Do you believe that some Canon lenses have the resolution ...

Silverback1988

Leading Member
Messages
695
Reaction score
12
Location
Surabaya/East of Java/Indonesia, ID
that could maximize 50 MP in FF, so it will have the same resolution as the 50 MP MF with its lens ?

Or is it just too difficult (impossible) to make a such lens that sharp ?

So the only way at the moment is just to make the sensor larger, thus less stretch for the lens ?

I could understand if smaller format (such APSC or mFT) will have another problem, DOF control problem, even if someone could make a lens sooo sharp for them.

But there is no DOF control problem in 35mm FF, is there ?

Maybe just resolution problem in wider aperture ? in f/2 perhaps ? But not resolution in f/2.8 ?

Some one who know these, please chime in ...

Thanks

-

Brian
 
Even wider than 35mm allow dof control albeit at F/1.4 e.g. It gets shallower at wider lenses even on FF but it is very doable.

The example is made with a Sigma 35mm DG HSM Art F/1.4 and a Canon 5D Mark III @ F/2.

If it had been made @ f/1.4 the difference in shallowness of dof would have been to great to get the object in focus.

I think it can do very well at high sensor densities.



71971a014afb4561b591bdf1e7065208.jpg
 
This is not talking about the resolution of each systems (Full Frame & Medium Format camera),

but only for DOF.

The FF is in 35mm f/1.4, then the MF (in this case is Pentax) will use (if it exist at all) 44 mm at f/1.8

Is the Full Frame system (35mm FF) has more option (regarding DOF) than the MF ?

Yes, high density sensor will have the potential to reveal the DOF more,

example the same 35mm f/1.4 at f/1.4 will look to have more DOF with 5D (12 MP) rather than with 5DII (21 MP) or hypothetical 50 MP Canon FF.
Even wider than 35mm allow dof control albeit at F/1.4 e.g. It gets shallower at wider lenses even on FF but it is very doable.

The example is made with a Sigma 35mm DG HSM Art F/1.4 and a Canon 5D Mark III @ F/2.

If it had been made @ f/1.4 the difference in shallowness of dof would have been to great to get the object in focus.

I think it can do very well at high sensor densities.

71971a014afb4561b591bdf1e7065208.jpg
 
This is not talking about the resolution of each systems (Full Frame & Medium Format camera),

but only for DOF.
From memory the fastest is F2.8

25mm F4 , 55mm F2.8 ( that is very nice) , 90mm F2.8 , 75 F2.8 , 200 F4, 150 F2.8 ,45mm F2.8, 120mm F4, 55-110mm F5.6, 35mm F3.5 300mm F5.6, 80-160mm F4.5 ,25mm F4, 300mm F4, 33-55 F4.5

these I know are in production

I don't think there is anything faster other than used, so FF should give you more as far a selection goes for fast shallow DOF
rob Ernsting, post: 53569992, member: 581372"]
Even wider than 35mm allow dof control albeit at F/1.4 e.g. It gets shallower at wider lenses even on FF but it is very doable.

The example is made with a Sigma 35mm DG HSM Art F/1.4 and a Canon 5D Mark III @ F/2.

If it had been made @ f/1.4 the difference in shallowness of dof would have been to great to get the object in focus.

I think it can do very well at high sensor densities.

71971a014afb4561b591bdf1e7065208.jpg
[/QUOTE]
--
The Camera is only a tool, photography is deciding how to use it.
The hardest part about capturing wildlife is not the photographing portion; it’s getting them to sign a model release
 
Last edited:
that could maximize 50 MP in FF,
The answer is NO. Digital sensor can get more and more pixels but hit bottle neck of physical glass. No lenses today can even maximize 36mp sensors in D800/E and A7R. The two best primes, Zeiss Otus 55/1.4 and Sony FE 55/1.8 only have 29 mpix respectively out of 36mp sensor. The new Sigma 50/1.4 Art should have similar mpix. The other lenses have much less mpix on D800/E and A7R.
so it will have the same resolution as the 50 MP MF with its lens ?
No. 50mp MF sensor resolves more than 50mp FF sensor with comparable lenses. The same true 50mp FF sensor resolves more than 50mp APS-C sensor with comparable lenses.
Or is it just too difficult (impossible) to make a such lens that sharp ?
Absolutely. Unfortunately the bottleneck is still glass, the very old martial unless there is a revolution breakthrough with a brand new alternative material. We see sensors getting better and better but not much in lenses. It's impossible to make such APS-C lens to have the same resolution power as Otus 55/1.8 and FE 55/1.8 on APS-C. FF lenses are designed and optimized for FF sensors and perform much less on crop sensors. For best performance, you'd have to design and optimize lenses natively on respective sensor crop but it's an upperhill battle to compete to FF lenses. For example Sigma tried hard to have 18-35/1.8 APS-C zoom that roughly is eq to FF 28-56mm F2.8 zoom, close to but still inferior to Canon 24-70L/2.8 II in IQ, and shorter in both ends but even larger/heavier.
So the only way at the moment is just to make the sensor larger, thus less stretch for the lens ?
Unless there is a big breakthrough in sensor QE, to resolve more, you'd have to increase sensor size and design native lenses that are also bigger and heavier.
I could understand if smaller format (such APSC or mFT) will have another problem, DOF control problem, even if someone could make a lens sooo sharp for them.
Smaller sensor, less control in DOF and lenses are much difficult to be designed or even impossible to have the same sharpness of FF counterpart. No such lenses that can eq to FF 35/1.4 or 50/55/1.4 prime, they would be enormous, larger/heavier but likely still inferior to FF counterpart on FF sensors.
But there is no DOF control problem in 35mm FF, is there ?
Relatively. MF then can achieve even shallower DOF than FF. But there is a limit that our eyes are comfortable and can see. 35mm FF is shallow enough with respective lenses. If you think 50/1.4 is not shallow enough, then there is 50L/1.2 and old Canon 50/1.0 lens, or even a third party 50/0.95 lens.
Maybe just resolution problem in wider aperture ? in f/2 perhaps ? But not resolution in f/2.8 ?
To get better sharpness or resolution, you'd have to stop down a bit if this is your question. Will see if the new Sony curve sensor could help in this matter.
Some one who know these, please chime in ...

Thanks

-

Brian
--
http://www.flickr.com/photos/55485085@N04/
http://qianp2k.zenfolio.com/
 
Last edited:
then MF is the only way to achieve 50MP+ resolution.

Does the 35mm FF hit the resolution wall in 36 MP ? I hope not.

Peter,
sometime I do stitching. I was quite sure that when I stitch 2 or 3 photos from 7D,
the result was significantly better than what my 5DII can do in 1 shot photo.

Was that a sign that FF lens is still capable to deliver about 40 MP or even 50 MP in FF sensor ?

Okay, Stitching with APSC is not 100% identical with using single shot FF.
I was using the center area of the lens, which has the more resolution.
Even though we could equalize the light gathering, still it's different when using the part of the lens.

I wonder if in FF lens, it's difficult to have uniform resolution corner to corner,
how difficult that it could be if the lens designer have to make a larger one for MF ?
Would corner resolution would be more problem with larger format ?


-
Brian
 
From memory the fastest is F2.8
In FF, that would be F/2.2 ? I rarely use f/2 or f/1.4 with FF
25mm F4 , 55mm F2.8 ( that is very nice) , 90mm F2.8 , 75 F2.8 , 200 F4, 150 F2.8 ,45mm F2.8, 120mm F4, 55-110mm F5.6, 35mm F3.5 300mm F5.6, 80-160mm F4.5 ,25mm F4, 300mm F4, 33-55 F4.5

these I know are in production

I don't think there is anything faster other than used, so FF should give you more as far a selection goes for fast shallow DOF
Thanks a lot for the info :)

-

Brian
 
With the new curved Sony sensor, does anyone know if it will affect DOF?
No it shouldn't do. But a curved sensor will need lenses designed to give a curved field at the focal plane and of course most of today's lenses have been designed to give a flat field (or as close to it as the manufacturer can achieve).

So if this curved sensor idea is to take off for interchangeable lens cameras there will have to be an agreement on standardisation of the curvature.

Or sensors which can have their curvatures altered in camera.

Whoopee! Another endless series of threads for some of us to argue over:
  • Which lenses best match the curvature of my xyz sensor?
  • What is the best method of adjusting MCA (Micro Curvature Adjustment)?
  • Will Canon / Nikon / Sigma adjust my 50/1.8 purchased in 1992 to match the new curved sensors?
  • I think my sensor is curved the wrong way. All my edges are fuzzy at f1.2!
  • How many sensor bending actuations can I get from my sensor before it gets sensor fatigue and snaps?
  • etc.
 
With the new curved Sony sensor, does anyone know if it will affect DOF?
No it shouldn't do. But a curved sensor will need lenses designed to give a curved field at the focal plane and of course most of today's lenses have been designed to give a flat field (or as close to it as the manufacturer can achieve).

So if this curved sensor idea is to take off for interchangeable lens cameras there will have to be an agreement on standardisation of the curvature.

Or sensors which can have their curvatures altered in camera.

Whoopee! Another endless series of threads for some of us to argue over:
  • Which lenses best match the curvature of my xyz sensor?
  • What is the best method of adjusting MCA (Micro Curvature Adjustment)?
  • Will Canon / Nikon / Sigma adjust my 50/1.8 purchased in 1992 to match the new curved sensors?
  • I think my sensor is curved the wrong way. All my edges are fuzzy at f1.2!
  • How many sensor bending actuations can I get from my sensor before it gets sensor fatigue and snaps?
  • etc.
Agree, Sony cannot survive another 'mount' or FC (full curved) mount...they would lose any shred of loyalty they have left. Most likely it is for a RX1 successor where they come up with faster, sharper, smaller full-frame compacts.

The other possibility is that some FE or E mount lenses are already geared for field curvature and are not well corrected for flat field to start with. Who knows as some of the Sony lenses are pretty horrid when it comes to borders/corners.

Since there are some FE lenses like the 55mm 1.8 which is just tack sharp everywhere, again its not likely.
 
Last edited:
that could maximize 50 MP in FF,
The answer is NO. Digital sensor can get more and more pixels but hit bottle neck of physical glass. No lenses today can even maximize 36mp sensors in D800/E and A7R. The two best primes, Zeiss Otus 55/1.4 and Sony FE 55/1.8 only have 29 mpix respectively out of 36mp sensor. The new Sigma 50/1.4 Art should have similar mpix. The other lenses have much less mpix on D800/E and A7R.
A couple of random thoughts on the above that should not be viewed as criticisms. I respect your knowledge on these matters.

1.) I've heard the exact opposite many times on this forum. True or not. I dunno. Maybe I am not picking up on a nuance of your argument.

2.) Are the MPix results noted above due to the lenses or the sensors?

3.) So, the MPix results noted above may not indicate 100% efficiency but should that negate the jaw-dropping value of the images?

4.) Will there ever be a perfect 36 MPix for the D800 or whatever camera? Not sure how the tests are calibrated.
so it will have the same resolution as the 50 MP MF with its lens ?
No. 50mp MF sensor resolves more than 50mp FF sensor with comparable lenses. The same true 50mp FF sensor resolves more than 50mp APS-C sensor with comparable lenses.
Or is it just too difficult (impossible) to make a such lens that sharp ?
Absolutely. Unfortunately the bottleneck is still glass, the very old martial unless there is a revolution breakthrough with a brand new alternative material. We see sensors getting better and better but not much in lenses. It's impossible to make such APS-C lens to have the same resolution power as Otus 55/1.8 and FE 55/1.8 on APS-C. FF lenses are designed and optimized for FF sensors and perform much less on crop sensors. For best performance, you'd have to design and optimize lenses natively on respective sensor crop but it's an upperhill battle to compete to FF lenses. For example Sigma tried hard to have 18-35/1.8 APS-C zoom that roughly is eq to FF 28-56mm F2.8 zoom, close to but still inferior to Canon 24-70L/2.8 II in IQ, and shorter in both ends but even larger/heavier.
So the only way at the moment is just to make the sensor larger, thus less stretch for the lens ?
Unless there is a big breakthrough in sensor QE, to resolve more, you'd have to increase sensor size and design native lenses that are also bigger and heavier.
I could understand if smaller format (such APSC or mFT) will have another problem, DOF control problem, even if someone could make a lens sooo sharp for them.
Smaller sensor, less control in DOF and lenses are much difficult to be designed or even impossible to have the same sharpness of FF counterpart. No such lenses that can eq to FF 35/1.4 or 50/55/1.4 prime, they would be enormous, larger/heavier but likely still inferior to FF counterpart on FF sensors.
But there is no DOF control problem in 35mm FF, is there ?
Relatively. MF then can achieve even shallower DOF than FF. But there is a limit that our eyes are comfortable and can see. 35mm FF is shallow enough with respective lenses. If you think 50/1.4 is not shallow enough, then there is 50L/1.2 and old Canon 50/1.0 lens, or even a third party 50/0.95 lens.
Maybe just resolution problem in wider aperture ? in f/2 perhaps ? But not resolution in f/2.8 ?
To get better sharpness or resolution, you'd have to stop down a bit if this is your question. Will see if the new Sony curve sensor could help in this matter.
Some one who know these, please chime in ...

Thanks

-

Brian
 
then MF is the only way to achieve 50MP+ resolution.
50mp is 50mp regardless from what crop format. The only issue is that how much resolution your eyes can resolve. If we use DXO perceptual sharpness/resolution, then at current sensor and lens technology, the rumored 54mp Sony FF sensor will not reach 50 mpix, but much below. DXO has not tested 80mp sensors in PhaseOne and Mamiya that have larger sensor than the one in Pentax 645Z BTW. But if they tested with the best prime from respective companies, I'd guess will be above 40 mpix and very close to 50 mpix. Let's wait and see DXO test on 645Z with Pentax best prime MF lens that certainly will be higher than even future 54mp FF sensor on Otus 55 or FE 55.
Does the 35mm FF hit the resolution wall in 36 MP ? I hope not.
Let's see on rumored 54mp FF sensor from Sony and not sure if its curve sensor will help in resolution (mainly help at edge/corner sharpness from what I read). But 36 mpix is a barrier a FF sensor with current technology could pass even with a 54mp sensor.
Peter,
sometime I do stitching. I was quite sure that when I stitch 2 or 3 photos from 7D,
the result was significantly better than what my 5DII can do in 1 shot photo.
True, but you also can do stitching with 5D2 that has even higher resolution. Here are two samples hand-held stitching from 5D2. They are only 1/6 of original size due to DPR post limit.


Venice, Italy with 70-200L/4.0 IS pano


Florence, Italy with 24-105L pano
Was that a sign that FF lens is still capable to deliver about 40 MP or even 50 MP in FF sensor ?
If you mean MPIX that human eyes can actually resolve not on paper MP, then it's very difficult if possible with current sensor technology and largely bottleneck of imperfect lenses.
Okay, Stitching with APSC is not 100% identical with using single shot FF.
I was using the center area of the lens, which has the more resolution.
Even though we could equalize the light gathering, still it's different when using the part of the lens.
I wonder if in FF lens, it's difficult to have uniform resolution corner to corner,
Let's see Sony new curve sensor development. But lenses need to be redesigned to work with curve sensor I believe.
how difficult that it could be if the lens designer have to make a larger one for MF ?
My understanding is that it's relative easier to design lenses natively on bigger sensor than on smaller sensor to achieve the same sharpness/resolution. It's an upperhill battle for a smaller sensor to overcome crop penalty if possible.
Would corner resolution would be more problem with larger format ?
No if you compare at the same output size at equivalent DOF. If we compare FF resolution with crop, usually you find you just need to stop down one stop (not even need to stop down to eq DOF stop), FF set is sharper than crop from edge to edge.
--
http://www.flickr.com/photos/55485085@N04/
http://qianp2k.zenfolio.com/
 
Last edited:
From memory the fastest is F2.8
In FF, that would be F/2.2 ? I rarely use f/2 or f/1.4 with FF
I think, but am not sure, that there's an 80 / 2 for MF. Used on the 645Z, that would be equivalent to a 62 / 1.5 on FF.
25mm F4 , 55mm F2.8 ( that is very nice) , 90mm F2.8 , 75 F2.8 , 200 F4, 150 F2.8 ,45mm F2.8, 120mm F4, 55-110mm F5.6, 35mm F3.5 300mm F5.6, 80-160mm F4.5 ,25mm F4, 300mm F4, 33-55 F4.5

these I know are in production

I don't think there is anything faster other than used, so FF should give you more as far a selection goes for fast shallow DOF
Thanks a lot for the info :)
The 150 / 2.8 on the 645Z would be equivalent to a 115 / 2.2, which would be a pretty good lens shallow DOF portraits.
 
that could maximize 50 MP in FF,
The answer is NO. Digital sensor can get more and more pixels but hit bottle neck of physical glass. No lenses today can even maximize 36mp sensors in D800/E and A7R. The two best primes, Zeiss Otus 55/1.4 and Sony FE 55/1.8 only have 29 mpix respectively out of 36mp sensor. The new Sigma 50/1.4 Art should have similar mpix. The other lenses have much less mpix on D800/E and A7R.
A couple of random thoughts on the above that should not be viewed as criticisms. I respect your knowledge on these matters.

1.) I've heard the exact opposite many times on this forum. True or not. I dunno. Maybe I am not picking up on a nuance of your argument.

2.) Are the MPix results noted above due to the lenses or the sensors?
Basically due to imperfect lenses as DXO said on a given sensor format. Sensor size is still #1 factor. You will need a much better lens on a smaller sensor to overcome a large sensor. Physical glass is still the largest bottleneck. To dramatically improves IQ you'd have to increase sensor size unless you have a dramatically better sensor on QE and better lens that both are unlikely.
3.) So, the MPix results noted above may not indicate 100% efficiency but should that negate the jaw-dropping value of the images?
Not fully understand this question. As I said sensor size is still the #1 factor follow by lens quality.
4.) Will there ever be a perfect 36 MPix for the D800 or whatever camera? Not sure how the tests are calibrated.
If you talk 36mp D800/E, then with today's best lenses, it's still 29 mpix. I doubt future lenses as slong as still made from the material as we know as glass will reach to lossless 36mpix on D800E. Higher MP, higher diminishing return.

I believe these factors impact IQ in such sequence,

Sensor format (size), lens optical quality and amount of pixels provided others are the same.

--
http://www.flickr.com/photos/55485085@N04/
http://qianp2k.zenfolio.com/
 
Last edited:
…how moire and aliasing is possible on the D800E. If no lens resolves more than 29mp, then moire and aliasing should be practically impossible on a 36mp sensor, shouldn't it, since the lens is effectively acting as a rather strong AA filter?

I've seen lenses like the Nikon 85mm f/1.4G produce both aliasing and moire, so that's why I'm puzzled.

SB
 
In dxo otus delivers more sharpness on d7100 than on 6d, and about similar level with d600. Isnt this enough to say that it is able to resolve 3.9 micron puxel sized sensor? Well at least in center. That means if we have had 3.9 micron FF, we will have increase in resolution in 1.5 times. This in mind I think otus with FF of similar technology as d800e, but about 54mp sensor will deliver roughly sharpness of 40-45. Which is the same step as from d7000 to d7100.
From d7000 to d7100, the resolution increases linearly with mp(I know they dont have the same tech sensor, but d7000's sensor is a very good one), so the calculation works, with ff it should also work.
Correct me if I am missing something.
 
In dxo otus delivers more sharpness on d7100 than on 6d, and about similar level with d600. Isnt this enough to say that it is able to resolve 3.9 micron puxel sized sensor? Well at least in center.
Yeah this Otus 55/1.4 is too good so a crop sensor in D7100 is able to leverage more pixels and overcome crop penalty but only in center area, also you can refer 5D3 vs D7100 with Otus 55/1.4

That means if we have had 3.9 micron FF, we will have increase in resolution in 1.5 times. This in mind I think otus with FF of similar technology as d800e, but about 54mp sensor will deliver roughly sharpness of 40-45. Which is the same step as from d7000 to d7100.
http://www.dxomark.com/Lenses/Compa...-on-Canon-EOS-6D___1242_865_1242_834_1241_836
Unfortunately diminishing return is higher and higher when pixel amount increase. I doubt a 54mp Sony sensor will have 45 mpix if it has the same QE as current 36mp sensor but roughly will around 35-40 mpix with Otus 55/1.4. With other much inferior lenses, they will lose much more.
From d7000 to d7100, the resolution increases linearly with mp(I know they dont have the same tech sensor, but d7000's sensor is a very good one), so the calculation works, with ff it should also work.
http://www.dxomark.com/Lenses/Compa...-ZF2-Nikon-on-Nikon-D7000___1242_865_1242_680
Correct me if I am missing something.
Otus 55/1.4 is too good and able to resolve very well to 24mp but will not increase linearly when sensor pixel amount grow. Still 24mp with Outs lost 3mp to 21 mpix while 36mp sensor in D800 lost 7 mp to 29 mpix, and 54mp sensor will lost more.
 
…how moire and aliasing is possible on the D800E.
Are moire patten and Spatial aliasing are the same thing? Any cameras have moire, more or less. D800E with AA filter removed, supposedly subject to more moire.
If no lens resolves more than 29mp, then moire and aliasing should be practically impossible on a 36mp sensor, shouldn't it, since the lens is effectively acting as a rather strong AA filter?
I am not an expert in this topic but I don't know amount of pixels have much to do with moire and aliasing. With imperfect lens, you get higher diminishing return when amount of pixels increase as lens' optical glass is not good enough to resolve all pixels that human eyes can see, as the two articles I linked below, a large part of those pixels are below MTF 20 that human eyes virtually unable to resolve. Moire/aliasing are different things from MTF resolution.

http://www.normankoren.com/Tutorials/MTF.html

http://www.imatest.com/docs/sharpness/
I've seen lenses like the Nikon 85mm f/1.4G produce both aliasing and moire, so that's why I'm puzzled.
On D800E? Maybe Nikon doesn't implement moire removal in D800E so you'd have to remove in software. In comparison Sony A7R does in this area very well. So far except one or two photos among thousand photos I have not seen any moire in A7R photos, as also said in DPR A7R's review. As a matter of fact, AA-filter-less A7R subjects less moire than AA-filter-equipped 5D3 as observed by Fred Miranda.

http://www.fredmiranda.com/A7R-review/

As I said I don't know moire/aliasing topic well but I have not read they have much related to the amount of pixels.
--
http://www.flickr.com/photos/55485085@N04/
http://qianp2k.zenfolio.com/
 
Last edited:
From memory the fastest is F2.8
In FF, that would be F/2.2 ? I rarely use f/2 or f/1.4 with FF
I think, but am not sure, that there's an 80 / 2 for MF. Used on the 645Z, that would be equivalent to a 62 / 1.5 on FF.
The nearest comparison would be 50mm. Not exactly the same of course.
25mm F4 , 55mm F2.8 ( that is very nice) , 90mm F2.8 , 75 F2.8 , 200 F4, 150 F2.8 ,45mm F2.8, 120mm F4, 55-110mm F5.6, 35mm F3.5 300mm F5.6, 80-160mm F4.5 ,25mm F4, 300mm F4, 33-55 F4.5

these I know are in production

I don't think there is anything faster other than used, so FF should give you more as far a selection goes for fast shallow DOF
Thanks a lot for the info :)
The 150 / 2.8 on the 645Z would be equivalent to a 115 / 2.2, which would be a pretty good lens shallow DOF portraits.
It seems the MF will still serve you nicely Joe :)

You love shallow DOF.

-

Brian
 
that could maximize 50 MP in FF,
The answer is NO. Digital sensor can get more and more pixels but hit bottle neck of physical glass. No lenses today can even maximize 36mp sensors in D800/E and A7R. The two best primes, Zeiss Otus 55/1.4 and Sony FE 55/1.8 only have 29 mpix respectively out of 36mp sensor. The new Sigma 50/1.4 Art should have similar mpix. The other lenses have much less mpix on D800/E and A7R.
Keep in mind DxO does weird things with their PMPIX and so on.

Also that even if it might be a struggle for some lenses do much better near the edges, they might still do much better in the center.

Don't forget that a 7D is like 50MP FF equivalent and in it's aps-c crop area it pulls noticeably more detail in than a 5D3.

And don't forget that overall sharpness is based upon a formula involving both sensor and lens and it's not a linear boom the sensor it outresolved thing.

And don't forget that DxO also uses weird criteria to decide what a lenses best point is. For instance they list their PMPIX values for zeiss and canon 50mm at f/1.4!!!! Do you think those lenses really deliver peak sharpness across the field at f/1.4 (or even in the center for that matter)??
 

Keyboard shortcuts

Back
Top