What's your favourite f-stop? - DPChallenge Forums

Canon EF 24-70mm f/2.8L USM

01/23/2007 11:16:22 AM · #51

Originally posted by kirbic:

The majority of what I like to shoot keeps me shooting between f/1.4 and f/4. I'm an existing-light kinda guy, I hate resorting to flash, and do so only when necessary.
I do like shooting landscapes, so occasionally do shoot at f/11 or f/16... except that night landscapes require f/4 or below, yet again!

Interestingly I read that for a camera like the 5D, you become defraction limited beyond f9 and start dropping resolution from there on out.

For something like a 20D, it happens at f7.1, where each further stop you go, you loose effective resolution.

So for a 5D at f/22 it becomes roughly equivalent in quality to a 2Mp camera.

More maths

Message edited by author 2007-01-23 11:16:50.

hanneke

Canon EOS-5D Mark II

01/23/2007 11:27:27 AM · #52

Originally posted by Gordon:

Originally posted by kirbic:
The majority of what I like to shoot keeps me shooting between f/1.4 and f/4. I'm an existing-light kinda guy, I hate resorting to flash, and do so only when necessary.
I do like shooting landscapes, so occasionally do shoot at f/11 or f/16... except that night landscapes require f/4 or below, yet again!

Interestingly I read that for a camera like the 5D, you become defraction limited beyond f9 and start dropping resolution from there on out.

For something like a 20D, it happens at f7.1, where each further stop you go, you loose effective resolution.

So for a 5D at f/22 it becomes roughly equivalent in quality to a 2Mp camera.

More maths

huh?

kyebosh

Canon EOS-20D

Canon EF 80-200mm f/2.8L

01/23/2007 11:32:18 AM · #53

2.8 I think.

Rooster

Canon EOS-400D Rebel XTi

Sigma 28-105mm f/2.8-4.0 Aspherical for Canon

01/23/2007 11:35:41 AM · #54

2.8

lowonenergy

Canon EOS-5D Mark II

01/23/2007 11:38:26 AM · #55

Depends on the situation, but I really like 1.8 alot

DjFenzl

01/23/2007 11:49:13 AM · #56

Originally posted by Gordon:

I admit I don't understand most of what that article was saying but according to it, what would be the range I would have to use to get full resolution out of my d200 or how would I figure that out?

Canon EF 17-40mm f/4.0L USM

01/23/2007 11:58:12 AM · #57

Originally posted by Megatherian:

I admit I don't understand most of what that article was saying but according to it, what would be the range I would have to use to get full resolution out of my d200 or how would I figure that out?

Somewhere around f8 or below. Anything more stopped down than that will start introducing defraction effects (although along with increased depth of field) So you'll have more depth of field, but globally less resolution. It depends a lot what you care about. Though it does mean for landscape shots, unless you really need the DoF, it makes more sense to focus at the hyperfocal point and shoot at f8 or so, rather than go to f16, for example.

Most of this really matters if you want to print large, it isn't so much of an issue for things like web display, that only really need about .3Mp

Message edited by author 2007-01-23 11:59:39.

soup

Canon EOS-30D

01/23/2007 12:02:55 PM · #58

this is pretty good advice even with a film camera. maximize the quality the optics are capable of by shooting within the lenses sweet spot - using the hyperfocal distance to maximize DOF if that is your goal.

Originally posted by Gordon:

it makes more sense to focus at the hyperfocal point and shoot at f8 or so, rather than go to f16, for example.

Message edited by author 2007-01-23 12:07:26.

Canon EF 17-40mm f/4.0L USM

01/23/2007 12:08:42 PM · #59

Originally posted by soup:

Yup. Though most interesting fact on that article was about the Canon G7, which gets defraction limited at f2 or above, but ships with an f2.8 max aperture lens, so though it claims a particular resolution due to the sensor size, it can never actually use it.

I'd been aware of the defraction limits on these cameras, I had just never realised that they were so close to the limits in terms of the sweet spots of the lens.

Message edited by author 2007-01-23 12:09:51.

soup

Canon EOS-30D

01/23/2007 12:14:55 PM · #60

hmm - that's pretty silly.

i wonder if - by design - the f:2.8 being more like f:8 for instance - weighs into the resolution loss. lessening it, in effect.

i admit i haven't read the article yet...

Originally posted by Gordon:

Nikon MF 180mm f/2.8 ED AIS

01/23/2007 12:22:43 PM · #61

Originally posted by soup:

hmm - that's pretty silly.

I suspect it is just marketing trumping engineering.

M: We need a 10Mp camera
E: But it'll only ever shoot as good as an 8Mp camera
M: Who cares, we need to have a bigger number!

mcmurma

Canon EOS-300D Rebel

01/23/2007 01:17:52 PM · #62

I dunno about the math, I'm not good with figures... but I have noticed that diffraction limitations are worse with some lenses than others. I believe it has to do with what the lens was optimally designed for. Some lenses (portrait lenses like the 85mm) are designed to be sharp wide open, and these tend to get diffraction limited quickly.

For example, with the 300D and an old Tamron 90mm f2.5 (an excellent macro/portrait lens) I can see diffraction starting to become an issue as soon as f5.6. It was the first lens that really made me start looking for diffraction limits, and the reason I never stop down below f11 any longer with any lens.

DjFenzl

01/23/2007 02:16:01 PM · #63

Originally posted by Gordon:

Originally posted by Megatherian:

I admit I don't understand most of what that article was saying but according to it, what would be the range I would have to use to get full resolution out of my d200 or how would I figure that out?

Somewhere around f8 or below. Anything more stopped down than that will start introducing defraction effects (although along with increased depth of field) So you'll have more depth of field, but globally less resolution. It depends a lot what you care about. Though it does mean for landscape shots, unless you really need the DoF, it makes more sense to focus at the hyperfocal point and shoot at f8 or so, rather than go to f16, for example.

Most of this really matters if you want to print large, it isn't so much of an issue for things like web display, that only really need about .3Mp

Just wanted to add that I found an interesting page about this that has a calculator to help figure out what print sizes you can get out of different apertures and resolutions based on typical sensor dimensions.
Diffraction Calculator

Spork99

OM System OM-1

Canon EF 100mm f/2.8 USM Macro

01/23/2007 02:41:56 PM · #64

Originally posted by Gordon:

I read the article and I can't find the numbers you're referring to, I did find the table that indicated the smallest aperture that would not degrade the image by diffraction. For a FF sensor, N=21, for an APS-C sensor, N= 14.

Originally posted by Luminous Landscape:

"The important point is that the value of N computed for a given sensor size is the largest value that can be used without degrading the image"

When I read that, it tells me that for a 5D, I lose image quality stopping down smaller than f22 (closest available aperture to N = 21) and for a 10D, I lose image quality stopping down smaller than f14.

Am I missing something?

jmsetzler

Sony Alpha a6000

Sony 16-50mm f/3.5-5.6 E-mount

01/23/2007 03:07:36 PM · #65

This is called measurebating.

violinist123

Canon EOS-6D

Canon EF 24-70mm f/2.8L USM

01/23/2007 03:15:25 PM · #66

From TFA:

Camera Max f/stop Max practical f/stop

Canon 5D f/8.6 f/9
Canon 1Ds f/9.3 f/9
Canon 1Ds Mark II f/7.6 f/8
Nikon D2X f/5.8 f/5.6
Canon 20D f/6.76 f/7.1
Canon G7 f/2.06 f/2

Spork99

OM System OM-1

Canon EF 100mm f/2.8 USM Macro

01/23/2007 03:29:28 PM · #67

** Warning: This post has been hidden as it may content mature content. Click here to show the post.

Canon EF 17-40mm f/4.0L USM

01/23/2007 03:49:01 PM · #68

Originally posted by jmsetzler:

This is called measurebating.

When it becomes a 7x reduction in effective resolution, its a wee bit more important than just talking about the numbers for the sake of it, particularly if you ever want to print larger than say an 8x10

Message edited by author 2007-01-23 15:49:41.

soup

Canon EOS-30D

01/23/2007 03:52:30 PM · #69

so i guess i answered the question - almost correct ;}

kirbic

Canon EOS R5

Canon RF 24-70mm f/2.8 L IS

01/23/2007 05:19:31 PM · #70

Originally posted by Gordon:

Originally posted by jmsetzler:
This is called measurebating.

When it becomes a 7x reduction in effective resolution, its a wee bit more important than just talking about the numbers for the sake of it, particularly if you ever want to print larger than say an 8x10

Gordon,
From direct experience with a 5D and top-quality glass, I must say I don't know where the "2Mpx" equivalence @ f/22 comes from. I can say that you will not see *any* detectable image degradation below f/16, and at f/16 it is slightly detectable with the sharpest glass. At f/22, it is readily observable, but equating the result to a 2Mpx camera is ludicrous.
These "experiential" results agree very closely with your luminous landscape link, with one caveat: the CoC for the 5D must be taken to be close to the theoretically ideal (Nyquist) value of 2*px pitch = approx. 16Âµm, not the 29Âµm given for 35mm sensors. If the larger CoC is used, the usable f/number actually rises to f/21. Per my observations, this does not square with reality.