Given the evolution from EF 11-24 to RF 10-20 I am getting afraid that more of this "getting wider" is thanks to the fact that lens corrections are moving from the glass to the firmware. So the glass is really wider, but the firmware corrections will make it suffer.
MintMar wrote:
Given the evolution from EF 11-24 to RF 10-20 I am getting afraid that more of this "getting wider" is thanks to the fact that lens corrections are moving from the glass to the firmware. So the glass is really wider, but the firmware corrections will make it suffer.
This isn't unreasonable to consider, but at the same time it may turn out better than we expect (or actually need), and obviously we can't do anything beyond speculate until third-party reviews hit.
The main concern though should be about size and weight... I can see such a lens being essentially perfect for travel, but man I hope it's manageable!
I am thrilled that Canon is focusing on producing lenses that are above $2000. We certainly don't want common, working class peasants to buy our same beloved Camera gear.
snegron7 wrote:
I am thrilled that Canon is focusing on producing lenses that are above $2000. We certainly don't want common, working class peasants to buy our same beloved Camera gear.
Signed,
A Working Class Peasant
You must really hate the 28-70/2.8 and 45/1.2, then? Or does the existence of higher-end lenses that push the boundaries of lens design negate their existence?
johnctharp wrote:
You must really hate the 28-70/2.8 and 45/1.2, then? Or does the existence of higher-end lenses that push the boundaries of lens design negate their existence?
I currently own the RF 16-28mm f2.8 STM, and yes, I strongly dislike it. If I had the money, I'd buy the RF 15-35mm f2.8L, but I don't. Therefore, I'm "stuck" with what I can afford, which is the RF 16-28mm f2.8 STM.
johnctharp wrote:
This isn't unreasonable to consider, but at the same time it may turn out better than we expect (or actually need), and obviously we can't do anything beyond speculate until third-party reviews hit.
The main concern though should be about size and weight... I can see such a lens being essentially perfect for travel, but man I hope it's manageable!
I would never buy a lens like that. 10mm is not wide enough for everything, f/2.8 is much faster than I want, and 70mm leaves a huge gap to the 100-xxx lenses. I suppose if you always used a 70-200 then it makes more sense as a 2-lens setup. If it has a stupid level of barrel distortion at 20mm, then definitely not worth it.
snegron7 wrote:
I currently own the RF 16-28mm f2.8 STM, and yes, I strongly dislike it. If I had the money, I'd buy the RF 15-35mm f2.8L, but I don't. Therefore, I'm "stuck" with what I can afford, which is the RF 16-28mm f2.8 STM.
I still use the 16-35/4 IS that works fine on R5/R5 II.
MintMar wrote:
Given the evolution from EF 11-24 to RF 10-20 I am getting afraid that more of this "getting wider" is thanks to the fact that lens corrections are moving from the glass to the firmware. So the glass is really wider, but the firmware corrections will make it suffer.
If it is anything like the VCM wides, the 14-35, or 24-105 Z, it will be fine.
What people forget is that in order to fully correct for distortion optically, you need additional elements which introduce their own aberrations, like coma and astigmatism. Allowing for some distortion enables a simpler lens design with greater sharpness edge-to-edge. Even after corrections, not much is lost.
I consider the corrections to be part of the optical design, taking the place of several lens elements that would otherwise have to be present. This was not possible in DSLR lenses, due to the possibility that they could be used with film cameras.
I think most lens reviewers have not really caught up with the new thinking, as lenses that are designed with digital corrections in mind should not really be judged for distortion (though I recognize why they still do this).
Distortion is not the issue; it's the loss of resolution if too much correction is needed. I wish they would publish the actual distortion and allow it to be turned off instead of people having to hack the files and do calculations. But if this is another video lens, then they are not caring much about that.
EBH
Mar 18, 2026 at 03:09 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
EB-1 wrote:
Distortion is not the issue; it's the loss of resolution if too much correction is needed. I wish they would publish the actual distortion and allow it to be turned off instead of people having to hack the files and do calculations. But if this is another video lens, then they are not caring much about that.
EBH
Yes, final resolution is of course important. The question, however, is whether the final resolution after allowing for higher distortion *and* correcting for it is higher or whether the final resolution is higher correcting the distortion optically. It is not obvious which approach will create the highest final resolution. Final resolution is of course not the only thing that is important either. It seems that allowing higher distortion and correcting for it can allow lenses to be substantially smaller. Would you prefer a lens that weighs twice as much but has 3% higher corner resolution? The answer is not obvious, but as in all things with lens design there are tradeoffs and striking a balance is not always easy. I say all of this even though my own general preference is for bigger lenses with more distortion correction built into the lens. Despite that preference I understand why many prefer smaller lenses with higher distortion that is corrected automatically.
And I am not always consistent in my preferences becasue of the tradeoffs. For example, I plan to get an RF 16-28 IS STM despite its high distoriton that is corrected automatically because I want to used it extensively for travel. For travel I will tradeoff high distortion that is corrected automatically to get smaller size. For that use size becomes most important.
EB-1 wrote:
Distortion is not the issue; it's the loss of resolution if too much correction is needed. I wish they would publish the actual distortion and allow it to be turned off instead of people having to hack the files and do calculations. But if this is another video lens, then they are not caring much about that.
EBH
At least in ACR / LR you can just turn off the corrections--they are not forced like they are in-camara.
Mar 18, 2026 at 03:44 PM
Steve Spencer Offline • • • • • • • Upload & Sell: On
Speaking of size, I would be a lot more interested in this lens if it was f/4 like the Sony. That would make it a really nice travel lens and slot it into a very nice f/4 L lens trio of the 10-20, 20-70, and 70-200, and it could basically be a revision of the EF 24-70 f/4 IS that never got made for RF. To me that would be a lot more compelling.
With the use of digital lens correction, this is quite possible. Although the more correction applied, the worse the result. I mean, you start stretching pixels and then you have to correct for the stretched corners with more correction. Not to mention the huge increase in noise when you start adding several stops of light to deal with vignetting.
Depends on the purpose of the lens, of course. In many cases, it's not a problem. But in others, it can and will be. There are, as mentioned above, tradeoffs.
Tom_W wrote:
With the use of digital lens correction, this is quite possible. Although the more correction applied, the worse the result. I mean, you start stretching pixels and then you have to correct for the stretched corners with more correction.
Not necessarily worse than having the pixels un-stretched by optical elements. It's a _different_ compromise. I certainly prefer the 10-20 "stretching" than my 11-24 "good old lens design."
Why, because it's smaller/lighter? Looking at the TDP data the 11-24 is sharper overall and has only a mild amount of light falloff. The RF lens is terrible for that with almost no improvement even at f/8 almost 3 stops. Maybe they got a bum 10-20 copy or they are not good up close or something else.
I'm not a photographer, but want even illumination in the final output. The 10-20 practically needs a center filter or you need to keep the ISO low enough not to have those radial looking differences in the outer areas when the edges/corners are jacked up. Wide open you expect it, but not stopped down. $2600 for that?
Like it or not, geometrical- and vignetting corrections seem to become normal. I have both the RF 10-20 and RF 24-105 f2.8, and I have to say that I’m really thrilled with the results from both of them, despite the heavy corrections applied.
