fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Sony Forum | Join Upload & Sell

1              3       end
  

Archive 2024 · Optical superiority vs Lens Profile Corrections

  
 
Grenache
Offline
• • • •
Upload & Sell: On
p.2 #1 · Optical superiority vs Lens Profile Corrections


mogul wrote:
Is the light new 300f2.8 optimized by software?


At least in LR, the only thing that the profile does for the 300GM is over-correct vignetting…which the lens has very little of in the first place.




Aug 19, 2024 at 02:14 PM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.2 #2 · Optical superiority vs Lens Profile Corrections


Let me debunk this silly notion. There is no intrinsic relationship between optical performance and size/weight. There is only intelligent user-centered design, and the others who want to load you up with a dependency on computer fixes or large lens bodies, both of which are highly undesirable and not in your best interests.

I only use small and light lenses that offer great performance with no fixes ('profiles'). Sometimes detailed information can help make a point with more clarity:

All Voigtlanders: 21/3.5 (230g); 28/2 (230g); 35/2 (350g); 50/2 (365g); 75/1.5 (350g). Nor are they particularly slow as a set - average aperture is f2.2. Total weight for five lenses is 1525 grams, average of 305 grams. I'm not giving up anything here, these are hand-picked for their image quality and weight/size. It wasn't hard.

It's a marketing meme - essentially a racket to hide lazy design practices and bloatware lenses. I haven't used a fix ('profile') in years. But strangely, I didn't see an option for me to select.



Aug 19, 2024 at 07:59 PM
Steve Spencer
Offline
• • • • • • •
Upload & Sell: On
p.2 #3 · Optical superiority vs Lens Profile Corrections


philip_pj wrote:
Let me debunk this silly notion. There is no intrinsic relationship between optical performance and size/weight. There is only intelligent user-centered design, and the others who want to load you up with a dependency on computer fixes or large lens bodies, both of which are highly undesirable and not in your best interests.

I only use small and light lenses that offer great performance with no fixes ('profiles'). Sometimes detailed information can help make a point with more clarity:

All Voigtlanders: 21/3.5 (230g); 28/2 (230g); 35/2 (350g); 50/2 (365g); 75/1.5 (350g). Nor are they particularly slow as a set - average
...Show more

Sorry, Phillip, but there is at least one obvious relation between physical size and optical performance: vignetting. And notably all those Voigtlander lenses have very high vignetting and this does affect the image and particularly the bokeh in certain predictable ways (swirly bokeh with cat's eye shapes further into the frame).

I have had the 28 f/2 II, the 50 f/2 APO, and 75 f/1.5 and the 15 f/4.5 III, 21 f/1.4, 35 f/1.7, 40 f/1.2, 50 f/1.2, and 90 f/2.8 APO and they are all good lenses, but it is not as though the small size comes with zero tradeoffs.



Aug 19, 2024 at 08:30 PM
chez
Offline
• • • • • • •
Upload & Sell: Off
p.2 #4 · Optical superiority vs Lens Profile Corrections




Steve Spencer wrote:
Sorry, Phillip, but there is at least one obvious relation between physical size and optical performance: vignetting. And notably all those Voigtlander lenses have very high vignetting and this does affect the image and particularly the bokeh in certain predictable ways (swirly bokeh with cat's eye shapes further into the frame).

I have had the 28 f/2 II, the 50 f/2 APO, and 75 f/1.5 and the 15 f/4.5 III, 21 f/1.4, 35 f/1.7, 40 f/1.2, 50 f/1.2, and 90 f/2.8 APO and they are all good lenses, but it is not as though the small size comes with
...Show more

Let’s also not forget the Voigtlander lenses are manual focus…not really an equal comparison against lenses that are AF.



Aug 19, 2024 at 08:50 PM
tctmp
Offline
• • •
Upload & Sell: Off
p.2 #5 · Optical superiority vs Lens Profile Corrections


Steve Spencer wrote:
Sorry, Phillip, but there is at least one obvious relation between physical size and optical performance: vignetting. And notably all those Voigtlander lenses have very high vignetting and this does affect the image and particularly the bokeh in certain predictable ways (swirly bokeh with cat's eye shapes further into the frame).

I have had the 28 f/2 II, the 50 f/2 APO, and 75 f/1.5 and the 15 f/4.5 III, 21 f/1.4, 35 f/1.7, 40 f/1.2, 50 f/1.2, and 90 f/2.8 APO and they are all good lenses, but it is not as though the small size comes with
...Show more

And heavy vignetting is equivalent to cheating, because the ISO needed will be higher than otherwise. For apple to apple comparison, T-stop should be used. That's why I don't buy the software fix for vignetting. It's like saying f/4 works as well as f/2.8 because one can raise the exposure by 1 stop in post. They are not the same.



Aug 19, 2024 at 09:07 PM
JohnDizzo15
Offline
• • •
Upload & Sell: Off
p.2 #6 · Optical superiority vs Lens Profile Corrections


Some very good points made thus far. I also agree, that from certain vantage points, the options for the poll have the potential to create some not entirely accurate data. However, I just wasn't sure how else to convey the points, as I didn't want the poll itself to get deep into the weeds by having 10 different options. Figured I'd leave the nuances to the actual discussion portion, as far as added insight goes.

Just to add a bit of background as to what prompted me to pose the question. I shoot Sony, Canon, and Fuji together, so I pay attention to what's going on with all three brands. One of the things I was really looking forward to for the RF system was the release of their fast 35mm prime. When it was finally released recently, what came to light fairly quickly was that the lens leaned very heavily on baked in lens profile corrections. If anyone is interested in further detail on this, Sam Hurd does a great job starting at around the 3 minute mark of his review on YouTube.

The initial selling points seemed to be that the lens was video-centric due to low focus breathing and the new VCM focus system, and the size/weight. However, the distortion is fairly high, and actually shaves off roughly a 400x260 resolution loss due to the shaving of the sides of the image in correction. This, before we even get to the vignetting. Over on the Canon forum, there seemed to be a decent number of members applauding the smaller size of the 35 without concern for the optical shortcomings, which really surprised me.

In contrast, the Sony GM 35 seems to do a much better job with regard to distortion sans having to rely on baked in corrections.

I've also been playing with the Viltrox 27/1.2 for the Fuji system for the last few weeks, which is another lens that is on the larger side, but is very well corrected without the need for any profiles (which I'm loving). But again, I've seen some members mention how the size of it relative to other Fuji lenses is off putting.

All of this got me thinking more about the philosophy behind these designs and how most of us around these parts feel about what is more important.

Personally, I'd like for there to be high IQ medium speed primes that are compact, alongside their larger siblings that are there to push the envelope of what is possible with modern engineering and design. What I don't like, is what Canon did with the RF 35 VCM. The only way I'd change my position on this, is if they also release their previously rumored 35/1.2 that happens to be a statement lens of epic proportions.



Aug 19, 2024 at 10:16 PM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.2 #7 · Optical superiority vs Lens Profile Corrections


Couple more points. The very large/heavy lenses have very large/heavy elements to move around, and consumer demand for ultra rapid focus speeds - a camera driven development - has led to heavy focusing hardware and multiple focus motors. So they get even heavier.

There are success stories. Sony replaced a 780g 50/1.4 with a 515g 50/1.4, with better technical performance and greater complexity (14/11 vs 12/9). No idea about profile dependence, probably not much.

But the major companies just pile on the complex designs. To the point where one wonders what the expected longevity is for these Rube Goldberg monstrosities. Please take a look at the tear down by lensrental of the Canon 50/1.2, and project forward 5-10 years of pro use to imagine what the repair costs or a CLA would cost your wallet.

https://www.lensrentals.com/blog/2018/12/disassembly-of-the-canon-rf-50mm-f1-2l/

"I wish we had enough time (and the guts) to disassemble further into the front internal and external rings, but we weren’t pushing our luck today..so we’re a little nervous about going further in this lens." "Time pressures (and fear, there was also fear involved) kept us from disassembling the inner focusing barrel"

So that is 'modern engineering and design' Canon style. Meanwhile Cosina just released an APO 50mm lens at a modest f3.5, that weighs 175 grams and is tiny, with eight elements and fantastic performance. How do they do it? They are glass specialists of a very high order. Their lenses will continue on decades after the Tesla-like Canon is gathering dust in the landfill. I still use 40 year old lenses that are fully serviceable and produce endearing images, made much better by digital cameras.

Steve, vignette is an artistic quality as you well appreciate. It's been a Zeiss specialty seemingly forever, Otus included. I have been reading critical commentary like yours for literally decades, lol. They have always reported it in their data sheets. It's usually very tasteful and saves many of us adding it in post to 'close off' the image edges. The outer frames are never as important as the content in the magic donut, out to IH of 12mm. Only photographers care. Some of them probably look at the corners first

People may be rejecting the bloater lenses and even the 'technical but boring' lenses (Nikon S in this case). See this important thread on the Nikon forum, for an indication of where popular photographer taste is going:

https://www.fredmiranda.com/forum/topic/1821301/



Aug 19, 2024 at 11:44 PM
j4nu
Offline
• • • • •
Upload & Sell: Off
p.2 #8 · Optical superiority vs Lens Profile Corrections


JohnDizzo15 wrote:
Some very good points made thus far. I also agree, that from certain vantage points, the options for the poll have the potential to create some not entirely accurate data. However, I just wasn't sure how else to convey the points, as I didn't want the poll itself to get deep into the weeds by having 10 different options. Figured I'd leave the nuances to the actual discussion portion, as far as added insight goes.

Just to add a bit of background as to what prompted me to pose the question. I shoot Sony, Canon, and Fuji together, so I pay
...Show more

Well said!
I also think there is room for both, but if I had to choose I'd probably go with a reasonably software-optimized lens like 35GM.
On the other hand, it really comes down to each case. Continuing with 35GM, I really like what it has to offer for its size & weight but there are properties which I like more on the considerably bigger and heavier Sigma 35/1.2DN. It's not really about faster aperture, but the quality of blur is simply nicer on the Sigma at equal apertures. Sharpness / contrast is similar, if not better, on the GM though ...



Aug 20, 2024 at 02:59 AM
tsdevine
Offline
• • • • •
Upload & Sell: On
p.2 #9 · Optical superiority vs Lens Profile Corrections


Who knows if Bryan had good copies of both the GM 35/1.4 and the Canon RF 35/1.4 VCM L, but it's worth looking at the distortion differences between the two. As well as the apparent performance at various apertures.

Note, FM truncates the URLs for TDP. You either have to manually pick the second lens (the Canon in this case), or copy and past the URL into your browser.

Distortion
https://www.the-digital-picture.com/Reviews/Lens-Distortion.aspx?FLI=0&FLIComp=0&Lens=1537&Camera=1538&LensComp=1694

Quite a bit difference no?

Now lets looks the the image quality.

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=1537&Camera=1538&Sample=0&FLI=0&API=0&LensComp=1694&CameraComp=0&FLIComp=0&APIComp=0

Wide open they seem to be reasonably in same ballpark. But as you stop both lenses down, the Canon never really improves in the corners, but the GM does. I think part of the reason is that the Canon just can't overcome the amount of digital distortion correction required for that particular lens design. Could be wrong, but I believe my reasoning is sound.

I'd pick the GM every time.

JohnDizzo15 wrote:
Some very good points made thus far. I also agree, that from certain vantage points, the options for the poll have the potential to create some not entirely accurate data. However, I just wasn't sure how else to convey the points, as I didn't want the poll itself to get deep into the weeds by having 10 different options. Figured I'd leave the nuances to the actual discussion portion, as far as added insight goes.

Just to add a bit of background as to what prompted me to pose the question. I shoot Sony, Canon, and Fuji together, so I pay
...Show more




Aug 20, 2024 at 05:32 AM
Kevner
Offline
• •
Upload & Sell: Off
p.2 #10 · Optical superiority vs Lens Profile Corrections


For what I prefer to shoot, which is mostly B&W architecture, I prefer to have lenses that have minimal optical distortion. I can live with greater vignetting. Color accuracy matters when I'm doing a professional shoot. I have found when I need to do perspectival corrections, that lenses that rely on digital corrections are harder to achieve good results with.

FWIW - I have quite a few lenses for a lot of different shooting situations.



Aug 20, 2024 at 05:56 PM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.2 #11 · Optical superiority vs Lens Profile Corrections


A reasonable coverage of the vignette and distortion debate here at FM:

https://www.fredmiranda.com/forum/topic/1742733/

To reiterate, the main Voigtlander lenses have tremendous corner resolution, which may have got lost in the chat. Better than Leica M, in fact. And superb color reproduction, tonal separation.



Aug 20, 2024 at 06:53 PM
JohnJ
Offline
• • • •
Upload & Sell: Off
p.2 #12 · Optical superiority vs Lens Profile Corrections


My complaint about software correction is with being forced to upgrade for profiles for new lenses alone. I have a relatively old version of Capture One ( just a few years old) which works just fine however the only way to optimise a new lens is to pay quite a lot for a new version of software. I'm not a fan of that approach, especially at the extortionate cost of Capture One.


Aug 21, 2024 at 07:04 AM
zugzwang2
Offline
• •
Upload & Sell: Off
p.2 #13 · Optical superiority vs Lens Profile Corrections


JohnJ wrote:
I have a relatively old version of Capture One ( just a few years old) which works just fine however the only way to optimise a new lens is to pay quite a lot for a new version of software.


I thought that Capture One did well with new lenses so long as (like new Sony lenses) information was encoded in the EXIF data. Although some of my older lenses like the Zony 55 can be treated to Capture One special sauce in my outdated version 22 of the program, the "Manufacturer Data" option works very well with the lenses that I use regularly (the Loxia 25 and the Voigtlander 35 and 50 APOs) that do not have specific C1 profiles but do include the necessary data.

(That said, my Capture One problem isn't with lenses, but that my version of the program cannot recognize new cameras, so I can't evaluate the new 24-50 G because all the raw files that I've found from that lens were captured with Sony's newest cameras.)



Aug 21, 2024 at 09:50 AM
JohnDizzo15
Offline
• • •
Upload & Sell: Off
p.2 #14 · Optical superiority vs Lens Profile Corrections


Reading through the replies and looking at the results of the poll, it actually begs another question in my mind. Where do we draw the line for what is acceptable, with regard to the level of reliance on software to get an image to a certain point? At what point would we consider something transitioning into electronic art versus a photograph?

In many ways, we are already there with regard to things like bokeh emulation, face cleanup, body shaping, sky replacement, fake flare/artifacts, haze, etc. Those things I just listed are just deeper into editing territory, but still stem from a lot of the more minimal things we all do in post with things like distortion correction/line bending, masking, color profiles, curve adjustments, etc.

I don’t believe most of us in the forum are huge fans of AI generated images, as we are photographers, and not graphics people. But are some of the images we see these days far off from a completely computer generated image? Just something I was sloshing around this morning while I was looking through the thread while sitting on the can.

Thoughts?

Side note - all of this begs yet another question. How will we feel about all of the software trickery in phone cameras/smaller sensor devices over the next 5-10 years? When these computing devices and software get even better, almost everything we do with hardware will be mostly replicable with processing. Just sliding a bit further down the slope.



Aug 21, 2024 at 01:02 PM
NonDecaf
Offline
• • •
Upload & Sell: Off
p.2 #15 · Optical superiority vs Lens Profile Corrections


Can you define "optically superior" ? Every time you correct something optically you trade off something else. For e.g. If you add aspherical elements, you trade off bokeh quality. If you add too many correcting elements you reduce light transmission, etc.





Aug 21, 2024 at 02:52 PM
tsdevine
Offline
• • • • •
Upload & Sell: On
p.2 #16 · Optical superiority vs Lens Profile Corrections


The proof is in the pudding, specifically talking about the attributes of the lens. How well does it perform for a given person's need? And each person sets their own bar, so you're not going to get some master consensus on this.

"Boy, this lens requires a lot of digital correction of distortion, but I like the images it produces for my portraiture work." - Can't say that person is wrong, right?

"Boy, for my landscape shooting, I wish the corners resolved more details" - Can't say that person is wrong, right?

Can you even say what factors drove the comments above?

Just like with all the tradeoffs in lens design, everyone looks at the lens and decides on its merits based on their value system.

Edited on Aug 21, 2024 at 07:24 PM · View previous versions



Aug 21, 2024 at 03:27 PM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.2 #17 · Optical superiority vs Lens Profile Corrections


'sloshing around this morning while I was looking through the thread while sitting on the can.'
Garbage in, garbage out?

Being serious, there is no 'we' here. Each will decide, that is why I pointed out that heritage manual focus lenses are still going strong - the modern MF lenses will be the same, so some of us (you could say) are set for a very long time. Many are user serviceable so they lie at the opposite end of the spectrum from say, Canon, the complexity kings.

But going forward, the eternal problem arises that the mass of people determine what the market produces. And the mass of people are so easy to manipulate. Speaking of trade-offs, the Leica Q series garnered some rave reviews, but not from Diglloyd:

'This is not a lens that can ever make a fully sharp picture, because distortion correction stretches pixels apart for at least 75% of the frame. As the analysis shows, the claimed resolution of the Leica Q is a false marketing claim, since the actual usable recorded pixels is only 22.4 megapixels (approx) and there are the quality losses from stretching pixels apart (due to the pronounced distortion of around 13%, and its correction).'

https://diglloyd.com/prem/s/LEICA/LeicaM/LeicaQ-distortion-Dolls.html?dglyPT=true

Then: 'What was barely adequate on the 24-megapixel Q with visibly poor outer-zone sharpness became obviously unsharp (large outer zone area) on the 45-megapixel Q2. '

A little later: 'Along comes the Leica Q3 with a 60MP sensor, and the lens is now nothing short of pathetic given the demands of the sensor. That is, extreme barrel distortion with distortion correction (required) guarantees poor sharpness over at least half the frame.'

'Leica never mentions the excessive lateral chromatic aberration or the fun-house-mirror level of barrel distortion, both of which require heavy lifting by computer software to fix.'

https://diglloyd.com/blog/2023/20230524_2100-LeicaQ3.html

One of the take away issues here is this: today's computerized fake-optical performance lens may become tomorrow's obsolete lens, in the wake of higher resolution cameras. Leica went even further down the highway to optical hell:

'Automatic Geometric Correction: When the camera is pointed upwards, it automatically detects the angle and produces geometrically corrected JPEG files without user input.' In for a penny, in for a pound. Where is the line between image remediation and image construction? It is open-ended and we know where that leads in the digital world.



Aug 21, 2024 at 06:33 PM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.2 #18 · Optical superiority vs Lens Profile Corrections


13% distortion of the image produced. I imagine that is also in-camera correction, so you see something in the EVF that looks OK, but it would look hideous in the real file (a.k.a. computer input). These issues are good to roll around in your mind.

I really enjoy seeing what the mounted lens provides to the final image by gazing through the EVF, in terms of rendering shown with pretty standard camera settings I use. It was a joy in optical viewfinders and it remains so in mirrorless.

Then it dawned that no user of this heralded Q camera series can *ever see what the lens provides* (if true re EVF display) because it is already radically computer-altered for the view in the viewfinder and s/ware. Users would be already viewing a simulation, a computer manufacture like a Marvel movie.

These lenses are computer image precursors, rather than optical lenses as we know them.They almost cannot be considered as 'lenses', because they are designed as entry points - as inputs - to feed computer programs, upon which they are now totally dependent if they are not to offend the senses.

People are falling in love with a simulation. The programmatic development preceded the lens design. The lens was designed so it could feed the computer they use. The less the lens is involved, the more the computer does. AI will have fun, soon enough.



Aug 21, 2024 at 07:00 PM
philip_pj
Offline
• • • • • •
Upload & Sell: On
p.2 #19 · Optical superiority vs Lens Profile Corrections


Returning to the OP's premise: To reinforce the claim that superior lens designers can economically produce simple, high optical performance lenses that are small and light - more evidence against the received wisdom (peddled by so many) that these characteristics require heavy large lenses of high complexity and high costs:

For distortion levels of the CV 35/2, 50/2 and 75/1.5 listed by me in this thread, from FM reviews in that order:

'The Voigtlander 35mm f/2 APO is well corrected for distortion. There is no profile for Lightroom yet but when available, it won't do much to the image. Outstanding.'

'The Voigtlander 50mm f/2 APO-Lanthar has negligible distortion.'

'The CV 75/1.5 Nokton is well corrected for distortion only showing a slight pincushion. It's visible in real world images but not very apparent and very easy to correct in post if desired.'

That's what happens when you know what you are doing.



Aug 21, 2024 at 07:03 PM
chez
Offline
• • • • • • •
Upload & Sell: Off
p.2 #20 · Optical superiority vs Lens Profile Corrections




philip_pj wrote:
Returning to the OP's premise: To reinforce the claim that superior lens designers can economically produce simple, high optical performance lenses that are small and light - more evidence against the received wisdom (peddled by so many) that these characteristics require heavy large lenses of high complexity and high costs:

For distortion levels of the CV 35/2, 50/2 and 75/1.5 listed by me in this thread, from FM reviews in that order:

'The Voigtlander 35mm f/2 APO is well corrected for distortion. There is no profile for Lightroom yet but when available, it won't do much to the image. Outstanding.'

'The Voigtlander 50mm
...Show more

Again, you are comparing manual focus lenses against auto focus lenses. Surely you can see that to add the ability to autofocus at the speeds demanded by today’s cameras, the bulk and weight of the lens must be increased.

The bottom line here is use whatever lens achieves your goals. I’ve seen enough boring simple photos taken by the so called magnificent optical designed lenses and amazing photos taken by those so called “lenses for the masses” to know today’s lenses are all capable of producing great photos…it all depends on the person pushing the shutter.



Aug 21, 2024 at 07:15 PM
1              3       end




FM Forums | Sony Forum | Join Upload & Sell

1              3       end
    
 

Welcome back
Log in to your account