Image Circle Database

[Matthew Duclos]

I feel bad for taking over the other thread that this was being discussed on.
So I figure I would drop it in a new thread for all to see.
I update the list as lenses pass through the shop here.
If there are any specifics you would like, just let me know and maybe I can throw one up to test.

Click Here! (http://web.me.com/ducloslenses/DataRef/ImgCir.html)

[Yousuf Abbasi]

Any information on image circle diameters for Zeiss T1.3 SuperSpeeds ?

Thanks for the awesome resource.

[Humberto Rivera]

Matthew, your Data Base is coming along GREAT keep up the good work, it isn’t everyday that something new is created, somehow it seems that the question of the image circle or diameter has been kept from all of us.

I do think that the Arri/Zeiss Master Prime and the Red Pro Prime are surprising similar with the Master Prime having the edged. That’s one of the reason that they deserved to be posted just below the Master Prime.

Humberto Rivera

[Matthew Duclos]

Any information on image circle diameters for Zeiss T1.3 SuperSpeeds ?

Thanks for the awesome resource.

Super Speeds coming soon.
I've had several sets through the shop, but they were rush jobs..
Not enough time to take a good reading.

[Humberto Rivera]

Yousuf correct me if I’m wrong. But the Arri/Zeiss Ultra Prime are listed on the Lens Image Circle Database. The 10mm is a T2, the 12mm is T2, the 14mm is T1.9 and all the rest are T1.9. The only T1.3 are the Arri/Zeiss Master Primes. The RPP are T1.8.

Humberto Rivera

[Humberto Rivera]

Yousuf, I stand corrected! There is a set of Arri/Zeiss Ultra 16mm Lenses made for the Super 16 format, that are on par with the 35mm Arri/Zeiss Master Primes. My apologies.

Humberto Rivera

[Steven Hilder]

Thanks Matthew! It's great that someone with access to all these lenses (not to mention the test equipment) is willing to share this kind of information.

Just a word of warning to anyone buying lenses for use on Scarlet and/or Epic - until RED release the final specifications of the production sensors, you shouldn't base lens purchases entirely on these figures. For example, it's unlikely that any recording mode on the S35 cameras will require an image circle of more than 31mm. The measurement of 33.54mm that Matt lists on his website refers to the size of the full sensor area (including lookaround).

[Matthew Duclos]

Just a word of warning to anyone buying lenses for use on Scarlet and/or Epic - until RED release the final specifications of the production sensors, you shouldn't base lens purchases entirely on these figures. For example, it's unlikely that any recording mode on the S35 cameras will require an image circle of more than 31mm. The measurement of 33.54mm that Matt lists on his website refers to the size of the full sensor area (including lookaround).

Very true!

The numbers I have for the future RED cameras are just numbers that I've seen published by RED. I don't know what their real numbers are or will be.
I chose to use the largest possible sensor size for the RED S35 frame just as a worst case scenario. That way if it's smaller, no problem.

[Evin Grant]

If you use the R1s total, active and useable image areas as reference you can approximate pretty accurately the maximum recordable image circle for the three primary S35 Epic formats. (These do not include look around)

2:1- 28.5mm
2.40:1- 27.5mm
16:9- 26mm

[Steven Hilder]

Evin, I think those approximations are a little low - by my calculations it'll be more like:

2:1- 30.91 (approx 31)
2.40:1- 29.95 (approx 30)
16:9- 29.94 (approx 30)

Full details in the full chart (http://reduser.net/forum/showpost.php?p=411234) from an old thread.

[Humberto Rivera]

Steven, I went back to your chart, and if I understand it correctly, one can get a 1.85:1 from the almost full Active Pixel Array of an S35 sensor with a resolution of 13+ Megapixels. The diameter of the lens would be slightly less than 31.26mm.

Humberto Rivera

[Evin Grant]

I'm pretty sure they're correct. The R1 sensor is 24.4 x 13.7mm full area is 4900(h) x 2580(v) but max record area is 4096(h)x2305(v). That's a diagonal difference of 5538 Vs. 4700 or about 15% 33.5mm-15%= 28.5mm.

[Steven Hilder]

Steven, I went back to your chart, and if I understand it correctly, one can get a 1.85:1 from the almost full Active Pixel Array of an S35 sensor with a resolution of 13+ Megapixels. The diameter of the lens would be slightly less than 31.26mm.

Humberto Rivera

Yes, I believe so. If my assumptions are correct, you'll be able to get a 4996x2700 frame (13.5 megapixels) with a diagonal of 30.67mm.


I'm pretty sure they're correct. The R1 sensor is 24.4 x 13.7mm full area is 4900(h) x 2580(v) but max record area is 4096(h)x2305(v). That's a diagonal difference of 5538 Vs. 4700 or about 15% 33.5mm-15%= 28.5mm.

Sorry Evin, what I meant was that comparing the ratio between full area and usable area of the RED ONE's sensor to the S35 would give an inaccurate estimate. The predictions in my chart are based on three key pieces of information that RED have disclosed; the pixel pitch (5.4 microns), the physical size (30x15mm) and active pixel area (5120x2700). With those three values in place, the rest pretty much falls in line.

[david farland]

Matt,
Do you know if the Zeiss compact primes use aspherics on their wides to get the 43+ coverage or can they get away with spherical design because of the relatively smaller aperatures (T2.9/3.6)?

[Derek Wan, HKSC]

Hi Matthew,

Thank you the great database. As we are shopping for lenses this chart is so invaluable to us.

It seems like no cine zoom actually covers the RED Epic's 33mm frame diameter at this point. Is that correct? What about those that are not on the list yet, like RED Pro zoom, Angenieux HP and HR, Cooke 10:1 Mark III, Cooke 18-100mm, 20-100mm and even Technovision lenses? Any luck at all?

Can you also include more classical frame sizes such as Full Gate 35mm, Academy, 16mm, Super 16, etc. for our reference?

By the way, is your RED FF35's graphic a little out of proportion? You might want to fix it to make it perfect since you have already done such an excellent job for the community.

Thanks again.

[Evin Grant]

Steven, the difference comes from the actual capture area, which once look around and non active area are subtracted is about 15% smaller than the physical size of the sensor.
At least on the R1.

[Matthew Duclos]

Hi Matthew,

Thank you the great database. As we are shopping for lenses this chart is so invaluable to us.

It seems like no cine zoom actually covers the RED Epic's 33mm frame diameter at this point. Is that correct? What about those that are not on the list yet, like RED Pro zoom, Angenieux HP and HR, Cooke 10:1 Mark III, Cooke 18-100mm, 20-100mm and even Technovision lenses? Any luck at all?

Can you also include more classical frame sizes such as Full Gate 35mm, Academy, 16mm, Super 16, etc. for our reference?

By the way, is your RED FF35's graphic a little out of proportion? You might want to fix it to make it perfect since you have already done such an excellent job for the community.

Thanks again.


I do have several of the zooms you mentioned, in the shop right now. I'll be adding them soon.
I have a feeling that this list will be a perpetual task.
Ill add a few more sizes to the sensor/film graphics.
I don't think their proportions are too critical. They are just there to give an idea of different sizes. The critical part is the diameter of the image circle produced by the lenses.

[Steven Hilder]

At least on the R1.

That's my point, there's no reason to suggest that Mysterium and Mysterium-X will share the same ratio between full sensor size and capture area. In fact, from the data we've already seen, it pretty much confirms that they won't.

As sensor size increases, that doesn't mean that lookaround increases at the same rate - obviously the 28K Mysterium Monstro won't have 15% of its area dedicated to lookaround, it'd be too wasteful.

[Matt Uhry]

Thanks Matthew - That's a great list - very useful !

Not many 10-1 Zooms are going to cut the mustard in this "Jumbo 35" jumble that we are headed into. The Optimo will squeak by but many Cooke 10-1, HR's and HP's are going to get shelved.

Matt Uhry
www.mattuhry.com

[Evin Grant]

As sensor size increases, that doesn't mean that lookaround increases at the same rate - obviously the 28K Mysterium Monstro won't have 15% of its area dedicated to lookaround, it'd be too wasteful.

Well then it'll be much less useful. There may very well be some difference in the unused area (+/- 2-3%) but without at least a 10% lookaround it's not going to do you any good because you won't be able to see mics or other elements entering your frame. After all that's the whole point of it.

[Steven Hilder]

Well then it'll be much less useful. There may very well be some difference in the unused area (+/- 2-3%) but without at least a 10% lookaround it's not going to do you any good because you won't be able to see mics or other elements entering your frame. After all that's the whole point of it.

I agree to some degree, but unfortunately the figures that RED has already disclosed suggest the S35's lookaround will be around 8% (depending very much on the chosen aspect ratio, of course). Any lookaround is better than none at all.

Anyway, to get back to the original topic, my calculations show that most 5K modes will require around 30-31mm coverage.

[Evin Grant]

Even assuming 8% look around and 2% unused sensor edge you would still only need the 30mm image circle coverage for the 2:1 aspect ratio which is non standard. The 16:9 (1.77) Aspect ratio will only require a 27.5mm image circle and the 2:40 aspect ratio will require 29.25mm of coverage. I think most lenses will work just fine in the standard ratios we're used to shooting. this is actually one of the key benefits of a 2:1 aspect sensor. I also think that with Epic's custom frame sizes most DP's (including myself) will choose to lose a few lines of res. rather than their favorite cinema optics.

[Steven Hilder]

The 16:9 (1.77) Aspect ratio will only require a 27.5mm image circle

A 16:9 crop from the 5120x2700 usable area will be 4800x2700, which at 5.4 microns per pixel, equates to an image circle of 29.74mm.


I also think that with Epic's custom frame sizes most DP's (including myself) will choose to lose a few lines of res. rather than their favorite cinema optics.

I don't disagree; people are free to use whatever glass and aspect ratio they choose, I'm just keen to make all the data readily available, and as accurate as possible with the information we have right now, hence the chart (http://reduser.net/forum/showpost.php?p=411234).

[Evin Grant]

First off the S35 Mysterium-X chip is 30x15mm, native 2:1 aspect ratio. Therefore the maximum record area would be 5120x2560. I believe the 2700 number was published earlier by Red but this would not be "active" as far as I can tell. That would give a max 16:9 resolution of 4531x2560 x 4.5 microns = 24.5x13.8mm and an image circle of 27.5mm.

If you use 2700 as your record height then I see where your numbers are coming from.
However, the pragmatic approach (especially for post) would suggest if you can't achieve 5K in 16x9 to window down to the standard format of 4096x2305. 5k will be especially useful for 2.40 and 2:1 projects but my gut feeling is the majority of 16:9 and 1.85 projects (the ratio 75% of professional media is shot in these days) will still be acquired at the 4K standard.

[Humberto Rivera]

“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.”

My understanding of: The Chart, that Steven is referring to list the full sensor with the look around area as 5556 X 2780, with 15.4 megapixels, the active pixel array as 5120 X 2700, with 13.8 megapixels, that’s were he gets the strange aspect ratio of 1.896:1, and the 2:1 aspect ratio of 5120 X 2560 with 13.1 megapixels.

It’s that correct?

Humberto Rivera

[Humberto Rivera]

One more thought, Matthew gets his Image Circle Diameter of 33.54 mm from the outside of the 5556 X 2780 image area. For the full Active Pixel Array it would be a diameter of 31.26.

Humberto Rivera

[Steven Hilder]

First off the S35 Mysterium-X chip is 30x15mm, native 2:1 aspect ratio. Therefore the maximum record area would be 5120x2560. I believe the 2700 number was published earlier by Red but this would not be "active" as far as I can tell. That would give a max 16:9 resolution of 4531x2560 x 4.5 microns = 24.5x13.8mm and an image circle of 27.5mm.

The maximum record area of the chip doesn't need to follow the same aspect ratio as the chip's exterior dimensions; I can't think of any requirement for a direct relationship between the two. Also remember that both horizontal and vertical pixel dimensions will generally be divisible by 4 to account for tiling in the wavelet codec.

True, the 5120x2700 area described as "Max Resolution" in the original DSMC announcement hasn't been mentioned in a while, but I'm going to assume it's still applicable until we hear otherwise. Remember the full height of the chip will be over 2700 pixels anyway - I wouldn't mind betting you'll be able to squeeze a larger 16:9 frame (ie. closer to "5K") into that area.

[Steven Hilder]

“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.”

My understanding of: The Chart, that Steven is referring to list the full sensor with the look around area as 5556 X 2780, with 15.4 megapixels, the active pixel array as 5120 X 2700, with 13.8 megapixels, that’s were he gets the strange aspect ratio of 1.896:1, and the 2:1 aspect ratio of 5120 X 2560 with 13.1 megapixels.

It’s that correct?

Humberto Rivera

Humberto, yes, that's exactly what I'm predicting. The 5.4µm dot pitch is the spec that's least likely to change (as it's pretty fundamental to the pixel design), and Jim has referred to the 30x15mm S35 sensor several times this year, so I reckon they've finalised the physical dimensions (and therefore the full sensor's resolution), too. The only figure that might still be subject to change is the maximum recordable resolution, which was originally defined as 5120x2700 pixels. The rest is just arithmetic.

[Humberto Rivera]

Matthew, I don’t know if I getting too obnoxious with my suggestions, but they are just that, suggestions. Anyhow I was thinking that the following format might be a bit more useful, again is just a suggestion. Your web site looks great, and it will be better from week to week, thanks.

Make/Model - Focal Length
Zeiss Master Prime Dia/mm T Stop AOV

14mm T1.3 81°
16mm 34mm T1.3 70°
18mm T1.3 67°
21mm T1.3 60°
25mm 35mm T1.3 51°
27mm T1.3 48°
32mm 36mm T1.3 42°
35mm T1.3 38°
40mm T1.3 34°
50mm 40mm T1.3 27°
65mm T1.3 21°
75mm 42+mm T1.3 18°
100mm 42+mm T1.3 14°
150mm T1.3 9°
100mm Macro T2/T4.3/T32 14.02°/13.52°/12.42°

RED Pro Prime

18mm T1.8
25mm 34mm T1.8
35mm 36mm T1.8
50mm 41mm T1.8
85mm 43+mm T1.8
100mm 43+mm T1.8

Humberto Rivera

[Humberto Rivera]

Well the page I had did not translate from to what came out on the Red screen, sorry.

Humberto Rivera

[joakim_hansson]

Any chance of seeing the Cooke 15-40 T2.0 CXX Zoom in the list? :)

[Humberto Rivera]

I was thinking that the S35 could actually work utilizing the cine lenses that already exist. The FULL AREA OF THE SENSOR of the S35 is 5556 mm X 2780 mm 15.4 megapixels with a 33.55mm image diagonal, the ACTIVE PIXEL ARRAY is 5120 mm X 2700 mm 13.8 megapixels with a 31.26 image diagonal, the 2:1 RECORDING MODE is 5120 mm X 2560 mm 13.1 megapixels with a 30.91 image diagonal, and the 16:9 RECORDING MODE is 4800 mm X 2700 mm 13.0 megapixels with a 29.74 image diagonal. The 1.85:1 RECORDING MODE would be 4996 mm X 2700 mm 13.5 megapixels with a 30.67 image diagonal. All well within the range of the Master Primes and the Red Pro Primes. That would mean that current glass can continue to be used according to Matthew Image Circle measurements. All the lenses would be slightly wider than previously in the Red One due to the fact that they go out further into the frame. Just a thought!

Humberto Rivera

[Evin Grant]

That's what I've been trying to tell everyone.
Even older glass like Zeiss T2.1s will cover the 16:9 and 2.40:1 aspect ratios, even 2:1 for some of the later designs and teles.

[Steven Hilder]

Me too, it should be a brilliant format to shoot on - loads of options!
it's unlikely that any recording mode on the S35 cameras will require an image circle of more than 31mm.

[Humberto Rivera]

Good Morning to you all. I’m glad that it took Matthew Duclos Image Circle Database, which is a fine work in progress. Steven Hilder cart which outlines the basic mathematical concept behind the measurements of the S35 sensor, until we hear different from Jim. And Evin Grant persistent in the understanding of the S35 as being an alternative camera that can use most of the current glass. The better of two worlds, the high resolution of cine lenses that took some time to evolve to their present configuration, and the slightly bigger S35 sensor with its new technology, and accessories. So I would say the S35 is just as good as the FF35 for both of them to succeed. I can see a world were both cameras would serve their purpose.

Humberto Rivera

[Humberto Rivera]

Matthew, I have a question. What is the Image Circle Diameter of the Duclos Lens PL Mount 11-16 mm f2.8? Thank you.

Humberto Rivera

[Matthew Duclos]

Matthew, I have a question. What is the Image Circle Diameter of the Duclos Lens PL Mount 11-16 mm f2.8? Thank you.

Humberto Rivera

The image circle is beautiful... kidding.
That will be included in my next batch update.
Without the numbers, it covers 5k from 12mm to tele.
It just barely crops on a "5k" at 11mm. Works perfect with the RED One.

[Michael Hastings]

Matt,
Do you know if the Zeiss compact primes use aspherics on their wides to get the 43+ coverage or can they get away with spherical design because of the relatively smaller aperatures (T2.9/3.6)?

The compact primes are based on the ZF lenses (which in turn were based on the Contax/Yashica lenses)

The zeiss site doesn't have datasheets on the compact primes that show the optical configuration but if you go to the photo section and look up the z series lenses you can get individual datasheets for each lens that show the actual lenses and optical contruction.


http://www.zeiss.com/photo

This link is the .pdf file for the 21mm/2.8

http://www.zeiss.com/C12567A8003B8B6F/EmbedTitelIntern/Distagon28-21ZFen/$File/E_2.8_21ZF.pdf

[Michael Hastings]

The image circle is beautiful... kidding.
That will be included in my next batch update.
Without the numbers, it covers 5k from 12mm to tele.
It just barely crops on a "5k" at 11mm. Works perfect with the RED One.

Matt:

Do you have a way to check SLR lenses like the Canon EF-S 10-22?

Along the same lines can you project those lenses? would like to compare that Canon 10-22 to the 11-16 Tokina. I have them both but with the projector it is a little easier to be objective.

[Matthew Duclos]

Matt:

Do you have a way to check SLR lenses like the Canon EF-S 10-22?

Along the same lines can you project those lenses? would like to compare that Canon 10-22 to the 11-16 Tokina. I have them both but with the projector it is a little easier to be objective.

Of course.
I can check:
PL
B
C (threaded)
Pana
Canon EOS
Nikon
OCT-19
BNCR
Aaton
B4 (proper)
And a few others I'm sure I'm forgetting...

[Humberto Rivera]

I was just thinking the other day; the FULL AREA OF THE SENSOR of the S35 frame is 5556 mm X 2780 mm with 15.4 megapixels of coverage and with a 33.55mm image diagonal with look around within the FULL AREA OF THE SENSOR.

However the ACTIVE PIXEL ARRAY itself, what is recorded in the camera sensor is 5120 mm X 2700 mm with a 13.8 megapixels and a 31.26 Image Diagonal, with the strange Aspect Ratio of 1.896:1 that Steven Hilder gave the sensor the name of the X-35 Aspect Ratio as he calls it.

So let’s say that the ACTIVE PIXEL ARRAY of 31.26 is the maximum area that is required to cover the Lens Image Circle outline by Matthew in his Diameter Database. Of course never going over the 31.26 mm Diameter of the S35 or S35-X.

Now the 1.85:1 could be cut off from the top of the frame (just slightly) and take it to the edge of the 5120 mm frame area leaving just a bit more for the lenses to go wider on its coverage, the megapixels should be similar from leaving the 2700 mm of the top (possibly with a slight change, but not much).

Now for the smaller 2.40:1 or the 2:1 frame area it’s just a matter of cutting the top of the 2700 mm frame area, but keeping the width of 5120 mm area active. The 16.9:1 can be arranged accord ling to fit within the ACTIVE PIXEL ARRAY.

So it can actually work with existing High Quality Cine Lenses. All the Aspect Ratio fit well within the Red Pro Primes (25 mm at 34 mm of diagonal size and bigger) and the Arri/Zeiss Master Primes (16mm at 34 mm of diagonal size and bigger). It’s all about how far the Cine lenses can go out to cover the 5120 mm frame area, and that’s where Matthew Lens Image Diagonal Circle Chart of the Diagonal Database comes in.

The more I think about it, the more convincing this option it’s the better of two worlds. I was just thinking out loud. One more stop of ISO/ASA, a larger sensor still at 5.4 microns, and the new technology of modular assembly all wrap into one package. It’s like the Red One evolved into the S35 with all of its evolutionary advancements, but the lenses, tripods, lighting and all the other things remained the same, just pull the Red One out and put in the S35 or S35-X.

Of course as Jim likes to say; "Everything in life changes... including our camera specs and delivery dates..."

Humberto Rivera

[Fredrik Callinggard]

Don't know if it has already been mentioned (my apologies if it has), but this thread should become a sticky!

Thanks for all the great info, this one is a keeper.

[Steven Hilder]

...with the strange Aspect Ratio of 1.896:1 that Steven Hilder gave the sensor the name of the X-35 Aspect Ratio as he calls it.

I just want to make it clear that I did not come up with the term 'X-35', and I do not recommend that anyone else use this term to refer to RED's upcoming S35 sensor format. The single most important specification concerning lens coverage requirements for the new sensors is the image circle / diagonal measurement. This is why I am against using 'X-35' or any other ambiguous label, and why I am very grateful that Matt Duclos is sharing his findings online in his Image Circle Database.

[Humberto Rivera]

Steven, my sincere apologies for saying that you did! But someone did, here is the proof.

Humberto Rivera

[Steven Hilder]

Steven, my sincere apologies for saying that you did! But someone did, here is the proof.

No worries, Humberto :smiley:

Yes, several people in this thread (http://reduser.net/forum/showthread.php?t=28932) thought that 'X-35' would be a good label to describe the 5K DSMC S35 30x15mm sensor - I, for one, hope it doesn't catch on.

[Humberto Rivera]

I just wanted to set the record straight by using the following quote; “Let's try to use the phrase "acts like" instead of "is". A 50mm is always a 50mm, no matter what the sensor size. A 50mm can "act like" a 40mm on an X35 size sensor.” Jim Jannard

The X35 label to describe the ACTIVE PIXEL ARRAY of the S35 seems just as good as any other, it’s got an Aspect Ratio of 1.896:1, call it what you may. It’s up to Red to name it what it wants, because they created it. You got to call it something, and thinking about some more, it makes sense to shoot the full ACTIVE PIXEL ARRAY.

The sensor is wider, and it’s covered by the current Pro Lenses, and it will yield a wider frame that where’re accustomed too. It was Corrado Silveri that came up with the X35 drawing, and Jim who suggest the X35 designation with lots of help from everyone else.

Steven Hilder said; “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.”

Nonetheless is still has Aspect Ratio of 1.896:1.

Humberto Rivera

[Humberto Rivera]

Matthew, from time to time I enter your web page; http://web.me.com/ducloslenses/DataRef/ImgCir.html to see if it was up-dated. I’m only just asking, when do you think that it would be up-dated? Or is there another web page we should be look at, I might have the name wrong? Don’t get me wrong, it’s a wonderful idea to do the database. Just Asking! Keep up the great work you are doing. Thank you!

Humberto Rivera

[Peter Plevritis]

This is a useful thread. But unless I am just missing it - what about the Red 18-50mm image circle? Will it work with an 5K EpicX?

Thanks
Pete

[Humberto Rivera]

I was just thinking. If this is what the Arii/Zeiss Master Prime see’s at the front end of the X35/S35 camera sensor, and they are all over an Image Circle of 33.54 then what we see at the other side of the lens is over 34 mm Diameter, for both the Red Pro Primes and the Master Primes will do just fine. The only difference is that it would be slightly wider from the Super 35 mm Film Camera to and X35/S35 Camera Sensor, thus slightly wider lens. Matthew keep-up the good work!

The is one question, what about the NEW Arri/Zeiss Distagon T2.8 Ultra Prime 8 mm R lens, what is the Image Circle for that one, it certainly a beauty. It makes the very close up shot look almost cartoonish without any distortion, with is a desirable property these days. I suspect that it will probably be ideal for the X35/S35 sensor, but the jury still out on that.

Humberto Rivera

[Humberto Rivera]

Matthew, congratulations on the ongoing improvement to your web page; http://web.me.com/ducloslenses/DataRef/ImgCir.html it’s an indispensable tool for all of us to look at from time to time. For example what is the true Image Diameter of the new Cooke Super Speeds 5/i lenses, I know they are not out yet, but when they do come out it would be nice to know. And what about the NEW Arri/Zeiss Distagon T2.8 Ultra Prime 8 mm R lens. Thank you!

Humberto Rivera

[Matthew Duclos]

I'll be adding a bunch this upcoming week.

[Humberto Rivera]

Matthew, like always your Image Circle Database is an invaluable resource that we can all turn too. Thank you!

I don’t understand why lens manufacturer don’t provide this information, knowing that the new mark to meet is at least 32 mm of diagonal image coverage for the full ACTIVE PIXEL ARRAY (31.26 Image Diameter) of the S35 Camera Sensor or 33.5 mm like you like to put it. If it meets that it will certainly cover the super 35 Film Image. The 2:1 image has a diagonal of 30.91 mm.

The Red Epic S35 is the new mark for every lens manufacturer to meet; it’s an important part of the Film Culture. We know that the Red Pro Primes and the Master Primes will do just fine. We are all waiting to confirm if the New Cooke 5/i Lenses will also cover this image circle, as well as the NEW Arri/Zeiss Distagon T2.8 Ultra Prime 8 mm R lens.

The Red Epic S35/X-35 is an important part of the Film Culture. That’s a significant statement, were using all the tools of Filmmaking with the new S35/X-35, the two worlds are slowly merging into one, they’re interchangeable. The lens manufacturers need to adapt to the new measurements if they want to survive.

Matthew, we’re all looking forward to your next update.

Humberto Rivera

[Michael Panfeld]

Matthew: can you please explain how you are measuring the image circles? Specifically, what equipment are you using. I ask because I have a bunch of cine lenses (and still lenses) that are not currently in your database and, if I can understand how to measure, I might be able to help add to the database.

I have a Richter autocollimator, but still am not totally familiar with it. I don't have a projector yet.

Thanks

[Humberto Rivera]

Thanks’ Matthew, your web page keeps on improving: http://web.me.com/ducloslenses/DataRef/ImgCir.html

Now that we had a peek at the new Red Epic X35/S35 your web page becomes more important, the actual real Image Circle needs to be covered without any real lens degrading. The database needs to show that a lens has at least the FULL ACTIVE PIXEL ARRAY of the X35/S35 frame of 31.26 Image Diameter. Eventually the Actual Camera will be on the market, now is the time to think about the glass.

We now know that the Image Circles in order to fit within the S35/X35 Image Circle has to be at least:

The FULL AREA OF THE SENSOR of the S35 frame is 5556 mm X 2780 mm with 15.4 megapixels of coverage and with a 33.55mm image including look around.
The FULL ACTIVE PIXEL ARRAY of the S35 Camera Sensor has a 31.26 mm Image Diameter with an Aspect Ratio of 1.896:1 and its 13.8 megapixels.

The 1.85:1 ASPECT RATIO is slightly less that the 1.896:1 Aspect Ratio, or if you want to go the other way it could also be 13.5 Megapixels, with a 30.67 Image Circle, and a 4996 mm X 2700 mm image.

The 2:1 IMAGE ARRAY has to be at least 30.91 mm covering the 5120 mm X 2560 mm with a 13.1 Megapixel area.

The 2.40:1 RECORDING AREA is 29.95 mm covering the 5120 mm X 2132 mm in a 10.9 Megapixels area.

Thanks’ Matthew, always looking forward to your new current updates! Thanks’ also to Steven Hilder for his chart.

Humberto Rivera

[Humberto Rivera]

Matthew I notice a little change in your Website with the introduction of appears to be a difference between the FULL AREA OF THE SENSOR and the FULL ACTIVE PIXEL ARRAY of the sensor. Keep-Up the good work!

Humberto Rivera

[R. Gonzales]

I feel bad for taking over the other thread that this was being discussed on.
So I figure I would drop it in a new thread for all to see.
I update the list as lenses pass through the shop here.
If there are any specifics you would like, just let me know and maybe I can throw one up to test.

Click Here! (http://web.me.com/ducloslenses/DataRef/ImgCir.html)

Hello Matthew,

Do you have the image circle of the Tokina 11-16mm? and since I am asking questions how much does it cost? you could always pm me with a price if that is ok?

best

[David Rasberry]

Haven't seen Matt post in while. He must be very busy hacksawing the backends off of Tokinas. :smile:

[Matthew Duclos]

Haven't seen Matt post in while. He must be very busy hacksawing the backends off of Tokinas. :smile:

Indeed. Among other inhibitors.

[Shawn Nelson]

Haven't seen Matt post in while. He must be very busy hacksawing the backends off of Tokinas. :smile:

Hey I have the trademark on applying the words "hacksaw" towards Duclos work! :sifone:

[Matthew Duclos]

Shawn is the inspiration behind this pic from a couple months back...
http://farm4.static.flickr.com/3659/3552067921_d7b0f92b3d_o.jpg

[Stephen Williams]

as well as the NEW Arri/Zeiss Distagon T2.8 Ultra Prime 8 mm R lens..

Humberto Rivera

Don't hold your breath, it portholes really badly wide open on S35 film.

[Joshua Brown]

Just wanted to drop a line letting folk know that Matt has been kind enough to get the database on the Wiki: http://red-wiki.net/wiki/Lens_Image_Circle_Comparison

Would love some more info there about how tests are/were performed as well as any lenses that you all have tried and positively know stats on.

-J.A.Brown

[Shawn Nelson]

Shawn is the inspiration behind this pic from a couple months back...
http://farm4.static.flickr.com/3659/3552067921_d7b0f92b3d_o.jpg

Haha! I love it! It needs me looking at lens with a thought bubble going up to that picture

[albert rudnicki]

Super Speeds coming soon.
I've had several sets through the shop, but they were rush jobs..
Not enough time to take a good reading.

Matthew,
Do you have any idea if superspeeds would cover fully Epick's 5K?
Thank you again:)

[albert rudnicki]

can anyone chime in please.
thanks

[Derek Wan, HKSC]

Matthew,

Can you get a set of Illumina S35 lenses in to measure their image circle dimensions? When you have a chance, can you add IC info of other lenses to your database to benefit the community even more.

By the way, I have reserved a Duclos 11-16mm lens for more than three months now, when can I pay a deposit or even pay in full? When should I expect delivery?

Thanks.

[Matthew Duclos]

Matthew,
Do you have any idea if superspeeds would cover fully Epick's 5K?
Thank you again:)

No. They will not.

[Chris Parker]

is there any lens out there that will cover the full 5K, INCLUDING the widest length.

[Michael Panfeld]

Matthew: Thanks for this service to the community. I don't see any LOMO lenses in the mix. Do you not get them in your shop or is there some other reason the data is not in your database?

I have about 40 or so LOMO's. If you can tell me how to accurately measure the image circle, perhaps I can augment the database.

Thanks

[Matthew Duclos]

is there any lens out there that will cover the full 5K, INCLUDING the widest length.

Zeiss Compact Primes.
RPP should cover 5K I think.
Everyone keeps asking if their lenses will cover 5K.
Im gonna wait to answer these questions until RED rolls out the actual camera.
Specs change...

[Matthew Duclos]

Matthew: Thanks for this service to the community. I don't see any LOMO lenses in the mix. Do you not get them in your shop or is there some other reason the data is not in your database?

I have about 40 or so LOMO's. If you can tell me how to accurately measure the image circle, perhaps I can augment the database.

Thanks

This is VERY difficult because there are so many different versions of Lomos and very little way to identify them without actually seeing them.
We see Lomos here fairly often but they are just too random and scattered.
In fact I have the MONSTER LOMO sitting in our vault.
It's a 25-500mm f/3.5 that weighs about 35 lbs. lol

[Michael Panfeld]

This is VERY difficult because there are so many different versions of Lomos and very little way to identify them without actually seeing them.
We see Lomos here fairly often but they are just too random and scattered.
In fact I have the MONSTER LOMO sitting in our vault.
It's a 25-500mm f/3.5 that weighs about 35 lbs. lol

Huh? I'm confused. They all have unique model #'s engraved on each lens (a model starts with OKC + design # + focal length + version). Its very clear how to identify them. For example:

OCK1-100-1 is a 100mm f2.3
OKC7-28-1 is a 28mm f 2.0
OKC10-28-1M is a 28mm f1.4
OKC1-300-1 is a 300mm f4.0
OKC12-35-1 is a 35mm f1.5
OKC11-35-1 is a 35mm f2.0
OKC1-50-6 is a 50mm f 2.0

All lenses within each model # should have the same size image circle. Each model is produced according to a strict Russian standard put out by GOST, the Russian Standards agency (similar to the U.S.'s SAE or ASTM standards). Are you saying that you have had two lenses of the same unique model # and that they had different image circle sizes?

Nevertheless, if you can please explain how you are measuring the image circle, I will do it myself, post it, and people can take it for what its worth.

Thanks

[Matthew Duclos]

I believe I was not clear enough in my explanation.
I'm aware that each lens has a unique serial number as with every manufacturer.
My predicament is in the identification on the list.
For example, it is easy to create a section for Zeiss Super Speeds.
For Lomo lenses there is, to my knowledge, no nomenclature for the Lomo lenses other than user adopted "square front", "round front" and production date.
If I am missing something in the Lomo brand that would allow me to better identify the lenses, please help me help you.
To further illustrate my dilemma I propose an example:
If I were to list a 25mm Lomo, there isn't really any way to determine which 25mm I would be referring to in the Lomo family other than its optical properties such as aperture speed, square/round front anamorphic, etc.

[Matthew Duclos]

I measure the effective image circle with a projector and a precisely marked reticle.
My measurements are relative to specific marks on the reticle.
There is usually a certain level of light loss around the edges of the image circle to which I determine a "usable" range based on my experience and measure from there.

[albert rudnicki]

No. They will not.

Thank you Matthew !

[Michael Panfeld]

For Lomo lenses there is, to my knowledge, no nomenclature for the Lomo lenses other than user adopted "square front", "round front" and production date.
If I am missing something in the Lomo brand that would allow me to better identify the lenses, please help me help you.
To further illustrate my dilemma I propose an example:
If I were to list a 25mm Lomo, there isn't really any way to determine which 25mm I would be referring to in the Lomo family other than its optical properties such as aperture speed, square/round front anamorphic, etc.

Sorry for the delayed response, but we had a bit of a snowstorm on the East Coast. Yes Matthew, there is a very specific nomenclature.

Unlike most cinema lens manufacturers, LOMO has a very large number of models. Indeed, I'd say that LOMO was the most prolific cinema lens manufacturer. Not necessarily in terms of the sheer number of lenses produced (although that may be true), but in terms of the number of models. LOMO lens models are not "rolled out" for all focal lengths as part of a series (e.g., Cooke S4s or Zeiss Superspeeds). Rather, each focal length has its own series of optical designs and each design usually has multiple versions and these were rolled out on an ad hoc basis over time.

I have been documenting LOMO lens models and have a list of LOMO 35mm cine primes alone that approaches 90 different lens models. That does not include the dozen+ 35mm zooms and the 10-20 anamorphic primes. And that is just 35mm cine use. There is a whole second set of primes and zoom lens models for 16mm usage.

So no, you are correct that you cannot just toss up on the list 25mm and have it make sense, any more than you can simply add Cooke 35mm and not distinguish between Cooke S4 35mm, Cooke S3, 35mm, Cooke S2 35mm, etc...

The nomenclature needs a bit of explanation. First of all, it uses the cyrillic alphabet. Some people use the characters as they are written, while some translate these using the sound that they make. For example, cinema primes have OKC imprinted on the lens. But, because a "c" in Russian is pronounced like an "s" some people identify the lens as OKS in postings online, ebay, etc. Once you get past this confusing part, the rest is pretty standard (Ha!).

In terms of the nomenclature, first note that lens model numbers are different from the lens' serial number. The serial # would appear as 8700012 (where the first two digits are the year of manufacture and the remainder is a serial number - so that would be the 12th lens of that model produced in 1987). The serial number restarts each year and for each model.

The lens model # for spherical primes is simplest to decipher and consists of several parts:

Part 1: indicator of film type (16 or 35mm). So it will have a 16 or a 35 as the first two digits with some caveats: first, lenses manufactured by EKRAN have the 35 printed while spherical prime lenses manufactured by LOMO leave this blank (anamorphic primes made by LOMO always start with 35). If the lens is intended for 16mm film, then the Part 1# always starts with 16 regardless of manufacturer.

Part 2: indicates the type of lens

OKC means spherical prime

Part 3: Optical Design

a 1 or 2 digit number that indicates the optical design. These follow a chronology so that 1 means the first (and earliest design, whereas a larger number like "12" indicates a later design). Later designs are not necessarily better.

Part 4: focal length

self-explanatory. The focal length is both preceeded and followed by a dash

Part 5: series number

there may be multiple iterations within an optical design as the Soviets were always improving and experimenting. Note that if there is an "M" at the end, this indicates a "fast" aperture. However, not all fast lenses had this additional designation.

So with this its pretty easy to decode:

OKC1-22-1 is a sperical prime made by LOMO, of the 1st optical design, of a 22mm focal length, in its first version. All OKC1-22-1's will have the same image circle. However, it may or may not have the same image circle as other 22mm LOMO spherical primes (e.g., OKC3-22-1, OKC5-22-1, or OKC6-22-1).

Zooms and anamorphic models have a similar nomenclature, except that it does not have Part 4 (i.e., the focal length) in the model#

Part 1: 35 or 16

Part 2: indicates the type of lens

OPF means spherical zoom
BAC means anamorphic prime (round front or square front - you just have to know which is which based on the design number)
HAC means anamorphic prime (old school 2-piece square front designs)

Part 3: unique design #

Part 4 - version - always 1

So 35 OPF-18-1 means a 35mm film zoom lens of the 18th design. The 18th design happens to be a 20-120mm zoom.

35BAS22-1 means a 35mm film round front anamorphic, series 22. Series 22 = 50mm focal length.

There may be multiple anamorphics and zooms of the same focal lengths, but each would have a unique design number. For example:

35 OPF-7-1 and 35 OPF-15-1 are both 25-250mm zooms.

So no, you can't just say LOMO 28mm and the image size. However, you can say OKC12-28-1 and give the image size and everybody with that model can rest assured that theirs has the same image size, but those with say a OKC4-28-1 will not expect the same image circle size.

I realize that posting LOMO image circles on your list may require either an asterisk with an explanatory note or an extra column to indicate the focal length and the model #. I'll be glad to help. Let me know. Thanks

[Nick Gardner]

Michael is completely accurate and correct in the above explanation. I would just like to add, that the same lens (ie OKC1-22-1) may be in any number of different housings. They were housed differently for different cameras, and different uses (studio versus doc/news). So you can have three completely different looking lenses that are optically identical.

Nick

[Matthew Duclos]

Michael,

Your 30 paragraph explanation is a perfect example of why I haven't even begun to list Lomo lenses.
I may start to do so, but in a separate page so that the page length isn't 10,000 px. ;)

[Ryan S]

Now, here's a curiosity.

I was reading the Cooke website, specifically about the new Panchros. They're advertised as having a maximum coverage of 30mm, which is slightly smaller than required for 5K 2:1, as I understand it.

I then looked at the page for the S4s, and they are advertised as having the same maximum coverage. Contrary to this, Matt's database shows the majority of S4s as having an image circle of more than 30mm.

Matt, can you shed any light on this discrepancy? I have no doubt that you're conducting your image circle tests properly, and yet this information is coming right off of Cooke's website. I'm just hoping that this means the Panchros (and the 5is) might actually cover some form of 5K, which would be great.

[Matthew Duclos]

Now, here's a curiosity.

I was reading the Cooke website, specifically about the new Panchros. They're advertised as having a maximum coverage of 30mm, which is slightly smaller than required for 5K 2:1, as I understand it.

I then looked at the page for the S4s, and they are advertised as having the same maximum coverage. Contrary to this, Matt's database shows the majority of S4s as having an image circle of more than 30mm.

Matt, can you shed any light on this discrepancy? I have no doubt that you're conducting your image circle tests properly, and yet this information is coming right off of Cooke's website. I'm just hoping that this means the Panchros (and the 5is) might actually cover some form of 5K, which would be great.

Not sure why Cooke would say that...
It may very well be a matter of opinion. They may consider 30mm the acceptable image circle.
I measured to what I considered reasonably usable in regards to light and focus falloff.
They may also want to keep things simple and not list a mere 2mm difference between some lenses and not others.
Just my guess.

[Ryan S]

Not sure why Cooke would say that...
It may very well be a matter of opinion. They may consider 30mm the acceptable image circle.
I measured to what I considered reasonably usable in regards to light and focus falloff.
They may also want to keep things simple and not list a mere 2mm difference between some lenses and not others.
Just my guess.

Well, then I look forward to somebody doing some tests with the Panchros and 5is on 5K. Hopefully you'll get to test one or both sets in the reasonable future.

[Richard Bradbury]

Has anyone measured the image circles of Canon K-35s? I know most of them (24, 35, 55, 85mm) are said to cover FF-35, but I'm specifically interested in the circle size of the 18mm T1.5.

[Steven Hilder]

If anyone's interested, I've updated my chart with the specs that Jim has confirmed (http://reduser.net/forum/showpost.php?p=560105).

Hopefully this and Matt's database (http://web.me.com/ducloslenses/DataRef/ImgCir.html) will be helpful for those buying new lenses for use on MX sensors.

http://img151.imageshack.us/img151/2553/mysteriumx.png

[Ivan G]

5k 2:1 is larger then 5k 16x9?

[Mark Andersen]

I don't see 5KFF on the list, but I think it's close to "active pixel array". I am shooting this a lot so I can also extract high quality stills from the footage.