IQ to subject distance test...

*FHPhotographer* · 10-10-2008, 10:28 PM

This is one of those things that feels like pushing on an ice cube... the more I push, the faster it slides away from me

I get the perspective changes, and the DOF changes, with subject distance, but I wanted to find out if there is an image quality change that comes with increased/decreased subject distance.

Specifically, assuming a constant aperture and lens focal length (in this case FA 50 f/1.4) does image quality vary with the subject distance (focal subject to lens, beyond the minimum focusing distance)?

The three images below were shot under the same conditions with the 50 f/1.4 mounted on a K100D. All were shot at f/4, with shake reduction off, tripod mounted and 2-second delay. The first battery is 18" from the front of the lens, the second is 30" and the third is 42":

So, is there an IQ difference that correlates with subject distance?
Brian (FHPhotographer)

SpecialK · 10-10-2008, 10:43 PM

Not unless you have a poor lens, I think.

nobbsie · 10-10-2008, 11:54 PM

This has got me thinking too. One factor you do need to account for is a closer subject will be larger in the frame, take up more pixels and therefore be more detailed (assuming you use the same focal length of course).

I believe most lenses are optimised for an intended use, providing the best resolution possible at a certain focal distance common to that use. For instance I know my M28 performs much better for landscapes than for narrower subjects, which makes sense since on film it was a wide-angle lens.

*FHPhotographer* · 10-11-2008, 12:12 AM

Originally Posted by nobbsie

This has got me thinking too. One factor you do need to account for is a closer subject will be larger in the frame, take up more pixels and therefore be more detailed (assuming you use the same focal length of course).

I believe most lenses are optimised for an intended use, providing the best resolution possible at a certain focal distance common to that use. For instance I know my M28 performs much better for landscapes than for narrower subjects, which makes sense since on film it was a wide-angle lens.

Just to make sure we're talking about the same thing and using the same terms: when you say "best resolution possible at a certain focal distance" do you mean subject distance ?

If we are talking about the same thing, I wonder if Pentax (or any lens manufactuer) would make that optimized set of conditions available to consumers? Just a thought,
Brian (FHPhotog)

nobbsie · 10-11-2008, 07:43 AM

Originally Posted by FHPhotographer

Just to make sure we're talking about the same thing and using the same terms: when you say "best resolution possible at a certain focal distance" do you mean subject distance ?

If we are talking about the same thing, I wonder if Pentax (or any lens manufactuer) would make that optimized set of conditions available to consumers? Just a thought,
Brian (FHPhotog)

Yes that' exactly what I meant. off course I have no evidence at all to back it up, just general impressions from the lenses I use.

Wheatfield · 10-11-2008, 10:10 AM

Originally Posted by nobbsie

Yes that' exactly what I meant. off course I have no evidence at all to back it up, just general impressions from the lenses I use.

I've noticed absolutely that lenses perform differently over varying subject distances.
Some worse than others. The M50/4 macro is a very good macro lens, but quite an iffy performer at infinity, IIRC, the FA*85/1.4 is excellent at about 10 feet, not so good at distance.
My FA200/4 macro is excellent at close focus distances (as well it should be), but even at 15 feet or so the bokeh is falling apart (I realize this isn't IQ specifically, but is still an important imaging characteristic)
Seems to me this would be something that lens testers should be interested in. For example, if the guy who does the photozone tests isn't shooting the lens at varying focus distances, then it is hard to know if his tests are valid.

Or am I misinterpreting the entire discussion??

Marc Sabatella · 10-11-2008, 02:19 PM

I think based on this and the similar on the other forum, you are making *far* too big a deal of this.

The obvious - and I mean *obvious* - answer is that the further away something is, the less detail you will be able to see. That's true of our naked eyes, that's true when looking through binoculars, when looking through a camera viewfinder, and it's true of the final image. And it's true for the *obvious* reasons - no complicated mathematical DOF forumlas should be required to see this.

I mean really, pretend you had never held a camera. Ask this question: can I see more details on an object that is further away or one that is closer? I think most 4-year-olds could answer this question, and this answer would be exactly the right answer for exactly the right reasons.

Now, if any given lens happens to have some quirks where it actually performs better in terms of color or contrast at some specific subject distance, that may well be, but it would be entirely lens specific. And it's pretty unlikely that any such effect would trump the *obvious* fact that the further away an object gets, the less detail you can see. A particular lens might have more contrast for a subject 20 meters away versus 2 meters away, but no way is it going to provide more resolution. That's as patently impossible as it would be for our unaided eyes.

*k100d* · 10-11-2008, 03:22 PM

i think wheatfield and marc have it right too
depends on the lens. portrait / macro lenses are more optimized for shorter focusing distances while may be pretty crap at infinity.

*FHPhotographer* · 10-11-2008, 09:19 PM

Originally Posted by Marc Sabatella

I think based on this and the similar on the other forum, you are making *far* too big a deal of this.

The obvious - and I mean *obvious* - answer is that the further away something is, the less detail you will be able to see. That's true of our naked eyes, that's true when looking through binoculars, when looking through a camera viewfinder, and it's true of the final image. And it's true for the *obvious* reasons - no complicated mathematical DOF forumlas should be required to see this.

I mean really, pretend you had never held a camera. Ask this question: can I see more details on an object that is further away or one that is closer? I think most 4-year-olds could answer this question, and this answer would be exactly the right answer for exactly the right reasons.

Now, if any given lens happens to have some quirks where it actually performs better in terms of color or contrast at some specific subject distance, that may well be, but it would be entirely lens specific. And it's pretty unlikely that any such effect would trump the *obvious* fact that the further away an object gets, the less detail you can see. A particular lens might have more contrast for a subject 20 meters away versus 2 meters away, but no way is it going to provide more resolution. That's as patently impossible as it would be for our unaided eyes.

Before I lay several hundred dollars on the line I want to know what the lens can do for me … not in theory, not in bench testing, not because of an MTF table, but from having drawn on the experience of others in the real world. That's what I thought the Forum was for, to get the feedback from people who have empirical knowledge of what a lens does well, and not so well.

Marc, I have great respect for you knowledge and willingness to share that knowledge with the Forum; however, I'm going to keep asking my question my way, making as big a deal as I feel the need, and asking those questions as often as I need to get either the answer I'm looking for or to become convinced there is no reasonable answer to be found.

So if I'm making a big deal of issues that may seem simplistic to some (but perhaps may be informative to others) I will continue to do so and in terms that have relevance to me. I suppose if some in the Forum community aren't interested in my questions or the responses, they will simply ignore my threads and responses, or will tell me to put a sock in it. In either case, I'll play that as it lays,
Brian (FHPhotographer)

*FHPhotographer* · 10-11-2008, 11:25 PM

Originally Posted by Wheatfield

I've noticed absolutely that lenses perform differently over varying subject distances.
Some worse than others. The M50/4 macro is a very good macro lens, but quite an iffy performer at infinity, IIRC, the FA*85/1.4 is excellent at about 10 feet, not so good at distance.
My FA200/4 macro is excellent at close focus distances (as well it should be), but even at 15 feet or so the bokeh is falling apart (I realize this isn't IQ specifically, but is still an important imaging characteristic)
Seems to me this would be something that lens testers should be interested in. For example, if the guy who does the photozone tests isn't shooting the lens at varying focus distances, then it is hard to know if his tests are valid.

Or am I misinterpreting the entire discussion??

... when you said "FA*85/1.4 is excellent at about 10 feet, not so good at distance." That's what I'm looking for because if I like to work at 15-20' subject distance, this may not be the lens for me. It's still a good lens, it just isn't a lens to meet my expectations.

I'm guessing these "best subject distance" numbers exist for most camera+lens combinations. If we compiled them, we'd have something for newcomers like me to use as an emperical baseline.
Thanks,
Brian (FHPhotographer)

Marc Sabatella · 10-12-2008, 12:46 AM

Originally Posted by FHPhotographer

That's what I thought the Forum was for, to get the feedback from people who have empirical knowledge of what a lens does well, and not so well.

Well, sure. It just really seems to me there can be no objective answer to your question aside from the one I gave. Every lens is different, everyone's opinion of image quality is different. It's not like DOF where there are very specific and objective ways in which it depends on subject distance. But when it comes to IQ, the only real absolute here is that more distance means less resolution. Beyond that, it's too lens-specific and too subjective to say much more.

Now, if you are *wanting* lens-specific and subjective info, that's fine, but your original question didn't make this clear - it seemed you were looking for a more general and objective answer. And I don't think there is one.

But in any case, please don't take offense at my post. I was just trying to make clear that what I was saying really wasn't the result of some complex and obscure mathematical formula, but was exactly as straightforward as it seemed. I wasn't trying to imply it was stupid question - just that the very simple answer is really the only answer to the question as I understood it.

Wheatfield · 10-12-2008, 08:10 AM

Originally Posted by Marc Sabatella

Well, sure. It just really seems to me there can be no objective answer to your question aside from the one I gave. Every lens is different, everyone's opinion of image quality is different.

Something I've learned over the decades is that there is nothing wrong with subjective opinions, providing the person making the opinion has enough experience to have a valid opinion.
For example, when I mention that the M50/4 macro is tack sharp at macro focusing distances, but iffy when used as a general purpose lens, I am making an admittedly subjective evaluation based on comparative experience with other 50mm lenses that I own, but an opinion I don't feel the need to quantify.
Rather than waste my time proving how bad the lens is when compared to something that works better, I'll just shoot with a K, M, A, or FA 50mm lens when wanting a general purpose lens, and save the macro for what it seems to do well.

Of course, I'm not a person who believes that to prove something you must collate results and quantify everything and come up with a completely objective summation.
My eyes still work pretty well, and can tell me when what I am looking at is good or not, and I trust my eyes to allow me to form good subjective opinions.

Marc Sabatella · 10-12-2008, 01:23 PM

Originally Posted by Wheatfield

Something I've learned over the decades is that there is nothing wrong with subjective opinions, providing the person making the opinion has enough experience to have a valid opinion.

I'd agree, but I'd also add it helps if the person *reading* the opinion has enough experience to put the opinions he is reading into context. For instance, I read that the M50/4 macro lens performs better up close than at infinity. Well, OK, I have enough experience to know that *all* lenses do that. So I interpret the opinion I am reading as meaning that the lens in question demonstrates this effect more than other lenses that the person expressing the opinion has compared it to. Fine. But, while I may know that the performance of the M50/4 at infinity is worse than at 1 foot, and I now know that the difference is "more" than it would be for another lens, I *still* don't know if the 50 macro is actually better or worse at infinity than any *paticular* other lens. Say, the M50/2, which my experience says is perhaps the least sharp 50 Pentax ever made. That is, I know the fall-off in resolution is greater for the M50/4, but it started from so much higher, it might *still* be better at infinity. So I still cannot make a conclusion unless someone offers that specific bit of comparative data as well. But then, I still wouldn't t know if the M50/4 is better at infinity than, say, the kit lens. Or if any of this actually varies with aperture. Or, for that matter, if the differences in resolution are large enough to be noticeable on my particular camera, or if they are enough to trounce other potential IQ advantages of one lens over another (such as bokeh, contrast, or color). All I have is one little piece of data, but ot enough to put it into the context I'd need.

Meaning, I can take you at your word that the M50/4 macro suffers at infinity, but that still doesn't give me enough information to determine if it might nevertheless be an improvement over anything I already at infinity. *You* might know what other lenses it performs better than, and in what ways, and at what apertures, but simply reading a post stating that the M50/4 suffers at infinity, *I* would still have no way of knowing what this might mean in practice.

Anyhow, again, my point isn't to say there is no worthwhile data to collect. Indeed, it could be quite interesting. But we are talking about a pretty big data collection & collating exercise in order to make the information useful.

*FHPhotographer* · 10-12-2008, 06:17 PM

Originally Posted by Marc Sabatella

Anyhow, again, my point isn't to say there is no worthwhile data to collect. Indeed, it could be quite interesting. But we are talking about a pretty big data collection & collating exercise in order to make the information useful.

Marc, no harm no foul. And I'd be happy to monitor, tabulate and present it to the forum... but only on condition that there is a consensus among posters for criteria, and that the criteria have to apply to all Pentax dslrs, i.e., no inclusion of data that says the 50 f/1.4 is best at 4-8 feet but only when used on the K20.

I will take the lead from Forum folk with more experience to set the criteria, but I would think we should as least ask participants:
1. tell us the lens and body
2. ID any settings that affect that combination (aperture, speed, ISO?)
3. and the question... what is an ideal specific subject distance range for that camera+lens , i.e., a response that says "2 feet to infinity" won't be tabulated, but responses that say 4-8 feet, or no closer than 10 feet will work.

If there is interest in this, I'll start a polling thread to gather the criteria points (by majority vote), then start the next polling thread to gather the data.
Brian (FHPhotographer)

*sewebster* · 10-12-2008, 07:21 PM

You could reproduce resolution charts in various sizes and place them at various distances from the camera. You adjust the size so that the chart always looks the same size in the picture. Of course the chart for "infinity" is going to have to be rather large.

It is perhaps interesting that all of the resolution tests that people do are at a certain intermediate subject distance. Often they point out that things could be different for macro work (and they could also be different for infinity) because the lens elements are physically in different positions for different focus distances.