What does "L" Quality Even Mean Anymore? - A Point in Focus S6E8

Two topics on today's podcast. First, I was contacted by another creator, @33rdframe, who wanted to share a 3D printed add-on for the R5C with me. I looked, and it looked interesting, especially the cable clamp that he built for it, so I'm passing it on. It's free, or at least he's giving away the STL files so you can print your own.
You can find his video here : • This Mod Turns your R5...
That said, the main topic for today is the question: What does L image quality really mean anymore?
Canon has always advertised their L lenses as having the best image quality they can muster. However, the marketing material for more and more lower tier lenses, like the recently released RF 28-70mm f/2.8 IS STM are claiming to have L image quality.
Simultaneously, and across the board, Canon's lens designers have embraced designs that rely more and more on extensive and significant software corrections; especially for distortion. And while this could be easily expected in entry level lenses, this design ethos has been used on many high end L series lenses as well.
This isn't to say that this is necessarily bad. Canon's RF 35mm f/1.4L VCM uses this extreme distortion design, and still delivers higher corner resolution than the EF 35mm f/1.4L II USM did using a traditional optical approach.
Admittedly, L image quality, was never something publicly defined by Canon. It's always just meant the best they could offer. But with more and more of what use to be done optically being handed over to software, does that change the calculus? Or is it just that my thinking about how a lens should be designed is antiquated, and the only thing that matters is how the images look, regardless of how you get to them?
══════════════════════
💵 Support the Channel and Content Like This 💵
══════════════════════
⯈ Click the Thanks Button under the video.
⯈ Use PayPal: paypal.me/poin...
⯈ Buy yourself something from the affiliate links below.
⯈ Amazon.com: amzn.to/4dloDzU
══════════════════════
Gear I Use to Make Videos
══════════════════════
(Affiliate links)
⯈ Canon EOS R5 - amzn.to/3uJNNVI
⯈ Canon EOS R5 Mark II -
⯈ Canon EOS R5C - amzn.to/4goyveX
⯈ SmallRig NP-F Battery Adapter Pro (power for extended shooting time)- amzn.to/3ybCY3h
⯈ Anker Prime 27650 mAh Power Bank - amzn.to/4dsP8o4
⯈ Canon RF 15-35mm f/2.8L IS USM - amzn.to/37SuLDw
⯈ Canon RF 24-70mm f/2.8L IS USM - amzn.to/3M9I0yQ
⯈ Rode NTG2 -
⯈ Rode Wireless Go -
⯈ Zoom F6 - amzn.to/2TWsSvo
══════════════════════
Chapters
══════════════════════

[@richardtoft6998]

As someone who cut his teeth shooting film in the 70's, I am bemused by your perceived dilemma. Surely the end image quality is what matters, not the technicalities of how we get there. Today's glass paired with modern mirrorless bodies produce a level of detail and quality I could only dream of a couple of decades ago. Part of that is surely because the lens manufacturers are no longer so constrained by the optical distortions that can now be predictably and more accurately corrected by software. Back in 2005 I got my hands on a 1Ds Mk II, which was the camera that triggered a lot of pros to go digital. It was a 16.7 mp camera and I remember a senior Canon rep saying that this was the highest megapixel camera they would ever make because the resolution of the sensor was at the limit of the resolving power of the available EF lenses at that time. It produced superb images. Now in 2024 I have picked up the 45 mp R5 mkII with some new RF L zoom lens and would not trade it with anything I was working with 20 years ago. The image quality is utterly superb. Would it be possible to get that level of image quality on those lenses if they still had no option but to correct the lens distortion optically? I don't know the answer, but I do know (a) I wouldn't want to pay for it if they could, and (b) if the end result I get is the same, I really don't care.

[@bojcio]

Yea, the dude's really overthinking it.

[@PointsInFocus]

A big part of the problem, at least as I see it, is that Canon's decision to shift to increased reliance on distortion correction in RF lenses has directly resulted in decreased in camera functionality. Specifically they've changed the in camera multi-exposure mode from using raw (what it does on my R5) to JPEG (every camera since the R5). I use to use the frame averaging function, to shoot cleaner "neutral density" images (up to 3.2 stops) instead of using ND filters. Now if I want to do that, I either have to do it all in post, or I have to take a significant hit to image quality that comes with using JPEGs. What L zooms are you using? The 25-105/4L is optically one of the best zoom lenses Canon has ever made, and it doesn't require distortion correction at all. The RF 15-35/2.8L IS USM, RF 24-70/2.8L IS USM and RF 28-70 f/2L USM are optically corrected and not dependent on software corrections either. And none of these problems are issues with any of the telephoto lenses. So yes, it's possible to get that level of image quality optically, and at a price that you've likely already been paying.

[@StephanBuchin]

DPP does a fantastic job at correcting almost any lens flaw be it L or regular so I don't really care. I get extremely good results out of a 35 years old EF 35-105mm f/3.5-4.5 once corrected.

[@richardtoft6998]

⁠​⁠​⁠​⁠@@PointsInFocus I notice that in-body processing in the R5 II for multiple exposures produce a JPEG at the end (regardless of which lens you take the images with), and so you are suggesting that this is a direct result of SOME lenses requiring too much distortion correction? Hmmm, maybe that is a contributing reason. I understand it is the same story with Nikon - their DSLR bodies could produce in-body RAW multiple exposures, but their mirrorless Z line only produce JPEGs. Same reason perhaps? My experience of blending images in-body is limited to focus stacking, and again you can shoot in RAW but the stacked image is a JPEG and that has always been the case for all brands as far as I know (I usually do focus stacking in post anyway). I presume this is an issue with the processing power required to focus stack huge RAW files and that the camera was only ever working internally with jpeg versions even though it was saving RAWs. Presumably then, they are now following a similar processing path with multiple exposures as well, so it does make sense that this is because of the potential for a lot of digital corrections being needed. I imagine if you had a particular use case and workflow (as you describe above) that it would be annoying to have that particular raw output feature taken away. And ok…fair point, the RF-L zoom lenses I actually use are optically very good anyway. Thanks for your detailed response - I didn’t pick up on your point about potentially losing features in new bodies from the video, but that was probably my aging brain 😂

[@PointsInFocus]

@richardtoft6998, On the R5, which does in body multiple exposure in raw, the availability of the function depends on the lens. For example, if I have an RF 15-35/2.8L IS USM (which doesn't require distortion correction) on the camera I can use it, but if I put an RF 14-35/4L IS USM (which requires distortion correction) then the feature is disabled. This is obviously not a very user friendly situation. On subsequent cameras that have switched to JPEG, it just works regardless of what lens you're using. Way more user friendly. Technically, at least if you want to be picky about it, raw files can't be distortion corrected and still be raw (meaning they contain the sensor's actual pixel data). The only processing that can really be done, while keeping the data raw, is the dark frame subtraction used by long exposure NR. Anything else would require the pixel data to be remapped in some way --- or the raw files to support some kind of demosaicied RGB format (though this also triples the file size). Re. focus stacking. The ability to generate an in camera composite is "new" to me. Off hand, I'm not 100% sure if it's new to the R5-2 or if it was included on an earlier model, but it's not supported on the original R5. That said, I haven't played with it yet, so I don't have any insight into what kind of processing it's doing. This also has a slightly more complex problem compared to the multi-exposure situation, since the algorithm to select what parts to use in the stack is, AFAIK, done using the localized sharpness of the demosaiced image, and not the raw data itself. That said, I'm doubtful that it's compute limited problem. More than likely it a mix of lens corrections and the usability problems noted above and needing to work with demosaiced RGB data to determine what parts of each image should be used. For that matter, I don't think I'd personally have as much of a problem with all of this, if there were better demosaiced RGB formats available in camera. Canon only offers JPEG and HEIF, but HIEF is only used when the camera is set to HDR-PQ mode. The HEIF files are 10-bit, though which is good, but that doesn't help if you can't choose to use them all the time. Alternatively, borrowing a page from Nikon's playbook and having 16-bit TIFFs wouldn't be horrible either --- though TIFF's compression efficiency is horrific and a 45 MP file would be somewhere around 200-250 MB. That said, ideally I think I'd like to see Canon support JPEG XL for non-raw images. Though keeping JPEG for the time being for legacy reasons would be a good idea. JPEG XL supports both lossy and lossless compression, up to 32-bit per channel color, and official support for wide and HDR color spaces (so it works for HDR-PQ), among other things (including lossless trans-coding with JPEG). Compression is also substantially better than JPEG. PetaPixel reports that in their testing on the iPhone 16, a 32 MB JPEG, would be about 24 MB in lossless JPEG XL, and about 5 MB in perceptually lossy mode. Then at least we'd have something that wasn't also limited to a max of 11-stops of DR, and limited tonal resolution.

[@jaysmoov4ever]

Dude, your videos are saving my life. Love this channel.

[@mikede2464]

According to Ken Rockwell, 'L' means "expensive as L"

[@paulhenry7]

Just his little joke.

[@33rdframe]

Hahaha, bro, subtilling calling me out for my horrible name changes... hahaha made me chuckle in pain XD. But I greatly appreciate your shout out, brother. Also I 100% second people to share any criticism they have of my design. There are many, mind you, so don't think you will hurt my feelings. I want to make it better!

[@alansach8437]

Premium weather sealing. Faster autofocus, as in dual USM motors. Better, multiple, optic coatings to reduce glare. But any lens made today is better than even the most premium lens made thirty or forty years ago.

[@Raytrace3Dee]

If Canon charges L prices for an L lens the L lens should stand on it's own in producing quality images on ANY body!. if the L lens needs specific camera functions to perform its duties, it should cost less and not be classified as such.

[@FamilyofTech]

Great discussion! I was thinking the same about the l glass series

[@Twobarpsi]

Right now, "L" stands for LAZY among other things!! It is completely unacceptable to produce LOW quality high end lenses that heavily rely on software to correct their inadequacies. Some of the files of uncorrected RF lens are actually LAUGHABLE! Canon has not sold me on their RF line as of now, and probably never will. Great video and topic!

[@ColinRobertson_LLAP]

Great thoughts. As an architecture photographer, I have the same concerns about distortion correction and I have to wonder where the more distorted lenses "pull from" to be able to outperform the less distorted versions (eg, the 24-105 f/2.8 vs. the 24-70 f/2.8). I also wonder about other aspects that make an 'L' lens-you touched on weather sealing (which is not always the case-see the TS-E lenses), but I'd suspect a lot of the color fidelity and flare characteristics come from the various lens coatings they use. I would push back on the notion that 45MP is "enough". While more is not always better, I have used the 100mp GFX cameras enough to understand the freedom to crop... Possibly more than wanting for more resolution, why are full-frame cameras always limited to 14-bit color? This seems like an easy option to grant to more demanding studio photographers or in situations where you want the most color fidelity possible. Do you have another video discussing diffraction limiting on high-resolution sensors? I have heard you mention this before but I don't quite understand this. Wouldn't the resolution depend on the lens more than the aperture?

[@PointsInFocus]

*Re: the more resolution argument...* Honestly, the argument about how much resolution is enough resolution is always problematic; from both sides. And it’s one that I’ve never been completely happy with it. There’s always an argument between “i need more to crop” and “stop cropping, your pictures will be better if they’re composed properly in camera.” Technically the latter is always true. You use more of the pixels (which yields the best possible image quality) and have less potential for atmospheric distortion to impact the image negatively. However, it’s also not always the most practical way to do things, and practicality is often the best argument for doing things one way or another. In the end that whole argument is very complex, and has intangibles aspects that can’t be cleanly wrapped up (what you shoot, the atmospheric conditions, stuff like that). As an architectural photographer, you may never be in a situation where cropping doesn’t work well. My experience in wildlife photography has been that cropping (and using a really long lens) almost always is worse than getting closer. As for adding more bits… That gets into engineering that that I’m not sure I’m qualified to talk about. Simplifying a lot, there are limits on what is actually necessary. Human vision can only resolve around 1-10 Million colors. A 14-bit file can store 4,398,046,511,104 colors. Plus there are limiting factors due to noise that make even the current situation somewhat over-specified. Sure, more wouldn’t be a deal breaker, but without better sensors capable of 2-4 times better dynamic range, I’m not sure it would be all that useful to just add more bits to current cameras. (I've got a couple of videos in the works that get into some of the problems here.) *Re. diffraction.* Unfortunately no. I use to have a link to a very good video on the topic from an optics expert and lens and telescope maker. You might want to check out: kzbin.info/www/bejne/iYSVcpWQi8meaKM, or kzbin.info/www/bejne/nYDTk4t4pdWIe7c. The short of it, is that since light exhibits wave like properties, when it passes through a small aperture, it "wraps" around the edge (like a wave around a break water) and which results in the wavefront spreading out and a sharp point becoming a larger softer point. The smaller the aperture, the more strongly this happens.

[@ZadakLeader]

You raise some very good points indeed. For me, the heavier/bigger the lens, the fewer issues optically. The fewer optical issues, the less processing needed, therefore a better quality overall.

[@thomastuorto9929]

Except for my longest lens, I purchased the best there was at the time & feel they produce a higher quality photo. I'm way behind in the new glass class thow. I agree about the less one had to play with the pixels, yields higher image quality.

[@ZadakLeader]

@@thomastuorto9929 What i mean is, an equivalent focal length lens, the more glass the better. But you also need to take into account when it was designed. The CAD simulators being used now are probably way way better than the ones used 20 years ago for making the EF ones.

[@Camedia74]

I'd love to see a video on whether you think these slight differences in sharpness, contrast...etc actually are distinguishable after full editing via software (not just the checkmark on lens corrections). Because more and more I am questioning whether L lenses even make a difference in terms of image quality - for example I'm finding it quite hard to distinguish the cheap 16mm RF vs the 14-35 RF. Especially with Topaz, even corner sharpness differences seem to get eliminated. Contrast is much easier to edit in just Lightroom. Thanks!

[@paulmerrick3733]

I like the fact that good results can be achieved with even quite modest equipment. Big issues arise after the camera stage, related to how the image is going to be used. I am keen on prints because I hope that images will be kept. The quality of the print is mostly down to the quality of the printer and I contend that the result is that that issues is ultimately more impactful than what happens in the camera. Of course the result depends on every stage of the process. Ultimately it must be the quality of the result is the only important issue. Anyone aside nerds will not ask what autofocus mechanism you camera uses for example.

[@mikede2464]

There is certainly a benefit to lens optimization vs. software-based corrections when it comes to fall-off. Software based fall-off lifts the shadows in-camera before exporting to raw/jpg thereby reducing the dynamic range and leaving less room for post processing (for the shadows lifted by the camera). I see people on youtube obsessed over iso noise and dynamic range, but totally fine with software based fall-off corrections. Never understood that seemingly obvious contradiction.

[@ultraveridical]

While Canon might be the most egregious currently, due to the prices they charge for their distorted VCM L prime lenses, they aren't the "innovators" in this area. In the m4/3 realm designing lenses with distortions and CA specifically for digital corrections has been the norm for longer than a decade. Canon is just catching up. 35mm GM 1.4 on the other hand has perfectly usable distortion without corrections, and a higher optical quality in general for less money. Too bad Sony refuses to compete with Canon in hybrid camera market on some crucial video features such as RAW video, oversampled 4k, or higher resolutions and frame rates with tolerable rolling shutter and at least 422 10 bit.

[@PointsInFocus]

Color me unsurprised. I think it's the inherent temptation of mirrorless technology. You can compromise on correcting distortion optically, fix it in the camera's software, and most people are non the wiser. Sure, reviewers who go out of their way to convert the raw files without correction will see and measure it. But even there, almost every reviewer has changed their tunes to significant distortion isn't a big deal and you shouldn't care/worry about it.

[@mipmipmipmipmip-v5x]

These days, L means "this Canon lenses' circle might actually cover the sensor area, instead of relying on in-body correction to turn the vignette into mush slightly resembling the corner of the scene like on our slightly less expensive models"

[@PointsInFocus]

Check out the RF 10-20/4L and 24-105/2.8Ls sometime. They do this as bad or worse than the entry level RF 24-240/4-6.3 does.

[@mipmipmipmipmip-v5x]

@PointsInFocus oh I had forgotten! Yeah Christopher Frosts tests were pretty damning! It's a lot of money for lenses that should have been sold as having a smaller focal range, for that 24-105 you can imagine the boardroom meeting between marketing and engineering ..

[@paulhenry7]

Like you, my instinct is to say that distortion and other aberrations should be fixed optically, especially in top-end lenses. I see ‘L’ as indicating both higher image quality AND better build quality. But then there’s the danger of being a Luddite in resisting change. Dylan was slated for moving from acoustic to electric guitar. Some purists would say we should only use film. But I don’t think many people will be shooting on 5”x4” or 10”x8” large-format cameras in the years to come. I think a danger is that the more that software can do, the more there will be a temptation to cut corners in lens design and construction. What also occurs to me is that this trend might be a way to prevent a lens made by one manufacturer being used on another manufacturer’s camera, i.e. enforced brand loyalty. Or am I missing something? What I do know is that I will be sticking with my EF 11-24mm f4 L, EF 16-35mm f4 L & EF 35mm f1.4 L II lenses. They do what I want and have the advantage of being able to put filters between the lens and my R5 (Mk.I). As for DoF, you could argue that it does not strictly exist. What it really means is how much is acceptably sharp in front of and behind a point that is focused on. And that of course is influenced by many factors.

[@PointsInFocus]

There's so much to unpack there.... Re. large format film/film in general... I think we might see more of that in the coming years, and film itself for that matter. With improvements in AI, and the general devaluing of photographs at large, having an actual artifact may become a much bigger value point. Re. brand lock in... First, if manufacturers wanted, they could follow Hasselblad (though I don't know if they still do this) and completely encrypt their lens communication protocol. There's already so much in modern lenses that's problematic in for platform conversions. Like you need to reverse engineer the lens communication protocols. The good ole days where you can stick a pre-G NIkon F mount lens on anything and control focus and aperture are gone. Everything now is focus by wire, which means without both power and a camera that knows how to communicate with the lens you can't even focus it. From an image correction perspective, as long as we can manipulate the images in post, we can fix any problems that can be fixed in software. LibRAW can decode any raw format without any adjustments, and so long as we can photograph a checkerboard target with the lens, we can build our own profiles for correcting the aberrations. The ability to do that is already out there in open source software like Dark Table, as well as most computer vision libraries like OpenCV. While the math is a bit beyond me these days, it's not that difficult of a task to implement from scratch if you had to. Really the same goes for decoding raw formats too (lots of OSS libraries have done it independently and without OEM support). Re. DoF. I've wanted to do a video on it, just because it is an illusion that so many photographers seem to take as if it was an actual concrete thing.

[@tara6664]

I have been debating is the EF35 1.4 II better then the the RF 35 because of looking at the price and the size of the front element and that depth of field is the reason that the outer edges are not as sharp . If a photo was taken that the distances on the outer edge of the EF35 were the same as the centre of the shot would they then be sharper ? If that is so then would at wide open not be the dreamy effect that would be desired ?

[@FlatWaterFilms]

Cheaper using software than perfecting optics?

[@PointsInFocus]

Yea, the lenses should be cheaper if aberrations are fixed in software instead of in the optical design itself. A better optical design means more design time (including compute time), more expensive materials, more complex manufacturing process and so on. All those costs add up pretty quickly, especially when you're talking about a very low volume part. Keep in mind, most lens models sell between 50,000 and 150,000 units across their entire production life. A really popular lens, like a 24-70/2.8 more like 1 million copies over a 8-10 year run. Keep in mind, that optical engineering has some real fun trade offs. Improving one aberration can create or exacerbate another, fixing that can cause a 3rd, and fixing that can reverse the corrections made in the first place. And of course, this is all predicated on the design software that's avaiable and it's strenghts and weakness. For example Canon's lens design software has always been really good at mid and telephoto designs, and not so great on wides. While Nikon's software has historically done better with wide angle designs and struggled more than Canon's has on the telephotos. In fact, in writing that, it's possible that Canon's distortion heavy wide angle designs is the most effective way they can deliver a better product with in the confines of the lens design software they use.

[@thelaststandhdr]

More stress towards video recording is the culprit bringing more and more distortion. I think they should keep in the mind the need of photographers.

[@PointsInFocus]

From the engineering side of things, I'm not sure what would actually drive that. My suspicion has always been it's related to wanting smaller lighter lenses, and the perception that that's what everyone really wants. I mean look at the RF-S lenses. Canon has compromised max apertures basically across the board (and zoom range in some cases too), to make a "smaller" lens than an EF-S lens was. E.g. An EF-S 18-55s use to be f/3.5-5.6, then it became a f/4-5.6, now the RF lens is an 18-45 f/4.5-6.3. The EF-S 55-250 was an f/4-5.6 lens, the RF-S replacement is a 55-210 f/5-7.1. And this is all with the 32 MP R7 being diffraction limited at f/5.2 and the 24MP APS-C R's being diffraction limited at f/6.0. In doing that, all the RF-S lenses have a silly step from the mount's large diameter to the lens's normal smaller diameter. Canon could have delivered faster lenses across the range without having to step them up/down in front of the mount, they'd be functionally as compact and likely only marginally heavier. But instead they choose to make them slower and smaller than the mount itself.

[@thedirtygot9570]

I can’t stand the state of photography, nothing but computer generated fantasy! Probably why all the iconic photos, were taken decades ago!

[@PointsInFocus]

I wonder how much of that sentiment is actually an indictment of modern society/culture and the times we live in as much as anything else?

[@thedirtygot9570]

@@PointsInFocus I actually took a photography class in high school, learned how to develop film in a dark room, learned composition! Sadly photography classes now are how to use Lightroom and photoshop, and the merits of fropack 4😂😂😂

[@TheChiggerBug]

This opinion I think comes from a resentment for technology. Just because there are crutches for modern photographers and because developing our images is quicker doesn’t mean that it’s easier. It just means the bar has been raised. We don’t have an excuse for our images to be event slightly out of focus, or for the exposure to be wrong. Instead, even if you nail both of those, your held to a higher standard on composition. Rightfully, since we have an easier time than past photographers and because we have essentially unlimited storage. Photography isn’t easier now, it’s better than ever before and the standards are higher too. Just saying.

[@doghot5374]

I don't care how much the distortion is , however I can't stand the black corners in these lens. I don't think Canon will design any wide lens that covers full sensor.

[@EddySawaya8637]

Distortion will become apparent once you shoot against straight lines and they look warped. This level of distortion is not tolerable for a not so wide lens like 35mm.

[@robertatwell7615]

I think that you are caught in going down a rabbit hole that is the result of changes in lens design that is being driven by both the larger rear element size of the RF mount vs the EF mount and the fact that some of the lenses that you are talking about are the Z lenses which are not only L lenses but also pseudo parfocal lenses that have extended zoom range and the ability to maintain focus while zooming. Canon and other top manufacturers have created hybrid stills/video cameras and lenses now that they have ditched mirror based bodies for mirrorless bodies. Mirrorless bodies employ electronic as opposed to optical view finders. The new viewfinders display what the sensor has captured as interpreted by the camera maker’s firmware and hardware choices. Whereas the older optical viewfinders just displayed what came through the lens and was reflected into the view finder via its mirror. Optical viewfinders don’t facilitate the additional modifications available to electronic view finders. Thus, lens designs can be chosen that make use of post processing of captured images or not. In short today’s camera and lense manufacturers have additional choices available to them that were not available in the past. Traditionally the greatest image distortion occurs at the highest and lowest focal lengths in a lens zoom range. Thus if you extend the zoom range from 24-70mm to 24-105mm at f2.8 you may well see the need to add software based image correction even on the best optics available. Also remember that true parfocal zoom lenses are not $3k lenses but $30-50k lenses. Today’s cameras are truly hybrid cameras that are designed to capture both video and stills images. Those two technologies are not the same and yet they are both desired by today’s buyers. Canon in particular has released lenses that are optimized for both types of shooting. Some are better optimized for stills and some for video. In general if used with all of the available post processing capabilities you can get wonderful stills or videos from all the new lenses. The idea of L is better in many ways the Z additional designation adds power zoom , pseudo par focal capability, and more video centric features. I think that we are seeing the merger of stills and video capture. Time will tell

[@kelb89]

'Lesser than GM'.

[@thomastuorto9929]

It is what it is! Why even think about it? Just do what you got to do to get the best image possible. When taking/making a photo , if you need that focal length, you need it. If you want to use a super wide zoom with that big bulb is front lens element, you need to understand when you would want to use it how to work with the photos that come with using it & any other lens. I had to stop less than halfway through this vid. The print is what I’m thinking about when going out to press the shutter. Nothing else.

[@PointsInFocus]

For starters, because Canon's choice in lens design has directly resulted in the removal of in camera functionality that I used. Specifically I'm talking about being able to do frame averaging in raw in camera, using the multi-exposure mode. I've used this for years as a better quality version of neutral density (frame averaging reduces noise and improves DR compared to a traditional ND filter). Now, I either have to shoot a pile of raws and stack them in post, or take a significant image quality hit doing in JPEG in camera.

[@thomastuorto9929]

@@PointsInFocus I shoot Nikon (DSLR) & never have used Multiple Exposures in camera. Just now looked it up. I do some Landscapes but not much of it in my area suburban area & don't travel much. Mostly Bird/Wildlife & Nature photos with an occasional fall foliage or sunset shoot. I just looked it up & remember seeing it in the manual when first purchased the equipment. Example was doing a water falls to smooth water in a vid I watched some time ago. Manual says it will increase the dynamic range. I have done exposure bracketing for post work but usually end up choosing one file to edit. I'll have to give it a try. I have been considering the Canon R5 or RII a RF long lens for wildlife along with the Sony line up. I'll keep my Nikon F-mount glass & update from the D810 (first DSLR) to the D850 or Z7II with adapter just not to take the $$ hit on my lenses considering my upgrade budget. Then a Z8/9 or the switch brand thing just for wildlife thinking the AF might be more accurate. Sorry for getting off your dilemma. I'd probably keep a Canon DSLR ( ? 5S or 5SR ? ) the 5omp around just for that if I were in the Canon shooting camp. If always shot Canon, Your DSLR cameras didn;t do what you needed needing the upgrade or the EVF benefits forced the move? Either way, thanks for the explanation.

[@PointsInFocus]

I've been banging the frame averaging drum for years now. It's one of those cases where the digital replacement is better in almost all respects; and faster sensor read speeds fix the one real problem (inter exposure time). Less noise, more DR, no need to carry more filters. Plus with the right implementation that you should be able to specify a time (e.g. I want to shoot for 2 minutes) instead of a filter factor, without having to figure out what filter you need for the exposure settings you're starting with. It's a win across the board in my book. In fact, as I understand it, PhaseOne (not in my budget, ever...) has implemented it in camera specifically for that purpose; not as a subset of a more general multi-exposure feature.