T-stops, shutter speed, video (originally posted in Panasonic is Developing a New, Smaller 35-100mm

Note from Amin: These posts were originally posted in this thread and moved here to their own thread.


The speed of the lens is dictated by the circumference of the aperture opening based on the f-stop formula. The purpose of the f number is to show the size of the opening of whole of the lens so to speak to allow light to expose the sensor to form an image. It is this opening that dictates how much light passes through. Which is why f/stop is so important in photography. But there seemed to be a drive to go with T-stops even with professional photographers and that's where the confusion comes in. Somehow you can bend physics by just magically raise or lower T-stops, but reality only dictates that the physical opening (the size of hole of the lens and the length of the lens) dictates the amount of light that hits the sensor.

For example. Let's take a close look at 2 lenses. 35mm @ f/2.8 and 200mm @ f/2.8. The 35mm @ f/2.8 has an aperture opening of 12.5mm (35mm divided by 2.8), whereas the 200mm @ f/2.8 has an aperture opening of 71.42mm. But then you wonder, how the heck you say 200mm @ f/2.8 allows the same amount of light as 35mm @ f/2.8 when you have a bigger opening with the 200mm lens?!? 71.42mm hole vs 12.5mm hole. If you look closely, you will notice that the 200mm lens is LONGER than a 35mm and this length acts like a long extension tube which reduces light, so the amount of light that 200mm lens lets through will eventually be the same as the amount of light as a 35mm lens of the same aperture opening. The T-stops rating basically identify the "APERTURE" blades efficiency by how accurately it stops down the lens to that specific aperture. The confusion comes, in which I'm noticing at lot in those so called marketing guys standing around in trade shows, that they use T-stops as though it replaces f/stops. And with this confusion, they don't realize why the lens is designed this way and why is it so big. There's no confusion if a guy actually knows what he is talking about. This stuff is even taught at least in photography classes 2 decades ago. It's common knowledge, it is immediate that the opening of the lens and length of the extension of the tube do not mirror a constant f/2.8.

Having said that and just by looking at the lens itself, this is more likely a 35-100 f/3.5 to 5.6 or f/4 to f/5.6 simply by doing the f/stop formula calculation and by knowing Panasonic has the 12-32 pancake version. The openings of 35 @ f/4 and 100 @ f/5.6 requires only a short extension tube like the photo illustrated. Another possibility is a f/2.8 to f/4 lens, but the length of the lens is too short -- hint hint take a look at the Olympus most hated 12-50 to give you an idea about lens length.

 

Going with Panasonic is very video focused company and that they are not a camera company until recently. My take is that, because there's a marriage between video and stills that a video centric representative may have implied that this new lens can have the same light transmission capabilities as the 35-100 f/2.8 and that he may confused the answers. Assuming he really knows what he's talking about. I've met people in trade shows that do express lenses in T-stops because they are assuming you are using a DSLR capable video (my D800) for video production purposes. Mirrorless lenses have a shorter lens flange than non-mirrorless. This has been discussed at CP+ by Sigma and Canon.

You can bend physics sure, but at a higher price and letting the camera does all of its automated optical corrections -- that's something Panasonic and recently Olympus has started to do. I'm not sure this is what Panasonic is targetting its market consumer for a GM1 telephone lens. A $2000 lens? Nah. f/stops is intrinsic in the service industry for repairs. Replacement overnight? Nah..

 

f/stop basically determines the amount of light transmitted by the lens based on a formula f number divided by aperture. The variation you talked about in regards to noticeable difference in exposure between 2 lenses of the same f/stop is due to
1, Different lens design
2, Varying focus
3, The use of accessories

Can all effect the actual amount of light projected. This variation of difference in exposure is automatically compensated by through the lens metering (TTL) in the camera.

T-stop or T-number is a f/number of a lens corrected for the light loss during transmission through the lens. T-stop is extremely important for film and movie making because the light loss of older lens design can significantly impact the depth of field needed to be used and the number is an accurate transmission reference rather than the stated theoretical max.

You loose about 1% per air to glass surface of a well corrected lens with coating, so say a 6 element prime lens (2 surfaces) loose approximately 8-12% of light. Now contrast that to a zoom lens with say 20 elements (2 surfaces) will loose approximately 40% to 50% light, but with excellent coating and element bonding might get you down to maybe 35%.

So a prime lens set to T/4 will be about f/3.8 to f/3.6 and a zoom lens set to T/4 will be about f/3.2 to about 2.9. The background will be a little bit sharper with a prime lens at the same T stop assuming both lenses are equal in optical sharpness. T-stop number is always slower than the f/stop number. So in this case, EVEN if you have 2 lenses set to the SAME T-STOP, the background sharpness due to depth of field will be different. Both T-stop and f/stop use the same exposure scale.

So why then, the film and movie industry rely on T-stop? The answer I was given was because TTL metering was not as commonly used compared to a film still camera.

 

To explain #2. Some consumer lenses focus by moving the focusing element in and out outside of the lens body itself and the length of the lens grows as a result. This effect is like an extension tube thus resulting in some slight light loss. But this loss is compensated by TTL metering. To explain #3. The accessories are those lens filters -- UV, Polarizer, Graduated, Neutral Density, color filters and the lot that I won't mention here. Anytime you attach a filter on to the lens, you will loose some light transmission because you are introducing a glass element in the front of the lens. If you add a Neutral Density 0.6 filter (2 stops), you will loose 2 stops of light. This is necessary in some photographic applications. But this is of no consequence, because again the TTL metering will compensate for this. How does the TTL metering compensate for this variation of light levels? By telling the camera to vary the shutter speed. So in the camera, the shutter is used to control ambient exposure as well as controlling motion blur. The shutter speed IS NOT FIXED. If you examine this issue, you will notice that it does not matter what light levels are, the metering will control the shutter speed to provide a balanced exposure assuming you lock into a specific aperture. Therefore, you can attach as much filters you want to your zoom lens without creating excessive vignetting and when you stop it down to f/8 or f/2.8, you WILL get the DOF equivalent to f/8 or f/2.8 no matter how many elements are on the lens whether they are in the lens or outside the lens. f/stop is also referred to as Aperture which is related to DOF.

BUT, when we talk about T-stops, the number of elements in different lenses (prime and zoom) with the same T-stop rating can affect the DOF as I had described in my last post.

A movie film camera creating a cinematic film look @ 24fps has a set exposure time of 1/48sec or 1/50sec for each frame. For 60fps, I believe it is 1/120sec. To control the duration of the shutter (changing the shutter speed), we vary the shutter angle in a rotary disc shutter. This variation of the shutter angle creates different shutter speeds. Shutter speeds in film and video cameras control the effect of motion blur.
Steven Spielberg's movie "Saving Private Ryan" is a fine example of a high shutter speed of 1/2000 sec to create that crisp moving motion, or they call the newsreel effect.

But a film still camera or digital still camera like our Olympus does not angle the shutter blades to create different shutter speeds and neither it sets its exposure time fixed at 1/48sec or 1/50sec for 24fps, because you are not making a movie.
With digital SLR, the movie making is via the electronic pulsing of the sensor array to mimic shutter angle and set at a fixed exposure time of 1/48 or 1/50sec, while the shutter blades stay open the whole time. There is also no rule in still photography where you must set your shutter speed to 1/50 sec or 1/120sec for every frame.

Even if your lens is loosing some light and it's not what it's rated to be.

Modern cameras has 2 ways to compensate this on top of TTL metering.

1, Shutter monitor sensor (on expensive pro bodies)
2, Sensor grid light fall off mapping (on expensive pro bodies)

The shutter monitor sensor is a special nuisance if you are doing infra-red photography, because it will show up on the sensor on long exposures as some sort of colored haze.  

 

You have the freedom to believe in what you want. If you want to work in T-stops with your photography, I'm not stopping you.

Cheers!

 

I said exposure time, not shutter speed that relates to T-stops. The exposure time of the movie film camera is always set to 1/48sec @ 24fps is it not? Or are you trying to confuse me that shutter speed sets the exposure time on a movie film camera?

 

Speedandstyle said the following -- Exposure time and shutter speed are the same thing. The length of time the sensor or film is exposed to light is determined by how fast the shutter is opening and closing.

This is very odd and that's not what I was taught at least when we are dealing with making movies and videos.

The shutter speed in a movie camera is known as the Frame Shutter Speed. Frame Shutter Speed controls the motion blur of the movie and is usually double that of frame rate and that's it to get natural looking motion blur. It does not control the exposure of light if you're shooting at very bright sunlight @ EV 15 (right at noon). Usually the Frame Shutter Speed rate can not change during capture dramatically due to frame rate limitation in HD-DSLR. I shoot movies with a D800 btw.
So to change exposure, I was told by the instructors @ Nikon USA that only ISO, aperture and filters can be altered to obtain a desired exposure. Are you implying that people at Nikon USA or even at Panavision are completely wrong?!?

 

Only in M (manual) mode that you have control on shutter speed and aperture with the D800; other modes seem to lock you in the frame shutter speed based on frame rate you use. The idea they convey is that, while you can make the shutter speed faster than the frame shutter speed, you will change the look of the blur. If you want consistent motion blur on your movie, then you would set the frame shutter speed constant. This practise is NO different than shooting HDR, where you would LOCK IN the aperture setting and vary shutter speed to bracket exposures. You can surely vary aperture settings and shutter speed to bracket your exposures, but the merged HDR photo won't look all that sharp; if that's the kind of look you go for. But in practise, if you want a sharp HDR photo, the bracketed photos are done via Tv (shutter speed) and not aperture. Same with movies. If you want to maintain the same consistent motion blur look on the entire movie length, wouldn't you then keep the frame shutter speed constant and change exposure elsewhere? Isn't T-stop f/stop with light transmission loss taken into account?
That's what Nikon USA was teaching. It's not a mistake. Some of this information is in the digi tutor site as well as shooting HD-SLR tutorial.

Here's the link..

http://cinema.nikonusa.com/produce-...to-consider-plus-need-to-know-pointers/#Frame Shutter Speed + Frame Rate