Question on shutter speed rule-of-thumb

[JohnnyRebel]

The old rule-of-thumb is that your min shutter speed s/b the inverse of your focal length - rounded up. (e.g. 50mm = 1/50 rounded up to 1/60 shutter speed)

I am often hearing that with crop sensor cameras you would need to multiply this by the crop value. (e.g. 50mm = 1/75 rounded up to 1/125 shutter speed)

Is this correct? If so, why? Since DX is a "crop" from the middle of FX and focal length doesn't really change (angle of view changes) why would the old rule not work for DX?

JR

 

[cvrefugee]

I was told that with APS-C and the 1.5/1.6X crop factor you have to do some math to find the adjusted focal length, which then in turn changes the minimum shutter speed. I don't know if that only applies when you're using full-frame lenses, though.

 

[finbarqs]

I don't believe so. Because you take a picture, and you crop it. That's essentially what cropping is. that means when they actually take their picture @ 1/50th on a full frame, it's out of focus already, and they needed to change their shutter to something faster. Full frame does NOT add any weight to how still you can hold it. It will not magnify it anymore either when you're on crop. You're just simply cropping the frame approx. 44%

 

[AkumaX]

I don't believe so. Because you take a picture, and you crop it. That's essentially what cropping is. that means when they actually take their picture @ 1/50th on a full frame, it's out of focus already, and they needed to change their shutter to something faster. Full frame does NOT add any weight to how still you can hold it. It will not magnify it anymore either when you're on crop. You're just simply cropping the frame approx. 44%

This sounds right in hindsight (you are just cropping the middle of a pic that's already 1/f-length) but a lot of people seem to suggest otherwise

"If you cropped in post-processing, and enlarged the resulting picture to the same size, the effect of camera shake would be magnified by the enlargement. As a result, you cannot tolerate as much camera shake in a cropped picture."

http://photo.stackexchange.com/ques...rule-of-thumb-apply-to-cropped-sensor-ca?lq=1

http://en.wikipedia.org/wiki/Crop_factor#Secondary_effects

http://www.techniphoto.com/wiki/index.php?title=The_shutter_speed_rule

 

[slashbinslashbash]

It all has to do with the angle of view. Let's take a P&S with a sensor 1/10th the size of full-frame, and put a 50mm lens on it. So the FOV is effectively that of a 500mm lens. Do you think that 1/50 will still give you a good shot?

All people/hands are shaky (note, the rule of thumb does NOT apply when shooting from a tripod or other sturdy mount). Just for the sake of illustration, let's say that in 1/50th of a second, the average person's hands will shake/move the camera in such a way as to move the camera through an angle of 1°. If the full angle of view of the lens is 100° like it might be with a wide-angle, this is a blurriness of 1% and not terribly visible. If the full angle of view of the lens is 10°, as it might be with a telephoto, this 1° blurriness constitutes 10% of the image and is probably visible. Again, these numbers are pulled from my ass (I have no idea of the average degrees of shake from an average person's hands) but the principle applies no matter what the actual numbers are. Ideally you want it so that the shake is completely not visible, i.e. it constitutes less than 1/2 of a pixel.

The best thing to do is to practice and take test shots with your own equipment and figure out your own limits. Take some test shots of things with fine details with some of your own lenses, and bump down the shutter speed with each shot. Keep on doing it and see where your shots start to get visibly blurry. The rule of thumb probably is not the same for every person or choice of equipment (i.e. a heavier camera with a large grip can probably be held more securely/stably than a P&S with no grip). If you have a good shutter finger and can gently press the shutter button instead of mashing it and moving the camera, you're another step up.

 

[twistedlogic]

Since DX is a "crop" from the middle of FX and focal length doesn't really change (angle of view changes) why would the old rule not work for DX?

I would say the opposite. If your planning on shooting a high MP FX camera with the intent of cropping away 44%, I'd probably wanna make sure I'm using a faster SS than the rule-of-thumb.

It all comes down to enlargements. If you start with a postage size stamp for a senor and you wanna make a big print, you will be stretching/enlarging that small file quite a bit, showing off diffraction and glass flaws. With FX or even MF, that stretching is significantly reduced.

 

[GoSharks]

Yes, you need to use the 35mm equiv.

 

[ElFenix]

photographically speaking there is no difference between taking a picture with a 35 mm camera and cropping the APS middle out of it vs. using an APS camera to begin with, ceteris paribus

 

[finbarqs]

if you think of it this way, if 1/60th is too slow for you to hold @ 60mm, then bump up the speed. If it looks fine on a full frame, but you can see the camera shake on a crop, that just means that your FF is also OOF.

 

[slashbinslashbash]

photographically speaking there is no difference between taking a picture with a 35 mm camera and cropping the APS middle out of it vs. using an APS camera to begin with, ceteris paribus

Yes, "taking a picture" is the same, it's what you do with it (and how you display it) that makes a difference. (You could say the same thing about a 35mm camera vs. a camera with a film size of 1mm*1mm; this is a meaningless statement.)

There is no definition of "sharp". There is only "acceptably sharp" i.e. the circle of confusion is small enough not to be seen at the desired print size. There is always some blurriness / lack of sharpness, whether caused by subject movement, camera movement, or optical problems. Nevertheless we get photos that we call "sharp" because the blurry parts are too small to be perceived at the size of image that we are viewing. For film photos this could be said to occur when the print was sharp to the naked eye, and for digital photos we often go "pixel peeping" and say that if the blurriness does not affect more than 1 pixel then it is a a perfectly sharp photo.

Taking a crop out of a photo, and then viewing that crop as a finished image in its own right, is to effectively enlarge the image. If you have a standard sized 4"x4" printed photo and you cut out the middle 1"x1" of it, and leave the 1"x1" crop as a tiny 1"x1" print then you are not adding any perceived blurriness. But if you take that 1"x1" crop and blow it up to a standard sized 4"x4" print then it is increasingly likely that you will see whatever blurriness was always there, but which was previously imperceptible. If you take that 1"x1" crop and blow it up to a 20"x20" print, any sharpness problems will be readily apparent.

You are effectively magnifying the image and therefore magnifying the effect of any shakiness in the camera. For APS-C crop factors (<2) it is questionable whether you "have to" multiply the focal length by the crop to get the minimum shutter speed. This is because it was a rule of thumb or first-order approximation to begin with. A factor of 1.5 or 1.6 is not going to make that much of a difference. For greater crop factors (P&S sensor sizes) it is more important. It is still the best to make actual tests with your actual equipment and your actual technique, of printed text or something else detailed, and make the determination yourself what is the lower bound of the shutter speed for hand-held photos of acceptable sharpness.

 

[blastingcap]

Yes, you need to use the 35mm equiv.

Few words, but correct words.

 

[SecurityTheatre]

I've done the math on this regarding the circle of confusion and the other issues that affect motion blur.

I think the final answer is that "for a given megapixel" you must crop to the "35mm equivalent".

Think of it this way. If you have a 12MP sensor full frame and you crop out the APS-C center, you have a 5MP crop frame with an equally sharp image (but lower resolution).

If you expect an image of the same sharpness AND resolution, you need to have a more stable exposure (hence using the 35mm equivalent for calculating the shutter speed).

It's pretty simple in this regard to compare the two. You need higher shutter speed to get the same sharpness AND resolution on a crop frame, as on a full frame with the same lens.

 

[JohnnyRebel]

I think the final answer is that "for a given megapixel" you must crop to the "35mm equivalent".

I think that we might be right to think in terms of 35mm equivalent.

Does MP matter? What if we are considering D4 vs D3200 with the same 50mm lens.

D4 min shutter speed would be 1/50 rounded up to 1/60
D3200 min shutter speed would be 1/75 rounded up to 1/125

The 100% image from the D3200 would be bigger than the 100% image from the d4, but that would not matter since it is angle-of-view that affects blurriness % (for the lack of a better term).

Is this right?

 

[ElFenix]

Yes, "taking a picture" is the same, it's what you do with it (and how you display it) that makes a difference. (You could say the same thing about a 35mm camera vs. a camera with a film size of 1mm*1mm; this is a meaningless statement.)

no, not even then. if you have a 100 mm lens and a 35 mm body, and then crop it down to APS, and if you have a 100 mm lens and an APS body, with sensors using the same technology and the same pixel density, there is no difference in the images. display wouldn't be different. an APS crop out of the middle of a 35 mm body and an image straight from an APS camera, using the same lens, same focal length, same f-stop, same shutter speed, same PoV, same tech, and same pixel density, will be the same. if you blow each up 20x they'll still be the same.

that's important to know.

everything else about your post i agree with 100%.

I think that we might be right to think in terms of 35mm equivalent.

Does MP matter? What if we are considering D4 vs D3200 with the same 50mm lens.

D4 min shutter speed would be 1/50 rounded up to 1/60
D3200 min shutter speed would be 1/75 rounded up to 1/125

The 100% image from the D3200 would be bigger than the 100% image from the d4, but that would not matter since it is angle-of-view that affects blurriness % (for the lack of a better term).

Is this right?

you'd use 35 mm equivalent because your FoV changes based on 35 mm equivalent.

 

[slashbinslashbash]

no, not even then. if you have a 100 mm lens and a 35 mm body, and then crop it down to APS, and if you have a 100 mm lens and an APS body, with sensors using the same technology and the same pixel density, there is no difference in the images. display wouldn't be different. an APS crop out of the middle of a 35 mm body and an image straight from an APS camera, using the same lens, same focal length, same f-stop, same shutter speed, same PoV, same tech, and same pixel density, will be the same. if you blow each up 20x they'll still be the same.

Yes, of course. But... ok, say we're sticking with film and prints so that megapixels and screen resolutions and all that stuff doesn't enter the discussion. Let's say we are holding the filmstock to be the same in all cases as well. Same film, just different sizes.

A 35mm photo comes out printed at 4"x6". We physically cut it down to what the APS-C size would be, so let's say we end up with a 3"x4.5" crop. No difference, obviously nothing has changed except the paper has been cut down; the photographic quality has not changed at all. So the APS-C sized piece of film is now represented as a 3"x4.5" piece of paper.

But if we take a photo with an APS-C camera and get it printed, it will be printed at the 4"x6" size, NOT the 3"x4.5" size. So the photo has been enlarged. We are taking the same APS-C sized piece of film and printing it on a 4"x6" piece of paper. It is enlarged relative to what it was before. You can see more detail with the naked eye; or the lack of detail, as the case may be. Stuff that might have appeared "sharp" on the 3"x4.5" print might look blurry on the 4"x6" print; but it's still from the same film, the same negative.

This may seem like an obvious point, but it is very important to the discussion at hand. If you take a 15MP photo and resize it down to 600x480 pixels for web viewing, the raw image can have a LOT of visual acuity problems that will simply disappear when you downsample to 600x480. So any discussion of "sharpness" has to begin with the fact that there is no actual "sharpness", only "acceptable sharpness" and that depends completely on how you are viewing the image, not how you took the image.

EDIT: I guess I should also note that it works both ways, as a practical matter. If you are taking a photo with the intention of cropping it, you will need to adjust your shutter speed accordingly. Say you are taking a photo of a bird with a 200mm lens, your longest lens. You know that you will be cropping this image down further to approximately a 400mm FOV. So you should halve your shutter speed if at all possible.

 

[ElFenix]

But if we take a photo with an APS-C camera and get it printed, it will be printed at the 4"x6" size, NOT the 3"x4.5" size. So the photo has been enlarged. We are taking the same APS-C sized piece of film and printing it on a 4"x6" piece of paper. It is enlarged relative to what it was before. You can see more detail with the naked eye; or the lack of detail, as the case may be. Stuff that might have appeared "sharp" on the 3"x4.5" print might look blurry on the 4"x6" print; but it's still from the same film, the same negative.

right, but that's not the situation i was presenting. i'm talking about taking an APS segment out of the larger negative (or sensor, doesn't matter) and comparing it to a native APS negative at the same print size. it comes out the same.

that's not obvious to a lot of people.

This may seem like an obvious point, but it is very important to the discussion at hand. If you take a 15MP photo and resize it down to 600x480 pixels for web viewing, the raw image can have a LOT of visual acuity problems that will simply disappear when you downsample to 600x480. So any discussion of "sharpness" has to begin with the fact that there is no actual "sharpness", only "acceptable sharpness" and that depends completely on how you are viewing the image, not how you took the image.

which is why pixel peeping is ridiculous

 

[twistedlogic]

I think that we might be right to think in terms of 35mm equivalent.

The focal length rule is designed with 35mm in mind. It is based upon the magnification of the subject when printed at a certain size and the acceptable sharpness needed.

Does MP matter? What if we are considering D4 vs D3200 with the same 50mm lens.

Only if you crop the image, the resulting image will have a narrower angle of view, changing it's perceived magnification. The more magnification the more careful you have to be with SS.

Just as with Macro photography. Focal length will not be the significant factor in determining the SS, it will be the amount of magnification you are reproducing and how much blur is introduced from just a slight shake or subject movement.

The 100% image from the D3200 would be bigger than the 100% image from the d4, but that would not matter since it is angle-of-view that affects blurriness % (for the lack of a better term).

Is this right?

You are correct. When printed to the same size, the blurriness will be the same regardless of pixel count. The higher pixel count camera will show more blur when magnified more, the lower pixel camera will still have the same blur, just not as visible.

100% viewing should not be used to determine sharpness.