LensTip tests the brand new Sigma 85mm f/1.4 Art - sharpest lens ever :-)

[pwp]

It was solved according to Lenstip.....

Great performance indeed. Still a non-starter in my studio unless AF consistency is proven to be fixed. Sigma can make very sharp lenses, but until AF consistency is completely solved and become a distant memory, their lens sales will never reach their potential.

-pw

I'll be looking for confirmation on the AF consistency from multiple reviewers and also people like yourself.
Still, the initial LensTip position is certainly promising.

-pw

 

[Jopa]

It was solved according to Lenstip. My lens is on backorder with ETA in beginning of December. There is no stock in Australia until then. I will be able to report my findings shortly after.

Great performance indeed. Still a non-starter in my studio unless AF consistency is proven to be fixed. Sigma can make very sharp lenses, but until AF consistency is completely solved and become a distant memory, their lens sales will never reach their potential.

-pw

Congrats Alex! This lens is really fun to shoot.

 

[Viggo]

Did I read it right that 75% of the shots sharp at f1.4 is a great and fantastic result? If so, then wow, I sure have a VERY different opinion of what is fantastic and not....

 

[Alex_M]

Virgo

Sorry, not quite right.

Focus error : 2% of all shots only. that's an excellent result.

75% was the number of shots that reached _Maximum_ MTF50 resolution. That's excellent again.
23% of photos with errors smaller than 20% of the maximum MTF50 <<<that's not a focus error.

Focus with errors less than 20% of maximum lens resolution are normal for a DSLR. Focus cannot be at max resolution at all times. The sharpness achieved will be no worse than 80% of MTF50 lens resolution.

As far as I remember, ahsanford mentioned that for Canon 85 1.2 L lens focus error rate was around 5% ?

P.S.

According to Lenstip and ahsanford:

Canon 85L: 3% misses on a 1Ds3.
Sigma 85A: 2% missed on a 5d mark III

Did I read it right that 75% of the shots sharp at f1.4 is a great and fantastic result? If so, then wow, I sure have a VERY different opinion of what is fantastic and not....

 

[Viggo]

Thanks for clearing that up for me. It sounds a bit better then. I must try to find what the 50 Art did in the same test, if it got similar results I know not believe them, lol ;D

 

[Eldar]

Thanks for clearing that up for me. It sounds a bit better then. I must try to find what the 50 Art did in the same test, if it got similar results I know not believe them, lol ;D

I´m with you on this one. Having tried a number of 35 and 50 Art lenses, all with inconsistent AF, I will be a hard one to convince. Another problem as I experienced it was that AF could be consistent for a while and then drift off.

Has anyone seen a firm statement from Sigma explaining that they actually fixed this problem?

Besides, fresh rumours about new 50 and 85 primes from Canon may keep me patient for awhile longer

 

[Alex_M]

No worries, Sir!

http://www.canonrumors.com/forum/index.php?topic=31294.msg635583#msg635583

From LensTip:

50L: 7% on a 1Ds3
50 Art: 6-7% on a 5D3 -- but 'oscillating' or toggling between two options was reported on the 1Ds3 (no hit rate given), and this is the nutty 'it won't repeat focus' lens that TDP called out

85L: 3% misses on a 1Ds3
85 Art: 2%

Thanks for clearing that up for me. It sounds a bit better then. I must try to find what the 50 Art did in the same test, if it got similar results I know not believe them, lol ;D

 

[Viggo]

No worries, Sir!

http://www.canonrumors.com/forum/index.php?topic=31294.msg635583#msg635583

From LensTip:

50L: 7% on a 1Ds3
50 Art: 6-7% on a 5D3 -- but 'oscillating' or toggling between two options was reported on the 1Ds3 (no hit rate given), and this is the nutty 'it won't repeat focus' lens that TDP called out

85L: 3% misses on a 1Ds3
85 Art: 2%

Thanks for clearing that up for me. It sounds a bit better then. I must try to find what the 50 Art did in the same test, if it got similar results I know not believe them, lol ;D

Yup, all my five Art lenses have 20 different degrees of sharpness in 20 shots in a controlled optimal setting for AF.

 

[candc]

Thanks for clearing that up for me. It sounds a bit better then. I must try to find what the 50 Art did in the same test, if it got similar results I know not believe them, lol ;D

I´m with you on this one. Having tried a number of 35 and 50 Art lenses, all with inconsistent AF, I will be a hard one to convince. Another problem as I experienced it was that AF could be consistent for a while and then drift off.

Has anyone seen a firm statement from Sigma explaining that they actually fixed this problem?

Besides, fresh rumours about new 50 and 85 primes from Canon may keep me patient for awhile longer

from the sigma site:

"The newly designed Hyper Sonic Motor (HSM) offers 1.3 times the torque of its predecessor. Even at low speeds, it offers exceptionally stable performance, and the updated AF algorithm helps ensure nimble autofocus performance."

we will have to see if sigmas newly designed af motor and algorithms do what they claim.

 

[Viggo]

Thanks for clearing that up for me. It sounds a bit better then. I must try to find what the 50 Art did in the same test, if it got similar results I know not believe them, lol ;D

I´m with you on this one. Having tried a number of 35 and 50 Art lenses, all with inconsistent AF, I will be a hard one to convince. Another problem as I experienced it was that AF could be consistent for a while and then drift off.

Has anyone seen a firm statement from Sigma explaining that they actually fixed this problem?

Besides, fresh rumours about new 50 and 85 primes from Canon may keep me patient for awhile longer

from the sigma site:

"The newly designed Hyper Sonic Motor (HSM) offers 1.3 times the torque of its predecessor. Even at low speeds, it offers exceptionally stable performance, and the updated AF algorithm helps ensure nimble autofocus performance."

we will have to see if sigmas newly designed af motor and algorithms do what they claim.

... and more importantly, if it makes it into the other two lenses...

 

[Alex_M]

My impression is that AF algorithm updates will be offered for 35A and 50A at some stage later. That may fix AF inconsistency issue but to some degree only. What is more likely is that Sigma will release revised version of the lenses with new HSM motor and new AF alforithm as generation / version update.

 

[risc32]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

 

[Viggo]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

And to think the 300 is only 25 times the money of the 85 8)

I've been using the 85 f1.8 for a couple of months and I agree with you, it's not very good at all! I use it only for portraits with light and 2.8. Everything else it's REALLY useless.
It

 

[Ryananthony]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

Any error focusing with a Canon lens is always 100% user error, nothing else.

Any error focusing with a sigma lens is always 100% sigma af issues, never user error.

Didn't you know this?

 

[Maiaibing]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

Problem is no one does real comparative controlled AF tests so we end up with a lot of "impressions" and very little facts. The result is that claims are all over the place.

Dpreview is trying with camera AF for their newest camera reviews. I give them credit for at least trying. But its not a very controlled test - to say it mildly.

With a lens you would also have to have several copies.

 

[Jopa]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

Problem is no one does real comparative controlled AF tests so we end up with a lot of "impressions" and very little facts. The result is that claims are all over the place.

Dpreview is trying with camera AF for their newest camera reviews. I give them credit for at least trying. But its not a very controlled test - to say it mildly.

With a lens you would also have to have several copies.

The controlled AF tests are quite easy with FoCal. My worst lens (inconsistent) so far is the 70-200 f/2.8L II @ 200mm on the 5dsr. Suuuuuuuper random! And the funny thing it's 99.5%+ consistent on the 1dx2. Need to send the lens + the 5dsr to Canon, just don't have much time for this...

 

[Maiaibing]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

Problem is no one does real comparative controlled AF tests so we end up with a lot of "impressions" and very little facts. The result is that claims are all over the place.

Dpreview is trying with camera AF for their newest camera reviews. I give them credit for at least trying. But its not a very controlled test - to say it mildly.

With a lens you would also have to have several copies.

The controlled AF tests are quite easy with FoCal. My worst lens (inconsistent) so far is the 70-200 f/2.8L II @ 200mm on the 5dsr. Suuuuuuuper random! And the funny thing it's 99.5%+ consistent on the 1dx2. Need to send the lens + the 5dsr to Canon, just don't have much time for this...

I use Focal but I do not think you can rely on it alone. Result consistency is simply lacking. I use it as a first indicator and test on from there.

 

[Loswr]

I use Focal but I do not think you can rely on it alone. Result consistency is simply lacking. I use it as a first indicator and test on from there.

I've gotten very consistent results from FoCal, but I don't use their automated or semi-automated methods. Rather, I capture the images myself (with substantial oversampling) and run FoCal's manual analysis.

 

[3kramd5]

And to think the 300 is only 25 times the money of the 85 8)

Which canon 300mm (cine lenses notwithstanding) costs $30,000?

 

[streestandtheatres]

you guys must have better luck with canon AF than i have. It's very good, but perfect? I have had many lenses, and many camera bodies and i can't say that my sigma 35 art is in anyway worse than my canon lenses. in fact, i had to sell my 85mm 1.8 as it was just so random. i had much better luck with my 300 at 2.8 at minimum focusing distance than i ever had with that 85.

Problem is no one does real comparative controlled AF tests so we end up with a lot of "impressions" and very little facts. The result is that claims are all over the place.

Dpreview is trying with camera AF for their newest camera reviews. I give them credit for at least trying. But its not a very controlled test - to say it mildly.

With a lens you would also have to have several copies.

The controlled AF tests are quite easy with FoCal. My worst lens (inconsistent) so far is the 70-200 f/2.8L II @ 200mm on the 5dsr. Suuuuuuuper random! And the funny thing it's 99.5%+ consistent on the 1dx2. Need to send the lens + the 5dsr to Canon, just don't have much time for this...

My experience is similar. My 35A is 99% accurate on my 5dsr, and 50% on my 7dii. (I don't have a dock, I do have focal.)