Canon 24-105L vs. Leica Prime Test

[Benjamin Hilton]

So I've been curious recently. I've been shooting with the GH4/5 Canon L 24-105 Metabones speedbooster for a good couple of years now, and have been pretty happy with the usability, color etc.

One problem though, I don't think my images are as detailed as they could be. Digital sharpness is there, kind of native to the GH processing, but the detail seems lacking. I've always put this down to the compression of my final render output, but I think there is something else going on. I assumed the 24-105 would be tack sharp, even at F4 as I've never seen anything amiss with DSLR photos and this lens combo. 

So, I did this quick test yesterday, comparing different focal lengths of the Canon to some Leica Primes, and also the Helios 56mm I believe it is. Nothing too scientific, just a quick cycle through apertures pointing at the same background. The focus point is the same double checked for each lens, and the exposure was just eyeballed using shutter speed to keep ND out of the equation. The Helios isn't too helpful for comparison as it couldn't focus far enough to match the other lenses, so take that one as you might. 

To add even more confusion, I labeled the Canon by the F number on the back of the GH5 screen (after the speed booster stop gain) and the Prime lenses are labeled by what was marked on the lens. So the F2.8 on the Leica would really be closer to a F1.9 after the speed booster, as compared to the Canon F2.5.

Anyone have any experience with these lenses or thoughts on the test or sharpness in general? 

[Adam Kuźniar]

I think you should eliminate the speedbooster from this comparison first. Whether it impacts the sharpness (and the amount of so) or not, it's a variable that makes this comparison canon+speedbooster vs leica, not canon vs leica.

[Benjamin Hilton]

That's a valid point. With having the GH5 as an A cam right now though, these lenses really don't do me much good without a speedbooster, meaning I'm really only interested in seeing how they perform in combo with a speedbooster and GH5. Lenses look very different depending on which sensor you pair them with.

 

[UncleBobsPhotography]

It's really worth testing these things with the exact setup you're planning to use. I found my Viltrox speedbooster+Canon L35mm1.4 to be unusably soft, while the speedbooster works great with the Canon 85mm1.8.

[Kisaha]

3 hours ago, UncleBobsPhotography said:

It's really worth testing these things with the exact setup you're planning to use. I found my Viltrox speedbooster+Canon L35mm1.4 to be unusably soft, while the speedbooster works great with the Canon 85mm1.8.

What camera?

There are a lot of variables influencing such an adapters performance. Sensor stack, lens's exit pupil, adapters materials and craftmanship. Usually adapters are reaching their limits with brighter and wider lenses. That is the rule.

That is why Metabones has so many different versions optimized for various cameras.

Viltrox EF-M2 ver.II costs185€, Metabones Speedbooster costs 850€ and Lens Regain used to cost around 550€ when it was available here. Guess which one performs best!

[Esguelha]

It looks quite sharp to me, even wide open. I am looking at small screen though, it's still 1080p. 

[Avenger 2.0]

The Canon 24-105mm f/4.0 L lens just isn't a sharp lens. Bought my first one second hand and thought it was just too used because at 100% the image most of the time looked slight out of focus. Sold it and bought a new 24-105mm f/4.0 ii L and this one is even worse. Bought the Sigma 24-105mm f/4.0 and it is way sharper. The only good thing about the Canon is that it's weather proof when needed. Maybe for video still good, but for photography not sharp enough for professional use.

[Emanuel]

Photography depends on variables. There are no soft nor sharp tools per se ; ) just soft or sharp combinations, respectively.

[KnightsFan]

3 hours ago, Emanuel said:

Photography depends on variables. There are no soft nor sharp tools per se ; ) just soft or sharp combinations, respectively.

That is so true. Even on paper, you have to look at combinations and not individual pieces. Leica uses a very thin filter stack on their sensor, whereas the standard MFT filter stack is very thick. So a lens designed for maximum sharpness on a Leica digital camera will not be as sharp on a MFT sensor. Canon is somewhere in the middle. (LensRentals has a lot of nice articles about it https://wordpress.lensrentals.com/blog/2014/06/sensor-stack-thickness-when-does-it-matter/)

[UncleBobsPhotography]

17 hours ago, Kisaha said:

What camera?

There are a lot of variables influencing such an adapters performance. Sensor stack, lens's exit pupil, adapters materials and craftmanship. Usually adapters are reaching their limits with brighter and wider lenses. That is the rule.

That is why Metabones has so many different versions optimized for various cameras.

Viltrox EF-M2 ver.II costs185€, Metabones Speedbooster costs 850€ and Lens Regain used to cost around 550€ when it was available here. Guess which one performs best!

I use it on the GH5. My suspicion is that it has trouble with wide angle lenses, but my plan is to just shoot some test charts with each new combination to verify their performance before using it. 

[Benjamin Hilton]

16 hours ago, Avenger 2.0 said:

The Canon 24-105mm f/4.0 L lens just isn't a sharp lens. Bought my first one second hand and thought it was just too used because at 100% the image most of the time looked slight out of focus. Sold it and bought a new 24-105mm f/4.0 ii L and this one is even worse. Bought the Sigma 24-105mm f/4.0 and it is way sharper. The only good thing about the Canon is that it's weather proof when needed. Maybe for video still good, but for photography not sharp enough for professional use.

I think the score for the mk ii was less sharp than the original in official chart tests