PDA

View Full Version : 70-300 VR question



JBO
10-27-2007, 09:39 AM
Hey,

I have one question. If I put the 70-300 VR lens on my D80, would I have 70-300 or would it be like 105-400 or whatever?

BionicSniper
10-27-2007, 10:11 AM
105-450

I used one last night for a football game (slow lense made it hard... had to go 250-320 at iso 3200)

I must say that the VR is AMAZING! and the zoom is awesome. I did find my self wishing for 50mm bottom end on this lense but i just took a few steps back and i never missed anything because of it.

erichlund
10-27-2007, 10:15 AM
Hey,

I have one question. If I put the 70-300 VR lens on my D80, would I have 70-300 or would it be like 105-400 or whatever?

You will have all the characteristics of 70-300 VR except for one. Your angle of view will be that of a 105-450.

It is angle of view that affects ability to hold by hand, so when you calculate the 1/focal length, you will need the 1.5x crop multiplier. So, if you are shooting at 300mm, you will need 1/450 of a second to be able to hand hold without VR. Since VR gives you up to 4 stops, you can try exposures as long as 1/30. :eek: No guarantees. :rolleyes:

K1W1
10-27-2007, 03:57 PM
It is angle of view that affects ability to hold by hand, so when you calculate the 1/focal length, you will need the 1.5x crop multiplier. So, if you are shooting at 300mm, you will need 1/450 of a second to be able to hand hold without VR.

Surely you jest.
The 1/focal length rule applied from the days when a 300mm lens was 300mm long from the focal plane to the front element. Using your theory are you saying that the 70-300 extends to 450mm when it's at 300mm?

e_dawg
10-27-2007, 11:33 PM
Eric's got it right. 1/(f * crop factor) = min shutter speed. Camera shake is manifested in an image as a sort of double image. This is because the shutter is not fast enough to capture the image projected onto the sensor where each pixel records only one point on said image. The shutter is open long enough that the image has time to travel far enough on the sensor such that each point on that image is now on a different part of that sensor. Often 1 or 2, but possibly 6 or 7 pixels away from the original one, thus recording a double image.

And it is the angle of view that determines how much of said image is being projected on the sensor. The smaller the angle of view, the greater the effect any movement from the camera or subject will have, as it will represent a proportionately greater linear / tangential movement in the image as projected on the sensor.

A 70-300 at 300 mm is still 300 mm no matter what body you use it on. Doesn't matter what sensor you have. It's the angle of view that changes. So a 300 mm lens on a 1.5x crop sensor has an AoV that's equivalent to a 450 mm lens.

BTW, AFAIK, focal length is not measured from the focal plane to the front element. For a technical explanation, see this article:
http://www.bobatkins.com/photography/technical/measuring_focal_length.html

K1W1
10-28-2007, 12:42 AM
So what you guys are saying is that if I have a 300mm telephoto lens on my D3 at full frame (or my film camera) I can take photos hand held at 1/300 sec in theory but if I use the same lens on my D50 I can only take the same shot at 1/450 sec. :confused::confused:

erichlund
10-28-2007, 12:50 AM
Remember. It's only a rule of thumb that 35mm photographers came up with. If you have steady hands, you can do better than the rule. If you have unsteady hands, you won't do as well. But, essentially, you are correct. On a full frame camera, the 300mm lens shows less shake than it does on an APS-C camera.

And no, I don't jest.

Rooz
10-28-2007, 12:50 AM
in principal that is right, 1/450s cos teh d50 will magnify 1.5x wheras the D3 won;t. but its really only ever been a rule of thumb.

TheWengler
10-28-2007, 12:53 AM
So what you guys are saying is that if I have a 300mm telephoto lens on my D3 at full frame (or my film camera) I can take photos hand held at 1/300 sec in theory but if I use the same lens on my D50 I can only take the same shot at 1/450 sec. :confused::confused:

Sounds right. Camera shake comes from the focal length (35mm equiv.) not the size of the lens, though maybe sometimes the size of the lens can make it worse.

e_dawg
10-28-2007, 06:30 AM
So what you guys are saying is that if I have a 300mm telephoto lens on my D3 at full frame (or my film camera) I can take photos hand held at 1/300 sec in theory but if I use the same lens on my D50 I can only take the same shot at 1/450 sec. :confused::confused:

Yep! You've got it K1W1!

(but I'd think you'd want to use a 450 mm lens on the D3 if you needed the same amount of reach, defeating the purpose, so to speak)

K1W1
10-28-2007, 02:59 PM
I still believe that you guys are wrong. The crop factor is a function of the smaller sensor size relative to a 35mm negative it's not extra focal length or anything like that.
If I were to take this magic 300mm lens and shoot an image with a D50 then take a film body and shoot the same image with the same lens BUT I masked the 35mm film so it was effectively the same size as an APS-C sensor then the image on both cameras would be the same (i.e. "cropped") and when they were developed they would be identical in content. "Cropping" relates to the field of view nothing else.
Why would I need to take the film image at 1/300 and the digital one at 1/450 in the example above which is what you are saying?

Rooz
10-28-2007, 03:05 PM
your crop theory is right if you can get the shutter speed high enuf to eliminate any chance of camera shake. but lets say the max shutter speed you can get is 1/100s. you will have much less chance of camera shake on the D3 at 300mm then you will on a D50 at the equivalent magnification of 450mm. the higher the magnification of your subject the more pronounced the subject blur due to shake.

i think thats right anyway.

achuang
10-28-2007, 03:22 PM
Up until now I would have thought that you'd need the higher shutter speed on the crop body like the D50, but now that I think about it I get what you mean K1W1. Now I'm really not sure of what's right. Hard to make a decision based on my personal experience as everyone can handhold at different shutter speeds. I've shot at 300mm on my D70s at 1/320 and they were sharp. And I've shot my 60mm at 1/13 and was reasonably sharp. So regardless of the answer it's different for each person.

K1W1
10-28-2007, 03:39 PM
you will have much less chance of camera shake on the D3 at 300mm then you will on a D50 at the equivalent magnification of 450mm. the higher the magnification of your subject the more pronounced the subject blur due to shake.

That is where I believe that you guys are wrong.
There is no magnification involved in a digital crop. You are not magnifying the image by 1.5x you are simply recording a smaller portion of the same scene.
If you take an example of a black circle against a white background. The actual black circle on the APS-C sensor and the film is exactly the same size. the only difference is that there is more white background on the film because it's recording a wider field of view.

Rooz
10-28-2007, 04:21 PM
sorry this is too technically advanced for me. i'm out of my depth. lol

XaiLo
10-28-2007, 05:53 PM
It's just an illusion... hmmmm my math is totally under used but the field of view and the picture element are both larger in a (35mm or full frame) ergo a smaller sensor will proportionately (in relation to camera shake) will move a greater distance across the sensor than a larger sensor.

let's say these are the approximate size for a full frame and a cropped sensor
24x36 mm sensor = full frame
15x23 mm sensor = cropped sensor

while 2.4mm of vertical movement on a (35mm or full frame) sensor equals approx 10% of the area of the sensor along the vertical plane. On a cropped sensor that same 2.4mm covers more than 15% along the same vertical plane. And on top of that you have to figure in the greater density of picture elements.

so I must draw the conclusion that to offset the greater distance traveled accross the plane of the sensor a higher shutter speed would be necessitated to stop motion..... then again I could be wrong.:confused:

erichlund
10-28-2007, 06:23 PM
I still believe that you guys are wrong. The crop factor is a function of the smaller sensor size relative to a 35mm negative it's not extra focal length or anything like that.


Maybe if you think in different terms it will strike home. The Fujifilm S6000fd has a 6.2mm - 66.7mm lens that's 300mm equivalent on the long end. What you are saying is that 1/60th ought to do it, but that's wrong. This is a high powered zoom lens. However, the sensor is so small, that it doesn't have to be extremely, physically long to generate all that power.

What we are saying is that no matter what the actual physical characteristics of the lens, in order to put an equivalent amount of image on the sensor as another lens at the same distance, it will require the same hand holding limits.

Perhaps another thing to understand is that the issue is not the length of the lens at all. Think in terms of one pixel. That pixel views a certain area of the image (very small, but still defineable). Lets say that that pixel is representing 1mm x 1mm of area in the image at a defined distance. The higher power the lens, the farther away that 1mm sq area will be, so any shake by the user will be magnified by the arm (distance from the sensor to the subject area). Since the 66mm lens and a true 300mm lens are seeing that same 1mm area at the same distance, camera shake will impact them in an identical manner. (Now, I've used equivalent megapixel cameras for this example, so 1 pixel should represent the same 1mm square for each camera).

Here's another example that should just blow your mind. You put a 50mm lens on a camera with huge mega pixels and a full frame sensor. Let's say that it has enough pixels that a 300mm equivalent crop fits in the same mega pixels as our APS-C sensor with a 200mm lens (300mm equivalent). You need the same hand holding limits for both shots, because you have to treat that 1 pixel exactly the same on both.

K1W1
10-28-2007, 07:08 PM
Perhaps another thing to understand is that the issue is not the length of the lens at all. Think in terms of one pixel. That pixel views a certain area of the image (very small, but still defineable). Lets say that that pixel is representing 1mm x 1mm of area in the image at a defined distance. The higher power the lens, the farther away that 1mm sq area will be, so any shake by the user will be magnified by the arm (distance from the sensor to the subject area). Since the 66mm lens and a true 300mm lens are seeing that same 1mm area at the same distance, camera shake will impact them in an identical manner. (Now, I've used equivalent megapixel cameras for this example, so 1 pixel should represent the same 1mm square for each camera).

This is exactly where I can't follow the logic.
A 1mm square (your example) on a film camera is still a 1mm square on a 1.5x crop digital camera (assuming same lens and all that stuff). So assuming that we were photographing a 1mm square why using your logic is it necessary to have a shorter exposure time on digital using the same lens?
Digital crop is not magnification it's simply a smaller viewable area.

BTW. I know that we are talking theories here and I think it's terrific that this sort of debate can go on without the resorting to name calling that generally characterises the Internet forum world.

Rooz
10-28-2007, 07:14 PM
why using your logic is it necessary to have a shorter exposure time on digital using the same lens.

i didn;t think we were talking about correct exposure ??.:confused:

i thought we were talking about the ideal shutter speed to prevent camera shake due to the focal length as opposed to the shutter speed affecting the exposure which won;t change regardless of DX or FX format.

K1W1
10-28-2007, 09:07 PM
Bad phrasing on my part. When I said exposure I meant the shortest exposure able to be hand held without evidence of shake based on the rule of thumb as per the last couple of pages of this thread..

K1W1
10-28-2007, 09:38 PM
let's say these are the approximate size for a full frame and a cropped sensor
24x36 mm sensor = full frame
15x23 mm sensor = cropped sensor

while 2.4mm of vertical movement on a (35mm or full frame) sensor equals approx 10% of the area of the sensor along the vertical plane. On a cropped sensor that same 2.4mm covers more than 15% along the same vertical plane. And on top of that you have to figure in the greater density of picture elements.

so I must draw the conclusion that to offset the greater distance traveled accross the plane of the sensor a higher shutter speed would be necessitated to stop motion..... then again I could be wrong.:confused:


That would be fine if both sensors recorded the same scene i.e. the complete scene on the digital sensor was compressed into a smaller space. This does not happen all the digital sensor does is record a portion of the scene (roughly 66%) that the film or full frame does. The actual components of the scene are the same size on both sensors (those components that are recorded that is) so any movement (shake) should affect the image clarity the same way.

For my final (I think) example on this subject I will use the D3.

Let's say you have a D3 with this magic 300mm telephoto lens.
What other are arguing is that with the camera in FX mode you will get a steady shot at 1/300 sec using "the rule" but turn the camera to DX mode (same camera, same lens, same scene, same photographer) and you will need to select 1/450 sec to get a steady shot.
That simply does not make sense to me.
Nobody has yet explained to me how by simply changing a menu item (or is it a button on the D3?) you will need a shorter exposure time to get a steady shot using the same equipment.

fionndruinne
10-28-2007, 09:51 PM
The 1.5x crop factor, however, means that you will be viewing the image at 1.5x magnification as compared to a 35mm sensor/film. Thus, the effects of hand shake on an image will be 1.5x more visible/obvious than it would be on a full frame - we all know how cropping an image, as you come closer to its max resolution, emphasizes imperfections.

[/monkeywrench]:p

BionicSniper
10-28-2007, 10:02 PM
ok if you have a camcorder or anything with a digital zoom go grab it.

hold it as still as you can and zoom in. Notice that it seems to get harder and harder to hold it still.

Now that image was shaking just as much as it was when you were zoomed all the way out. But the image was covering a broader area. This is why theres the 1/ focal length rule. and yes on the d3 d300 if you switch to dx mode you are going to need to drop the increase the shutter speed because the image is being cropped so you now have the same image except it is being cropped.

achuang
10-28-2007, 11:12 PM
ok if you have a camcorder or anything with a digital zoom go grab it.

hold it as still as you can and zoom in. Notice that it seems to get harder and harder to hold it still.

Now that image was shaking just as much as it was when you were zoomed all the way out. But the image was covering a broader area. This is why theres the 1/ focal length rule. and yes on the d3 d300 if you switch to dx mode you are going to need to drop the increase the shutter speed because the image is being cropped so you now have the same image except it is being cropped.

This makes sense now, i remember with my old olympus C-750 with 10x zoom equivalent to 38-380mm. It had 4x digital zoom and when i tried to use digital zoom once at max zoom of both optical and digital i could barely keep the subject in the frame and the effect was much more pronounced.

erichlund
10-29-2007, 07:08 AM
This is exactly where I can't follow the logic.
A 1mm square (your example) on a film camera is still a 1mm square on a 1.5x crop digital camera (assuming same lens and all that stuff). So assuming that we were photographing a 1mm square why using your logic is it necessary to have a shorter exposure time on digital using the same lens?
Digital crop is not magnification it's simply a smaller viewable area.

BTW. I know that we are talking theories here and I think it's terrific that this sort of debate can go on without the resorting to name calling that generally characterises the Internet forum world.

The 1mm square is at your subject and is represented by 1 pixel on your sensor. So, if you have a 300mm lens on a an FX camera and a 300mm lens on a DX camera, both with the same megapixels, the 1mm will be 1.5 times as far away (or whatever the math works out to, I'm not sure that the crop factor works to direct distance transferrece, but it would be a greater distance). That's because the angle of view is narrower on the DX. As a result, any small movement of the camera is magnified by the increased distance to the subject. Since the 1 pixel is affected by this, and it could be any pixel, then all pixels are affected. So, that means that the 300mm lens is more sensitive to camera movement on the DX camera.

Jason25
10-29-2007, 07:44 AM
I still believe that you guys are wrong. The crop factor is a function of the smaller sensor size relative to a 35mm negative it's not extra focal length or anything like that.
If I were to take this magic 300mm lens and shoot an image with a D50 then take a film body and shoot the same image with the same lens BUT I masked the 35mm film so it was effectively the same size as an APS-C sensor then the image on both cameras would be the same (i.e. "cropped") and when they were developed they would be identical in content. "Cropping" relates to the field of view nothing else.
Why would I need to take the film image at 1/300 and the digital one at 1/450 in the example above which is what you are saying?
I agree with you K1W1. It's still 300mm, it's just cropped, not magnified. There's no difference otherwise.

erichlund
10-29-2007, 11:26 AM
I agree with you K1W1. It's still 300mm, it's just cropped, not magnified. There's no difference otherwise.

This, I think, is the biggest part of the problem you guys are not getting. A lens is a magnifying glass. The bigger the lens, the more it magnifies the image.

Let's say I'm trying to take a picture of a bird. The bird is a seagull that is about 5 feet away. It's a trained bird, so it stays exactly where we want it to and does not move at all.

Our cameras (we're loaded), are a Nikon D3 and D300 and a highly modified Panasonic S6000fd. On all three, we have mounted a Nikkor 300mm f4 (No VR). I told you the S6000fd was highly modified (Due to the modification, the lens is equivalent to 1500mm on and 35mm/FX camera). The time of day is just before sunset and it's cloudy, so the light is not great. In fact, the meter in all three cameras is saying that the shutter speed is 1/300 (how convenient).

This lens lists angle of view as 8deg, 10minutes for FX, 5deg, 20min for DX. If you divide 8.166667 x 1.5, you get pretty close to 5.3333 (actually 5.44444...). Remember, the 1.5 crop is actually an approximation. If you divide the FX Angle of View (AoV) by the DX AoV, you get the actual true crop value of about 1.5313. Close enough. That makes the modified Panasonic a 5x crop.

With the D300, the bird just fills the frame. With the D3, there's some crop space around the bird. With the Panasonic, we're talking head shot. This is what you see in the viewfinder of each camera. So, while the D3 and D300 appear similar, the Panasonic gives you an extremely different view, yet it is the exact same lens.

If I breath while taking the shot, I will move the lens angle 1 degree (just for giggles). If I don't breath, I will feint, so I decide to breath (good choice). This is the simulation of camera movement based on human biometrics. It is by no means meant to be a particularly accurate measurement of the human condition, but a convenient reference value. Now, with the FX that's only about 1/8 (about 12.5%) of the angle of view. With the DX, it's up to 20 percent. With the Panasonic, it's about 55 percent of the angle of view, so you can see that each time I take a breath, the Panasonic is much more sensitive to that movement than the other two cameras, and the DX is slightly more sensitive than the FX. My breaths occur over time, so the movement occurs at a periodic rate. However, the larger the format, the more "in control" the image appears in the viewfinder, even though I'm using the exact same lens. It appears the 1/300 will be sufficient that only a very small motion blur (essentially undetectable) will remain in the D3 image. It's there, and with the proper equipment, we could probably measure the image degradation. But, we cannot visibly detect it on the 8x10 print. However, there will be a slight blurring of the D300 print, and the Panasonic print will be horribly disfigured. This is because the angular momentum of my breathing was too great on these images for the 1/300 shutter speed to overcome. The pendulum of movent on the Panasonic image was so large that even at 1/300, the distance of the image covered is still to large for the shutter speed to overcome.

Note, this periodic movement that I've described is actually probably much smaller than the 1 degree I used. However, for the sake of this argument, it does describe how the human body affects the camera when holding for the photo. Even if you hold your breath during the shot, your heart pumps, your neurons fire, muscles twitch, your body sways. All these things add up to a periodic motion that affects the camera. The faster the shutter speed, the less this motion can affect the image, but the greater the magnification of the image, the more the image is affected by the periodic motion. The magnification is purely based on angle of view of the lens, not on the lens that is mounted. The Angle of View is affected by the size of the sensor in direct relationship to the crop factor. So, when making your calculation for hand holdability of the shot, APPLY CROP FACTOR.

This is my last attempt. So, if you don't get it now, you will have to actually go take some geometry and physics in school.

XaiLo
10-29-2007, 01:33 PM
The problem maybe derived from the use of the word crop an APS-C sensor does not crop anything. It's physically incapable of capturing or displaying the same FOV as a full frame sensor. So while a 300mm lens that was designed for a "full frame sensor" is still a 300mm lens on a APS-C sensor and the fact that it's still in effect capturing the same scene is made irrelevant by the physical size of the APS-C sensor. The APS-C sensor can only relate to the light that's hitting it and as such the comparable image FOV is 450mm and not 300mm consequently it's subject to the same limitations in movement as a "full frame sensor" with a 450mm lens. The FOV needs to be the constant not the lens.

Think of it this way, say your at a rifle range and you're set up to hit dead center (which is equal to the FOV of the APS-C sensor) and the outer most ring is equal to the ( FOV of the "full frame sensor"). Now while a small fraction of movement will not cause you to miss the target completely you will miss the bullseye.

VTEC_EATER
10-29-2007, 02:48 PM
Isn't this simply an issue of resolution?

Take for example Canon's 5D. Its close enough to full frame, 12.8 megapixles, and has a resolution of 4368 x 2912.. Now take a D300. Its a 1.5x cropped sensor, 12.1 megapixles, and has a resolution of 4288 x 2848.

Those numbers are almost the same. But wait, how can that be? One is full frame, the other is a cropped sensor. The full frame should much larger than the cropped sensor. But its not. Why?

Because it is not a literal "crop" of a full frame camera's resolution. A "crop" is literally taking away portions of the image and leaving it as is. If a "cropped" sensor was a literal "crop", the resolution size of the file would be considerably smaller than its full frame counterpart. What a "cropped" sensor does is take that same "crop" and magnify, or up-sized, it to the same resolution that it would have on the full frame sensor. That is why you need to add shutter speed to a "cropped" sensor. Because its not a crop, its a magnification ratio. The term "crop" is a bad term to throw around.

Now, on a camera like the D3, I can only assume that in "DX" mode it will be a literal "crop" of the image. Since its resolution is bumped down from 12 megapixles to 5. I have no idea what its resolution at 5 megapixles will be. If it somehow up-sizes the image resolution, then a D3 user would probably benefit from increasing their shutter speed. If it is literally a crop, then there will be no magnification effect, and increasing shutter speed will not be needed.

e_dawg
10-29-2007, 10:48 PM
Thank you Eric. I didn't know how to explain it any better than I already did on page 1, but I'm glad you did.

I would like to offer another (oversimplified, scientifically inaccurate, numbers made up for argument's sake) way of looking at this in the hopes that it will help people understand.

When you take a picture, the shutter is open for a certain amount of time. As soon as the shutter opens, the image is formed on the sensor at time t = 0. To get zero blur, the image must not move at all while the shutter is still open. That way, the sensor picks up one image and one image only.

However, this is not realistic, especially with handholding. The camera shakes due to breathing, heart pumping, shaky hands, etc. The image moves around on the sensor at a certain speed. At any given point on this sensor, we can measure the linear velocity tangent to the arc described by the angle of view and a function of the angular velocity Eric alluded to in the above post.

Let's forget about cropping and how big an actual object is on different sensor sizes for a second. The point is that the image is moving at a certain linear speed along the plane of the sensor for whatever period of time the shutter is open for. Let's say that with a 300 mm lens, your shaky hands causes the image to move across the sensor at a rate of 100 mm/sec. Doesn't matter what size the sensor is, the same amount of shaking always causes the image to move across the sensor at the same speed.

With a shutter speed of 1/200 sec, the image is able to move 0.5 mm across the sensor before the shutter closes. This 0.5 mm shift in the image represents 0.5 out of 16 mm in the vertical dimension of an APS-C sized sensor, or 3% of the image height. With a full-frame camera, the vertical dimension of the sensor is 24 mm, so the 0.5 mm shift only represents 2% of the image height.

While you may not detect too much blur from the full-frame camera at 1/200 sec, you will from the APS-C camera, because the image shift is 50% greater (3% vs 2% of the image height) than that compared to the full-frame sensor.

In order to keep the blur from becoming detectable on an APS-C camera, you need to keep image shifting to under 2% of image height while the shutter is open. To pull that off, you need the shutter speed to be 1/300 sec. For an image moving at 100 mm/sec due to camera shake, it will move 0.33 mm in the 1/300 sec the shutter is open, or 2% of the 16 mm vertical dimension of an APS-C sensor.

e_dawg
10-29-2007, 11:39 PM
BTW, I'm all for debate and open-minded discussion here, but there really isn't much to debate here. The rule of thumb is your minimum recommend shutter speed handheld is 1/(f * crop factor). This is an accepted "rule" just like the rule of thirds, the traditional 1/f shutter speed rule, monitor gamma should be 2.2 on PCs, printers are usually calibrated to D50 whitepoint, monitors are usually calibrated to 6500K WP, etc. (although it is catching on a little slowly due to the legacy of the 135 film shooters out there). Why do so many people argue that 1/(f * crop factor) is not true?

Eric and I have explained all the reasons why, but if you don't believe us, here are some links on the subject taken from the first page of google results:

http://hannemyr.com/photo/crop.html
(near the bottom, "Hand-holding" section)

http://en.wikipedia.org/wiki/Focal_length_multiplier
(at the bottom, "Secondary Effects" section)

http://www.berniecode.com/writing/photography/beginners/
(awesome general photography tutorial, see three-quarters of the way down in the Accessories > Tripod section)

http://software.canon-europe.com/files/documents/EF_Lens_Work_Book_5_EN.pdf
(see page 4 of the Canon EF Lens Workbook, comment *2 in the 17-55 IS lens section)