Emulation of Super 8 on a digital slr

[Christopher Guerrero]

Anyone have a good idea how I could be able to emulate certain Super 8 conditions on my Olympus E-300 Digital SLR. I have a Canon 814xl-s no external meter yet. But I think the shutter speed at 220 degrees comes out at about 1/40th of a second. And the stock is about 40iso for kodachrome I suppose. How would I be able to punch those certain specs into a SLR for gaining exposures before I shoot with film. Any tips or ideas on this would be a great help. Thanks

[Nate Downes]

Physically impossible. Your DSLR doesn't have the resolution. With a Nikon D2X you might be able to pull it off for resolution, but it doesn't have the color-depth nor contrast needed for K40. You can approximate it by pushing your saturation sharply yellow and pushing your contrast to maximum, but then you'll blow our your highlights badly, ruining the effect. A digital sensor just can't hack it.

[Christopher Guerrero]

thanks I appreciate the info. But I'm not trying to capture an effect just the exact exposure before I began shooting scenes on Super 8. It doesnt have to look wonderful just close to what the exposure would be on my camera.

[Filip Plesha]

Physically impossible.  Your DSLR doesn't have the resolution.  With a Nikon D2X you might be able to pull it off for resolution, but it doesn't have the color-depth nor contrast needed for K40.  You can approximate it by pushing your saturation sharply yellow and pushing your contrast to maximum, but then you'll blow our your highlights badly, ruining the effect.  A digital sensor just can't hack it.

<{POST_SNAPBACK}>

 

 

What are you talking about? A DSLR doesn't have the resolution of 8mm Kodachrome?

[Nate Downes]

What are you talking about? A DSLR doesn't have the resolution of 8mm Kodachrome?

<{POST_SNAPBACK}>

 

Technically, perhaps, but realistically it lacks the lattitude, even against Kodachrome, which doesn't have much lattitude for film. I did a test, shooting a standard 8mm frame with a D2X using a slide copier, and the D2X still pulled out more details of that old Kodachrome than the D70 on the same frame of film, which says that the 8mm frame does have more resolution than a consumer DSLR.

 

But, that's an arguement for another day.

[Nate Downes]

thanks I appreciate the info. But I'm not trying to capture an effect just the exact exposure before I began shooting scenes on Super 8. It doesnt have to look wonderful just close to what the exposure would be on my camera.

<{POST_SNAPBACK}>

 

Oh, ok! Misunderstood your question.

 

Yes, 1/40th of a second, but the Olympus doesn't go to that fast of an ISO. Set your ISO to 200 and use some ND filters, ND 2 should get you to pretty close what you'd get with the Canon.

[Filip Plesha]

Technically, perhaps, but realistically it lacks the lattitude, even against Kodachrome, which doesn't have much lattitude for film.  I did a test, shooting a standard 8mm frame with a D2X using a slide copier, and the D2X still pulled out more details of that old Kodachrome than the D70 on the same frame of film, which says that the 8mm frame does have more resolution than a consumer DSLR.

 

But, that's an arguement for another day.

<{POST_SNAPBACK}>

 

 

well then, remind me to use 8mm next time I want to make a 10x15

[Filip Plesha]

If you take a look at the MTF curve you can see that it reaches about 90 lp/mm when the contrast of the lines gets consumed by grain structure.

 

When you put that together with the lp/mm of the BEST lenses at center, you get something around 70lp/mm, which is something around 1k for super8 frame

 

Are you saying your Nikons can't even reach true 1K?

Edited August 5, 2005 by Filip Plesha

[Nate Downes]

If you take a look at the MTF curve you can see that it reaches about 90 lp/mm when the contrast of the lines gets consumed by grain structure.

 

When you put that together with the lp/mm of the BEST lenses at center, you get something around 70lp/mm, which is something around 1k for super8 frame

 

Are you saying your Nikons can't even reach true 1K?

<{POST_SNAPBACK}>

 

I'm just going by what I found, a D2X pulled more information out of it than a D70 did. It's an impractical system in any case, waste of a D2X. 8)

[Filip Plesha]

I do believe what you saw, I'm just saying you are misinterpreting it.

 

Scanning film and beating it in resolution is not the same thing.

It takes a lot more resolution to scan film than to capture image with same resolution.

 

 

I'll explain your mistake on example of lenses.

 

let's say you have a lens that can resolve 800lp/mm and one that can resolve 300lp/mm at some specific aperture in the center.

 

You test both of them with the same film. You get better results with a 800lp/mm lense, and you conclude that your film must be able to capture more information than 300lp/mm

 

wrong.

 

Even if your film captures 50lp/mm maximum, using 800lp/mm lense will give improvement, why?

 

 

because the formula for resolwing power of all systems is:

 

1/(resolving power of component 1) + 1/(resolving power of component 2) + ...... = 1/(resolving power of full system)

 

 

So, a 100lp/mm film in combination with 100lp/mm lens will produce a system that can resolve 50lp/mm, and not 100

 

 

It took the resolution of your DSLR to capture the detail from the super8 frame because each component of this complicated capture system reduces resolution, and diagonal detail can only be sampled with higher frequency than the one of film.

 

But that does not mean that your dslr can not capture the same amount of detail from actual scene. In fact, it can probably capture 3 times more.

[Nate Downes]

You missed my point, I shot it twice, with two different DSLR's, a 6MP and a 12MP. Same lens on both. How about I post the samples for you?

Edited August 5, 2005 by downix

[Filip Plesha]

no, I didn't miss your point.

 

two different sensors are two different components in the equation.

 

 

But please do post the examples

[Filip Plesha]

let me put it this way:

I'll use film terms only

 

this is your logic:

if you copy an 8mm frame to 35mm, you will get more detail than if you copied it to another 8mm frame, therefore 8mm frame must have resolution similar to 35mm

 

which of course is not true.

 

35mm would probably be enough to capture every little detail from 8mm in copying.

But copying 8mm to 8mm will get you half the resolution you had already.

 

 

same goes for film to digital, or digital to film, whatever combination you like.

 

scanning film is just another form of analog image copying. You have to use a DSLR that has 3-4 times more resolution, to accuratly copy a small 8mm frame.

That does not mean they have equal resoluving power.

 

Copying medium MUST have at least 2 times more resolving power to make a faithfull copy

 

Your tests prove only that high-quality Nikon DSLR is good enough as a copying medium to sample a 8mm frame.

 

here is the explanation:

 

let's say that there is a sharp line in nature, which gets photographed and is recorded as a blurry line on film. In order to copy it to another piece of film (or digital capture) so that it stays the same, you need much more resolution. why?

because if you had the same resolving power as the original piece of film, your blurry line would turn out even more blurry, or it would dissapear compleatly. why?

because for this resolving power it takes a sharp line to "draw" a soft line on film, and your original piece of film does not have a sharp line but a soft line, therefore, the copy will have an even softer line. This is not due to optical problems, or drawbacks in technology, it is the nature of all capture mediums.

 

that is why print films and intermediate films are always made so that they can have many times the resolution of avarage camera negative film (they say it's usually 3 times more for print stock, but I don't know this). If they had the same resolution

as camera stock, even the perfect contacts print would lose resolution.

 

To get to the point:

super8 frame has around 1K reoslution in best conditions, but to get a PERFECT copy you need 2-3 times more resolution. That is why your better DSLR still makes a difference.

[Stephen Williams]

that is why print films and intermediate films are always made so that they can have many times the resolution of avarage camera negative film (they say it's usually 3 times more for print stock, but I don't know this). If they had the same resolution

as camera stock, even the perfect contacts print would lose resolution.

 

To get to the point:

super8 frame has around 1K reoslution in best conditions, but to get a PERFECT copy you need 2-3 times more resolution. That is why your better  DSLR still makes a difference.

<{POST_SNAPBACK}>

 

 

Hi,

 

IMHO Intermediate and print films have little in common resoloution wise!

 

The reason people scan the negative and not a print in telecine is there is far more detail in a low contrast Negative than in any print.

 

Stephen

[Filip Plesha]

what does that have to do with what we are discussing here?

[Stephen Williams]

what does that have to do with what we are discussing here?

<{POST_SNAPBACK}>

 

 

Hi,

 

I was disagreeing with your claim that print stocks had 3 times more resoloution than negative stocks. A print will never have the resoloution of negative.

 

Stephen

[Jonathan Spear]

"You missed my point, I shot it twice, with two different DSLR's, a 6MP and a 12MP. Same lens on both. How about I post the samples for you?"

 

Please do. This is an interesting thread.

[David Mullen ASC]

Hi,

 

I was disagreeing with your claim that print stocks had 3 times more resoloution than negative stocks. A print will never have the resoloution of negative.

 

Stephen

<{POST_SNAPBACK}>

 

A print off of a negative will have less resolution than the negative, but the print STOCK may possibly have more resolution than the negative stock, being so slow-speed, but of course, you'd have to photograph the image into the print stock directly in the camera, with all the attendant problems (lack of color mask, lack or remjet, lack of speed, etc.)

[Filip Plesha]

A print off of a negative will have less resolution than the negative, but the print STOCK may possibly have more resolution than the negative stock, being so slow-speed, but of course, you'd have to photograph the image into the print stock directly in the camera, with all the attendant problems (lack of color mask, lack or remjet, lack of speed, etc.)

<{POST_SNAPBACK}>

 

 

Yea, the print stock itself has greater resolution to compensate for the loss of resolution in copying. If it had same resolution as camera film, a lot of resolution would be lost. Any copying material (print or intermediate) has to have more resolution than the original medium in order to capture all its detail.

 

 

Here is how it is in numbers.

 

if the original film stock captures say 100lp/mm, and you use print film that can capture 10 times more (1000lp/mm), you would end up a copy degraded down to about 90lp/mm.

And now consider the reality, print stocks can probably go up to 200lp/mm or something like that, which degrades 100lp/mm film down to a 66lp/mm copy.

That is why print stock MUST have more resolution, otherwise the copying process would be ever worse than it is now.

[Stephen Williams]

if the original film stock captures say 100lp/mm, and you use print film that can capture 10 times more (1000lp/mm), you would end up a copy degraded down to about 90lp/mm.

And now consider the reality, print stocks can probably go up to 200lp/mm or something like that, which degrades 100lp/mm film down to a 66lp/mm copy.

That is why print stock MUST have more resolution, otherwise the copying process would be ever worse than it is now.

<{POST_SNAPBACK}>

 

 

Hi,

 

Where do your figures come from?

 

Stephen

[Filip Plesha]

Hi,

 

I was disagreeing with your claim that print stocks had 3 times more resoloution than negative stocks. A print will never have the resoloution of negative.

 

Stephen

<{POST_SNAPBACK}>

 

and a print in theory can not have the full resolution of the negative unless the resolving power of print stock is infinite, which is impossible, or unless you use a larger format to print.

 

that's not the point here. I was talking about technical capabilities of the print stocks itself. If you used a laser recording system to record test lines on it or something, you would be able to record more resolution than on a camera film.

[Stephen Williams]

and a print in theory can not have the full resolution of the negative unless the resolving power of print stock is infinite, which is impossible, or unless you use a larger format to print.

 

<{POST_SNAPBACK}>

 

 

Hi,

 

So we agree on that!

 

Stephen

[Filip Plesha]

Hi,

 

Where do your figures come from?

 

Stephen

<{POST_SNAPBACK}>

 

 

from this formula (I've already posted it a few times):

 

1/(resolution of component #1) + 1/(resolution of component #2)=(resolution of the system)

 

for example:

 

you have a lens that captures 400lp/mm and film that captures 150lp/mm.

You calculate the maximum resolving power of the combination of these two with this formula.

 

1/400 + 1/150 = 1/ x

 

x= about 110

 

Same can be applied with two (or more) different film stocks in process of contact printing. In case of optical printing you must include the resolving power of the printer lens too.

 

 

try it yourself, and you will see what I'm talking about.

[Stephen Williams]

from this formula (I've already posted it a few times):

 

1/(resolution of component #1) + 1/(resolution of component #2)=(resolution of the system)

 

for example:

 

you have a lens that captures 400lp/mm and film that captures 150lp/mm.

You calculate the maximum resolving power of the combination of these two with this formula.

 

1/400 + 1/150 = 1/ x

 

x= about 110

 

Same can be applied with two (or more) different film stocks in process of contact printing. In case of optical printing you must include the resolving power of the printer lens too.

try it yourself, and you will see what I'm talking about.

<{POST_SNAPBACK}>

 

 

Hi,

 

The Formula I accept. Where does your data for the film stocks come from ?

I have been looking at Kodak's site and don't see those numbers!

 

Stephen

[Filip Plesha]

I was just using general examples to explain what is going on in the process of copying, I haven't been using specific examples from actual MTF curves.

 

But since you brought it up, here is the actual data form some films from Kodak's site.

 

here is an example of a slow film and a fast film

 

EXR 50D: goes over 200lp/mm before the response hits 10% (not suprizing considering the speed, many slow still films from the past could capture that much resolution)

 

Vision2 500T: not charted after 80lp/mm, but it looks like the curve would fall to 10% somewhere around 120lp/mm, which is typical.

 

it seems Kodak doesn't measure beyond 80lp/mm for any of their films, which makes those MTF charts almost useless for extracting maximum resolving power.

 

As for lenses, there is a natural limit of resolving power of glass accoarding to two different criteria, one is Rayleigh criteria, the other is Dawes criteria.

they are similar though. In theory a perfect lens (pure glass with no optical problems) can at f4 resolve 500lp/mm accoarding to Dawes, or 410lp/mm accoarding to Rayleigh.

 

Needless to say, only the best lense reach this theorethical resolving power