Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

RPPs... 4K vs. 5K

Okay, here goes . . .

When I put the very first PL mount on the famous Peleng 8mm lens, I would get phone calls asking me if you could use it with a super 16 camera, and what would the lens turn into, or act like - a 12mm? a 16mm?

I would explain that the lens is ALWAYS AN 8MM LENS, IT DOESN'T TURN INTO ANOTHER FOCAL LENGHT OR ACT LIKE A DIFFERENT LENS.

This was usually met with silence on the other end of the line. This happened again and again, so I finally posted a photo taken with the 8mm and then showed the different formats you could use the lens for, as frames. I think this helped, and I will try to dig it out but I don't sell the converted lens anymore so it's gone from my website.

Anyhow, here's what works much better: the "normal" lens for any given format. Now I know that David Mullen doesn't like the idea of a normal lens, and I respect that. But for purposes of figuring out what is a normal, wide, and telephoto lens for the format you are using, I think it's very useful and eliminates this notion that a lens turns into another lens or acts like another lens, which is just a bad way of communicating.

So, here goes:

A normal (not wide, not telephoto) lens for

16mm and 2/3" video is about 16mm
3k RED about 25mm
4k RED and most 35mm formats about 35mm
X35 about 40mm
still aka full frame 35 aka Vistavision about 50mm


If you can remember the above, then you will know that on X35, a 25mm is a wide angle. But on 3K it's more of a normal. On 16mm it's a bit telephoto, but not much: good for shooting close-ups of actors because it flattens faces instead of making them look bloated, as you would get if you used, say, a 12mm lens but moved the camera closer.

Now why do I say 50mm is normal for full frame/still 35? Probably it's more like 65 or 75mm but that's the lens that you get when you buy an old Nikon still camera, or most other still cameras. They come with that 50mm.

Also, you just MIGHT NOT LIKE these numbers. You might decide that 25mm is more of a normal focal length with RED 4k or especially RED 4k HD which is a bit smaller. You just might like the look of the 25mm as a normal lens, and shoot most of your dramatic feature with it. Or you might feel that 40mm is more "normal". That's up to you. Different focal lengths have different feels, and so your normal might be someone else's wide angle.

But if you can remember the the numbers in bold above, or your version of it, then you won't go down the wrongheaded path of thinking that a 50mm becomes a 75mm when you go from a larger format to a smaller one, which I think is a bad way to operate.
 
Image circle chart

Image circle chart

Agree a hundred billion percent. That's what I've been trying to convey, it just seems to get lost in translation when I say it...

A concrete answer from RED of just how much of the chip gets used in the main aspect ratios would be extremely helpful right now.

Thanks for the vote of confidence, Häakon! I think we're fighting a losing battle though :(

Okay, here's my best estimates for the sizes of the main shooting options on the Mysterium and Mysterium-X sensors. Of course, I could be way off, but there is logic behind all of my estimations, and I'm pretty confident that the production chips will be very close to the numbers below.

mysterium.png


So now my question is this:
To everyone who's been in favour of using 'X35' or other marketing terms; which mode(s) is the proposed term supposed to relate to?

Or, to put it another way:
If I design a new lens with a 30.10mm image circle, can I market it as an 'X35'-compatible lens? Sure, it wouldn't cover the full 5K active sensor area (which is a non-standard aspect ratio of 1.896:1), but it'll comfortably cover 5K at 2.40:1 and 16:9 aspect ratios.

So what is the purpose of a single term to describe sensor coverage, when the sensor has different coverage requirements depending on which aspect ratio you choose?
 
The point of any label is so that you don't have to print that chart on the side of the lens, or post it in every discussion. If you don't have a shorthand term to describe some format, then everyone starts to create one anyway -- when I was at NAB, everyone kept asking me about this new 5K lens format and I had no idea what they were talking about. "5K" became the shorthand term to describe a lens that covered the 30mm-wide Mysterium-X sensor.

So you could say that labels are unnecessary or inaccurate, but that doesn't negate their usefulness -- people need a term that represents something too long to type out in every post.

Which comes back to my other question, is this "5K" area a new format or not? If not, then a label isn't really necessary. But I suspect that we will be using some sort of one-word term to describe this inbetween size between Super-35 and FF35. And if no one tries to be proactive, that terms will end up being "5K".

"Are you shooting your next feature in Super-35?" "No, I'm shooting 5K". "What size is 5K?" "Well, it depends on what sensor you are using."

The point of a label for a format is not to be accurate necessarily -- in fact, it's useful if the label is a bit vague because it has to cover a small range of variation.

Look at the label "4K", which seems to describe a range from 3840 to 4096 pixels across, just as "2K" sometimes covers the 1828 to 2048 pixel range. Which is fine, for "4K" to not be exactly 4000. I'm just objecting to the label "5K" to describe a lens coverage area or sensor size because now we're getting into somewhat mixed metaphor territory.

There is no perfect solution. There will always be a need for charts of exact figures to describe something. How many times have we discussed the size of Super-35 or FF35, or the size of the 4K recorded area of the RED ONE? That need will never go away because you will always need shorthand labels and they will always therefore not contain the actual dimensions they are referring to, but sometimes in conversion, all you need is the shorthand and other times, what you need are real dimensions.

So you can say that a term or label is unnecessary, but I suspect one will arise naturally. If it's not a term we create, it will become the one everyone is throwing around casually out of convenience, so I suspect it will become "5K" over the next coming year. "Do these lenses cover 5K?" etc. will become the most common discussion topic. Which is fine when the question really is about whether the lens covers a 5K area of a specific sensor. But when it starts to become that "5K" becomes the term to describe a format that is 30mm wide, then it will just bug me because we let this term become established. But it's not the end of the world, it will just be another misleading label.

What you're not getting is that a vague term like "X35" can't really be misleading or inaccurate because it is meaningless; it represents something else without trying to refer to actual dimensions in the title. Otherwise you get into that little bit of inaccuracy like with formats like "Multivision 235" to describe a format meant for scope release, which is more like 2.39 : 1.

To get to your specific question, it seems that the point of a label like "X35" or "5K", whatever, is to describe a lens that covers more than Super-35 but less than FF35. Otherwise the conversation runs "does this lens cover Super-35?" "Yes." "Does it cover FF35?" "No but it covers [insert whatever term or descriptor you want]."

Sure, maybe a new term isn't needed; we can just wait and see. I'm just guessing that a term will arise naturally, that's all. People find all sorts of ways of putting labels on things.

Afterall, we use the term "Super-35" to describe a vague range of different aspect ratios that are all 24mm-ish wide. Super-35 isn't 4x3 or 16x9 or 1.85 or 2.40 necessarily.

Maybe for marketing purposes, RED should call this new size "5K" -- just because "4K" has become a vague marketing term and 5K sounds better than 4K. It would bug me of course, but maybe no one else.
 
Another way to look at it

Another way to look at it

David, the reason for their being so much talk about the size of the 30mm X 15mm sensor is; can I fit a 16 X 9 frame in the “5K’ sensor, and by default a 1:85 frame with it! The answer is yes!

Why, because weather we like it or not, the next evolution of the Epic camera will be “5K” without leaving the expensive glass and matt boxes behind and all the other accessories behind.

The question has manifested in many different words, but the bottom line remains the same.

Humberto Rivera
 
I appreciate your input, David. I agree that "5K" will probably become the default label, unless Jim continues to refer to "X35", in which case I'm guessing that'll catch on in the forum.

Which comes back to my other question, is this "5K" area a new format or not?
I'm still leaning towards 'No'. I prefer to think of it as a framework that makes provision for several smaller formats.

"Are you shooting your next feature in Super-35?" "No, I'm shooting 5K". "What size is 5K?" "Well, it depends on what sensor you are using."
My analogy is this: "Are you shooting your next feature in Super-35?" "No, I'm shooting X35". "What size is X35?" "Well, it depends on what aspect ratio you are using."

Obviously, when referring to footage in a post/workflow context, "5K" has a distinct semantic meaning to most people. But when referring to sensor dimensions, lens coverage, and the subsequent effect on FOV, I'd much rather people say: "does this lens cover Super-35?" "Yes." "Does it cover FF35?" "No but it covers 31mm at 2:1."
 
So now my question is this:
To everyone who's been in favour of using 'X35' or other marketing terms; which mode(s) is the proposed term supposed to relate to?

Or, to put it another way:
If I design a new lens with a 30.10mm image circle, can I market it as an 'X35'-compatible lens? Sure, it wouldn't cover the full 5K active sensor area (which is a non-standard aspect ratio of 1.896:1), but it'll comfortably cover 5K at 2.40:1 and 16:9 aspect ratios.

So what is the purpose of a single term to describe sensor coverage, when the sensor has different coverage requirements depending on which aspect ratio you choose?

Great chart! Thanks. Looking forward to the Monstro version.

X35 would be shorthand for the sensor size. Lens choice would follow aspect ratio, as you observe, but if we have a starting point (a label everyone recognizes, like S35), then these choices become more clear. One can quickly distinguish between X35 2.4x1 and X35 16x9, whereas 5K 16x9 could be shot on other sensors with different lens requirements.
 
Then you're back to the non-parallel labelling of "5K" being the size between Super-35 and Full-Frame 35mm...

... besides, some of that "expensive glass" you are referring to won't cover that new size.

There are two very different issues here, and trying to treat them as one causes some confusion.

At first, i was wondering whether new photographers were less intelligent than those 50 or 100 years ago, because nobody back then would dream of asking what a 100mm lens "worked like" in a different format. You just knew what would suit your purpose in 4x5", 6x6cm or 24x36mm.

Then it dawned on me that the entire confusion comes from rather clueless marketing people trying to sell the new breed of digital cameras into a mass market.

The next step was to start questioning why we see different lenses as "normal" for different negative sizes, and the answer appears to be that they would give approximately the same field of view when you look at the finished picture as they have on the camera.

Further consideration led me to believe that it is not very useful to define "normal" by the image diagonal now that we have so many different aspect ratios. The interesting part is the horizontal field of view.

Nor is the size of the imager the important factor. With film, one could assume that the resolution and grain was the same for all negative sizes. With digital, it is the number of horizontal pixels that defines how much of the spectator's field of vision one can fill.

This means that if you have a 2k release target, your normal lens would be one that gives the same field of view over 3k as the typical viewer will see his display in if you originate in 3k. If you orignate in 5k, the "normal" lens should have the same field of view over that width, no matter how big or small your pixels are. It is no longer the number of mm that define your normal lens, but the number of horizontal pixels in your target format.
 
How about refering to the diagonal measurement with an X (as if an x were drawn on sensor) then a / preceding the max resolution i.e

MYST X 2/3 X12/3K
RED ONE X30/4K
MYSTX S35 X35/5K
MONSTRO 645 X70/6K
MONSTRO 617 X180/28K
 
With new choices and opportunities comes complexity. It is in our nature to want to simplify things as much as possible so we can keep track of all this stuff.

We don't have (and haven't heard) a good solution to this yet. But something will come up that makes sense.

One thing for sure... "Professional" will mean something. He is the guy (or girl) that understands all this...

Jim
 
A 50 mm is only a 50 mm on FF35, at least by historical standards. .

That made me laugh, historically people realised you used different focal length lenses on different formats. If by historically you mean the last 5 -10 years you might have a point.
 
Some Thoughts

Some Thoughts

Let say that Steven Hilder’s chart was right, or at least close; in a “5K” camera or whatever you chose to call it, the one with the 30mm X 15mm sensor. OK, he says that at “5K” (sorry if I offend any-one with using the term “5K”) which in a 16:9 recoding mode is 4800 X 2700 vertical and horizontal resolution, at resolution of 13.0 megapixels and uses 25.92 X 14.58 of the “5K” sensor at a 29.74 diameter. Then that fits just right with either the RPPs or the Arri Master Primes. And let’s say the within that 16:9 space I frame it for 1:85 aspect ratio, then I’m Ok. I have a slightly larger image, I can still use the only two lenses that I would consider, and I can get a “4k” print when releasing to 35mm film.

I’ll be using a newer model camera with all of its modular advantages, it’s a stop faster, and if and when I’m ready in the future I can just change the brain. I’m not really sure what Jim is referring too, but we all have our opinions.

Humberto Rivera
 
I remember reading that human and rodent vision are about the same at an equivilant of 38mm in the 24x36 format.
 
I’m not really sure what Jim is referring too, but we all have our opinions.

What we're discussing is that '5K' doesn't refer to physical sensor size or format but to how many pixels are on a sensor. And pixels/photosites are variable in size, so if we start saying 'this lens covers 5K sensors' and tomorrow someone releases a camera with a 5K sensor of a totally different size (which that lens doesn't cover) it could get confusing.
 
If I write an article on digital cine cameras with "Super-35" or "35mm cine" sensors, I can list the Arri D21, the F35, the Genesis, the Red One, the Phantom HD... but they all have different pixel dimensions -- the Arri D21 is 2.8K RAW, etc.

So labeling the next major jump in sensor size as "5K" is, well, confusing. If other companies adopt this 30mm-wide sensor size, I'm going to have to make up a name for that size, because they won't all be 5K in horizontal pixel resolution. But if the idea is that this odd size will be unique to RED's Mysterium-X and will always be 5K across, then it's less of an issue if they want to call it 5K because it only refers to one sensor.

I'm writing the new chapter on digital cameras for the next edition of the ASC Manual -- actually I wrote it some time ago but now and updating it before publication -- and my overview of the Dalsa Origin (now obsolete) reads "a slightly larger than 35mm-sized single sensor", an awkward phrase. But I guess that's how I'm going to have to describe the Mysterium-X cameras as well.

There should be a term for standardized sizes of sensors and then the information regarding their pixel dimensions. I just don't think the physical size should be labeled by their pixel dimensions. "APS-C" is a name of a sensor size, but it doesn't tell you how many pixels are in that sensor. The fact that it's called "APS-C" is less relevant -- they could have called it "Fred" for all that matters as long as everyone used the term and agreed as to what it covered.

I guess we could shelve this discussion until it looks like more cameras will need lenses that cover this area.
 
http://en.wikipedia.org/wiki/APS-H#Multiplier_factors

I just noticed that Canon makes a sensor that is 28.7mm across in the "APS-H" format. Haven't heard of it, which I guess means there aren't a lot of cameras using it. But here's a case where the format name, whether or not it catches on, does not indicate pixel dimensions.
 
If this format would/should/might be used by other camera making companies, then Red should be omitted from the name not to scare them away. Since it is a pure digital format I would suggest names like DigiRama or PixelRama or PixelVision or PixelScope or, one of my favorites, TrueScope ...

If not, then Red word (or its product names) just shouldn't be omitted, so something like ScarletRama, RediculousVision, ...

Just FYI, we've already had PixelVision for about 20 years. It was a pretty awesome little machine, though don't know if anyone would actually want to shoot a feature on it.

http://en.wikipedia.org/wiki/PXL-2000
 
How about using the diagonal? It would certainly make it easier to match the sensor size with the lens. So, Mysterium-X sensor would have R33.5* format and 5k resolution. No confusion. I can immediately work out my angle of view in diferent aspect ratios and angle of view on different focal lengths and crops.

*30mmx15mm gives approximately 33.5 diagonal
 
I don't mean to dumb-down the conversation, but generally, I've always hated the marketing around "pixels". It's one of the reason that even in the beginning, Red drove me a bit crazy. I never really cared how many pixels, but the physical size of the chip!

It's even worse when you buy a camera at Best Buy or others. They can never tell you the physical size. Hell, no one at Henry's can ever tell you either. And in the cheaper cameras, more pixels was probably a bad thing as it reduced speed and could add noise. How ironic.

For most of the indie guys coming from the video world, 35mm DOF was the holy grail--not resolution!

Which leads me to the current Red offerings in the works. I can tell you the pixel size, but I get confused on the physical size, and relationships to industry standards and the impact on lenses and DOF.

I am now a dumber customer.
Chris
 
Back
Top