810 was superior to the A7r for total IQ.

Now equipped with a D850, a good selection of Nikon, Sigma and Zeiss lenses including a full Otus set, I find myself less than enthused with a Nikon mirrorless and refuse to buy a Sony because of unethical business practices.

IMHO - Nikon doesn't need a mirrorless camera - what they need is the Highest Quality EVF in photography. I am tired of using a clunky 3x loupe to get critical focus out of my Otus lenses and the D810/D850. I'd easily pay $1000 for a well designed EVF than I would pay any money for a whole new infrastructure of camera gear.

I still have all of these Leica lenses, Contax Zeiss, some M43, the f-mounts.... enough. I about ready to trade in all in for that little Lumix with the tilting EVF.

Nikon should think more like Porsche instead of follow the leader. Make enhancements every year to the 911 and eventually arrive at the line up they have today.

In another thread you were just exalting the fact that you were the proud owner multiple expensive manual focus wide aperture lenses.

So what's your point for starting this thread?

ggbutcher

> 2. Modern digital ISO can address low light situations better than a fast lens

I don't buy into this one. And I'm a software guy.

Both here and at work, I see all sorts of bravado about algorithimically making the best of sub-optimal measurements of the physical phenomena. And in my experience it never (and I choose that word carefully) has adequately compensated for not putting the same money into whatever it would take to get a better measurement from the sensor. And, regarding imaging, a stop of light collection can be significant in pulling those low measurements out of the noise regime of the sensor.

Yes, if you're in a studio or outside in the daylight, maybe not worth the weight or fiddling. But, if you're in a concert venue, or trying to tease an image the Andromeda galaxy out of the dark sky, those lenses will be like gold...

MunchmaQuchi

Nikon Noct.

I don't think of 1.4 as ultrafast but as standard lenses. The Nikon Noct will more than likely be a waste of money and will do more things wrong than it will do right.

spraynpray

Well, I figure Nikon must have a pretty good business case for making it, so let's see.

FWIW I have seen nightscape/starscape photographs from Zeiss lenses that have been no better than those from a cheapo Rokinon. I make that point to illustrate that I am not impressed by the price of gear, only by the performance. The company that brings out an ultra low aberration ultra-wide angle fast lens will get my money. Sigma are the closest, but they are not sufficiently better than my Irix to warrant the cost uplift (4 times). The Nikon lens will be too long for my use.

@ggbutcher : Agreed. Algorithms don't make photons, they only plaster over the cracks that not having enough photons give.

MunchmaQuchi

ggbutcher said:

> 2. Modern digital ISO can address low light situations better than a fast lens

I don't buy into this one. And I'm a software guy.

Both here and at work, I see all sorts of bravado about algorithimically making the best of sub-optimal measurements of the physical phenomena. And in my experience it never (and I choose that word carefully) has adequately compensated for not putting the same money into whatever it would take to get a better measurement from the sensor. And, regarding imaging, a stop of light collection can be significant in pulling those low measurements out of the noise regime of the sensor.

Yes, if you're in a studio or outside in the daylight, maybe not worth the weight or fiddling. But, if you're in a concert venue, or trying to tease an image the Andromeda galaxy out of the dark sky, those lenses will be like gold...

# 1 - these lenses never really provide a full stop of light in T-stops but just fractional and calculated and not nearly as easy as other methods that do (ISO, n-axis VR etc) that provide multiple stops of latitude.

# 2 - something like a Sigma 14mm 1.8 ART lens may offer some improvement over Zeiss's and Nikon 2.8 offerings but this is about the difference between f1.0 and 1.2 when compared to a 1.4.

In those situations you offer (concert venue or dark sky) there are far more robust and logical choices to be made other than just shoot an ultrafast wide open that will result in higher keeper percentages and greater IQ overall.

WestEndFoto

MunchmaQuchi said:

In those situations you offer (concert venue or dark sky) there are far more robust and logical choices to be made other than just shoot an ultrafast wide open that will result in higher keeper percentages and greater IQ overall.

Intriguing. Tell us more.

CraigR

2. Modern digital ISO can address low light situations better than a fast lens

Nonsense.

spraynpray

CraigR said:

2. Modern digital ISO can address low light situations better than a fast lens

Nonsense.

Well, to be fair, if you don't capture the photons, you can't invent them by signal amplification later so a fast lens with low T stops is fundamental. I mentioned the T stops because so many fast lenses absorb too much light so aren't that good for low light situations.

BVS

MunchmaQuchi said:

I don't get too excited about fast lenses (< 1.4) and Nocts and here's why.

1. They don't resolve very well wide open - even the best ones.
2. Modern digital ISO can address low light situations better than a fast lens
3. Size, weight, and cost (and this is from someone that owns all 3 Otii)
4. Depth of field nonsense - play around with DOF calculator and you'll see - studio and field work will even drive this point home further. Lastly, razor thin DOF is unnatural to the depth we typically see.
5. Poor transitions into OOF.
6. LOCA and SLOCA when wide open and typically for the next 3-4 full stops.

Each one of these is reason enough but in reality, all of them occur at the same time when trying to use an ultra fast lens wide open or within a few stops of WO.

I wonder how the old Noct Nikkor held up concerning these points? How was the WO resolution, aberrations, and OOF transitions?

moreorless

MunchmaQuchi said:

4. Depth of field nonsense - play around with DOF calculator and you'll see - studio and field work will even drive this point home further. Lastly, razor thin DOF is unnatural to the depth we typically see.

That's making a pretty sweeping claim isn't it? that any kind of photography that doesn't represent human vision has no worth to it. I mean your F/1.4 lens already have potential for far thinner DOF than human vision.

Really this is going to be the big selling point of any ultrafast lens I would say just as it is for say Leica's 50mm 0.95. Yes you can obviously get thinner DOF by focusing closer but some people will want the ability to get the thinnest DOF possible with a certain framing.

Performance and price wise its really wait and see I'd say although even then I would argue many people probably aren't going to demand extreme resolution wide open.

HankB

I can't help wondering if Nikon isn't just exuberantly rejoicing that after several centuries — or maybe just half a century — with a new, wider mount, they will finally be as unconstrained as Canon when engineering lenses. Releasing an f/0.95 could be part of that celebration.

And the mirrorless technology will make exact, precise, perfect focus a reasonable expectation.

MunchmaQuchi

spraynpray said:

CraigR said:

2. Modern digital ISO can address low light situations better than a fast lens

Nonsense.

Well, to be fair, if you don't capture the photons, you can't invent them by signal amplification later so a fast lens with low T stops is fundamental. I mentioned the T stops because so many fast lenses absorb too much light so aren't that good for low light situations.

Capturing photons caused by aberrations doesn't positively impact overall IQ.

WestEndFoto

You should shoot the new 105. You might change your tune.

MegapixelSchnitzel

If I am to believe that amplifying a sensor signal is better than having glass with an extra stop of light-gathering capacity, then you can probably sell me on believing that digital zoom, or cropping down, is better than having the actual telephoto lens to do the job.

I don't think so.

spraynpray

MunchmaQuchi said:

spraynpray said:

CraigR said:

2. Modern digital ISO can address low light situations better than a fast lens

Nonsense.

Well, to be fair, if you don't capture the photons, you can't invent them by signal amplification later so a fast lens with low T stops is fundamental. I mentioned the T stops because so many fast lenses absorb too much light so aren't that good for low light situations.

Capturing photons caused by aberrations doesn't positively impact overall IQ.

True, but we have no idea of the actual performance of this Noct lens yet. Also, you may be a daylight shooter so your idea of aberrations may be completely different to mine as I shot a lot at night and have yet to find any aberration-free lens at any price.

MunchmaQuchi

I am speaking from experience and using the best lenses available. #1 - I try to stay on base ISO no matter what. But if I have to use a higher ISO because of low light or motion, I am convinced that higher overall IQ is achieved through a compromise of slightly stopped down (Otus lenses 1-2 stops) and noise reduction in post. Fast lenses that are wide open bring in a number of IQ deficiencies that can not be fixed in post or as readily as compromising the capture.

pictureted

MunchmaQuchi said:

spraynpray said:

CraigR said:

2. Modern digital ISO can address low light situations better than a fast lens

Nonsense.

Well, to be fair, if you don't capture the photons, you can't invent them by signal amplification later so a fast lens with low T stops is fundamental. I mentioned the T stops because so many fast lenses absorb too much light so aren't that good for low light situations.

Capturing photons caused by aberrations doesn't positively impact overall IQ.

but, it's easier to make software corrections for known lens aberrations than lose those photons, never to be available.

Lens speed and DOF are just tools to use. Like words for a writer and I always prefer a larger vocabulary and larger tool set.

spraynpray

MunchmaQuchi said:

I am speaking from experience and using the best lenses available. #1 - I try to stay on base ISO no matter what. But if I have to use a higher ISO because of low light or motion, I am convinced that higher overall IQ is achieved through a compromise of slightly stopped down (Otus lenses 1-2 stops) and noise reduction in post. Fast lenses that are wide open bring in a number of IQ deficiencies that can not be fixed in post or as readily as compromising the capture.

Stopping down when I am desperate for photons kind of rubs me badly, but yes, I agree the aberrations I am talking about are not fixable in post other than by cropping them off! Zeiss are no better than a lot of lenses IME>

Symphotic

On the other hand, being conventional and not pushing the envelope is not what many of us here want from our camera suppliers. The D800 pushed the envelope, but had its problems at the start. But if there were no D800, there would be no D850. Anybody that doesn't think the D850 is one of the best digital cameras ever made isn't thinking. I took a gamble on the D800 when it first came out and purchased two for a client. That gamble paid off quite well.
The f/0.95 is a technical challenge, as was 36 MP for the D800. But I support it, and will likely buy one. I will be trying to figure a way to get it underwater on my video cameras, as we are light starved and an extra stop means a lot. Fortunately, Nikon has made it easy with the FTZ adapter: in order for the adapter to work, the Z and F communications format has to be the same. No new engineering.
We also have two other devices that will benefit from a f/.95 lens. We have experimented with the Noct f/1.2 and it works OK, but of course it is no longer available.
Besides that, I can't wait to put the new Noct on my Z7 and take pictures of my cat. ;-)
So count me in.

WestEndFoto

Symphotic said:

On the other hand, being conventional and not pushing the envelope is not what many of us here want from our camera suppliers. The D800 pushed the envelope, but had its problems at the start. But if there were no D800, there would be no D850. Anybody that doesn't think the D850 is one of the best digital cameras ever made isn't thinking. I took a gamble on the D800 when it first came out and purchased two for a client. That gamble paid off quite well.
The f/0.95 is a technical challenge, as was 36 MP for the D800. But I support it, and will likely buy one. I will be trying to figure a way to get it underwater on my video cameras, as we are light starved and an extra stop means a lot. Fortunately, Nikon has made it easy with the FTZ adapter: in order for the adapter to work, the Z and F communications format has to be the same. No new engineering.
We also have two other devices that will benefit from a f/.95 lens. We have experimented with the Noct f/1.2 and it works OK, but of course it is no longer available.
Besides that, I can't wait to put the new Noct on my Z7 and take pictures of my cat. ;-)
So count me in.

One little detail about the 0.95 will sabotage that plan Symphotic. The lens is manual focus.

That said, you point about the D800 and D850 is well said and matches my experience.

And I suspect that this noct will make it into my bag at some point.

Symphotic

WestEndFoto said:

...
One little detail about the 0.95 will sabotage that plan Symphotic. The lens is manual focus.

That said, you point about the D800 and D850 is well said and matches my experience.

And I suspect that this noct will make it into my bag at some point.

Manual focus should be OK for me. I went until 2012 without an autofocus lens except on my point and shoots. I only have one Noct now, and it is manual focus.
Under water we use motor drives that connect to a hand controller on the surface. Autofocus isn't designed for use at depth: although some cameras have it, they tend to focus on bubbles and not on the target of interest. For my other two applications we need manual focus that is done by peaking a signal on an oscilloscope.
Sometime let's get together for a beer. You are in the UK?

WestEndFoto

Ah! I see. I just learned something new.

MunchmaQuchi

The Nikon Noct may be $6000 + and manual focusing in low light with a wide open ultra-fast has its challenges. Read about the compromises and discipline when Kubrick used the 0.7 for Barry Lyndon.

IMHO moving pictures are far more forgiving than stills.

http://neiloseman.com/barry-lyndon-the-full-story-of-the-famous-f0-7-lenses/

MunchmaQuchi

Even more -