Copy to copy variation

FEChariot wrote in post #16137578
I am assuming that he isn't using any filters since this is from a sharpness test though. Or are you talking about contamination inside the filter housing?

He means the gelatin drop in filters. He is right that contamination can reduce contrast and sharpness. That being said the filters we used were brand new.

What does everyone think of the samples I posted? I dont think you need MTF tests to see the difference.

FEChariot wrote in post #16137578
I am assuming that he isn't using any filters since this is from a sharpness test though. Or are you talking about contamination inside the filter housing?

There's still a disc in the filter housing (52mm drop-in) and this forms part of the optical formula of the lens.

Bob

Both images are underexposed by about a stop, but yes there is some difference between them. The one you identified as coming from his lens isn't as sharp as the one you identified as coming from your lens.

Try swapping the filters from one lens to the other and see if it makes any difference.

Be sure to MFA the lenses accurately. And lock them down on a tripod. Use mirror lockup or Live View. Either use a 2 second delay or use a remote release so you aren't touching the camera at the moment of exposure. Turn off or be sure any in-camera sharpening is set the same. Best to use the same body for both lenses, which it sounds as if you are already doing (a dirty sensor can effect resolution, for example).

In other words, try to elminate everything possible aside from the optics themselves, in your testing. There are a ton of things that can throw of sharpness!

If there's still lots of difference, send the weaker lens back (if it's new and the retailer will exchange it) or send it to Canon for a tune-up.

There are many things that can happen to a lens... ranging from some problem missed by quality control to a dropped box during shipping to the buyer... Why fret over it? Send it back and get it fixed.

amfoto1 wrote in post #16138013
Both images are underexposed by about a stop, but yes there is some difference between them. The one you identified as coming from his lens isn't as sharp as the one you identified as coming from your lens.

Try swapping the filters from one lens to the other and see if it makes any difference.

Be sure to MFA the lenses accurately. And lock them down on a tripod. Use mirror lockup or Live View. Either use a 2 second delay or use a remote release so you aren't touching the camera at the moment of exposure. Turn off or be sure any in-camera sharpening is set the same. Best to use the same body for both lenses, which it sounds as if you are already doing (a dirty sensor can effect resolution, for example).

In other words, try to elminate everything possible aside from the optics themselves, in your testing. There are a ton of things that can throw of sharpness!

If there's still lots of difference, send the weaker lens back (if it's new and the retailer will exchange it) or send it to Canon for a tune-up.

There are many things that can happen to a lens... ranging from some problem missed by quality control to a dropped box during shipping to the buyer... Why fret over it? Send it back and get it fixed.

Changing the filters was not something we tried. That being said, as I was so curious about this, I just drove up to his house and tried running the same tests with:

* My filter in both lenses
* His filter in both lenses
* No filter in both lenses (just for the heck of it)

We also used live view with remote control and mirror lockup (as we did last time). The results are the same, identical. His filter was worse than mine, but anyways, it made no difference.

At F2.0, the center of the image is pretty similar, in fact I could say identical. At F2.8 mine looks a tad better. At F4.0 and above using the extender, there is big difference.

The corners is a whole other story, even at F2.0, there is big difference and it gets worse as you add the extender. All corners exhibit this issue, I cannot say if it's equal or not, but all of his corners are somewhat softer than mine.

Interesting how there is such variation between lenses. The sad thing is, 95% of the people buying an L lens won't have the opportunity to compare it with another one side by side like we did and doing this with the extremely expensive telephoto primes is even less likely. Paying all that money for a lens, you expect to get the best, seeing this much variation was not something I expected.

Note that with an extender, you do get a calibration issue that does not exist when just testing the lens. I.e. with just the lens, a manual focus or a contrast-based auto-focus will short-circuit any calibration issues in camera or lens.

With an extender, there will be calibration issues between lens and extender. If the light doesn't enter the extender optimally, then there will not be any optimal image to focus onto the sensor.

That is also why some lenses just do not work well with extenders - or at least with some extenders. Trying extenders from different manufacturers can end up with an extender that very well matches a specific lens, giving spectacular results.

Understood.

One thing that made me wonder is the when we were testing the focusing in both lenses, we haven't seem to be able to get MFA to actually do anything. What I mean is, we took 41 shots from -20 to +20 in both lenses, and both lenses had pretty much identical focusing no matter what MFA point was used.

Now I know that my MFA is not broken because a) it used to work wonders with my good old Sigma 35 (/sarcasm) and b) I know it throws of my other lenses when say I use the wrong MA like +20 (i.e. I can see the image blur).

I'm not sure if this is because this is a 200mm lens or something else is up. That being said, I shot the standard focus chart using both Single AF and manual focus live view, and both lenses are identical and focus fine right on the center line using a 45 degree angle. So I believe the focus is fine.

One other interesting fact is, not that it matters much, my lens has a 190.2 cm MFD, his is 186 cm. Seems like maybe his glass is a bit misaligned.

Any comments regarding why any MFA makes no difference as explained in my previous post? Please note that it doesn't make it better or worse - just no change.

TheLensGuy wrote in post #16144031
Any comments regarding why any MFA makes no difference as explained in my previous post? Please note that it doesn't make it better or worse - just no change.

If you are testing using live view (contrast detect AF) then MFA which is not used for live view (only phase detect AF) doesn't have any input.

So how can one MFA a lens and accurately manual focus that is something 10 meters away without 10x liveview?

TheLensGuy wrote in post #16144382
So how can one MFA a lens and accurately manual focus that is something 10 meters away without 10x liveview?

Can you reword that question? It doesn't make sence to me. MFA and liveview are completely independent or each other. You basically use do MFA by trial and error although the new focal software you can buy makes it easy by doing the trial and error in software. This does not take away from using 10x in liveview.

FEChariot wrote in post #16144410
Can you reword that question? It doesn't make sence to me. MFA and liveview are completely independent or each other. You basically use do MFA by trial and error although the new focal software you can buy makes it easy by doing the trial and error in software. This does not take away from using 10x in liveview.

What I mean is, how will I know, when I focus on the chart 10 meters away, that it is in fact focusing on the correct point and I am getting the result I get? I may dial in MFA +3 and may get tack sharp results but that very well may be because that tiny dot focused on something it shouldn't have been. With liveview 10x, I can manually focus the lens with more precisely.

I don't use the live view when testing lenses under 85mm, because that you can manage and you are a lot closer with little room for error. But when you are testing 200 + 2x III @ 400mm 10 meters away, it gets very tough without liveview 10x.

TheLensGuy wrote in post #16144466
What I mean is, how will I know, when I focus on the chart 10 meters away, that it is in fact focusing on the correct point and I am getting the result I get? I may dial in MFA +3 and may get tack sharp results but that very well may be because that tiny dot focused on something it shouldn't have been. With liveview 10x, I can manually focus the lens with more precisely.

I don't use the live view when testing lenses under 85mm, because that you can manage and you are a lot closer with little room for error. But when you are testing 200 + 2x III @ 400mm 10 meters away, it gets very tough without liveview 10x.

Like I said, liveview doesn't use MFA. Manual focusing does not have anything to do with either contrast detect (through the viewfinder) or phase detect,( live view auto focus. ) You seem to be confusing so many things I don't know where to start.

If you want to test AF accuracy ( through the viewfinder AF that is) pthen use MFA. If you are testing lens sharpness, then use manual focus or life view AF. I doubt you would see much better results in manual focus or live view AF. Read this article: (See Rogers results in live view MF versus live view AF)

http://www.lensrentals​.com …oint-single-shot-accuracy

Maybe google phase detect and contrast detect AF and read on how they differ.

FEChariot wrote in post #16144511
Like I said, liveview doesn't use MFA. Manual focusing does not have anything to do with either contrast detect (through the viewfinder) or phase detect,( live view auto focus. ) You seem to be confusing so many things I don't know where to start.

If you want to test AF accuracy ( through the viewfinder AF that is) pthen use MFA. If you are testing lens sharpness, then use manual focus or life view AF. I doubt you would see much better results in manual focus or live view AF. Read this article: (See Rogers results in live view MF versus live view AF)

http://www.lensrentals​.com …oint-single-shot-accuracy

Maybe google phase detect and contrast detect AF and read on how they differ.

Thanks I knew about the difference, just didn't know live view did not take MFA in to account. I guess all of our MFA testing was useless and we need to re-do everything pretty much as the reason to all of this could be incorrect focusing.

That being said, using live-view, when I put both lenses on the same camera on a tripod with a 45 degree angle looking at the focus chart on the floor, they both focus on the center line correctly. Not sure if that means anything.

With live view, it is the image sensor that captures your photo. And it is the image sensor that gives the image on the LCD. So if you do manual focus until max sharpness on the display, it will also be max sharpness on the photo. And if you ask the camera to do a contrast-based AF, it will do it using the sensor - so no missmatch between focusing and then taking the photo.

Phase-detect on the other hand, is a totally separate optical path. And that is the reason for a need to calibrate the camera/lens so the phase-detect AF will compute correctly, i.e. so that best guess from the phase-detect optical path will result in an image that will also be in focus on the camera image sensor. And there is where the MFA comes in - to make sure the phase-detect AF doesn't front- or back-focuses.

You have to be careful or at least knowledgable when testing via the 45 degree angle stuff. Two things about Canon AF systems that are very important to know here. First Canon AF systems always focus on the closest thing with contrast. And 2 the AF points are actually bigger than the rectangles in the camera viewfinder.

So say your shooting a tape measure. You are at a 45 and aim for 12". The picture comes back and 11.5" is in focus. So you think your front focusing, but the camera is actually doing its job. You then dial in BF MFA and your pictures are back focusing!

It's far better to shoot a target with contrast parallel to the sensor plane and run the tape at a 45 far outside the active AF point but still in the frame.