Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1      
2
       end
  

Archive 2014 · Yet Another Tamron 150-600 Thread! Interesting Observations --

  
 
fraga
Offline
• • • •
Upload & Sell: Off
p.2 #1 · p.2 #1 · Yet Another Tamron 150-600 Thread! Interesting Observations --


Well, I didn't actually say that with a higher resolution crop sensor it would be worse than a FF sensor with less pixel density.
In fact, it is not, look at the last side by side pictures of the OP.

What I said (or at least tried to) was that a higher pixel density sensor will put more stress on a lens.
It will show the lens weaker points a lot more. Do you agree?
It is, after all, a magnified image of what would be a shot taken with a sensor with less pixel density.
That is why people often say here that with a high MP sensor, you shouldn't inspect the files at 100% and expert very sharp results. This has been repeated ad nauseum in this forum.
With a 8MP camera, results are often sharp at 100%. The exact same lens on a 18MP camera, and the results at 100% might not appear as sharp, though, in reality, you do have more detail (resolution) there.

So the problem here could be that the 5D II sensor does not have the potential to extract more detail that the one provided by the lens wide open. The lens, wide open, out resolves the sensor. So when you close down the aperture, you get more DOF but not much more detail.
This does not invalidate the fact that "The optimal aperture of a lens is the same for all pixel densities". It is, of course. The lens is independent from the camera. But what if the sensor cannot resolve more than what it already does with the lens wide open?
On the other hand, the 7D, with more pixel density, is able to extract more detail as the lens approaches the optimal aperture.

As an example of what I was trying to say, if the problem is indeed related to pixel density, a D800 should yield a similar result to the 7D, that is to say, it will display a more noticeable difference between wide open and at optimal aperture.

I could be completely wrong here, as I am far from being an expert, like the previous posters before me.
This is just an explanation of what I was trying to say, as I never intended to imply that we would be worse with a higher resolution sensor.
I think the last picture on the OP proves that.

As an analogy, if you have a race car and suddenly put a lot more power in the engine, it will go around a circuit faster but it will also show that the brakes, suspension, chassis and tires need improvement, even if they appeared fine for the previous HP output.
The more powerful engine will reveal the shortcomings of the rest of the car, because the less powerful version was not enough to stress those components enough.
All thing being equal, the driver of the more powerful version will complain about the brakes and suspension and tires. The other, not so much or not at all.
Not a very good analogy, sorry, I hope you can understand what I was trying to say, since English is not my first language.



Feb 18, 2014 at 07:09 AM
alundeb
Online
• • • • •
Upload & Sell: Off
p.2 #2 · p.2 #2 · Yet Another Tamron 150-600 Thread! Interesting Observations --


Thanks for the explanation, fraga, I agree with most of what you say. It is just that f/14 should not be found to be the optimal aperture, as the OP did find.
It should be around f/8.



Feb 18, 2014 at 07:21 AM
arbitrage
Offline
• • • • • • •
Upload & Sell: On
p.2 #3 · p.2 #3 · Yet Another Tamron 150-600 Thread! Interesting Observations --


I don't find Focal to be all that reliable for any of these type of tests. I also find it to be most unreliable as the focal length increases. 600 is at the edge of its usefulness. When I test at 840 and 1200 it falls apart and results are not repeatable. The OP did mention that f/11 and f/8 were very close also.


Feb 18, 2014 at 07:23 AM
pKai
Offline
• • •
Upload & Sell: Off
p.2 #4 · p.2 #4 · Yet Another Tamron 150-600 Thread! Interesting Observations --


Shutterbug2006 wrote:
If it's any consolation, all the shots in the first post are sucky ...


.... those are more than 100% crops of 100% crops. Here is one of the images in full size. This is the 5D2 at F14.










Feb 18, 2014 at 08:04 AM
pKai
Offline
• • •
Upload & Sell: Off
p.2 #5 · p.2 #5 · Yet Another Tamron 150-600 Thread! Interesting Observations --


greenfield wrote:
I assume camera(s) has been on a tripod while conducting your test.
Try again and do focus via LV and switch VC off unless done so ...
Also use mirror lockup and remote release (or internal timer release).
Make more than one shot for each f-stop and select the best of them.

This way you'd get more consistent and reliable results.



Yes, the cameras were on a very sturdy tripod plus had a 3lb soft weight on top to dampen vibration.



Feb 18, 2014 at 08:08 AM
pKai
Offline
• • •
Upload & Sell: Off
p.2 #6 · p.2 #6 · Yet Another Tamron 150-600 Thread! Interesting Observations --


PixelPerfect, fraga, et al..... Thanks for your analysis... in theory, I tend to agree with it, hence my puzzlement.

I did repeat the both the 7D test and 5D2 that day thinking the same thing and got the same results. When I get the chance, I will do it again and see what happens.

One thing I didn't consider... when I ran the test, the lens was brand new with barely a dozen shots through it. Now it has a couple of thousand.... maybe there's a break-in variable at play?

In the field, I used the lens exclusively with the 5D2 and what you see in the 2 images I posted is representative of the lot regarding accuracy of focus.

Remembering that the 5D2 got a +8 adjustment which is NOT typical of other lenses with that body..... so maybe the 7D adjustment is off. All other lenses I have give similar adjustments on those 2 bodies... they may be off by one or 2 from each other but not 6.

Arbitrage: I did notice that Focal struggled more with this lens to complete the tests compared to my other lenses. I chalked that up to this being a non-Canon lens but long FL makes sense too. Previously, the longest lens I've calibrated with FoCal is a 400.



Feb 18, 2014 at 08:21 AM
Pixel Perfect
Offline
• • • • • • •
Upload & Sell: On
p.2 #7 · p.2 #7 · Yet Another Tamron 150-600 Thread! Interesting Observations --


fraga wrote:
So the problem here could be that the 5D II sensor does not have the potential to extract more detail that the one provided by the lens wide open. The lens, wide open, out resolves the sensor. So when you close down the aperture, you get more DOF but not much more detail.
This does not invalidate the fact that "The optimal aperture of a lens is the same for all pixel densities". It is, of course. The lens is independent from the camera. But what if the sensor cannot resolve more than what it already does with the lens wide
...Show more

I don't believe this result has anything to do with pixel density and is due to slight misfocus on the 7D result. The sensor resolution is fixed, so the only change we can make to total system res is by altering the aperture of the lens. f/14 is well past the optimal aperture of the lens and the res would be much lower than at say f/8. The MTF curves for a lens get worse as we stop down (if we have a perfect lens free of aberrations), as we are limiting the ability of the lens to capture high spatial frequency detail.

As an aside, it is interesting why a lens' MTF curve shows a rapid fall off with increasing spatial frequency. The MTF is basically giving us the transmission function of a lens as a function of spatial frequency. Effectively this is the
contrast vs spatial frequency curve. More abstractly the MTF curve is the magnitude of the optical transfer function of the lens, which is in turn the Fourier transform of the point spread function of the lens. The PSF is the response of the lens to a point source.

So why does a lens have trouble capturing higher spatial frequencies? Well when light is scattered from a object the it turns out that if we perform a Fourier analysis of the scattered light it consists of a superposition of many spatial
frequencies propagating at different angles and that the higher the spatial frequency the more obliquely the scattered wave travels. To perfectly reproduce the subject through an imaging system we need to capture all these spatial frequencies. However, this is just not possible as every lens has a maximum acceptance angle which is related to it's f-number (in turn related to the numerical aperture). Thus the higher the spatial frequency the less able the lens is able to capture it due to its large obliquity factor. If we had an infinitely large lens then yes we could capture all spatial frequencies. Stopping a lens down further reduces it's ability to capture high frequency detail, but on the other hand it may greatly reduce aberrations that lower resolution and improve contrast of lwer spatial frequencies.

The maths behind all this is treated by Fourier optics which looks at diffraction theory. The maths gets complex fast, but all we really need is the MTF curve which can be measured by imaging subjects of varying spatial frequencies.



Feb 18, 2014 at 08:39 AM
cameron12x
Offline
• • • •
Upload & Sell: On
p.2 #8 · p.2 #8 · Yet Another Tamron 150-600 Thread! Interesting Observations --


Pixel Perfect wrote:
All we really need is the MTF curve which can be measured by imaging subjects of varying spatial frequencies.


+1. Agreed.

And the UOM output for MTF curves is usually expressed in lp/mm for lenses.

I believe that empirical testing of this lens has shown it to be sharpest between f/8 and f/11.

Usually a lens is sharpest, relative to its testing medium, about 2 stops away from its maximum open aperture. The empirical data for this lens would seem to corraborate that generalization.



Feb 18, 2014 at 10:22 AM
alundeb
Online
• • • • •
Upload & Sell: Off
p.2 #9 · p.2 #9 · Yet Another Tamron 150-600 Thread! Interesting Observations --


cameron12x wrote:
Usually a lens is sharpest, relative to its testing medium, about 2 stops away from its maximum open aperture. The empirical data for this lens would seem to corraborate that generalization.


All long Canon L zooms and all the big whites are sharpest in center at about 1 stop from maximum aperture.



Feb 18, 2014 at 10:31 AM
greenfield
Offline

Upload & Sell: Off
p.2 #10 · p.2 #10 · Yet Another Tamron 150-600 Thread! Interesting Observations --


alundeb wrote:
All long Canon L zooms and all the big whites are sharpest in center at about 1 stop from maximum aperture.


I wouldn't claim that for the 800L which at f5.6.already reaches its best sharpness (only tiny difference compared to f8 IMO)



Feb 18, 2014 at 10:41 AM
alundeb
Online
• • • • •
Upload & Sell: Off
p.2 #11 · p.2 #11 · Yet Another Tamron 150-600 Thread! Interesting Observations --


greenfield wrote:
I wouldn't claim that for the 800L which at f5.6.already reaches its best sharpness (only tiny difference compared to f8 IMO)


I agree about that.

The point was that regarding long and especially long and slow lenses, only very poor lenses need to be stopped down as much as 2 stops to reach maximum sharpness in the center. And the problem is that then they are so limited by diffraction that they never get really sharp.



Feb 18, 2014 at 10:44 AM
1      
2
       end




FM Forums | Canon Forum | Join Upload & Sell

1      
2
       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.