[Ian Stuart Forsyth]

What you mean by "level of resolution" is ambiguous

Captured resolution

Ands is the big reason 45 MP is not much better than 36 MP. As you add more an more MP, you get less and less for them.

Correct and this is why you need a lens that can deliver more resolution, If you have 2 sensors that both contain the same amount of pixels but on one of the sensors those pixels are half the size you need that lens to resolve more for that sensor to capture the same resolution as the sensor with larger pixels

if and only if the lens referred to isn't sharp enough already, which any lens made in the last 20 years is, as far as I know.

Even with the K5 sensor you can see a difference in the captured resolution, so much so that you can use a antique lens like push pull m 80-200 f4.5on the ist ds you capture more resolution per pixel than you can with the dfa 100mm marco on the pentax k5


Here is the comparison both 100% crops and shot at the same DOF and held the same FOV

Here is the K5 along with the dfa100mm and the ist ds with the dfa 100mm

Held the same framing by moving the ist ds closer
Here you can see how much captured resolution is degraded when the pixel size decreases

If you would like another resource
Canon EF 85mm f/1.4L IS USM Lens Image Quality
same lens on 2 different size sensors

Ever wonder why a 24mp FF camera needs a AA filter while a 24mp cropped body does not. The cropped body is using the lower resolution that the sensor can capture from the lens as a AA filter.
If

[normhead]

Held the same framing by moving the ist ds closer

Your methodology is seriously flawed in so many ways. Moving closer to the subject creates more resolution on the subject, but has no relationship to either lens or sensor resolution. I actually quite like some of my Sigma 70-300 images taken from 8 feet away. Taken from 15 feet away, I need a better lens. Handicapping the better sensor by backing it up makes the whole comparison inaccurate. Resolution is measured in lw/ph. Moving the *ist up makes the image smaller. Even though the images may look similar, the picture real world height on the *ist is much smaller, so in lw/ph it still loses even in your example.

When you find a test subject for these things you have to make sure you have detail one lens can resolve that the other can't. A concept explained here, Understanding folks saying a TC doesn't ad more detail. - PentaxForums.com

But first, to test one lens or one sensor against another you have to keep the distance from the subject identical for both systems. Your slight of hand, saying a *ist can resolve at 10 feet the same K-5 can at 20 pretty obvious if you just state all the numbers.

"If you don't like your pictures move closer" has been around since forever. Closer is better. If you need to get closer to draw even, that just proves how much worse you are. Sometimes I seriously wonder how you come up with these things.

[Kunzite]

If you allow me to chime in - perhaps you're talking about slightly different things?
One thing I've noticed in Ian's posts is that he is constantly varying the frame size. He's trying to prove that a 24MP FF sensor can resolve more detail than a 24MP APS-C sensor, and tries to approximate that by comparing a 6MP APS-C image with a 6MP crop from a 16MP APS-C image.

I wouldn't say it's the approach I'd be using, first and foremost because it doesn't measure what we want to know. Here's something that does:
Studio shot comparison: Digital Photography Review

Changing the conditions would just introduce unwanted variables; for example, the K-5's quite strong AA filter - Pentax abandoned AA filters in their latest cameras, which gives them a slight edge.

IMO the difference between a 24MP FF and a 24MP AA-less APS-C is minimal, in good light (and of course with a good lens). Sort of what we can see in the link above.
In the past you'd need a few extra MPs for an APS-C to match a FF (again, in good light and with a good lens). Or a sharper lens.

I guess nobody's denying that a higher resolution sensor gives you in general better detail than a lower resolution sensor of the same format.

[Nicolas06]

That's nonsense. We aren't talking about capturing at the same rate, there is a law of diminishing returns but that in no way means the sensor out revolves the lens. We are talking about whether a sensor out resolves a lens. When the sensor out resolves the lens, adding more pixels on the same sized sensor will produce no increase in lw/ph. Not a reduced rate of capture, no increase at all. From what i've read in actual scientific papers, for APS-c that will happen somewhere around 50 MP. With 24 MP there is still a lot fo possibility of more resolution by a sensor with more pixels. There is a 28 MP APS-c sensor out there somewhere that has already proved that. Rate of capture has nothing to do with anything. A sensor with four pixels positioned in a square, could probably reproduce 2 lw/ph with 100% efficiency. It all goes downhill from there. The higher the resolution the lower the capture efficiency.

That the sensor outresolve the lens or the lens out resolve the sensor is just words and has no practical impact.

What is practical is that with the same sensor, usually a better lens give you better results. Typically, more accurate/better AF, better constrast/micro constrast, more pleasing overall images and less time spent in post processing. Displayed or printed, a given shot at iso 100 with K5 and FA31 has more change to be pleasing than a K3 shot with 18-135 @31mm. And you can get it at f/2.5 if necessary.

What is pratical is with a larger sensor you can get away with smaller appertures for the same lens getting much better constrast, micro constrast and performance in borders/corners. This means you can get with zooms results that were only achievable with primes (like a 24-70 or 70-200 on FF vs APSC primes) or that a 150 euro/dollar entry level prime on FF may give better results than an expensive 1000 euros/dollard prime on APSC...

What is practical with a sensor with more pixel density if that in good light you can crop more if you are limited in reach by your lenses. This apply mostly to wildlife and macro.

Now in all cases, this isn't the most visble with quite similar sensor size like APSC vs FF.

[following.eric]

i would say:
if 2018 goes by without significant improvements in AF
and burst speed and capacity... its totally legal to jump ship. i will also leave if nothing comes for us. the SONY A9 is my body of choice... maybe the Sony A99II.
or just a used Sony A6500 to have a good start in "e-mount lens addiction". and 1 year afterward upgrading to a FF body.

Funny you mention Sony, I have decided to sell all my gear, even though I just got new lenses and will hopefully be making the switch to Sony. After getting the Tamron 70-200mm I really realized how heavy it can get and I want to go lighter and more compact. I also keep buying different lenses likes 50mm, 15mm, 28mm and find I never use them much or at all. I just really want a decent wide for landscape and a decent lighter zoom for landscape/everything else.

I have also really been thinking of whats to come, if I will always be behind the game, buying older system when the price comes down or second hand systems, then I want to put myself in a situation where good glass can just move from body to body while still improving. I look at Sony or Canon and if I went with a 5DIII or Sony A7II for what I can afford, I can already see a progression in a year or so as the next body starts to decrease in price (5D4 or Sony A7Rii & eventually A7Riii)

at the moment I dont see the KP as the next body for me, the K3iii is a rumour and the K1 would still mean a complete system change and new glass plus less compact like Canon.

Since I don't have money to switch systems every year, I am not a pro and still learning, I am thinking of it as an investment and I think Sony will be a better investment at the moment, They are already leading the way for mirrorless and I know the future at the moment looks good A7Riii and A9. who knows what they release next year and I could be set for the next decade.

I couldnt have asked for a better system to teach me about photography. The K50 taught me so much and had so many features that other entry level cameras only dream of, and because of that I think I am more knowledgable then I am had I been using another system. But a few years down the road things change.

So at the moment, selling gear and looking at the A7II, FE16-35 & FE70-200mm and that's it. and obviously a few extra batteries. Then maybe in two years get the A7Riii. I know every system has it pros and cons and that understandable. But loking at future models already released and the direction they are going and sony seems to be a leader in mirorless and the next move.

But first need to get some cash back from current gear

[Alex645]

I wouldn't say it's the approach I'd be using, first and foremost because it doesn't measure what we want to know. Here's something that does:
Studio shot comparison: Digital Photography Review

Changing the conditions would just introduce unwanted variables;

I guess nobody's denying that a higher resolution sensor gives you in general better detail than a lower resolution sensor of the same format.

As much as I find the dpreview studio scene comparison somewhat helpful, to some degree it's not comparing apples to apples; for example, with a Nikon FF they use an Nikkor 85mm and with the Pentax K1 they use the Pentax 77mm. In RAW, we are seeing both the result of different sensors, but also thru different lenses.

So it's a realistic comparison using similar primes as would be used by Pentaxians or Nikonians, but not identical primes unless you're comparing only Nikon FF sensors. And anyone's expectations of dpreviews studio comparison tests will change if you take a superior sensor with a Tamron 18-200mm zoom @ 55mm vs. an inferior sensor with a DA* 55mm prime.

The current debate, which seems off thread to the OP, is whether and how a lens can or can't outresolve a sensor and why. This seems to belong to it's own thread.

---------- Post added 12-30-17 at 01:07 PM ----------

Funny you mention Sony, I have decided to sell all my gear, even though I just got new lenses and will hopefully be making the switch to Sony.
I am thinking of it as an investment and I think Sony will be a better investment at the moment, They are already leading the way for mirrorless and I know the future at the moment looks good A7Riii and A9. who knows what they release next year and I could be set for the next decade.

Eric, glad you came to some conclusions and hope our posts were helpful. My only words of advice is that you make your decision based on what gear you need or want today, and not try to foresee the next decade; the future.

I love Sony products, but they love to innovate and have dumped products faster than any other large company I can think of.

In the past for video it was betamax to 8mm to Hi-8mm to miniDV, etc. My school invested about $3000 on two brand new Sony UP-DR80 dye sublimation printers. One year after the purchase, the required paper and ribbon was outsourced to Dai Nippon Printing. And then one year after that, DNP only made paper and ribbon for their own DS80 models and the two Sony printers became fully functional bricks. Sony corporate did nothing and we ate it. It's the Sony culture to ax any technology that isn't hot. One reason I appreciate Pentax is because I have found them, more than any other brand, fairly future-proof.

I only mention this because you used the word "investment". I still have burn scars from my Sony experiences.

[Kunzite]

The DPR link responds to the "24MP APS-C vs. 24MP FF" problem raised some posts ago - much better than cropping from APS-C and comparing with obsolete equipment, I'd say
You cannot really discuss "outresolving" by referring to obsolete 6MP sensors. If it's "a" lens and "a" sensor, sure, you can't make it happen either way... but that's not what we want, right?
Besides, what do we mean by "outresolving"?

I guess its own thread would not suffice!

---

Last edited by Kunzite; 12-30-2017 at 02:17 PM.

[following.eric]

As much as I find the dpreview studio scene comparison somewhat helpful, to some degree it's not comparing apples to apples; for example, with a Nikon FF they use an Nikkor 85mm and with the Pentax K1 they use the Pentax 77mm. In RAW, we are seeing both the result of different sensors, but also thru different lenses.

So it's a realistic comparison using similar primes as would be used by Pentaxians or Nikonians, but not identical primes unless you're comparing only Nikon FF sensors. And anyone's expectations of dpreviews studio comparison tests will change if you take a superior sensor with a Tamron 18-200mm zoom @ 55mm vs. an inferior sensor with a DA* 55mm prime.

The current debate, which seems off thread to the OP, is whether and how a lens can or can't outresolve a sensor and why. This seems to belong to it's own thread.

---------- Post added 12-30-17 at 01:07 PM ----------



Eric, glad you came to some conclusions and hope our posts were helpful. My only words of advice is that you make your decision based on what gear you need or want today, and not try to foresee the next decade; the future.

I love Sony products, but they love to innovate and have dumped products faster than any other large company I can think of.

In the past for video it was betamax to 8mm to Hi-8mm to miniDV, etc. My school invested about $3000 on two brand new Sony UP-DR80 dye sublimation printers. One year after the purchase, the required paper and ribbon was outsourced to Dai Nippon Printing. And then one year after that, DNP only made paper and ribbon for their own DS80 models and the two Sony printers became fully functional bricks. Sony corporate did nothing and we ate it. It's the Sony culture to ax any technology that isn't hot. One reason I appreciate Pentax is because I have found them, more than any other brand, fairly future-proof.

I only mention this because you used the word "investment". I still have burn scars from my Sony experiences.

The worst thing to leave Pentax will be the forum. I have learnt a lot here and everyone has been very helpful. I mean until I actually sell my k70 I am not gone yet. I am also only going to buy something I can have a go at first. Since buying a decentered 15mm lens recently and the unavailability of decent priced Pentax items, I am tired of trying to buy second hand items online hoping they will be fine. Unfortunately for me if I am spending 2k-3k over the next year or so on a new system I do have to look at it as an investment.

If I bought for now I think I would lean towards the 5diii. Proven body even for it's age, well built. better AF and other specs. Would have access to Canon glass which is a bit more accessible here in the UK. But it doesn't solve the size problem. Glass and camera body will be much larger than I have now with a k70. Not sure it will be as travel friendly as I would like.

Until I actually handle future cameras and see how I feel when the time comes I am not truly gone, as aslways also still figuring out what kind of photographer I am and what I want to shoot. I am kind of a floater at the moment as I don't feel I have much subject matter in my current location. Going to see if anyone at my camera club has a system I can borrow and try.

[Alex645]

Besides, what do we mean by "outresolving"?

Here's a short article that discusses if sensors can outresolve lenses:

Will we get sensors that outresolve our lenses? | byThom | Thom Hogan

The bottom line is having a work flow from capture to processing to printing that eliminates the weakest link.
The benefits of the best lenses, sensors, huge RAW files, high bit depth, is nullified if the image is viewed on a bad monitor or printed on subpar media.
And using an 8k monitor isn't going to help an over processed jpeg or badly processed dng.

So to answer the question, what do we mean by outresolving? It goes two ways. Either you have a low res lens and the high res sensor is amplifying its flaws, or you have a high res lens but the lower res sensor can't benefit from the higher resolution.

I may have over simplified this, as one may be under-resolving, but it's the essence of my understanding of the term.

[Nicolas06]

if I am spending 2k-3k over the next year or so on a new system I do have to look at it as an investment.

An investment is something that out to make you money. Here you are just spending more. It is a cost. I agree you can justify it in a number of way like people justify a new/better car, a bigger/nice house or what ever else they want for themselve but you are just raising the cost of your living.

In exchange, hopefully, you live better today, justifying to not actually save the money on a real investment that allow you to live even better tomorrow. That's the tradeof.

But that photographic gear ultimately will loose its value even if it is still working and in the long run will just stop working will a final negative value taking some space in your house. Quite likely that in a few years, a new product will appear that make the current gear look ancient and outdated.

I am not saying this is an issue, I dot it myself and I think we can all be fine to spend money for our pleasure, but there no way to consider it as an investment. Spending the money to get yourself a training that will allows you to get a new job or putting the money for your retirement fund are investments.

[Alex645]

The worst thing to leave Pentax will be the forum. I have learnt a lot here and everyone has been very helpful.
If I bought for now I think I would lean towards the 5diii.

Until I actually handle future cameras and see how I feel when the time comes I am not truly gone, as aslways also still figuring out what kind of photographer I am and what I want to shoot. I am kind of a floater at the moment as I don't feel I have much subject matter in my current location. Going to see if anyone at my camera club has a system I can borrow and try.

I own and use two systems for a reason. Digital cameras almost always depreciate, so you need to make sure if you go with a single system (like Canon), will it do everything you need? For me, Nikon has its pros and cons over the Pentax system. When I couldn't get what I wanted out of one, I expanded into the other, but not jettison the original, unless the first system had no redeeming value relative to the new one.

And you don't have to be a "pro" to make money with your camera. There are a lot of folks looking for a family portrait, head shots, real estate for sale, autos for sale, food or catering photography, etc, that can help pay for your gear on weekends. One of my best paying jobs was shooting batteries for a catalog. Incredibly boring, but $50 per shot in an extensive catalog of batteries.

In your original post you wrote, "I love my K70, for entry level camera is boasts features I know I will miss. But it's more on the lens side and the cost of Pentax Lenses." When I started photography as a teenager, I somehow thought I needed every focal length and spent a lot of money on lenses are rarely used. Maybe we all have to go through that to discover our own tastes and style and preferences.

Now, 40 years later, I am liberated to believing and knowing that with two primes, I can get what fulfills my vision. I do have my lazy wishy-washy days and then will use a 16-85mm or 18-135mm zoom, but if I'm wearing my artist hat, the zoom stays home.

You won't be the first or the last to switch away from Pentax and you won't be the first or last to return. Honestly if Nikon made a medium format camera, I may have never gone with Pentax. But as a photo teacher, once the K5 and K30/K50 came out, I knew there were legit alternatives to Canikony, and with the K-1, it was a sign Ricoh was moving forward.

[following.eric]

I own and use two systems for a reason. Digital cameras almost always depreciate, so you need to make sure if you go with a single system (like Canon), will it do everything you need? For me, Nikon has its pros and cons over the Pentax system. When I couldn't get what I wanted out of one, I expanded into the other, but not jettison the original, unless the first system had no redeeming value relative to the new one.

And you don't have to be a "pro" to make money with your camera. There are a lot of folks looking for a family portrait, head shots, real estate for sale, autos for sale, food or catering photography, etc, that can help pay for your gear on weekends. One of my best paying jobs was shooting batteries for a catalog. Incredibly boring, but $50 per shot in an extensive catalog of batteries.

In your original post you wrote, "I love my K70, for entry level camera is boasts features I know I will miss. But it's more on the lens side and the cost of Pentax Lenses." When I started photography as a teenager, I somehow thought I needed every focal length and spent a lot of money on lenses are rarely used. Maybe we all have to go through that to discover our own tastes and style and preferences.

Now, 40 years later, I am liberated to believing and knowing that with two primes, I can get what fulfills my vision. I do have my lazy wishy-washy days and then will use a 16-85mm or 18-135mm zoom, but if I'm wearing my artist hat, the zoom stays home.

You won't be the first or the last to switch away from Pentax and you won't be the first or last to return. Honestly if Nikon made a medium format camera, I may have never gone with Pentax. But as a photo teacher, once the K5 and K30/K50 came out, I knew there were legit alternatives to Canikony, and with the K-1, it was a sign Ricoh was moving forward.

Thanks for the info. I suppose at the end of the day gear is gear and only small differences appear between different companies. New equipment and new glass won't turn me into a better photographer and the money probably could be spend elsewhere. I am for starters selling lenses and stuff that I barely use and probably will never use. Take that money and use it somewhere else.

I think as well I have to go through natural growing pains I suppose and in 40 years we may not know who is still around and who wont be. Maybe some new company will revolutionize the industry and we may just realize a cellphone is the best option.

First thing is trip to the camera store to have a little hands on experience and actually touch and hold the items and see how I feel after that.

[surfar]

The worst thing to leave Pentax will be the forum.

Theres a section for "leavers"...lol....Go to the back of the class,sorry,i mean the bottom of the page!
Also that caters for people who use multiple systems which is necessary for some.

[Alex645]

First thing is trip to the camera store to have a little hands on experience and actually touch and hold the items and see how I feel after that.

Good move. Just be ready for the heavy sell on Canon or mirrorless as camera stores have tons of inventory of both and many sales staff get a commission or bonus for sales. Brick & mortars (without a store online) hate the web, but need to realize their sales tactics can be a total turnoff.

At most camera/electronic stores, I feel like I've entered an auto dealership...but you've got to test drive it. Just remember not to be WOWed by the touch screen, the blue tooth, and all the name dropping. You came in to buy a camera, not a cell phone with interchangeable lenses.

[Ian Stuart Forsyth]

The DPR link responds to the "24MP APS-C vs. 24MP FF" problem raised some posts ago - much better than cropping from APS-C and comparing with obsolete equipment, I'd say

It wasn't meant to compare obsolete equipment but it is rather useful to use the 6mp and 15mp cameras as a 24mp FF sensor would have pixels that fall mid point between those to Mp count sensors

You cannot really discuss "outresolving" by referring to obsolete 6MP sensors.

Never have I said that the current sensors out resolve lenses, my point was to show that as pixels get smaller they are less able to resolve equal resolution to that of the increased resolution of the senso r.So if we double the pixel count we don't double the captured resolution, going from 6mp to 15mp with the captured resolution we only gain half of that.
To show this I used 2 cameras with pixels that very in size very close to the difference in size we would see between a FF and a cropped body. A FF 16mp camera would have pixels very close to the size of those found in the 6mp ist ds
To only compare the resolution captured from a lens that is equally sharp I used the same lens on the bodies

To do this I needed to move the camera closer to the target, This way we can compare and show that even at the pixel size of the DS and the K5 we do see a decrease in captured resolution.


There is enough of a decrease in captured resolution that one can switch out a lens with poorer performing lens ( for this I used a m 80-200 push pull)


For this shot I held the FOV so that both cameras can be shot at the same distance the K5 using the 100mm and the ds the 80-200mm at 150mm( same FOV)
Again the 2 test was to show that even at the pixel density of outdated equipment (VERY LARGE PIXEL SIZES) that we do see a decrease in captured resolution per pixel size decreases as those pixels become smaller.

Because pentax has no FF 24mp cameras its hard to compare what the difference would be between the K3

If we look at other cameras that have these differences
Cropped 24mp camera

same lens on a FF 24mp

If we take into consideration that the FF has a AA filter that is purposely blurring the image before it reaches the sensor we still can see that the size of those pixels do indeed play a role in the captured resolution
. But this removal of the filter in the cropped body comes at a cost, false artifacts ( see bottom)

24mp FF


Similar to the testing of using a lower quality lens on a sensor with larger pixels
24mp cropped with a 24mm prime

24mp FF and using the old and outdated AFD 24mm

Other than wide open the AFD on FF will give you better results.

False artifacts
Thank to Image resources and using there resolution test let see what happens when the AA filter is removed and compare it to one that has the filter This then can give us an insight into what resolution is being captured.

Removing the AA filter does give us increased resolution but is that resolution even real