| Author | Thread |
|
|
03/01/2006 12:10:28 AM · #1 |
Here is one to ponder...
Does a good lens pass more light at a given aperture than a cheap lens?
In other words, you have 2 identical camera pointed at the same subject. One has a $50 lens, the other a $1000 lens. Both are at f5.6, same ISO.
Do their internal meters read the same, and therefore show the same shutter speed?
Does a lens with fewer elements pass more light?
|
|
|
|
03/01/2006 12:13:30 AM · #2 |
| That's a really interesting question. I was just wondering that as well, as I've been researching different qualities of filters... and the more expensive, MC/SMC ones claim to allow more light into the lens... hmmm... |
|
|
|
03/01/2006 12:14:36 AM · #3 |
If both lenses are in proper working order, they will give you the same meter readings.
|
|
|
|
03/01/2006 12:21:44 AM · #4 |
Originally posted by jmsetzler:
If both lenses are in proper working order, they will give you the same meter readings. |
Then the 'value' of L glass just dropped. LOL
What you're saying is, the quality of the lens elements make no difference - so the number of them, the thickness of them, be they bottle glass or lead crystal, be ther 1/4" of total lens material or 3" of it, regardless of coatings, a given amount of light at the big end of any lens at F4 will give the exact same meter reading at the other end?
So then a 12mm at f4 and a 500mm at f4, a mirror lens, a macro lens, a 15 element zoom lens (assuming you could point them at the exact same amount of illumination) would all yield idential meter readings?
|
|
|
|
03/01/2006 12:26:01 AM · #5 |
Originally posted by Prof_Fate:
Originally posted by jmsetzler:
If both lenses are in proper working order, they will give you the same meter readings. |
Then the 'value' of L glass just dropped. LOL
What you're saying is, the quality of the lens elements make no difference - so the number of them, the thickness of them, be they bottle glass or lead crystal, be ther 1/4" of total lens material or 3" of it, regardless of coatings, a given amount of light at the big end of any lens at F4 will give the exact same meter reading at the other end?
So then a 12mm at f4 and a 500mm at f4, a mirror lens, a macro lens, a 15 element zoom lens (assuming you could point them at the exact same amount of illumination) would all yield idential meter readings? |
The amount of light that passes through is not the purpose of the various arrays of elements and they glass they are made of. The quality of the light, however, is another story...
|
|
|
|
03/01/2006 12:28:23 AM · #6 |
Originally posted by Prof_Fate:
So then a 12mm at f4 and a 500mm at f4, a mirror lens, a macro lens, a 15 element zoom lens (assuming you could point them at the exact same amount of illumination) would all yield idential meter readings? |
The answer would be yes in the rare instance that both of those lenses would 'see' the same thing. if you pointed both of them at a white object that filled the frame, then yes, you would get the same meter readings.
|
|
|
|
03/01/2006 12:31:03 AM · #7 |
| While the amount of light is the same, the sharpness and contrast you get from a good lens will be much better than from cheaper quality glass. |
|
|
|
03/01/2006 12:33:23 AM · #8 |
Depends on the lens. Some will be slightly off on exposure. That's why serious photogs doing the zone system tested each lens individually to get the exposure compensation down for the lens for the most accurate exposure.
Crazy, I know...
|
|
|
|
03/01/2006 03:10:39 AM · #9 |
there is a BIG difference..
lets say you have a 55-200 sigma at f6.3 @ 200mm and you have the 70-200L f2.8 set at f6.3 @ 200mm
same settings on both lenses but the 70-200 will still pass mych more light just up to the time you take the picture.
the aperture blades only closes when you press the shutter so you have maximum aperture to focus on your subject, and all lenses focus much faster and better @f2.8 than they do @ f6.3, usually thatis the turningpoint when you have to go manual focus.
so there is a big difference in using the $50 lens with max aperture f6.3 or the $1000 lens with max aperture f2.8 stopped down to f6.3, you still have f2.8 while looking through the lens :)
|
|
|
|
03/01/2006 03:19:49 AM · #10 |
I think you are confusing people Dan.
I do agree with you on some points, or well nothing you say is false, except that price has nothing to do with it.
The $100 2.8 50mm canon is going to look a lot brighter through the viewfinder than the $1000 4.0 Canon 70-200mm .. How bright the lens is while looking through it, has only to do with the aperature.. which IS what you are saying, but price has nothing to do with it.
HOWEVER.. what the lens meters, as has been said, is the same no matter what type, style, price the lens is.
|
|
|
|
03/01/2006 06:08:35 AM · #11 |
Hi-ho,
Yes. And No.
Different lenses do pass different ammounts of light when stopped down to the same apeture.
But.
The difference in levels is smaller than the graduations on most camera light meters. Certainly a 1/3rd stop light meter will not pick it up, but you could by experiment if you were bored.. Oh, hang on, I'm bored.. Please hold....
|
|
|
|
03/01/2006 06:51:37 AM · #12 |
OK..
My 'Method':
Take three lenses.
Canon EF 24mm 1:2.8
- prime 10 elements, 10 groups 6 bladed apeture.
Canon EF 17-40 1:4 L USM (set to 24mm, reported in EXIF as 24mm)
- Zoom 12 elements, 9 groups 7 bladed apeture.
Canon EF 24-70 1:2.8 L USM (set to 24mm)
- Zoom 16 Elements, 13 groups, 8 bladed apeture.
Stick the 20D on a tripod, about 6inches from a white bit of A3 paper, backlit with an incandescent bulb that has been on for 2+ hours (not warming up, so no light level change).
Settings: ISO200, Manual focus at infinity, manual exposure F/8.0, 1/50th of a second. Incandescent white balance. JPEG fine, 8.2Mp.
Take one photo with each lens, changing nothing but the lens in the setup.
The light meter reports correct exposure for all shots (set to 1/3rd stop steps), camera histogram is very slightly different, but not apparently 'offset' either way.
Take the three JPEG's and on screen they look identical, sort of a muddy yellow (manual incandescent wb is way off for this bulb!) with a slight gradient towards darker on the right side.
Gaussian blur the pics with a 100 pixel radius to 'average' out any noise/texture issues in the paper target.
The result:
The 24mm prime gives a center pixel RGB value of 201,143,76
The 17-40 gives 186,131,67
The 24-70 gives 184,128,65
(Center pixel being 1752,1168 in the jpeg)
Difference is lighter in favor of the Prime by 15,12,9 (R,G,B) over the 17-40, and slightly more (17,15,11) for the 24 prime to 24-70 difference.
And what does this tell me? The 24mm prime gives a slightly brighter image overall than the zooms, and the prime is slightly 'warmer' (more red/green) than the zooms as well. The two zooms are closer to each other, presumably as they are engineered to fit the 'L' colour profile.
How much brighter? Well, the exact dynamic range of dSLR's is a bit of a gray area, but if we take the dpreview test of the 5D as an indicator it's about 8 stops. The shift is (average of RGB diff for the prime to 17-40) 12 gradations which reperasents 4.6% of the range or about 0.375 stops. This assuming the camera is a linear device, which it isn't, but I don't have the faintest idea what the transfer curve (light level to rgb value) looks like.
Anyway... What does this all mean?
Yes, different lenses have different light transmission qualities, both in terms of overall average light level and colour balance...
Why else would people comment that a particular lens is 'warmer' than another, or that a good 'L' Lens has better colours than a kit lens? I imagine you could plot a light spectrum vs transmission graph for each lens and find they were like finger prints of that design...
It stands to reason that if you've got a bunch of glass bits in one lens with SLD coating and florite elements you'll get different light transmission levels than you would get from standard glass with simple anti glare coatings. These coatings are metals and organic substances which are not clear by nature.
The other variable (Which is why I mentioned it) is apeture shape. The apeture for a given setting in a given lens may not give you exactly F/8 etc. This is due to mechanical tollerances, differenes in blade profiles, and wear on the edges of the blades.
My parting statement to those who say 'no' is that any bit of flat glass, no matter how perfectly manufactured will still cast a visible shadow on a sunny day if you can see the edge of it's shadow compared to full sunlight... Why on earth should camera lenses be any different?
Hmmmm, I should get out more.
As always, my 2c worth, this is entirely non-scientfic, but a simple experiment anyone could try...
|
|
|
|
03/01/2006 06:57:38 AM · #13 |
Originally posted by KiwiChris:
As always, my 2c worth, this is entirely non-scientfic, but a simple experiment anyone could try... |
Anyone with time on their hands and a nice lens collection, that is...
Interesting work. I liked the rigourous set-up.
|
|
|
|
03/01/2006 07:16:22 AM · #14 |
Originally posted by legalbeagle:
I liked the rigourous set-up. |
:-). ROFL
I taped a bit of paper to some table legs, in front of my desk lamp!
edit: You could try it as well, you've got four overlapping lenses....
Message edited by author 2006-03-01 07:17:48.
|
|
|
|
03/01/2006 07:34:45 AM · #15 |
They will be different.
One way to prove it is to add another element in front of the lens for testing.
Make one a cheap single coated UV filter and the other a better multi-coated UV filter.
The cheaper filter with the single coating will reflect and/or refract X amount of light rather than let it pass through to the sensor.
Infact don't do the test....just read the results here.
Furthermore you can read a little intro on the role of lens coatings here.
cheers,
bazz. |
|
|
|
03/01/2006 07:43:41 AM · #16 |
Lenses for video and movie applications commonly had (still have?) T-values printed right on the lens. These are light transmission values, through which one can calculate the exposure compensation necessary to offset the light loss that happens in the lens.
Stil camera lenses are no different, and the more elements you have, the more light loss. As posted previously, for most lenses, this difference will be well under a half-stop.
|
|
|
|
03/01/2006 07:49:44 AM · #17 |
The bigger differentiator is the sharpness and quality of the image they throw onto the film or sensor. The total amount of light recorded is surely secondary to that.
|
|
|
|
03/01/2006 08:18:15 AM · #18 |
Originally posted by sir_bazz:
They will be different.
One way to prove it is to add another element in front of the lens for testing.
Make one a cheap single coated UV filter and the other a better multi-coated UV filter.
The cheaper filter with the single coating will reflect and/or refract X amount of light rather than let it pass through to the sensor.
Infact don't do the test....just read the results here.
Furthermore you can read a little intro on the role of lens coatings here.
cheers,
bazz. |
A more extreeme test would be an uncoated UV filter vs a multicoated one. The uncoated filter will reflect more light, so the multicoated one will have a higher transmission index.
One site claims you lose 4% of the light in an uncoated lens. Others have given higher numbers.
Does anyone know if coatings used on lenses intended for digital are not as neutrally colored as coatings on lenses for film? Do the lens manufacturers assume that WB settings on a digital camera can compensate for the color cast? |
|
|
|
03/01/2006 09:43:07 AM · #19 |
Originally posted by KiwiChris:
OK..
My 'Method':
Take three lenses.
Canon EF 24mm 1:2.8
- prime 10 elements, 10 groups 6 bladed apeture.
Canon EF 17-40 1:4 L USM (set to 24mm, reported in EXIF as 24mm)
- Zoom 12 elements, 9 groups 7 bladed apeture.
Canon EF 24-70 1:2.8 L USM (set to 24mm)
- Zoom 16 Elements, 13 groups, 8 bladed apeture.
Stick the 20D on a tripod, about 6inches from a white bit of A3 paper, backlit with an incandescent bulb that has been on for 2+ hours (not warming up, so no light level change).
Settings: ISO200, Manual focus at infinity, manual exposure F/8.0, 1/50th of a second. Incandescent white balance. JPEG fine, 8.2Mp.
Take one photo with each lens, changing nothing but the lens in the setup.
The light meter reports correct exposure for all shots (set to 1/3rd stop steps), camera histogram is very slightly different, but not apparently 'offset' either way.
Take the three JPEG's and on screen they look identical, sort of a muddy yellow (manual incandescent wb is way off for this bulb!) with a slight gradient towards darker on the right side.
Gaussian blur the pics with a 100 pixel radius to 'average' out any noise/texture issues in the paper target.
The result:
The 24mm prime gives a center pixel RGB value of 201,143,76
The 17-40 gives 186,131,67
The 24-70 gives 184,128,65
(Center pixel being 1752,1168 in the jpeg)
Difference is lighter in favor of the Prime by 15,12,9 (R,G,B) over the 17-40, and slightly more (17,15,11) for the 24 prime to 24-70 difference.
And what does this tell me? The 24mm prime gives a slightly brighter image overall than the zooms, and the prime is slightly 'warmer' (more red/green) than the zooms as well. The two zooms are closer to each other, presumably as they are engineered to fit the 'L' colour profile.
How much brighter? Well, the exact dynamic range of dSLR's is a bit of a gray area, but if we take the dpreview test of the 5D as an indicator it's about 8 stops. The shift is (average of RGB diff for the prime to 17-40) 12 gradations which reperasents 4.6% of the range or about 0.375 stops. This assuming the camera is a linear device, which it isn't, but I don't have the faintest idea what the transfer curve (light level to rgb value) looks like.
Anyway... What does this all mean?
Yes, different lenses have different light transmission qualities, both in terms of overall average light level and colour balance...
Why else would people comment that a particular lens is 'warmer' than another, or that a good 'L' Lens has better colours than a kit lens? I imagine you could plot a light spectrum vs transmission graph for each lens and find they were like finger prints of that design...
It stands to reason that if you've got a bunch of glass bits in one lens with SLD coating and florite elements you'll get different light transmission levels than you would get from standard glass with simple anti glare coatings. These coatings are metals and organic substances which are not clear by nature.
The other variable (Which is why I mentioned it) is apeture shape. The apeture for a given setting in a given lens may not give you exactly F/8 etc. This is due to mechanical tollerances, differenes in blade profiles, and wear on the edges of the blades.
My parting statement to those who say 'no' is that any bit of flat glass, no matter how perfectly manufactured will still cast a visible shadow on a sunny day if you can see the edge of it's shadow compared to full sunlight... Why on earth should camera lenses be any different?
Hmmmm, I should get out more.
As always, my 2c worth, this is entirely non-scientfic, but a simple experiment anyone could try... |
Dang, you must be really bored...
|
|
|
|
03/01/2006 11:09:55 PM · #20 |
Originally posted by Spazmo99:
Originally posted by KiwiChris:
OK..
My 'Method':
|
Dang, you must be really bored... |
Ya, slow night...
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 09/23/2025 08:16:32 AM EDT.
