Shutter Speed vs Resolution

well that's because for that 1/400th of a second you had the shutter open, the train did move.

If its movement was recorded over say, 2 pixels of the sensor, and you had a sensor with twice the density (that would be a 40mp aps-c sensor), it would blur across 4 pixels, but the image would look the same (provided the senor was equally good).

I don't think this issue has anything to do with camera technology. It has to do with movement vs how long the shutter is open.

Just to illustrate my point... if that train moves across the scene in two seconds , it traveled 1,944 pixels a second (half the width of your sensor in pixels).

this means at 1/400 shutter speed, the train traveled 4.86 pixels while the shutter was open.

AJSJones wrote in post #13620577
So anyone who makes large "fine art" prints is actually only making "posters" huh?

poster = big print. What you want to print big is up to you.

babel_fish wrote in post #13621235
^ curious as to what focal length was used?

That was at 40mm on the 17-40L, f/11.

mafoo wrote in post #13621245
well that's because for that 1/400th of a second you had the shutter open, the train did move...

Thanks mafoo, I know that. I was addressing the OP, trying to make the point that the 1/focal length rule of thumb for sharp shots goes out of the window if you really want sharpness at the pixel level.

tkbslc wrote in post #13621371
poster = big print. What you want to print big is up to you.

At least you didn't say "billboards"

If you meant big print, then fine. Often people think of billboards and low ppi resolution, low IQ when they use the term "poster".

mafoo wrote in post #13621245
well that's because for that 1/400th of a second you had the shutter open, the train did move.

If its movement was recorded over say, 2 pixels of the sensor, and you had a sensor with twice the density (that would be a 40mp aps-c sensor), it would blur across 4 pixels, but the image would look the same (provided the senor was equally good).

Let's say the 2 pixels were a 10µm blur in the 20 MP example, then in the 40 MP example, the blur is still 10µm but over 4 pixels. If you print to the same small size for each example, the image will indeed look the same. The bigger you print, however, the more apparent it will become that the blur is bigger relative to the pixels. If you had great glass and could resolve details in those 4 pixels and printed big enough to see those details, the blur would cause you to lose some of the improvement - because the blur got printed bigger too. If you print at a constant, let's say 240 original image pixels per inch of print, then you will get a bigger print but the blur will be more obvious. In other words, it depends - sometimes you'll notice, sometimes you won't. The blur here can come from any source - diffraction, motion, DoF considerations etc

Madweasel wrote in post #13621483
trying to make the point that the 1/focal length rule of thumb for sharp shots goes out of the window if you really want sharpness at the pixel level.

Point noted, that makes good simple sense.

I shoot as fast as I possibly can, although I don't see much need for shooting faster than 1/500s. This is to freeze ANY movement of anything that can be in the frame, with any of my lenses (nothing longer than 200mm).

I shoot concerts and once I get below 1/160s there is a great possibility that the subject will show evidence of motion blur unless I concentrate on timing the shot for a "quiet" moment. Nothing wrong with some motion blur, if that is your intention.

Madweasel wrote in post #13621483
Thanks mafoo, I know that. I was addressing the OP...

My comment was to the OP as well, sorry if that was unclear.

AJSJones wrote in post #13621567
Let's say the 2 pixels were a 10µm blur in the 20 MP example, then in the 40 MP example, the blur is still 10µm but over 4 pixels. If you print to the same small size for each example, the image will indeed look the same. The bigger you print, however, the more apparent it will become that the blur is bigger relative to the pixels. If you had great glass and could resolve details in those 4 pixels and printed big enough to see those details, the blur would cause you to lose some of the improvement - because the blur got printed bigger too. If you print at a constant, let's say 240 original image pixels per inch of print, then you will get a bigger print but the blur will be more obvious. In other words, it depends - sometimes you'll notice, sometimes you won't. The blur here can come from any source - diffraction, motion, DoF considerations etc

I Completely agree with all of this...

However, if you are getting to a size of print where you can start to tell a difference, doesn't that happen at the same point where image quality in general starts to suffer on the lesser dense sensor?

mafoo wrote in post #13621974
I Completely agree with all of this...

However, if you are getting to a size of print where you can start to tell a difference, doesn't that happen at the same point where image quality in general starts to suffer on the lesser dense sensor?

Yes, the less dense sensor just doesn't capture the detail. Another way of saying that is - lower resolution (aka less dense, bigger pixels) sensors will not detect the degradation from the blur that a higher resolution sensor will. So yes - it depends. Some folks want to use all their pixels to the max (for large prints or to crop out smaller subjects) and run the risk of making what would have been an invisible blur become somewhat visible.

Ah! Finally an interesting subject that is thought provoking!! Now all I need is someone who can think and give me some answers!

Not too sure about how resolution comes into play, but I'll take the word of people who know more about this than I do - I understand film...digital, I'm learning - big adjustment after a half century of using film. But I digress.

"Mr. Jones" (AJS to his friends) - gave an interesting explanation in his computation of how many pixels the train would be moving at a certain speed - clear enough even for me to understand (good job AJS!). This would result in motion blur - and as stated the larger the print, the more apparent the blur would be, although if printed at billboard size, it should look similar from a realistic viewing distance as the web size picture does - viewing from a block away or a foot or two away is quite different - despite how the hard-core pixel peepers may think. LOL

If the trestle that the train were on were perfectly perpendicular to the camera's lens, and if we assume the lens was used at 40mm then we can be close enough to assume the horizontal field of view is very close to 50 degrees (maybe a bit less, but in that neighborhood) - a "normal" perspective on a 35mm camera...if shot with a crop camera, then a bit narrower, but no sense making it more confusing. And really since we don't have any other info, it really is a moot point.

It would not be moot if we had one more factor to use but we don't - if we did, we could then split that 50 degree FOV in half (creating two congruent right angle triangles, each with a horizontal FOV of 25 degrees, then if we had one of the distances of the triangle created by the point of the camera's location, or the distance to either edge of the frame, we could rely on our junior high trigonometry and know how far the trestle is or how wide it is - we could also then time the train's crossing (front of lead car entering and leaving the frame) and know the exact speed. That would make it relatively simple to know exactly how fast the shutter speed would need to be if the camera were still (on a tripod ideally).

Why is this interesting to me? Because we could actually choose to forget about shutter speed completely since we would know at what rate to pan our camera to completely avoid motion blur. But even if we had all the info, we'd need some high tech motorized pano tripod head to follow the train at the exact right speed - (astronomers have them but they move quite a bit slower than what we'd need to follow a train at any normal speed) -

So we are left with the "skeet shooting" method as an alternative. (can be hand held, but easier with a pano head and a nice long handle). Follow the "target" and shoot.

If we do this and shoot at a slow(ish) shutter speed, which for the most part would be preferable to a fast SS because we will have the desired bit of blur of the turning wheels - if we shot so fast that we'd freeze the wheels, then it would look as if the train were stopped on the trestle - which would be fine if that was our intent, but we have to decide if we want the train to look as if it's moving or not . Can't have it both ways. And a train stopped on a trestle would probably mean some kind of trouble. We also can't go too slow or the boat in the foreground will be blurred. Goldilocks shutter speed - not too fast, not too slow, just right. If only photography were that easy and if only we had the time to calculate ideal settings for every shot - but we don't so we have to rely on experience which brings a certain "intuitive" basis and that seems to work. But it's fun to go through the "what ifs?"..

So even with the "just right" shutter speed for the train, then what happens to the (seemingly motionless) boat in the foreground? We shoot the train and the boat becomes a (mostly) white blur if we shoot slow enough. We have other possible concerns/problems..if it's windy, then the trees moving from the wind will look blurry - and most likely the water will look blurry or unnatural if there's a "chop" from the wind, etc., etc. - I'm sure there'd be other problems too. The slower the shutter speed, the more problems I would assume. (or maybe we'd get great abstract art?)

So ultimately we want a fast shutter speed, but not so fast as to freeze the train's wheels.

We can really make it much more interesting - the boat isn't a moored (or stalled) sailboat, but a Cigarette boat going 100mph. We try to "skeet shoot" the boat. Since the boat is closer to us than the train (assuming they are both going in the same direction..even the same speed) - if we shoot at a slow shutter speed (say the same shutter speed we used to make the wheels on the train look as if they are turning) - we pan at just the right speed to keep the boat from blurring - what happens to the train? The wheels will still look like they are going round and round, but since even though the boat and train are going the same speed (and direction to keep it simple) - that will make the train look completely blurred. Possibly even to the extent that the blur gives us the impression that the train is going backward (sort of like not using second-curtain sync when it's needed) . Why does the train appear to the camera to be going slower than the boat if they are both going the same exact speed? DISTANCE to the camera.

And it’s the distance that I don't think was mentioned earlier. The arc of our image (just for an example - we can't know without knowing the real distances - unless we bring a golfer's laser ranger finder ) - well, let's just say that the arc of the horizontal plane that the train needs to cover is triple in actual length of that of the boat - (we can know the degrees - it will remain constant with the rest of the image) - So for example, let's say the boat (going 100MPH) can cross our field of view if we do not pan the camera in let's say two seconds....the train is in this example, three times further away but also running at the same 100mph as the closer boat then it would take about 6 seconds to cross our frame - (I'm not completely 100% sure about this, but if I remember my rifle training by Uncle Sam in the bad old days, then that should be right .."minutes of arc double at double the distance, triple at three times, quadruple at 4 times, etc." IIRC).

OK - so if we pan at all, I don't know how fast we would need to have our shutter speed to get both the boat and the train in focus. I'd guess at 1/8000th of a second, we'd have the 100mph boat in perfect focus and the train looking like a train (how much like a train would be interesting) - What we really would need would be a slow shutter speed with the right speed pan and a bullet train going 300MPH and we'd be catching both the 100mph boat and the 300mph train both in perfect focus - and a nice smooth spray/wake from the boat and the blurred wheels on the train so we know it's not stopped on the trestle.

What if the boat is the same 1/3 of the distance from the camera that the train is, the boat is going it's 100MPH and the train is going it's 300MPH but they are going in opposite directions? If we pan to "skeet shoot" the boat, the train would probably look like part of the structure of the trestle. If we pan to follow the train, the boat would look (I would think) like an unrecognizable white streak on the water.

If we shoot at 1/8000th of a second (assuming we use a camera with that shutter speed) - I would wonder if we'd get a clear picture of a boat (with a "roostertail") and a clear picture of a train? Even at 300MPH, I'd guess we'd still see blurred wheels, but I wouldn't bet big on it.

Sort of mind boggling to me - we have the technology to have something like the Hubble telescope moving at close to 20,000MPH and staying focused on a star system or a nebula that takes up an infinitesimally small arc of the sky and gets perfectly focused images -

And we struggle to figure out how to get our 7 year old daughters doing a pirouetted at a ballet recital without motion blur.???

It's a good thing for guys like me that we have brain surgeons and rocket scientists –(hooray for my son!)… Imagine if everyone was like me - we'd be stuck in the stone-age forever! (and I even managed to sneak into an Ivy League school for grad work in journalism - with an emphasis on photo journalism and I still hit the wall with complex motion photos).

My parents were right - listening to the Rolling Stones and the Beatles (and the rest) rotted my brain!!

LOL


Not true… more likely it was basic training - I haven't been able to hear anything anyone says since firing on the range before they decided that earplugs would be a good idea.

I went to see the Dahli Lama and he told me the secret of life, but my hearing failed me and I have no idea what he told me (other than I had no chance of ever hitting a 2 iron and keeping the ball on the fairway).

Madweasel wrote in post #13621159
Here's the example I had in my mind when I made my previous post. The train was travelling something like 40-50mph across the field of view at some distance from my position. The camera was tripod-mounted and fired by remote switch. The sensor (40D) is only 10MP and the shutter speed was 1/400s and at the pixel level you can see the train has moved.

Resolution has little to do with this as much as the speed and direction of travel in relation to the shutter speed used.

Bosscat wrote in post #13623227
Resolution has little to do with this as much as the speed and direction of travel in relation to the shutter speed used.

They control only the amount of blur in the image the sensor will record. How well it is recorded and how visible it becomes, are factors under the influence of the resolution and print size/viewing distance respectively and respectfully

Can I get a glass or toke of whatever Delija is having