27 replies to this topic

[randapex]

I'm curious as to how the reproduction ratio would be figured on the following photo. Any comments regarding the lighting (or anything else for that matter) would be apprecitated as it's an area I'm trying to improve in.
The Dorid is about 1.5 inches long. I used two macro lenses on my Tetra housing and a snoot on the strobe. This is the full image.
Thanks,
Rand

[james]

I think that's a beautiful shot Rand.

Eric and I have had many "discussions" on this topic. My opinion is that something can be called 1:1 if it is 36mm wide and 24mm tall. That is a "traditional" film definition.

But what do you call it when your sensor is only 11mm across????

The nudi does look a little bit soft at the lower right to me though. Controlling DOF for this kind of reproduction ratio is really difficult. That's why it's important that you can tell the camera EXACTLY where to focus - not easy to do w/ a consumer cam.

I usually try to make sure foreground elements are in focus so focus on something in the front 1/3 of the frame as a bare minimum.

Cheers
James

[echeng]

Here we go again!

I maintain that a reproduction ratio of 1:1 is when the image projected on the sensor is life size. That means that you should be filling the frame with an area equal to your sensor size. The "traditional" film definition that James is claiming only applies to 35mm film.

Otherwise, it is not a defining definition. Let's say we use James' definition:

I fill my frame with an image 36mm by 24mm. But then I mask out portions of the sensor, so only the center half is exposed. Now I claim that I have just recorded a 2:1 image. That's just plain wrong. I could immediately go back and mask out 99% of the image, and claim that I'm shooting at 100:1.

This is what the crop factor does, by the way. It effectively masks out the sensor. Just because you crop out XX% of the image doesn't mean that you can decide that the reproduction ratio is greater.

The only definition that doesn't lead to confusion is that 1:1 means a life sized projection onto film or sensor plane.

[scorpio_fish]

I think I would go with Eric's definition. Don't forget, although we talk about 1:1 from our 35mm film days, film isn't just 35mm (36mm x 24mm). A 4" x 5" object on a 4x5 camera is 1:1. I'm pretty darn certain this is how macro lenses and reproduction ratios are defined in the MF world.

[randapex]

So without know the sensor info, it's not possible to make a determination? I was hoping someone could give me at least a guesstimate of the ratio.

[james]

Oh fine - well it's easy for Eric since he has a full-frame camera. :-) I see the point though.

Cheers
James

[james]

For Eric: I don't think it's the same as cropping when you compare 1:1 with a digital SLR camera to 1:1 with a consumer cameras.

Let's take the example of a full frame DSLR camera with a hypothetical 5 megapixel sensor. You can shoot a critter 36mm across and call that 1:1 by your definition.

But what if you have a consumer camera that has a 12mm wide sensor but ALSO has the same 5 recorded megapixels. Take a photo of the exact same 36mm wide critter. There's no crop there, the same amount if information is recorded just with smaller pixels. But the difference is that the consumer cam has a lens that is designed for use with the smaller sensor.

So I think we still haven't answered Rand's question. How do you define reporduction ratio for a consumer camera with MUCH smaller pixels and a DSLR?

Cheers
James

[echeng]

I think it's already defined.

Consumer camera vs. SLR -- it's just sensor size, just like it was with 35mm vs 4x5 vs 6x7. The "consumer" vs "SLR" nomenclature is inconsequential. The "smaller pixel" distinction isn't useful, either. A 6MP 10D probably has smaller pixels than an old 1.3MP consumer camera.

To get 1:1 with a 4x5 camera, you need to frame the same area as the 4x5 slide. Same with a consumer camera. The Oly 5050 has a sensor size of 1/1.8 " (7.18 x 5.32 mm). So you need to fill 7.18mm x 5.32mm to get 1:1.

It's a moot point, though. The "reproduction ratio" is having less and less of an interesting impact in photography. We aren't looking at plates of film. We're looking at some version of it on a screen or on paper. So while 1:1 for a small sensor camera may mean that you have captured a really tiny subject full frame... we're going to look at it on the same size monitor, and print it out on the same size paper.

[echeng]

Another thing to think about:

There are different size sensors in different consumer cameras (1/1.8" and 2/3", these days) Are you saying that if I have two cameras that are nearly identical, but sport different sized sensors, the one with the smaller sensor magically gets a higher reproduction ratio for the same framing?

[james]

Are you saying that if I have two cameras that are nearly identical, but sport different sized sensors, the one with the smaller sensor magically gets a higher reproduction ratio for the same framing?

No, because they will be at different focal lengths. They each have a specially designed lens to match their sensor.

We are not getting anywhere toward answering the question and it's really hard to form a new definition of reproduction ratio through a discussion on a BB.

<Edit by James> I think a definition IS needed. Put cropping aside for a minute - I think I will stick to saying that if an uncropped image (however many megapixels it is) shows a subject 36mm across, I'll call that a 1:1 reproduction. Otherwise I have nothing but "Hey, how big was that critter?" To me, the term reproduction ratio is very useful - since I can't go around measuring everything I shoot. If I know the RR for my lens and camera then it does the measuring for me.

But that's just my differing opinion of course.

Cheers
James

[echeng]

I disagree. "We" should definitely not stick with that definition, since there is no consensus. I just don't want people to be confused, so you can't say that we "stick with" a specific definition. [EDIT -- this paragraph no longer makes sense because original wording in the post changed.]

If you want to use the film definition, then it is the one that I have written out -- not the 35mm-centric one.

Did some searches and found:

"Reproduction Ratio - a number that represents the size of an image on film compared to the size of the original subject. A 1:1 ratio means that the image is the same size on film as the original object. 1:2 means that the image is half the size of the original object" (http://www.daveread....o/glossary.html)

" Scale of reproduction: Describing the ratio between object size and size of the resulting picture, one distinguishes the reproduction scale on the film from the final result on the print. The latter is often blown-up. " (from Agfa)

"The ratio between the size of a subject and the size of its image on film"
(Nikon Japan)

"This ratio is simply defined as the image-on-film size X to the actual subject size Y and is represented in the form of X:Y. For a magnification ratio of 1:10, the subject size would be ten times larger than the captured image on the actual film or the image on film is a tenth of the actual size." (Nikonians.org)

(etc. etc.)

I don't believe you will find any authority on photography that will say that the repro ratio is fixed to 36mm x 24mm. My point is just that you can't arbitrarily change the established definition, just because we now have digital cameras. Film = sensor. Sensors with smaller pixel pitch = film with smaller grain size.

I've stated my opinion enough, I think.., and won't write anymore about this subject. :)

[craig]

Of course Eric and George are correct. 1:1 means life-sized projection so a full frame of 36x24 is 1:1 only for 35mm film. Different formats cannot be directly compared using magnifcation alone.

Trouble is that we want to take pictures of a certain frame size regardless of the camera we use, so it's also useful to compare performance at fixed subject sizes. We can't call that magnification or reproduction ratio, though, since that's clearly defined already. I personally call it "35mm equivalent magnification". A "1:1 35mm equivalent" would be 36x24 regardless of the camera used to shoot it. That's also a useful concept when comparing DOF.

[Ryan]

I agree with James, and Craig. While I usually detest equivalent anythings, this situation, for uniformity, is necessary.

For example, I'm tired of seeing "macro" shots entered in contests that are no where close to being an actual macro shot (1:1 on a 35mm film plane is the definition I find in most texts, which is very challenging to shoot underwater). I can shoot a whale shark in macro mode on a digicam behind a 130deg Inon dome, is that entitled to be entered? How about a crop of the same whale shark's pupil? Does that count? ...but I digress.

Heck, I don't care if you choose your definition to be the medium format plane (it makes my shots sound better, and would get Chris Bangs into the triple digit reproduction ratio he desires ). I do feel some standard is necessary.

[Ryan]

Since all we can do is estimate, and you estimate your dorid is 1.5 inches long, I estimate the frame to cover 2 inches, which is 50mm, which would yield a reproduction ratio of 1.43:1. The only definite answer would be to include a ruler, which would likely reduce your chances of winning anything in any contest

[echeng]

Ryan -- if you agree with Craig, you must say, "... yield a 35mm equivalent reproduction ratio of 1.43:1"

My big problem is using "reproduction ratio" ambiguously. I don't mind it if it is qualified (which is Craig's point, too).

Um. How can you agree with James and Craig, if Craig agrees with me and George?

Quite a conundrum.

[james]

I agree.

[herbko]

I thought about this a little last year when I was drafted by NCUPS to help write the rules for the digital part of the California Beach Dive Photo Competition. The dividing line between macro and wide angle categories in the photo contest is 6:1. I decided to translate the 35mm equivalent into a length. The rule I wrote for the contest was: Long dimension of the framed area should be under 8.5" or 21.6 cm for macro entires. I felt a little silly giving the length to 3 digits, but did it to maintain consistency.

The answer to Rand original question is that the 5050's sensor is 7mm in the long direction. I would guess the magnification is around 6:1 .

[Ryan]

or 1:6?

I see your point in my second post, Eric. That is exactly the problem, while I intended to state equivalencey, i didn't. I agree with you, and James, and George, and Craig, and will try to make sure I include that (equivalency) information.

[edit-1:6 and agreeing with everyone are me being sarcastic about a confusing subject]

[herbko]

or 1:6?

You're right. It's 1:6 .

[echeng]

Is the step from 1:1 to 1:2 a doubling in one direction, or is each dimension multiplied by the square root of 2? (linear vs by area?)