MTF Charts vs. Sharpness in the field

[Cryve]

I have a 135mm f1.8 lens and a 400 f2.8 Lens.

While the 400 f2.8 has "perfect" mtf charts and the 135mm f1.8 does not, the 135mm f1.8 shows way more sharpness in a field test.

Im talking about the sony lenses here. I tested this with the 61mp a7r iv.
I shot a testsubject and framed it the same with both lenses, meaning i had to walk further away with the 400mm lens.

Tests were shot with a tripod and a remote shutter. The results also align with my experience with these lesnes from months of use.

How can this be? Shouldnt the 400mm f2.8 be clearly sharper according to the MTF charts?

 

[EricN]

I have a 135mm f1.8 lens and a 400 f2.8 Lens.

While the 400 f2.8 has "perfect" mtf charts and the 135mm f1.8 does not, the 135mm f1.8 shows way more sharpness in a field test.

Im talking about the sony lenses here. I tested this with the 61mp a7r iv.
I shot a testsubject and framed it the same with both lenses, meaning i had to walk further away with the 400mm lens.

Tests were shot with a tripod and a remote shutter. The results also align with my experience with these lesnes from months of use.

How can this be? Shouldnt the 400mm f2.8 be clearly sharper according to the MTF charts?

I think you have a copy of the 400mm that slipped through Quality Control. There could be another reason, but I'm not sure what would be the next likely possibility.

 

[Cryve]

I think you have a copy of the 400mm that slipped through Quality Control. There could be another reason, but I'm not sure what would be the next likely possibility.

Im not sure. Tests from Websites like this show, that the 400 has worse image quality in a comparison. What do u think?

Sony FE 135mm F1.8 GM Lens Image Quality

View the image quality delivered by the Sony FE 135mm F1.8 GM Lens using ISO 12233 Resolution Chart lab test results. Compare the image quality of this lens with other lenses.

www.the-digital-picture.com

 

[EricN]

Im not sure. Tests from Websites like this show, that the 400 has worse image quality in a comparison. What do u think?

Sony FE 135mm F1.8 GM Lens Image Quality

View the image quality delivered by the Sony FE 135mm F1.8 GM Lens using ISO 12233 Resolution Chart lab test results. Compare the image quality of this lens with other lenses.

www.the-digital-picture.com

It's interesting. I've read some companies make their mtf based on a hypothetical perfect copy from their data without preforming any actual tests with a real lens. It seems like the next most likely explanation.

Off topic, but are your purchasing the new 300mm? If I had a Sony camera, I think the weight would be too good to pass up, although you already have 400mm.

 

[Cryve]

Jeah safed up a long time to buy the 400 and thought its my endgame lens, but i am quite disappointed in it.
First and foremost because of the "bad" sharpness and the chromatic abberations. Dont get me wront. Its "sharp" but im a Pixel peeper peeping at 400% in a 61 mp image ;D

Second reason for me is, that i actually find it shows too little background. Attached is a comparison between 400mm and 135 with same framing. The 400 "zoomes" the background in too much. I do Environmental wildlife photos and i prefer more environment in the photos.
Im probably selling it to afford the 300, even tho i just bought it 2 months ago.

Do vendors sometimes take lesnes like these back for a reduced price? Otherwise i have to sell it on ebay or sth. Similar.

I hope the 300 has better sharpness

It's interesting. I've read some companies make their mtf based on a hypothetical perfect copy from their data without preforming any actual tests with a real lens. It seems like the next most likely explanation.

Off topic, but are your purchasing the new 300mm? If I had a Sony camera, I think the weight would be too good to pass up, although you already have 400mm.

 

[EricN]

Jeah safed up a long time to buy the 400 and thought its my endgame lens, but i am quite disappointed in it.
First and foremost because of the "bad" sharpness and the chromatic abberations. Dont get me wront. Its "sharp" but im a Pixel peeper peeping at 400% in a 61 mp image ;D

Second reason for me is, that i actually find it shows too little background. Attached is a comparison between 400mm and 135 with same framing. The 400 "zoomes" the background in too much. I do Environmental wildlife photos and i prefer more environment in the photos.
Im probably selling it to afford the 300, even tho i just bought it 2 months ago.

Do vendors sometimes take lesnes like these back for a reduced price? Otherwise i have to sell it on ebay or sth. Similar.

I hope the 300 has better sharpness

I think 2 months is going to be longer than most stores will take an item back, but there's no harm in trying. You could also try contacting Sony and see what happens there.
If you have to go with Ebay, I recommend checking with Mpb and Keh and finding out what they will buy it for, then start your minimum bid at that price.
Good luck

 

[Cryve]

To get back to the topic, does anyone know what could couse the discrapancy between MTF charts and results in the real world?

It doesnt seem specific to this sony combo of lenses either. If you compare the MTF of the canon rf 135 and the rf 400 and then the tests for those lenses on "the digital picture" image comparison tool it shows the same story.

So whats the reason the MTF doesnt translate to the real world?


Im kind of afraid, because the MTF for the new sony 300mm is worse than the MTF for the 400 and i really dont wanna be cought in an eco system with "unsharp" lenses for big prints.

 

[koenkooi]

MTFs tend to be calculated for the lens focused at infinity, they can change when focused closer. The classic example at the Canon 100mm macro lenses that are very sharp at macro distances and less sharp when focused further away.

And since the MTF plots are simulations they could be different from reality. Canon changed the way they calculate the plot a few years ago to match reality better. A lot of existing lenses suddenly had worse MTF plots.

 

[Cryve]

That could explain what im seeing.

Still surprised tho, as the RF 135 and RF 400 are newer lenses and should have the newer MTF Charts by canon and still the 135 seems to be sharper, despite worse MTF, in the comparison tool on "The digital Picture".

Also The MTF from my Sony doesnt match reality regarding sharpness in the center of the frame vs. at the edge of the frame. The MTF shows "perfect" sharpness across the frame, while in reality i see much better sharpness in the center.

Thats very weird. I want to be able to trust MTFs when buying a lens. Seems we are reliant on reviewers like christopher frost or our own testing, to asses real world sharpness?

 

[koenkooi]

Don’t look at the Canon USA site for
MTFs, they mix up old and new plots, make sure the plots you’re comparing have the same colours and number of lines.

 

[neuroanatomist]

Still surprised tho, as the RF 135 and RF 400 are newer lenses and should have the newer MTF Charts by canon and still the 135 seems to be sharper, despite worse MTF, in the comparison tool on "The digital Picture".

Based on Canon’s MTFs, the RF 135/1.8 should be sharper in the center than the RF 400/2.8, and slightly less sharp in the corners. That seems consistent with TDP’s testing.

Also The MTF from my Sony doesnt match reality regarding sharpness in the center of the frame vs. at the edge of the frame. The MTF shows "perfect" sharpness across the frame, while in reality i see much better sharpness in the center.

Looking at Sony’s MTFs for the 400/2.8, it seems too good to be true.

Each manufacturer calculates MTFs as they choose, so they’re not comparable across brands (and you’re not doing that here). Only Zeiss actually shows empirically measured MTFs from an actual lens (at least, they used to), all others use theoretical MTFs based on the optical design of the lens. That means real world performance may differ from the MTFs, and importantly that one copy of a lens may differ from another.

 

[Cryve]

Thanks for the help in understanding this topic a bit better.

Honestly im just emotionally exhausted after having a lens like this be a dream goal for most of my photographic life. It didnt turn out to be what i had imagined.
Im really hoping now that the new 300 will be. As i explained in the thread earlier, its a better lens for my photographic style anyway, but if its not sharp im probably gonna be very devastated.

The absolute holy grail for my style of photography would be a 200mm f1.8. I would instantly switch to any brand that releases that. Its basically a 300 f2.8 blur wise, just wider.

But we havent had a 200 f1.8/f2 for a long time. Im wondering why. Is the market for such a lens just too small?

 

[AlanF]

There are lies, damn lies and Sony MTF charts.

 

[AlanF]

Based on Canon’s MTFs, the RF 135/1.8 should be sharper in the center than the RF 400/2.8, and slightly less sharp in the corners. That seems consistent with TDP’s testing.


Looking at Sony’s MTFs for the 400/2.8, it seems too good to be true.

Each manufacturer calculates MTFs as they choose, so they’re not comparable across brands (and you’re not doing that here). Only Zeiss actually shows empirically measured MTFs from an actual lens (at least, they used to), all others use theoretical MTFs based on the optical design of the lens. That means real world performance may differ from the MTFs, and importantly that one copy of a lens may differ from another.

Sigma claims to measure MTFs using a 46 Mpx Foveon sensor. They do actually show the measured ones with diffraction and the theoretical.

 

[EricN]

Thanks for the help in understanding this topic a bit better.

Honestly im just emotionally exhausted after having a lens like this be a dream goal for most of my photographic life. It didnt turn out to be what i had imagined.
Im really hoping now that the new 300 will be. As i explained in the thread earlier, its a better lens for my photographic style anyway, but if its not sharp im probably gonna be very devastated.

The absolute holy grail for my style of photography would be a 200mm f1.8. I would instantly switch to any brand that releases that. Its basically a 300 f2.8 blur wise, just wider.

But we havent had a 200 f1.8/f2 for a long time. Im wondering why. Is the market for such a lens just too small?

I don't know of any figures stating how many of those 200mm f/1.8 or 2.0 were made by Canon and Nikon, but if I remember the first EF version didn't come until about ten years after the mount was introduced. So, I wouldn't hold your breath, but I think it could be a possibility for RF after most other lenses are available. I don't know enough to even offer a guess about Nikon.

 

[Cryve]

I don't know of any figures stating how many of those 200mm f/1.8 or 2.0 were made by Canon and Nikon, but if I remember the first EF version didn't come until about ten years after the mount was introduced. So, I wouldn't hold your breath, but I think it could be a possibility for RF after most other lenses are available. I don't know enough to even offer a guess about Nikon.

For me a 200 f1.8 is like 600 f4, just for medium range photogrqphy enjoyers.

I doubt the market is big enough tho. A lens like this would cost about 6000 euros and most people can probably make do with a 70-200.

I do really hope tho, that a company will produce this legendary lens again .

 

[AlanF]

For me a 200 f1.8 is like 600 f4, just for medium range photogrqphy enjoyers.

What does that mean?

 

[Del Paso]

There are lies, damn lies and Sony MTF charts.

How dare you criticise holy Sony?

 

[Cryve]

How dare you criticise holy Sony?

Tbh they have plenty you can cricise them for, this being one of the things.
I still like their camaras for their customizability and honestly im just in love with 61 megapixels and noone else has offered that yet.

But canon do have extremly good lens design, arguably better in many cases, in my opinion and i value its community way more, having been a canon shooter myself a bit ago, which is also why im posting here and not elswhere.

 

[Cryve]

What does that mean?

If 600 f4 is the "racecar" for long range photography, then 200 f1.8 is the "racecar" for medium range in my opinion.

Of course telephoto lenses are nice to get closer and get extreme blur and i value them for that aspect, but if you want to show more of the enviornment in the background of the picture, then 135mm and 200mm are more ideal.

Some examples attached.

The first one is a swan with 400 f2.8. Its nice, but also you dont really see much of the environment and even if you wanted to show it, the swan would need to be like 80 meters away which is not practical in many situations. With a 600 this would be worse.

On the other hand with 135 you can show the envionment (see attached picture), but sometimes its not blurry enough or the animals are too far away. 200 f1.8 is the perfect middle ground here and gives maximum blur like nothing else, while still showing the environment.