Canon Announces the RF 16-28mm F2.8 IS STM

[Exploreshootshare]

foto-erhardt.de was even faster. My inbox says 5:16 CET ;-)

Mine as well

 

[Zadak]

I think Canon very much wants to only release L-series lenses above $2,000 US from now on. Having a prosumer line in the $1100, 1400 etc range seems to be their plan. I'm fully expecting to see a $4,000 L-series lens within the next few years.

Inflation + weak JPY

 

[HeavyPiper]

Just in time as I've been looking to replace my EF 16-35mm f4.

 

[Exploreshootshare]

Thinking about the (ultra)wide angle options from Canon, there now is a really nice portfolio:

1. RF 10-20mm F4 L
2. RF 14-35mm F4 L
3. RF 15-35mm F2.8
4. RF 15-30mm F can´t remember
5. RF 16mm F2.8
6. RF 16-28mm F2.8

7. RF 24mm F1.8
8. RF 24mm F1.4 L
9. RF 28mm F2.8

(hopefully) still to come:
10. F2 UWA zoom
11. RF 12/ 14mm f2 L (something along that)
12. RF 20mm F1.4 (oh please Canon, please!) ---> the one lens I am waiting for here. If you don't want to make it, let Sigma do it!

That sounds like a great UWA portfolio to me

Edit: Omg, I forgot a few lenses added the 10-20mm, 24mm and 28mm. Don´t know where exactly to draw the line here. I once learned: 20mm - 35mm and lower --> (moderate) wide angle
10mm - 19mm --> ultrawide angle

I don´t know if everybody agrees to it.

 

[john1970]

Thinking about the (ultra)wide angle options from Canon, there now is a really nice portfolio:

1. RF 14-35mm F4 L
2. RF 15-35mm F2.8
3. RF 15-30mm F can´t remember
4. RF 16mm F2.8
5. RF 16-28mm F2.8

(hopefully) still to come:
6. F2 UWA zoom
7. RF 12/ 14mm f2 L (something along that)
8. RF 20mm F1.4 (oh please Canon, please!) ---> the one lens I am waiting for here. If you don't want to make it, let Sigma do it!

That sounds like a great UWA portfolio to me

I would also add the Canon RF 10-20 mm f4 L IS STM lens to that list.

 

[candyman]

Thinking about the (ultra)wide angle options from Canon, there now is a really nice portfolio:

1. RF 14-35mm F4 L
2. RF 15-35mm F2.8
3. RF 15-30mm F can´t remember
4. RF 16mm F2.8
5. RF 16-28mm F2.8

(hopefully) still to come:
6. F2 UWA zoom
7. RF 12/ 14mm f2 L (something along that)
8. RF 20mm F1.4 (oh please Canon, please!) ---> the one lens I am waiting for here. If you don't want to make it, let Sigma do it!

That sounds like a great UWA portfolio to me

You can add the Rf 28mm f/2.8 STM as well

 

[AlanF]

Review by The Digital Picture
Edit: This review does not yet include de image quality comparisons.

Yeah, it’s just a bland preview. Wait until he tests it.

 

[candyman]

Yeah, it’s just a bland preview. Wait until he tests it.

Yes, Better the reviewers use the term preview or announcement or introduction

 

[Brown]

When is Canon going to return to internal zooms for their ultra-wides?
Is a hybrid Z lens going to be the only one that\'s non-extending?

 

[AlanF]

For amusement, I collated the MTFs for the RF 16-28mm and the lowly RF 16mm/2.8, and its centre cropped to 28mm field of view (14.6 Mpx crop from R5). The zoom at 16mm is remarkably good, but the el cheapo prime can be adequate for less demanding shots at 16mm and cropped to 28mm.

 

[AJ]

Review by The Digital Picture
Edit: This review does not yet include de image quality comparisons.

Thanks!
I found the MTF mouse-over comparison to be quite useful. At 28 mm the new lens is a vast improvement over the 28-70/2.8

 

[AJ]

From the TDP preview:

16-28 promises to rival the 15-35's image quality, but it will have considerably stronger geometric distortion.

Which makes me think: do the published MTFs take into account digital distortion correction? If so, it would probably put at ease some of those who are still hesitant about this.

 

[Viggo]

What makes it not an L? It’s even weather sealed. It lacks both MF and clickable control ring, can that be it? It seems like a better option than my 14-35 L now, and a stop faster..

 

[neuroanatomist]

What makes it not an L? It’s even weather sealed. It lacks both MF and clickable control ring, can that be it? It seems like a better option than my 14-35 L now, and a stop faster..

Same reason the RF 28-70/2.8 is not an L, despite 'L-level optical performance' and weather sealing. There's already a 15-35/2.8L, just like there's already a 24-70/2.8L.

 

[Berowne]

Just in time as I've been looking to replace my EF 16-35mm f4.

I will not replace mine.

 

[neuroanatomist]

From the TDP preview:

Which makes me think: do the published MTFs take into account digital distortion correction? If so, it would probably put at ease some of those who are still hesitant about this.

Yes, the MTFs are shown with distortion correction applied. I don't know that Canon has explicitly stated that, but MTFs are determined along a diagonal line from the center of the image to the corner:


Lenses like the RF16/2.8, RF 14-35/4L, and RF 24-105/2.8L Z require distortion correction to 'fill the corners' which are otherwise black on a RAW image. On an MTF plot for one of those lenses without distortion correction applied, the lines would not reach the right side of the graph...but they do, therefore distortion correction is included in the MTF.

 

[HeavyPiper]

I will not replace mine.

The only reason to replace is the added weight of the adapter. I'm not as young as I once was.

 

[AJ]

Yes, the MTFs are shown with distortion correction applied. I don't know that Canon has explicitly stated that, but MTFs are determined along a diagonal line from the center of the image to the corner:
View attachment 222091

Lenses like the RF16/2.8, RF 14-35/4L, and RF 24-105/2.8L Z require distortion correction to 'fill the corners' which are otherwise black on a RAW image. On an MTF plot for one of those lenses without distortion correction applied, the lines would not reach the right side of the graph...but they do, therefore distortion correction is included in the MTF.

Right. Seems to me there are two aspects to this. (1) determining what constitutes the corner. It would have to be the corrected corner because the uncorrected corner is black and thus has zero lp/mm. (2) an adjustment of resolution because of the stretching required. This may hit the lp/mm by a few percent.

 

[neuroanatomist]

Right. Seems to me there are two aspects to this. (1) determining what constitutes the corner. It would have to be the corrected corner because the uncorrected corner is black and thus has zero lp/mm. (2) an adjustment of resolution because of the stretching required. This may hit the lp/mm by a few percent.

The resolution hit is likely fairly minor. Since the optical performance of the lens is modeled in silico, and the MTFs are generated from those models, distortion correction can be done in an ideal way.

Note that real-world performance is probably going to be less than that ideal, things get worse toward the corners and things like moderate recentering will have the biggest effect away from the center of the image.

 

[AJ]

The resolution hit is likely fairly minor.

Agreed. That's why, if it is included in the MTF curves, it would allow apples-for-apples comparisons and it would put to rest complaints about digital distortion corrections.