Images not Sharp Enough

[DacrimL]

Ok have been reading some info from a reputable photo magazine as to why some images are not sharp. Their #1 point was to do with shutter speed and I quote -


In the days of film photography there was a general rule that in order to get sharp images from a hand-held camera the shutter speed needed to be at least one second divided by the focal length of the lens.So if you were shooting with a 100mm optic the shutter speed needed to be at least 1/100sec, which because of the way shutter speed is set usually translates to a setting of 1/125sec or faster.
This rule still holds today, but it is somewhat complicated by the focal length magnification factors of sub-full-frame sensors and image stabilisation systems.
For example, if a 100mm lens is mounted on a Nikon APS-C format SLR like the D5200, which has a focal length magnification factor of 1.5x, the photographer would need to set a shutter speed of at least 1/150sec.
Canon APS-C format DSLRs like the EOS 650D have a 1.6x focal length magnification factor, so the shutter speed would need to be at least 1/160sec.

Will post link to article if that is allowed.

Basically my question is how does one know what the focal length magnification should be for ones individual camera and how does the conversion equation work. So far I have sort of been guessing what shutter speed to use relative to which lens I am using or am I just making things too difficult for myself?

[tduell]

Hello DacrimL,
An interesting story. Was the article referring to any specific subject type? That may be good advice for a passive subject (landscape etc) but I wouldn't think it is applicable for shooting kids, motorsport or other subjects that are moving rapidly, if you want a sharp image.
Regarding your question on the focal length magnification (also known as 'crop factor'), this should be readily found by searching on that subject and including your camera make/model.
The crop factor is simply a ratio of the size of a full frame sensor (35mm film) to the size of the sensor in your camera. Why it can change the effective focal length of your lens is probably explained in great detail with diagrams, on a number of web sites...and probably also on this site somewhere.

Cheers,
Terry

[I @ M]

Terry is on the money with his thoughts on moving subjects, a faster than focal length to shutter ratio may be needed in order to get a sharp shot but in cases where the photographer wishes to include plenty of motion in ( for example ) a racing car they will frequently photograph at shutter speeds well under that rule.

In the case of hand holding the camera and photographing a stationary subject especially with a camera / lens that doesn't have some form of stabilisation built in, I am firmly in the camp of exceeding the focal length / shutter speed rule. This is especially relevant when using longer focal lengths and or high mega pickle bodies. I frequently aim for 2x shutter speed to focal length when using any lens on a D800 body, part of which is due to my hand holding not being as good these days as when I was younger.

Yes, I agree with the "crop factor" being taken into account when calculating speed, it is the minute change in angle that shaking hands impart that causes the blur and a "crop factor" is magnifying the amount of movement that occurs in an image.

[ameerat42]

Usually the "crop factor" is mentioned SOMEwhere in the manual - or in specs found on-line - but for APS-C ~1.5 should help.

[ricktas]

Certainly as we move to the higher pixelcount camera sensors it seems to be an issue raised more often. When Nikon released the D800/e there were numerous threads post and general discussion on the net about the clarity of the images produced. These dense sensors were more prone to show slight camera movement than those that had gone before. Many comments were about having to re-learn technique.

shutter speed became much talked about. As was the relationship between focal length and shutter speed.

I think it has made some current photographers stop and think, and go back to simple rules, defined many years ago.

[DacrimL]

Have just referred to the manual and the crop factor for my particular Canon is 1.6X. So it does seem that some of the images I shoot need this conversion particularly since I shoot mainly hand held, perhaps this may help with sharpness, among other factors.

Rick I certainly remember the old rule and the simple logic to get the shutter speed in relation to the focal length of a lens.........it was just when I went digital the images taken with an old 35mm film camera somehow always seemed better and sharper than now.

[Mark L]

.........it was just when I went digital the images taken with an old 35mm film camera somehow always seemed better and sharper than now.

Do you sharpen your photos in PP?
And there's no reason you can't post links to anywhere now.;)

[ricktas]

http://www.ausphotography.net.au/forum/showthread.php?8579

[fillum]

The thing to keep in mind is that the 'one-over-focal-length' (1/FL) shutter-speed (s/s) 'rule' is a 'rule-of-thumb' - i.e. an approximation. Strictly speaking the 'crop-factor' should be considered, however the actual minimum s/s that should be used will vary between individuals and shooting situations. For example other factors that might affect minimum s/s are:-
- the photographer's own level of firmness / 'shakeyness'
- the photographer's hand-holding technique
- equipment characteristics and ergonomics (for example a camera without a viewfinder held out in front of the face can be less stable than holding a DSLR 'tucked in' against the face)
- whether the photographer is able to brace themselves (against a fence, wall etc)

What's important is to work out what works for YOU - i.e whether you can still get sharp images shooting a bit slower than 1/FL or whether your shooting 'style' requires a s/s a bit faster than 1/FL.



Cheers.

[flashc]

No one has mentioned the effect of image stabilization built into lenses where manufacturers state that it can allow you to handhold at up to 4 stops slower.

What does IS or VR in a lens really do to the 1/FL rule when you're hand holding the 100mm lens on a stationary scene

Under normal conditions, we should now be able to shoot at 1/10 sec instead of 1/160 using a Canon APS-C sensor

IS doesn't work this miracle very well for me, especially since I moved from a 6MP DSLR camera back in 2003 to 21MP now

[farmmax]

I guess I'll never be a serious photographer. By the time I tried to think and calculate all the things I'm supposed to, the moment to take the photo probably would have passed :(

Luckily I didn't own a tripod for many years, so just had to hand shot. I think that has been a good thing and now don't hesitate to shoot stationary objects down to around 1/15 expecting them to be OK. Moving objects definitely need higher shutter speeds ... at least I've worked that much out :D

It is all right to have theory, but life is full of examples where theory and practice don't meet, so I don't get too hung up on photographing "correctly". I know other people like to try and do it by the book, and that works for them. So long as new photographers don't feel they "have" to do it the technical way, and give up because it is too hard.

[Lance B]

The 1/FL rule is only a guide and has only ever been a guide as it depends on a number of factors and these other factors that are to be taken into consideration are:

1) How large you are going to print/view the image. A small image from just about any camera viewed on your phone screen can look sharp, but when you enlarge it to the size of your computer screen it may look quite unsharp and blurry depending on the image. This is where the "myth" about the new very high Mp cameras needing higher shutter speeds needs to be put to rest. The fact is, a high Mp camera will not look any worse then a low Mp camera when viewed at the same size and all other things being equal, ie physical camera sensor size (FF compared to FF for example), aperture, shutter speed, lens used , ISO etc. Where you may need to employ better camera technique and possibly employ higher shuuter speeds is when you decide to usitlise those extra Mp's and are going to display the image larger than that from a lower Mp camera or where you are going to crop significantly. It is only then that you may detect camera shake more so than from a lower Mp camera. However, you may have been able to hold the camera well enough for that not to be an issue.

2) Whether you are going to crop your image significantly when compared to a lower Mp camera. However, if you need to crop the high Mp cameras image, then you'd also need to crop the lower Mp camera's image the same amount to have the same resultant subject size thus making them both look just as blurred as each other.

[ameerat42]

No one has mentioned the effect of image stabilization built into lenses where manufacturers state that it can allow you to handhold at up to 4 stops slower.

What does IS or VR in a lens really do to the 1/FL rule when you're hand holding the 100mm lens on a stationary scene

Under normal conditions, we should now be able to shoot at 1/10 sec instead of 1/160 using a Canon APS-C sensor

IS doesn't work this miracle very well for me, especially since I moved from a 6MP DSLR camera back in 2003 to 21MP now


Good point, Flashc. I guess it hasn't been mentioned because the discussion has been about basic techniques so far, but now that you've raised it...

You'd probably know about it, but IS is stated in terms of "stops". For instance, "4 stops" at a certain focal length. (As I understand it) that means that
the image is stabilised to 4 exposure stops (not f-stops, but in this case) of exposure time, ie, shutter speed. So, if your camera is set to 1/15sec, the
image will be stabilised to the equivalent of 1/250sec. (Sequence: 1/ 15, 30, 60, 120, 240.)

[Hamster]

The 1/FL rule is only a guide and has only ever been a guide as it depends on a number of factors and these other factors that are to be taken into consideration are:

1) How large you are going to print/view the image. A small image from just about any camera viewed on your phone screen can look sharp, but when you enlarge it to the size of your computer screen it may look quite unsharp and blurry depending on the image. This is where the "myth" about the new very high Mp cameras needing higher shutter speeds needs to be put to rest. The fact is, a high Mp camera will not look any worse then a low Mp camera when viewed at the same size and all other things being equal, ie physical camera sensor size (FF compared to FF for example), aperture, shutter speed, lens used , ISO etc. Where you may need to employ better camera technique and possibly employ higher shuuter speeds is when you decide to usitlise those extra Mp's and are going to display the image larger than that from a lower Mp camera or where you are going to crop significantly. It is only then that you may detect camera shake more so than from a lower Mp camera. However, you may have been able to hold the camera well enough for that not to be an issue.

2) Whether you are going to crop your image significantly when compared to a lower Mp camera. However, if you need to crop the high Mp cameras image, then you'd also need to crop the lower Mp camera's image the same amount to have the same resultant subject size thus making them both look just as blurred as each other.

Yes, I couldn't understand this "higher Mp needing better technique" view that appeared in discussions. I'm with you regarding the need to compare images at the same size. People seem to have this view that at an un cropped image on a high Mp sensor will magically show hitherto unseen levels of camera movement. You explained it well.

[Tony Jay]

The debate about shutter speed to focal length rears its head every now and again. It is rather a horses-for-courses sort of thing. (Obviously this is an issue of handholding.) If one is only posting images to flickr then motion blur (camera/photographer as opposed to subject) is much less of an issue compared to printing that image at size several feet by several feet.

This kind of blur is absolutely a function of sensor resolution and also sensel pitch (the two are related). In the (old!) days of slide photography a shutter speed of 1/focal length was generally regarded as adequate, provided one's handholding technique was good. Now, in the days of rapidly increasing resolution (or sensel pitch) of modern camera sensors, the reciprocal of focal length is usually not good enough. The rule that I use now (using a 36 MP sensor camera) when shooting handheld is at least the reciprocal of twice the focal length of the lens and preferably thrice.

Clearly, the way to go, if possible, is to use a tripod (or whatever one uses as a substitute). Then, shutter speed can probably used as a much creative tool, unencumbered, as it were, by the above constraints.

In summary, it will be obvious to anyone reading this post that I am a newbie to this forum. Nonetheless, IMHO anyway, the info is sound.

Tony Jay

[DacrimL]

Do you sharpen your photos in PP?
And there's no reason you can't post links to anywhere now.;)

Yes I do, mainly using LR 5.7 as I cannot get my head around PS.
Basically I attempt to get most images to a point where not to much PP is required, so now using the above information am going to start from scratch again and try to improve the images even further.

Just one other quick question regarding faster shutter speeds.......Will they impact any on the amount of noise per image or does that remain generally from higher ISO's?

[ameerat42]

Well, "impact". Try an earth-hitting asteroid, but for this...

Shutter speed DOES affect exposure, as part of the idea in the Exposure Triangle (see AP Library). Too little exposure and you get
underexposed images which will need boosting by s/w, usually with attendant noise, which is often treatable, but at cost of IQ.

However, ISO increase can add signal noise to images.

[Analog6]

It's a good rule. So if your camera's crop factor is 1.6 then for a 100 mm lens the equivalent is 160 mm, so you'd shoot at 1/160 - OR FASTER. It does depend on subject. Also remember modern cameras/lenses have image stabilisation - that wasn't available to us back in the 'old days'. The manufacturers say you can claw back 2 stops! :)
When shooting surfing with a 400-500 mm lens I used to shoot in shutter priority and start at 1/1000 shutter sopped and dial up accordingly, always on at least a monopod.
Experiment with your own camera and lens combinations, you'll be able to read off the EXIF data what speed you used when you look at the images onscreen.

[Mark L]

The amount you crop a photo can impact the noise.

[Cris]

Ok. i am now confused, surely you don't have to factor in crop factor when using this rule, The lens is still the same focal length lens regardless of what sensor it sits in front of, the only thing that changes is the field of view- from the sensor. So why would you need to factor in any adjustment because you are using a smaller sensor. The rules of physics surely don't change just because you are using a smaller sensor. Does someone who uses a D810, which I think has an option to shoot in dx mode all of a sudden have to increase their minimum shutter speed whilst still using the same lens. Seems odd, still the same sensor, just using less of it.

[Kym]

Ok. i am now confused, surely you don't have to factor in crop factor when using this rule, The lens is still the same focal length lens regardless of what sensor it sits in front of, the only thing that changes is the field of view- from the sensor. So why would you need to factor in any adjustment because you are using a smaller sensor. The rules of physics surely don't change just because you are using a smaller sensor. Does someone who uses a D810, which I think has an option to shoot in dx mode all of a sudden have to increase their minimum shutter speed whilst still using the same lens. Seems odd, still the same sensor, just using less of it.

Yes you do. The FL is always the same, but the image area is smaller thus any movement is more obvious (needs a diagram) -- but believe me it is needed to be accounted for (basic geometry)

[ameerat42]

...still the same sensor, just using less of it...

Yes, that is true, but what are you doing with the resulting image? Cropping it. To look at it "closer".
Cropping, whether done in-camera via a crop sensor, or out of camera via software, will render a closer
look at an image.

The rest is

- - - Updated - - -

Humph! What happened to my post??? I thought I had deleted the last line.
It led onto something else that I deleted. OK, now gone.

[Analog6]

Ok. i am now confused, surely you don't have to factor in crop factor when using this rule, The lens is still the same focal length lens regardless of what sensor it sits in front of, the only thing that changes is the field of view- from the sensor. So why would you need to factor in any adjustment because you are using a smaller sensor. The rules of physics surely don't change just because you are using a smaller sensor. Does someone who uses a D810, which I think has an option to shoot in dx mode all of a sudden have to increase their minimum shutter speed whilst still using the same lens. Seems odd, still the same sensor, just using less of it.

And if you haven't been factoring it in that may explain why you feel your images are not sharp enough. Digital photography IS a very steep learning curve, when I first started using a digital SLR (after 40+ years of film cameras) it took me two years to get my head around expose for highlights. I'd nearly always shot slide (transparency) film, where you expose for shadows. And there's a myriad of other small 'hooks' you'll discover along the way.

[arthurking83]

....

Rick I certainly remember the old rule and the simple logic to get the shutter speed in relation to the focal length of a lens.........it was just when I went digital the images taken with an old 35mm film camera somehow always seemed better and sharper than now.


The major (and significant) difference between then and now is that back then you probably never really viewed your (film) images at 100% zoomed view .. ie at the pixel level.

Now the software used to view your images makes it too easy to do this to assess image quality, so it's done as a matter of course.
There's nothing wrong with this, and makes it easy to sift through your images to organise which to archive and which to delete.

I'm not a true believer in the crop factor multiplier rule(of thumb) theory.

[swifty]

Ok. i am now confused, surely you don't have to factor in crop factor when using this rule, The lens is still the same focal length lens regardless of what sensor it sits in front of, the only thing that changes is the field of view- from the sensor. So why would you need to factor in any adjustment because you are using a smaller sensor. The rules of physics surely don't change just because you are using a smaller sensor. Does someone who uses a D810, which I think has an option to shoot in dx mode all of a sudden have to increase their minimum shutter speed whilst still using the same lens. Seems odd, still the same sensor, just using less of it.

Hi Cris,
you are right in that things haven't changed really just by cropping a smaller portion out of a D810 sensor for the same scene. But in effect what you are doing is magnifying what you're seeing when you consider the image as a whole ie. the whole FF image vs the whole DX image.
On the pixel level, nothing has changed. Viewing the FF image at 100% on your computer is the same as viewing its DX crop at 100% because they are viewed at the same magnification when both images come from the same sensor.

Read post #12 of what Lance said.

Essentially what we need to consider with this rule is the field of view and pixel density.
Forget the actual focal length of the lens for a moment and just consider that the wider the field of view, the slower the minimum shutter speed that is required to hand hold for stationary subjects.
The higher the pixel density, the greater the ability to discern movement when you magnify the resulting image.

The 1/FL rule was only convenient because the shutter speeds corresponded relatively nicely to minimum shutter speeds required on the 135 format (FF). It gives people a nice starting point. So for the rule to apply we always need to convert it back to FF equivalent focal lengths. And that's why the crop factor does apply to this rule.


Using the D810 with 36MP as an example. It has a FF sensor and hence no crop factor applies. Compared to a D700 with 12MP (another FF camera), things haven't changed except the pixel density. So there is a greater ability to discern movement when you magnify a D810 image compared to a D700 image but there are actually the same amount of movement for a given lens and shutter speed.
Furthermore, using the same lens (a prime lens for simplicity sake) now on a D7000 (a DX sensor camera with a crop factor of 1.5X) with 16MP. The field of view has changed but the pixel density is very similar to the D810. So with the same shutter speed, there's again the same actual amount of movement. But viewing the D7000 image as a whole is different to viewing the full D810 image. It is more magnified on the D7000 and hence any movement will be more discernible at the whole image level compared to the D810. Hence the need to increase shutter speed correspondingly to compensate. But at pixel level, because the two cameras have very similar pixel densities, essentially the same amount of movement can be detected in both images.
The same prime lens is now on a D5500 with 24MP on a DX sensor. The field of view has not changed when compared to the D7000 so for the same shutter speed, on a picture level they will appear similarly sharp with respect to hand movement because when viewed at the same sized whole-image level they are at the same magnification.
But when viewed at 100% on a computer screen, because of the greater pixel density on the D5500, 100% on screen with a D5500 is at a higher magnification than 100% on screen for a D7000. So there is a greater ability to detect movement on the D5500, but not because there is actually more movement. There are the same amount of movement.

Not sure if I managed to clear anything up or added to the confusion LOL.

[arthurking83]

Great reply Swifty and even tho it may have confused some .. it's basically a perfect reply as to what 'happens'.

My next comment applies only to myself and is not a scientific explanation of the exact physical properties of focal length vs crop factor vs 1/focal length rules and regulations .. but!

Since I've gotten my D800E, I noted in many situations where I can actually shoot at a lower shutter speed(using the same, equal or inferior lenses) than I used too with the D300(lower pixel density) and still see clearer images to a certain degree.
(this is why I don't buy the 1/focal length rule thingy, and especially not the crop factor addendum to it)

The reason is that I very rarely crop an image, and shoot for the entire pixel count to count in the final image.

So where I was using the D300, and wanted a 2000x3000 pixel variant of the final image(the D300 having a pixel count of 4200x2800) but now with the D800E(having 7300x4900 pixels) .. if I still only want/need/use that same 2000x3000 pixel level for a final image, with the D300 the ratio of pixels used compared to pixels created(negative magnification .. or whatever you want to call it) was always higher than it now is with the D800.

So with the D800's 7300x4900 pixels being scaled down(de-magnified, or extended, or reduced .. etc) more than the D300's were .. I see less unsharpness(or blurriness).
The actual image viewed at 100% may be less sharp, or more out of focus .. but because of the reduction of the total number of pixels used to a set number of pixel output .. the image can actually appear more usable!

This is why I laugh when people write up stupid remarks on cameras with ever increasing pixel counts on the basis that we don't need more pixels!(and also make comments to the contrary).

If the photographer is a habitual cropper ... whether by necessity or due to a lack of diligence(this is not a scorn of such practices!!!) .. they probably won't see this effect.
These habitual croppers tho, always benefit from more pixels .. just as they have since time immemorial.
I see this benefit too, up to a point and I don't crop.

And part of the proof of my assertion, is that over the time I've posted images up for review/viewing .. I have never read any comments that my images are unsharp or blurry .. even tho half of them probably are(at the pixel level).
People may not necessarily like the images I've posted, but not one comment made that the images are not sharp.

If you view your images at 100% to discern every pixel in all it's glory .. then the rules regarding focal length and crop factor multiplication probably applies .. even tho it doesn't make any sense.

Andrew(I@M) made a perfectly appropriate point (reply #3) about handholding technique.
Things to take note of: as you age you do lose the ability to maintain a level of steadiness that you may have previously been gifted with. This is a given .. a fact of aging.
It's recently happened to me too .. I used to easily solder the most minute electronic components even just a couple of years back .. now I can barely hold two ropes steady to tie them together :p
And the other thing to be weary of is over confidence in your technique.
I guess as keenly enthusiastic types about our chosen hobby/job/interest(that is photography), we're probably trying many things to improve it in some way.
Handholding technique is almost certain to be one of those aspects that we all have tried to improve upon in some way or another .. but I also think that we probably get a bit over confident in our ability and (for lack of a better term) 'slack off' a bit ... (guilty as charged!! :D).
Or, simply forgetting the basics of what has helped us to achieve this specific improvement in our photography.
That is you've done it .. you've improved your technique, but after some time with this improvement ... which by now a natural movement to you .. is it actually the way you originally started the process of improvement.


To the OP .. working out what shutter speed works best for you is a process of elimination that only you can work out for yourself!
The reason this is, is simple .. we're all too different to have a general broad based rule of thumb that just works!
Some of us are neurosurgeons .. and others probably shake like Shakin Stevens with a degenerative muscular condition.
You can't expect these two disparate types to follow the same set of physical expectations.