F or T?

[Stephen Webb]

I know this must have been asked before, but a forum search isn't turning up an answer to this query so I thought I'd ask anyway:

Why are the Red lenses marked up in "f" rather than "t" stops? Also, looking at Jarred's photo of the rear panel of the camera it appears that the display is in f-stops too.

[Stephen Williams]

I know this must have been asked before, but a forum search isn't turning up an answer to this query so I thought I'd ask anyway:

Why are the Red lenses marked up in "f" rather than "t" stops? Also, looking at Jarred's photo of the rear panel of the camera it appears that the display is in f-stops too.

Hi Stephen,

I have asked, never got an answer from the Red team. I think 2.8 sounds better than 3.X, and many people don't seem interested.

Stephen

[Stephen Webb]

many people don't seem interested.


My DP's interested - his reaction was to dismiss them as "not professional lenses". It does seems an odd choice.

[Finner]

My DP's interested - his reaction was to dismiss them as "not professional lenses". It does seems an odd choice.

I had assumed the lenses were T and not Fstops. That is strange to me. I can see a lot of cinematographers reacting the same way yours did.

[Leo Ticheli]

F-stops are required for calculating depth of field. T-stops are used to set the iris with a light meter.

Since we have eyeballs and displays for setting the iris, T-stop markings are not required.

I think there are still some diehards out there who insist on setting the iris on HD camera with a light meter, but I believe it's crazy. The sensors are not linear and it's far faster, easier, and more accurate to set levels by looking at the displays.

In the HD world, light meters are useful in lighting sets and locations, but not for setting the iris. Unless you just want to look cool in front of unsophisticated clients! :-)

Still, it would be helpful to know the T-stop of the lens to judge it's true maximum aperture.

Good shooting and best regards,

Leo

[Finner]

Still, it would be helpful to know the T-stop of the lens to judge it's true maximum aperture.


This is the reason some may judge these lenses. Plus being such a good price if the lenses are good performers many would want to use them with film cameras and with them not being marked in T-stops would hinder that. I personally would want to use the lenses with both the RED and film cameras.

[Leo Ticheli]

Good point, Finner; it wouldn't be a big deal to add T-stops to the other side of the barrel.

I personally haven't shot film in five or six years, and have completely embraced HD, so it's not a problem for me, but the addition of T-stop markings would make the Red lenses more valuable.

Good shooting and best regards,

Leo

[David Mullen ASC]

Seems to me that, if as you say, with digital you're more likely to set the exposure based on a waveform and whatnot, then you're not going to pay much attention to either f-stop or T-stop marks, so you might as well put T-stop marks for the people who will be using them for exposure purposes.

T-stops have been fine for most depth of field calculations over the years anyway.

[donatello b]

"many would want to use them with film cameras and with them not being marked in T-stops would hinder that"

i don't think so ... pretty much any cine lens made in past 10 years - the difference between F & T stop is usually at MAX F stop & then over the next 1- 2 stops the f stop & T stop end up being equal ... and the difference at the max F stop is usually 1/3 stop at max - which FILM latitude can handle with no problems ....
i beleive the new cookes would be approx F 2.8 or T 3.0 (or T 3.1) ...

the older cine zoom lens made back in 60-70's - some had 3/4 to 1 1/2 stop difference bewteen F & T stop at max aperature and usually over the next 3 stops they became more equal ...

[David Mullen ASC]

One problem with lenses marked in f-stops is when you have multiple cameras running on a scene, like a stunt for example, and often a single exposure setting will be called out for all the cameras to be set to by the assistants... so if the lenses mismatch in light transmission, the exposures will be all over the map.

[Finner]

(David you beat me to some of my points. Ha.)

donatello

when you are matching pre shot stuff or using different lenses on different cameras or 2nd units F-stop will not be nearly as helpful as T-stops. Everything can be justified as "almost as good" if you go on that basis you could plan to shoot something with the highest quality 35mm package and then say well 16mm is "almost as good" as 35mm and it still uses cine lenses and then say well HD is "almost as good" so I will use a f900 and then say well the HVX is "almost as good".... before you know it your shooting with a $300 future shop special.

T-stop is the most accurate mesurement for lenses period. Anything else just becomes an "almost as good". This could be the reason RED is able to offer these lenses up for such a good price. I am very interested in seeing RED lenses tested head to head with some other pro cine lenses.

[roryhinds]

Isn't it the RED Digital Cinema Camera?

It should be T stop and not F stop surely.

[David Mullen ASC]

It would be a simple thing to remark the lens barrel.

[Finner]

It would be a simple thing to remark the lens barrel.

Ya, thats why I would find head to head tests interesting and I could see a lot of RED lenses being remarked to T-stops. It would just be nice if it was done ahead of time.

[Stephen Williams]

Hi,

Lightweight ENG zooms are marked in F stops. If you zoom in when wide open, the image gets darker. T stops do not work on all zoom lenses.

Stephen

[Chris Kenny]

Hi,
Lightweight ENG zooms are marked in F stops. If you zoom in when wide open, the image gets darker. T stops do not work on all zoom lenses.


This has nothing to do with F-stops vs. T-stops, really. You could take a lens which couldn't maintain its largest aperture when zooming and mark it with either. Such lenses are typically marked with an F-stop range, e.g. f/2.8-4.0. There's no fundamental reason you couldn't mark that same lens T3.1-4.5 (or whatever it actually was).

[Stephen Williams]

This has nothing to do with F-stops vs. T-stops, really. You could take a lens which couldn't maintain its largest aperture when zooming and mark it with either. Such lenses are typically marked with an F-stop range, e.g. f/2.8-4.0. There's no fundamental reason you couldn't mark that same lens T3.1-4.5 (or whatever it actually was).

Hi Chris,

A T stop is the light that is actually transmitted through the glass, not much good if it changes IMHO.

Stephen

[Chris Kenny]

A T stop is the light that is actually transmitted through the glass, not much good if it changes IMHO.


I agree that a lens which can't maintain the same stop throughout its zoom range is undesirable, particularly for motion picture work, but this is a completely separate issue from F-stops vs. T-stops.

F-stop takes into account aperture size, which is, as far as I'm aware, the only factor which can cause varying brightness when changing focal lengths on a zoom. (Light is passing through the same glass at any focal length, so you can't loose varying amounts of light there.) As such, a zoom which maintains the same F-stop throughout its range should maintain the same brightness.

In other words, Nikon's f/2.8 17-35 is just as bright at 35 as at 17, and you can know this even though it's not marked in T-stops. To know what that brightness actually is you'd need need a T-stop, but to know it remains constant you don't.

I'd be very interested to know if anyone has any information that contradicts this, e.g. data about lenses that maintain the same F-stop throughout their ranges but do vary image brightness. I've never run across any information suggesting this can happen.

[Stephen Williams]

In other words, Nikon's f/2.8 17-35 is just as bright at 35 as at 17, and you can know this even though it's not marked in T-stops. To know what that brightness actually is you'd need need a T-stop, but to know it remains constant you don't.



Hi Chris,

Have you measured it? A 2:1 zoom, possibly does change a little (or portholes ) but not enough to mention it.

FWIW the T1.4 SuperSpeed's aren't all T1.4 either, Most in the series are F1.3 but one is a F1.4, marketing is marketing! The only way you will know is to measure an individual lens.

Stephen

[Dreamcine]

Heck, we're already on an engineering hold, why not chage the f to t?

[Chris Kenny]

Incidentally, I'm not a fan of the current stop scales used in photography, conceptually. I know the geometry that leads to the F-stop scale, but IMO it would be more practical to mark stops based on the amount of light they (hypothetically) pass on a linear scale, e.g. 1, 2, 4, 8, 16, etc. with each passing half as much light as the previous one. Another way to say this is, it would make more sense if the stop number represented the area of the aperture, rather than the diameter.

Mind you, I have no illusion that this will ever happen. The current stop scales will be with us as long as the QWERTY keyboard, I'm sure.

[Jeff Kilgroe]

OK, I'm mostly coming from a video and still photo background in regards to lenses. But the way I understood this (and I'm probably wrong) is that there's no need for T-stop marks if the light absorption remains constant regardless of zoom or focal settings. The T-Stop represents the light transmission of the lens itself and is not directly linked to aperture. For example, If the RED 3000mm is a T1.8, then it remains a T1.8 regardless of which aperture you set.

T-stops would be most likely variable on a zoom lens. So it would make sense to mark t-stop marks on a zoom *if* they change with the aperture/f-stop settings or with a change in focal distance. But if a piece of glass can maintain a constant t-stop rating, then how and why would you mark t-stops?

[Adrian T.]

Jeff, you're mistaken.

The t-Stop is directly related to the f-Stop. For example t-Stop 5.6 is exactly where the f-Stop 5.6 would be on a perfect lens with no light absorption. So the aperture scale may show t-Stops instead of f-Stops. It's simply a matter of the marking. This has nothing to do with zoom lenses.

[Stephen Williams]

OK, I'm mostly coming from a video and still photo background in regards to lenses. But the way I understood this (and I'm probably wrong) is that there's no need for T-stop marks if the light absorption remains constant regardless of zoom or focal settings. The T-Stop represents the light transmission of the lens itself and is not directly linked to aperture. For example, If the RED 3000mm is a T1.8, then it remains a T1.8 regardless of which aperture you set.

T-stops would be most likely variable on a zoom lens. So it would make sense to mark t-stop marks on a zoom *if* they change with the aperture/f-stop settings or with a change in focal distance. But if a piece of glass can maintain a constant t-stop rating, then how and why would you mark t-stops?

Hi,

The T1.8 is only when the lens is wide open, close it down a tad and you have T2 setting on the lens

Cine Zooms have an extra iris that changes to keep the light transmission constant, so a still lens that was say f2.5-f 3.8 would turn into a cine zoom of possibly T4.5 throughout the range.

Some people don't understand why the F number increases as the aperture gets smaller, its because they are fractions 1/1.4, 1/2, 1/2.8 ,1/4, 1/5.6 ,1/8 etc.

Stephen

[Sean Michael Johnston]

Don't they have to run tests on the actual light transmission of the lens before marking the T stops? Who does these tests? The lens manufacturer?

[David Mullen ASC]

A T-stop is basically a "corrected" f-stop, the f-stop mark moved over to account for any light lost due to lens elements, etc. So if your meter says to use a certain f-stop, then you'd use the T-stop on the lens for that exposure. And if you switched to different lenses, the exposure should be the same if you use the same T-stop on each lens.

The thing that is unique to moving images as opposed to still images is that they are cut together in a continuous sequence, so shot-to-shot matching in color, contrast, and brightness is a lot more critical -- when you cut from a wider shot to a close-up, the person in the shot should not shift to green, get darker, or more washed-out, etc. And while many of those problems can be corrected in post color-correction, it's better if the lenses match each other and the exposures are consistent so you don't spend all your time in post trying to simply match things.

[Stephen Williams]

Don't they have to run tests on the actual light transmission of the lens before marking the T stops? Who does these tests? The lens manufacturer?

Hi,

Motion picture lenses should be individuslly marked and calibrated for T stops & Focus. Larger rental houses will be able to check lenses for light transmission.

Stephen

[Jeff Kilgroe]

OK, I got it now...

[roryhinds]

Anyone heard anything from RED on the subject?

[Stephen Williams]

Anyone heard anything from RED on the subject?


Hi Rory,

Yes.

QUOTE(Jim Jannard @ May 27 2007, 11:26 AM)
"How about cost? If you want T-stops, buy some Cooke or Zeiss glass."

Stephen

[Rob Lohman]

Would be good to include the full quote with a source

How about cost? If you want T-stops, buy some Cooke or Zeiss glass. Our 5 lens prime set costs less than one Zeiss Master Prime. Our zoom costs 1/6th the price of an Optimo. We are building a very sharp set of digitally corrected lenses with the idea of getting glass into the hands of many.
Source: http://www.cinematography.com/forum2004/index.php?showtopic=23269

[Stephen Williams]

Hi Rob,

Problem with a full quote is possible copyright issues.

T stops being more expensive than F stops is the interesting point IMHO

Stephen

[Rob Lohman]

Jim wrote that, and since this is his/RED's forum :)

Especially since he explains a bit more. Obviously you don't want to copy something verbatim from other sources.

[Stephen Williams]

Jim wrote that, and since this is his/RED's forum :)

Especially since he explains a bit more. Obviously you don't want to copy something verbatim from other sources.

Hi Rob,

He wrote it on www.cinematography.com and I thought its therefore copyright@ cinematography.com.

Stephen

Edit. Sorry I am wrong on that, "All posts are property of the poster." my apologies.

S

[Ivan G]

F-stops are required for calculating depth of field. T-stops are used to set the iris with a light meter.

Since we have eyeballs and displays for setting the iris, T-stop markings are not required.

I think there are still some diehards out there who insist on setting the iris on HD camera with a light meter, but I believe it's crazy. The sensors are not linear and it's far faster, easier, and more accurate to set levels by looking at the displays.

In the HD world, light meters are useful in lighting sets and locations, but not for setting the iris. Unless you just want to look cool in front of unsophisticated clients! :-)

Still, it would be helpful to know the T-stop of the lens to judge it's true maximum aperture.

Good shooting and best regards,

Leo

Does this apply to still photography as well?

[Matt Uhry]

I don't understand why T stops would cost more than f-stops except for the extra gas money for the camera PA to go pick them up from the rental house? ( joke )

Maybe this is part of the contractual de-tuning that the optics supplier does to prevent competition with it's high end products ? If what it takes to get a massive discount on some nice glass...

You can't attribute much added cost to measuring or engraving of T stops.

Matt Uhry
www.mattuhry

[David Mullen ASC]

We've already listed the reasons for T-stops in other posts, buy it's not merely about setting exposure by light meter -- it's also about speed of switching between lenses without resetting exposure by waveform to compensate for the transmission loss of every lens. It's about using multiple cameras and setting them all to the same T-stop for the same lighting situation without individually resetting exposure for each camera by waveform. Some people may have to resort to remarking the barrels themselves after testing, which is silly.

Plus if one just uses waveforms to expose, then it's not going to matter to them whether either f-stops or T-stops are used on the lens. In other words, the people who need the T-stop marks need them more than the people who need the f-stop marks. So why use f-stop marks on the lens?

A lot of professionals have taken their time to explain the reasons why T-stops are used on professional cine lenses and it's disconcerting to have our opinions constantly dismissed on this topic. The arguments here for not having T-stops seem to boil down to two: (1) we don't really need any accurate exposure marks because we expose by waveform anyway; or (2) it makes the lens cheaper to not take the time or effort to put T-stop marks on them.

Which is fine, I guess, but it does mean that this discussion will keep resurfacing as new people come in and question why the RED lenses don't have T-stop marks on them.

[Stephen Williams]

I am curious how many 'A' List cinematographers will be buying Red Lenses marked in 'F' stops to go with their Red One bodies.

Stephen


We've already listed the reasons for T-stops in other posts, buy it's not merely about setting exposure by light meter -- it's also about speed of switching between lenses without resetting exposure by waveform to compensate for the transmission loss of every lens. It's about using multiple cameras and setting them all to the same T-stop for the same lighting situation without individually resetting exposure for each camera by waveform. Some people may have to resort to remarking the barrels themselves after testing, which is silly.

Plus if one just uses waveforms to expose, then it's not going to matter to them whether either f-stops or T-stops are used on the lens. In other words, the people who need the T-stop marks need them more than the people who need the f-stop marks. So why use f-stop marks on the lens?

A lot of professionals have taken their time to explain the reasons why T-stops are used on professional cine lenses and it's disconcerting to have our opinions constantly dismissed on this topic. The arguments here for not having T-stops seem to boil down to two: (1) we don't really need any accurate exposure marks because we expose by waveform anyway; or (2) it makes the lens cheaper to not take the time or effort to put T-stop marks on them.

Which is fine, I guess, but it does mean that this discussion will keep resurfacing as new people come in and question why the RED lenses don't have T-stop marks on them.

[David Mullen ASC]

Well those on a budget might not have much of a choice... which is why I think the compromise might be to make it a purchasing option to order a set marked in T-stops at a higher price, depending on how much higher.

At least if there were an option to order a set marked in T-stops and no one ordered them because of the mark-up, then the controversy could go away because RED would have offered it as an option. No one could claim that it wasn't possible to get the lenses marked with T-stops and RED could claim that none of their customers wanted it anyway, not if they had to pay extra.

Just looking for some solution that satisfies everyone, rather than tell the people who want a lens with T-stop marks on them to take a hike.

[Stephen Webb]

So hang on a minute, after we've been banging on about this for weeks and discussing the pros and cons and trying to get to the bottom of why these lenses are "f" not "t" the answer is... 'cause it'd cost more???

And this reply was posted another forum, rather than the one we've all been asking the question on? Was it really that hard a question to answer?

[Jeff Kilgroe]

Jarred, if you're reading, can we merge these two threads on this subject? Same discussion going in two places...

I suppose T-Stop markings would cost more since they have to be tested and marked for each individual lens. Or at least that's my understanding.... It would make it difficult for any form of user-changeable marking ring to be made to allow for the swapping between T and F markings.

This is probably another reason RED is choosing to go with F markings -- they don't need to individually test and calibrate stop markings for every single lens. Also, a huge portion of their upcoming user base is coming from video and photography backgrounds and may not have lots (or even any) real experience with shooting cine style film systems. This group is used to F stops and not having T stops.

[Damien Molineaux]

...

I suppose T-Stop markings would cost more since they have to be tested and marked for each individual lens. Or at least that's my understanding.... It would make it difficult for any form of user-changeable marking ring to be made to allow for the swapping between T and F markings.

This is probably another reason RED is choosing to go with F markings -- they don't need to individually test and calibrate stop markings for every single lens. Also, a huge portion of their upcoming user base is coming from video and photography backgrounds and may not have lots (or even any) real experience with shooting cine style film systems. This group is used to F stops and not having T stops.

Is that a fact ? For T-stops, lenses need to be individually measured and marked ? Indeed that would bring the price up, although, if I'm not mistaken, all professional cine lenses are assembled by hand ? Therefore how much more would it be, while they're at it, to put the T-stop markings ?

If truely, having T-stops is much more expensive, than okay, say it, we can deal with cheaper quality lenses. But as David Mullen has explained, using various examples, T-stops would make life easier for professional DPs, it would also make the lenses more valuable and potentially interesting for film shoots.

I have another theory however, maybe, who ever is making theses lenses, or the glass at least, has cut a deal with Red, that they will only mark their lenses with f stops, so that Red lenses won't compete with that manufacturer's cine lenses. Just an idea, but hey, if the lenses are good, at this price, I'm not complaining.

Cheers,
Damien

PS But I would rather have T-stops.

[Poi Boy]

The quality of the lens will not change with the markings. I guess at the end of the day I would prefer a T to an F but it is of marginal value to me and I don't care about resale value. When my first camera ships I will make a choice given the reality of the moment.
Aloha
-A

[chuck colburn]

A few people here and over at Cine.com have stated that they believe T stopped lens are or have to be individualy calibrated. Has anyone sourced this "fact" to a given site? I would like to read it for my own edification.

Thanks,

Chuck

[Brook Willard]

Jarred, if you're reading, can we merge these two threads on this subject? Same discussion going in two places...

I'm no Jarred, but...

I would merge the threads, but at this point it would be too confusing. Merging the threads would create a new thread with the posts from each distributed chronologically within each other. That means that different replies would appear after different posts and all hell would break loose. If folks are comfortable with it, I'll lock one thread and link to it in the other. I don't want to cut off discussion if you guys want both threads open, but I won't merge them as it'll create too much confusion.

[Jeff Kilgroe]

Yeah, I suppose you're right, Brook... Probably should've done it a while back. Definitely too confusing now.

[Stephen Williams]

A few people here and over at Cine.com have stated that they believe T stopped lens are or have to be individualy calibrated. Has anyone sourced this "fact" to a given site? I would like to read it for my own edification.

Thanks,

Chuck

Hi Chuck,

This is a reply to my enquiry @ Cooke Optics:-

Hi Stephen,

Yes, Cooke individually calibrates, marks and engraves the scales on all of
our lenses. However, I believe this is not the case for all lens manufacturers!
Hope things are going well for you.

All the Best

Laura Hatton
Cooke Optics Limited

[Moir]

Would I be correct in assuming that all lenses have a theoretically fixed ratio between F and T stops, depending on the optical properties of the elements in the lenses and their coatings? So, for argument's sake, the 300mm lens might be f2.8 and nominal T3.1.

If I further assume that there will be tolerances in the manufacturing process that would make every element unique and no two lenses absolutely identical in their light transmission.

What would be a reasonable degree of variation from nominal? If it's say 1%, would this be noticeable in any practical real world situation?

My question is whether the QC process would likely result in lenses that are sufficiently similar that there is no need to individually calibrate.

[Stephen Williams]

Would I be correct in assuming that all lenses have a theoretically fixed ratio between F and T stops, depending on the optical properties of the elements in the lenses and their coatings? So, for argument's sake, the 300mm lens might be f2.8 and nominal T3.1.

If I further assume that there will be tolerances in the manufacturing process that would make every element unique and no two lenses absolutely identical in their light transmission.

What would be a reasonable degree of variation from nominal? If it's say 1%, would this be noticeable in any practical real world situation?

My question is whether the QC process would likely result in lenses that are sufficiently similar that there is no need to individually calibrate.

Hi,

There is no fixed relationship between F & T stop. The no of elements, the thickness of the glass & coatings will have an effect. A faster lens will have thicker glass, would in theory have a bigger difference than a slower lens with thinner glass.

A variation of +/- 5% is IMHO irrelevant.

If the QC process involved junking the proportion of the lenses that were not sufficiently similar yes. However without measuring lenses how would you know if they were sufficiently similar?

Stephen

[Moir]

There is no fixed relationship between F & T stop. The no of elements, the thickness of the glass & coatings will have an effect.

Stephen, thanks for the reply. If you re-read my original question, I asked whether there was a fixed relationship once you had taken the optics and coatings of the elements into account. In other words, I assumed that calculating the nominal T-stop for any lens would be possible without measurement, subject only to manufacturing tolerances.



If the QC process involved junking the proportion of the lenses that were not sufficiently similar yes. However without measuring lenses how would you know if they were sufficiently similar?


I don't know anything about cine lens manufacture or QC. For large production runs of manufactured goods, I believe it is typical to take samples from batches and not test each item.

You suggest that 5% variation in light transmission would not be noticeable, and for most manufacturing processes I suspect this would be considered a wide margin. If your sample variation is within 2%, you would assume that the batch is acceptable.

Perhaps wrongly, I would expect QC tolerancess for focus would be much tougher than for light transmission.

[Stephen Williams]

Stephen, thanks for the reply. If you re-read my original question, I asked whether there was a fixed relationship once you had taken the optics and coatings of the elements into account. In other words, I assumed that calculating the nominal T-stop for any lens would be possible without measurement, subject only to manufacturing tolerances.



I don't know anything about cine lens manufacture or QC. For large production runs of manufactured goods, I believe it is typical to take samples from batches and not test each item.

You suggest that 5% variation in light transmission would not be noticeable, and for most manufacturing processes I suspect this would be considered a wide margin. If your sample variation is within 2%, you would assume that the batch is acceptable.

Perhaps wrongly, I would expect QC tolerancess for focus would be much tougher than for light transmission.


Hi,

If all the optics were made at the same time, then possibly. I think some other manufactures probbly make that assumption.

If the focal length is exactly the same, once infinity is collimated the rest should work, if there is a slight varition (possible) then the scale can be computed from 3 accurte points as the gearing is known.

Stephen