My Canon 24-105L has 3x as many elements as my Ultron 40 f/2, 18 vs 6.
I'm gonna test again.
The OP's zoom has three times as many elements as his prime - 21 vs 7.
speters wrote:
It depends on the lens in particular, but as a rule, primes are often/mostly faster than zooms - just in part of how they are designed/constructed. But as I said, it depends on the lenses in particular. I have a zoom for example, that is pretty much the sharpest lens that I own ( and I do own a lot of primes), its an Angenieux 15x zoom lens. It is /1.2 throughout, so its very fast, I can shoot at candle light and get nice bright shots. It handles low light better (and with better results) than any of the "faster" primes that I do own. So in general, primes do let in more light, but there are exceptions!
It depends on the lens in particular, but as a rul... (
show quote)
Sounds like a movie camera lens. Is it marked in f/stops or T-stops?
I could not find a 15x f/1.2 but I found a 12x f/2.5, it's Transmission # is 2.8, only a 20% difference, so not a lot of improvement to be had with a "perfect" lens of any type.
T-stop numbers for the same lens are always higher. How much depends on how much the lens absorbs and scatters light. More elements and air to glass surfaces, absorb and scatter more light, so your 15x Angenieux, if marked f/1.2, will not transmit as much light as a high quality, prime lens with many fewer elements, assuming similar coatings and glass type. Absorbing only 20% is quite an accomplishment, for such a busy lens.
Could be. I tested using the Voightlander Ultron 40 f/2 as prime because it has relatively few elements, compared to the 24-105L.
dsmeltz wrote:
I do not use DxOMark for much since I am more interested in actual field performance, but T-stops are one thing I look at there when comparing lenses. However, I do not remember seeing that large a variation on a lens.
They show the 300 @ f2.8 with T=3.2 and the 100-400 at 200mm f5 with T=5.6.
When the OP ran the test, what f stop was used on each lens? He mentions both were set to the same length but did not mention the f.
I too am a little surprised at the 1-2 stop difference reported. Perhaps he meant one or two clicks on the adjustment wheel (1/2 or 1/3 f/stop per click).
Thanks Greg, Glad to confirm your findings with the tech explanation.
I have 60 year's experience with f/stops, but only realized this about a year ago.
OP was wondering why his zoom required more exposure than his prime at the same f/stop. f/stop vs T-stop for same lens construction (google Cine Lens) shows this. Different lenses at the same f/stop deliver different amounts of light to the camera. In-camera metering automatically compensates for this effect, so it does not become apparent to most photographers.
f/, F/, F: all refer to the focal length divided by the diameter of the aperture. They are dimensional measurements.
jerryc41 wrote:
Wouldn't that depend on the F/ specifications of each lens? Primes are usually faster than zooms. Pay the big bucks, and you can get an F/2.8 zoom, but most are F/3.4 - F/6.3, or something like that. Zooming in will decrease the aperture.
If the lens markings are transmitted light values they are marked T, if they are labeled f/xx, or F:xx, they are not transmitted light values.
wdross wrote:
I understand what you are indicating. But based off my meter readings and getting identical exposures from either my primes or zooms off those those reading, Olympus is putting on their lenses the actual transmitted light as the f-stop. Otherwise one of the lenses would deviate from the meter reading and give an incorrect exposure.
wdross-
I think you are being fooled by in-camera metering.
Set everything to manual, run a test shooting a graycard with a complex zoom and a simple prime, bracket 1/2 stop intervals for 2 stops and check which images match.
wdross -
No, do some research on T-stops.
Repeat your experiment, and read the RGB values in Photoshop.
f/numbers do not take into account the light that is absorbed or scattered by each element of the lens.
So zoom lenses with lots of elements absorb more light than primes with fewer.
T-stops are measurements of the light transmission of the lens, f/stops are derived from dimensions and do not reflect the actual amount of light transmitted.
I believe lens makers use f/numbers because they are needed for depth of field calculations. Cine lenses are marked with T-stops, for precise exposure settings using external metering.
Wrong wdross,
T-stop lenses transmit the same amount of light at the same setting, but f/stop lenses do not.
I confirmed this myself shooting a Canon 24-105L at 40mm f/4, and a Voightlander Ultron 40mm f/2 at f/4.
There was a little more than a stop difference. I had to set the prime to f/5.6+ to match the zoom .
Yes. f/stops are a dimensional calculation. T-stops are accurate light transmission values.
To achieve "perspective correct", print size and viewing distance should be adjusted to match the lens' angle of view (allowing for crop).