Re: Canon R5 (and R5M2) vs Nikon Z8 (low light AF performance)
Vento wrote: alundeb wrote:
DXO have not tested the RF.
There is no data nowhere supporting anything else than the RF 85 f/1.2 is T1.3
The DS variant loses 1.5 stops and is T2.2 according to Canon.
All we know about the non DS version of the RF 85/1.2 L is the f-stop, t-stop therefore unknown and that Canon specifies the DS version as T2.2.
Without measurements you can't say anything, so no T1.3 either.
DXO has at least measured previous models and the Canon EF 85mm f/1.2L II USM is at T1.5.
When you see that the Canon EF 50mm f/1.2L USM was measured at T1.5 by DXO Score and the current model Canon RF 50mm f/1.2L USM also only at T1.5, I wouldn't be surprised if it doesn't achieve a T1.3.
DXO mark has not tested very many lenses and some of them have been around for a few years, but none of the Canon f/1.2 L lenses tested so far has a better T-stop measurement than T1.5.
The now ancient Canon EF 85mm F1.2L USM, still identical in optical construction to my old Canon nFD 85/1.2 L, was still at T1.6.
So Bernardi is not entirely wrong in the tendency, none of the Canon f/1.2 L lenses tested with DXO Mark achieves a better T-stop value than T1.5 in their measurements.
The situation is definitely different at Nikon with the Z 85/1.2 S., which was measured with T1.3, or the Noct with T1.1.
It may be a coincidence that the RF 85/1.2 L performs better there than, for example, the Canon RF 50/1.2 L, or the predecessor models of the RF 85/1.2 L, but this is rather doubtful.
If you measure T-stop for very large aperture lenses using Canon cameras, there is a known problem with pixel vignetting with certain Canon cameras. When the exit pupil is very large and close, the light from the most oblique rays do not hit the photodiodes. So the low T-stop measurements could be a sensor problem and not a lens problem. I don't know the situation regarding this for the R5 or R3, but the R5 is said to be better in this regard than the R, which was used for the DXO measurements with the RF 50/1.2.
The large difference between f-stop and T-stop measured by DXO does not appear for lenses with aperture f/1.8 or smaller. The RF 135/1.8 looks fine.
It would be very strange if Canon lenses systematically had large transmission loss only for the f/1.4 and f/1.2 lenses.
It seems the original article cannot be found anymore, but there is a discussion about it here:
