Required Lens Quality versus Final Output.

[xpatUSA]

Firstly, let me say that, by "quality", I am talking about resolution, sharpness, acutance and possibly micro-contrast - rather than coloration, distortion, vignetting, or even CA to an extent.

For a long time, I've agonized over 'this lens versus that lens' based on, shall we just say, resolution. I've read many reviews where lens X does 3550 l/ph versus lens Y at a mere 2950, etc. Not to mention center vs. corner sharpness. Such considerations can seriously affect the wallet: witness my recent purchase of a Sigma 17-70mm 'Global Vision' lens - even though my existing 17-50mm took pretty good shots. In spite of it's glowing reviews, the 17-70mm didn't seem much sharper than the 17-50mm, which got me to thinking - because I've had similar experiences before as lenses came and went.

Then, last week, a light bulb finally came on, an epiphany of sorts. The missing clue is the fact that I don't print and I view only on an HD monitor, currently a NEC P232W. Should that affect my personal choice of lenses?

Is it a bit like the MP thing?: For example, I know that I can put any lens I like onto my 3.4MP Sigma and abberations become harder to see and I can go up to f/16 before diffraction "sets in".

So I did a simple calculation which showed that viewing an image on my screen is the equivalent of only about 40 lp/mm in the image plane because, even though my favorite sensor is good for 100 lp/mm, the image has to be down-sized to 1920px wide for viewing fully, thereby losing 59 lp/mm of resolution. (not talking about zooming in or out here).

Of course, 40 lp/mm does not tell the whole story, does it? After all, if a lens chart shows zero MTF at 40 lp/mm in the center, it's no good for a sharpness freak like myself or anybody else, I imagine. On the other hand, buying a lens with 100% MTF at 40 lp/mm could cost a buck or two. So what's an acceptable value between 0 and 100%?

Many articles about MTF and lenses quote an MTF of 0.5 or 50% as being acceptable - the so-called "MTF50" criterion. From which I deduce that if any of my lenses has an MTF of 50% or more at 40 lp/mm, it is good enough for my work. If it were significantly less than 50%, then it would need other highly redeeming features to stay in the collection.

Those who print large at 600 ppi will have different needs, I'm sure, but I always feel better when I can put numbers to something in order to support a gut feeling.

My apologies in advance to anyone who finds this post too technical or too confusing.

[rpcrowe]

I do understand the diverse complexities of lenses but, at times I think that many photographers get too technical when considering their photo needs.

I certainly agree with you that lens (and camera) needs can be impacted greatly by the intended output. IMO, if a photographer were intending to print large wall size images, then he/she would need a "better" lens than a person who only posts the images on FaceBook or emails images of the kids and Rover, the dog, to the grandparents.

However, there are venues in which "better" lenses are needed to get any type of good or even adequate imagery. These venues include, but are not limited to: sports (especially night or indoors), wildlife (needing a long focal length), and many low light shooting situations. Action shooting often requires very fast and accurate auto focus, although with skill and proper techniques, you could shoot action with manual focus. Heck, action was shot ONLY with manual focus cameas/lenses until the advent of the auto focus technology in the late 1980's...

However, if one can afford top line glass without taking food off ones table and if one gets enjoyment using top line equipment, GO FOR IT! However, buying increasingly better and more expensive equipment in seeking the Holy Grail of outstanding imagery is probably a waste of money. Using better techniques can often elevate the quality of one's imagery far better than can a better lens.

OTOH: there are times when better glass is the ticket to generally elevating image quality and to elevating skills. These often go hand-in-hand. An example of this is Brian's Philippines macro efforts. IMO, his images have increased in quality drastically since he started using the Tamron 90mm f/2.8Macro....

[Manfred M]

Ted - There is nothing to prevent you from pixel peeping, as in my view, that is going to be the only way that you are going to see a lot of difference in lens performance. I will suggest that this is generally true, even in prints. When viewed from a reasonable viewing distance the prints will look good, and lens differences will only really show up when you either look at them very close up or look at two images side by side.

That being said, I find it impossible to get these high quality results when hand holding. A locked down, anchored heavy duty tripod can end up giving me sharper images with a cheap lens than a handheld shot at reasonable shutter speeds and ISO settings. Frankly, I generally don't worry about what the MTF charts say, etc. When I buy a lens, I look for functionality (i.e. focal length range, minimum focus distance, maximum aperture, etc) rather than sharpness, just because my experience suggests I can get "good enough" performance out of just about any modern lens (and a few 40 years old ones too).

[chauncey]

I admit to using those MTF charts in purchasing of lenses, but...there are constant evaluations that
show that one cannot identify which lens was used on a particular print. Go figure.

[Shadowman]

Firstly, let me say that, by "quality", I am talking about resolution, sharpness, acutance and possibly micro-contrast - rather than coloration, distortion, vignetting, or even CA to an extent.

For a long time, I've agonized over 'this lens versus that lens' based on, shall we just say, resolution. I've read many reviews where lens X does 3550 l/ph versus lens Y at a mere 2950, etc. Not to mention center vs. corner sharpness. Such considerations can seriously affect the wallet: witness my recent purchase of a Sigma 17-70mm 'Global Vision' lens - even though my existing 17-50mm took pretty good shots. In spite of it's glowing reviews, the 17-70mm didn't seem much sharper than the 17-50mm, which got me to thinking - because I've had similar experiences before as lenses came and went.

Then, last week, a light bulb finally came on, an epiphany of sorts. The missing clue is the fact that I don't print and I view only on an HD monitor, currently a NEC P232W. Should that affect my personal choice of lenses?

Is it a bit like the MP thing?: For example, I know that I can put any lens I like onto my 3.4MP Sigma and abberations become harder to see and I can go up to f/16 before diffraction "sets in".

So I did a simple calculation which showed that viewing an image on my screen is the equivalent of only about 40 lp/mm in the image plane because, even though my favorite sensor is good for 100 lp/mm, the image has to be down-sized to 1920px wide for viewing fully, thereby losing 59 lp/mm of resolution. (not talking about zooming in or out here).

Of course, 40 lp/mm does not tell the whole story, does it? After all, if a lens chart shows zero MTF at 40 lp/mm in the center, it's no good for a sharpness freak like myself or anybody else, I imagine. On the other hand, buying a lens with 100% MTF at 40 lp/mm could cost a buck or two. So what's an acceptable value between 0 and 100%?

Many articles about MTF and lenses quote an MTF of 0.5 or 50% as being acceptable - the so-called "MTF50" criterion. From which I deduce that if any of my lenses has an MTF of 50% or more at 40 lp/mm, it is good enough for my work. If it were significantly less than 50%, then it would need other highly redeeming features to stay in the collection.

Those who print large at 600 ppi will have different needs, I'm sure, but I always feel better when I can put numbers to something in order to support a gut feeling.

My apologies in advance to anyone who finds this post too technical or too confusing.

I also look for reviews that very often use SQF charts (Subjective Quality Factors) and depending on which reviewer is quoting the data; there are times when print size/ISO settings are used in the evaluation. The use of SQF data is a personal choice, when I first got involved in digital photography doing large prints was a deciding factor. Now that I feel I have some of the basics of what it takes to make a good print I've stopped chasing that "perfect lens". By the way, I feel I've found the perfect lens, I won't mention which one it is as there's no need for arguing.

http://www.imatest.com/docs/sqf/

[xpatUSA]

I also look for reviews that very often use SQF charts (Subjective Quality Factors) and depending on which reviewer is quoting the data; there are times when print size/ISO settings are used in the evaluation. The use of SQF data is a personal choice, when I first got involved in digital photography doing large prints was a deciding factor. Now that I feel I have some of the basics of what it takes to make a good print I've stopped chasing that "perfect lens". By the way, I feel I've found the perfect lens, I won't mention which one it is as there's no need for arguing.

http://www.imatest.com/docs/sqf/

Thanks for your thoughts, John.

Bob Atkins has a good article about SQF too:

http://www.bobatkins.com/photography.../mtf/mtf4.html

It would probably pay me to return to the subject of SQF, in order to determine how it relates to people like me with bad eyesight who only view pics on their monitors.

[rpcrowe]

I bought my CF cards on sale... I have six Lexar UDMI cards of 400x or faster 16GB to 32GB. Since I download each day (even when traveling) that is plenty of memory when I shoot with a pair of 7D cameras or my 5Dii.

[dje]

Thanks for your thoughts, John.

Bob Atkins has a good article about SQF too:

http://www.bobatkins.com/photography.../mtf/mtf4.html

It would probably pay me to return to the subject of SQF, in order to determine how it relates to people like me with bad eyesight who only view pics on their monitors.

Yes Ted most of what I've read on SQF relates to how the eye/brain responds to viewing prints and it is based on angular resolution of the eye and extensive tests on real people. I've not seen anything on how this concept can be related to viewing a screen at 100% view at a certain distance but you would think that someone could work something out mathematically. (not me I'm afraid!)

Is it a bit like the MP thing?

Yes I think so. The pixel pitch needs to be fine enough so that the mtf caused by the sensor plus AA filter does not swamp the mtf of the lens. For those who own cameras like the D810, the lens mtf is quite significant.


Dave

[xpatUSA]

Yes Ted most of what I've read on SQF relates to how the eye/brain responds to viewing prints and it is based on angular resolution of the eye and extensive tests on real people. I've not seen anything on how this concept can be related to viewing a screen at 100% view at a certain distance but you would think that someone could work something out mathematically. (not me I'm afraid!)

Yes, there's this graph of contrast ratio versus angular frequency. Angular frequency is often shortened to "cpd" so no-one knows what it is . . didn't fool me though

So, if I view my monitor at about 12" (305mm), I can just barely make out the 0.265mm pixels and that would be about 10 cycles per degree . . . seems about right for an old geezer. Not much contrast though; for me the peak contrast is quite a bit lower than that, more like 2-4 cpd.

Yes I think so. The pixel pitch needs to be fine enough so that the MTF of the sensor plus AA filter does not swamp the MTF of the lens. For those who own cameras like the D810, the lens MTF is quite significant.
Dave

Indeed, the D810's 4.88um pixels with AA filter give 83% MTF at 40 lp/mm - not bad at all. As opposed to a humble 71% for my Sigma SD10 with it's 9.12um pixels. So my spread-sheet says, anyhoo . . .

[Shanghai Steve]

I rarely print myself and there is a basic reason for that. It costs a lot of money. From the specialist photographic printers I've seen, one 14 by 21 inch print costs 250 RMB or more. That adds up quickly. One can view one's best 100 shots quickly, easily and for free on a decent monitor, but to print them professionally at a good size would cost two and a half thousand pounds! Building up a good printed portfolio is one hell of an outlay.

That being said, I would not be happy looking at files I know do not have the resolution to print out well. I tend to follow my grandmother's maxim "Buy quality and treat it well." While there are some thoroughly decent affordable lenses such as the classic 50 D from Nikon, I feel it is often sensible to buy the best one can afford, or perhaps more accurately, the best one is comfortable owning. That said, I think with Nikon at least, many of their older "D" range lenses are both affordable and very good. I am not a fan of the G series lenses. The new Sigma Art line is great....if sharpness is your only criterion. But when one starts to consider properly high end glass it is as much a question of the character of the lens as anything else- technical problems (hopefully) being more or less dealt with. That extra you spent 8 years ago is easily paid for by the pleasure of ownership, wonderful bokeh, more accurate focussing, better build quality etc. As the years roll on, one seldom regrets owning quality equipment, as long as you can afford it in the first place. I know someone who's been married for 30 years and his wife STILL brings up the fact that he bought her a cheap engagement ring because he managed to get a great deal on a Gibson guitar the same month he proposed.

[JBW]

Firstly, let me say that, by "quality", I am talking about resolution, sharpness, acutance and possibly micro-contrast - rather than coloration, distortion, vignetting, or even CA to an extent.


My apologies in advance to anyone who finds this post too technical or too confusing.

Hi Ted, my apologies to everyone who doesn't find my reply technical enough.

I'm going to approach your question from the other direction. I care about visible IQ.. On the web when I look at medium format shots using great glass I can't see much if any difference between them and shots taken with full frame Good quality cameras (pick your brand) and throw on some good glass.

The major improvement for me comes when I compare what my Alpha a58 connected to my Tamron produces compared to a full frame connected to a Tamron.

Seems to me a Quality full frame connected to good quality glass is about as good as you can go for the web.
B.

[xpatUSA]

Hi Ted, my apologies to everyone who doesn't find my reply technical enough.

I'm going to approach your question from the other direction. I care about visible IQ.. On the web when I look at medium format shots using great glass I can't see much if any difference between them and shots taken with full frame Good quality cameras (pick your brand) and throw on some good glass.

Yes, Brian, I would expect that to be the case if viewed on the web with the whole image visible on (fitted to) the screen.

The major improvement for me comes when I compare what my Alpha a58 connected to my Tamron produces compared to a full frame connected to a Tamron.

May I ask how you view images for comparison. In particular, do you fit both images to the screen or do you zoom in to see the pixels at 100% or more?

Seems to me a Quality full frame connected to good quality glass is about as good as you can go for the web.
B.

Yes, if I had like a D810 with an Otus 55mm on it that would certainly would be as good as I could go for the web. Though I'm not convinced that such quality gear would show that much better sharpness on my monitor than say a D700 with a 55mm kit lens on it.

In this thread, I'm talking about whole images viewed on my screen with no zooming in to peep at detail - which is equivalent to stepping up close to a 'B' size (A3) print with magnifying glass in hand, eh? . . .

[JBW]

May I ask how you view images for comparison. In particular, do you fit both images to the screen or do you zoom in to see the pixels at 100% or more? . . .

Unless I am choosing a shot to work on I usually flip back and forth. I figure that if I have to pixel hunt side by side there's not that much difference.

Here's what I'm talking about. Bill is CiC member who uses a Tamron 90. But he has it attached to a much better ( and significantly more expensive ) camera. It doesn't require pixel hunting to see the iq difference between our two rigs.

[xpatUSA]

Unless I am choosing a shot to work on I usually flip back and forth. I figure that if I have to pixel hunt side by side there's not that much difference.

Here's what I'm talking about. Bill is CiC member who uses a Tamron 90. But he has it attached to a much better ( and significantly more expensive ) camera. It doesn't require pixel hunting to see the iq difference between our two rigs.

They're very nice, detailed shots. Were they cropped or significantly re-sampled or highly post-processed, do you think? That could make quite a difference between his'n and your'n when viewed on the same monitor.

My original post though was trying to avoid side issues like zooming in/out, cropping and such.

[JBW]

They're very nice, detailed shots. Were they cropped or significantly re-sampled or highly post-processed, do you think? That could make quite a difference between his'n and your'n when viewed on the same monitor.

My original post though was trying to avoid side issues like zooming in/out, cropping and such.

Is it possible to avoid these issues? IQ is all about what we see and and all of these change iq.

As to your questions... I can't imagine getting shots like his SOOC. But how much he does I don't know.

[xpatUSA]

Is it possible to avoid these issues?

Of course. Sometimes I get a bit carried away, sorry.

IQ is all about what we see and and all of these change iq.

As to your questions... I can't imagine getting shots like his SOOC. But how much he does I don't know.

Again, sorry for prying.

[JBW]

Of course. Sometimes I get a bit carried away, sorry.



Again, sorry for prying.

no apology needed. As for getting carried away I'm heading out for third try at a caterpillar if he's still there