[MadBill]

It seems to me the biggest photography advantages to mirrorless are that what one sees in the viewfinder is a closer approximation of the image that will be captured, and the sophistication and accuracy of autofocus. Although Ricoh has added eye detect autofocus to the K-3 iii, I doubt they will be able to duplicate the autofocus capabilities of mirrorless cameras like those from Sony and now Canon and Nikon (except perhaps in live view.) Improvement to size, reliability and cost are fairly minor except, perhaps, at the low end. Going mirrorless also is an excuse to dump 50 year old lens mount designs for something based on 21st century technology and lens design.

Personally, I find EVFs a little disconnected from the subject I am photographing, but then perhaps I have just not used a really good EVF. I like Pentax cameras. They more than meet my needs and I see no need to change systems.

[aldo taddia]

I state that in addition to the K1 I also use a Canon R and an Olympus OMD. Probably the market was now saturated with mirror cameras with the corresponding lenses, so the houses began to offer mirrorless to stimulate sales again. The mission was to create smaller machines with fewer moving parts, but the latest cameras are as big as DSLRs. If only DSLRs had a fast rear monitor focus system like mirrorless, they could still assert themselves. But as long as there are targets around for mirror machines there will be hope that DLSRs will still live.

[brainwave]

I found few reasons mirrorless have advantages over DSLR:
1. Lower flange distance means lighter and more compact lenses on the wide angle end.

2. Autofocus systems are more accurate due to them being based on what the sensor sees, and not an independent autofocus system sees. Additionally, content aware autofocus systems such as eye-tracking, human-tracking, pet-tracking, etc. are possible whereas they are limited on DSLRs. This is what makes the autofocus feel faster or more accurate on mirrorless.

3. Video focus "hunting" is eliminated, by using these advanced autofocus systems also for video. In DSLRs, when using live-view for video, the mirror is locked up, meaning the independant autofocus systems cannot be used and the camera is limited to contrast-detect which require hunting.

It is easily possible for them to incorporate the advanced sensor-based autofocus systems into DSLRs, by using the exact same technology as mirrorless when mirror is up. When mirror is down, they could go back to the traditional autofocus systems. However, that means the autofocus would be expensive and probably push the prices up by a chunk.

4. Lastly, with mirrorless systems, exposure etc. are possible to view immediately before exposing.

Yes, mirrorless is innovative no doubt. But for an amateur photography enthusiast like me, who takes only photos and doesn't like making videos, the mirrorless has no distinct advantages. Rather, I would miss the ka-ching of the mirror movements.

One thing I am excited about in future is computational photography on these dedicated cameras. Google, Apple, these companies have manage to create great, natural looking photographs using HDR+. Night mode, for example, outperforms a full-frame hand-held and its not even close. If I could get away with hand-held shots on -3 EV using computational photography, I would love that. Those developments would not be possible on a D-SLR and only possible on a mirrorless system.

[jersey]

I think this articke is, sadly (or not?) right. The issue here is not EVF but fact that mirrorless camera is able to compute image parameters based on the big sensor, while DSLR, if used in "native" way so with mirror and prism, needs to rely on secondary sensors. It was a plus for DSLR for some time as the tech was matured. But at some point on one side you have data from secondary sensors, that get this data from light that was reflected in some way and on the other data from big ass sensor that is getting it directly from lens.

Lack of mirror allows for computational photography advances that DSLR will never be able to achieve unless in liveview and then what is even the reason for DSLR. Also smartphones are mirrorless and this allows for sharing of software and some tech.

DSLR is now what horse was in the end of XIX century, being replaced with cars. Now almost no one is using horses for daily commuting or work. DSLR will be the same in few years, after most of proffesionals using them change the gear or leave the work.

[luftfluss]

As might be expected in a speculative article composed of inputs from several authors, it's a rambling opinion piece based on little research.

- Craig

What kind of research would be necessary? If manufacturers are (mostly) no longer producing DSLRs, the "Era of the DSLR" is over.

[UncleVanya]

Is the era of DSLR cameras coming to an end?

^ Don't mean to stir the pot, but it is an interesting read.


At the way things are moving in the camera industry, Pentax will more than likely end up being the only purveyor of DSLR cameras in the future.

The Death of DSLRs is Near | PetaPixel 2018…

We can probably find older articles with similar titles.

[BigMackCam]

The Death of DSLRs is Near | PetaPixel 2018…

We can probably find older articles with similar titles.

Bingo... and even if "The Death of DSLRs" really is near - which would require Ricoh to stop making Pentax cameras (and I don't see that happening any time soon) - it would only be in terms of development and manufacturing of new models. There's millions of DSLR cameras and even more compatible lenses in circulation, and people will continue to shoot with them for decades to come. Some folks still choose to shoot with SLR, TLR and rangefinder film cameras, and plenty will choose to continue shooting DSLRs whether-or-not new models are released. There's enough gear available to keep us all busy for many, many years...

As Mark Twain reputedly quipped, "The reports of my death have been greatly exaggerated"

---

Last edited by BigMackCam; 09-29-2021 at 12:13 AM.

[Rondec]

Of course camera brands are excited to move to mirrorless because it means a new mount. They are hoping to sell cameras, of course, but they are also hoping in the long run to sell all of the lenses that people already own in the F mount and EOS mount back to them again in a new mount. Yes, you can use an adapter and use your old lenses, but people will gradually move over and purchase the lenses that they use the most in a new mount.

Overall, other than better video performance and faster frame rates, the differences between SLR and MILCs is less than these sorts of articles make it. The cameras are not that much smaller, assuming you are shooting with a standard zoom in place, and the photos from them don't look that much different either (they use the same sensors). There is the downside of all those auto focus points on the sensor that they are actually much more prone to banding in backlit and high iso situations.

Regardless, I think Pentax will actually be in good shape selling SLRs in a market that doesn't cater to those who prefer optical viewfinders.

[bob5050]

What is the key difference between mirrorless and D-SLR that makes them the "future", so to speak?

Primarily computational photography (CP), and it's not really here yet. By 'computational photography' I mean the ability of the always-exposed sensor to capture a light information stream that extends both before and after you actually press the shutter, and use all that information to calculate or build up what the scene must have been, through frame stacking and other means.
The primary issues with this are that (1) there's no 'real' camera available today that does is as well as a top-notch cell phone, (2) it's still fairly primitive: 'artificially enhanced' reality isn't good enough yet, and (3) there's no camera available today that's 'future-proof' on this capability: current cameras lack the power to fulfill the promises of CP, even if the necessary software were available.
So future or not, there's no current reason to switch.
traveler

---------- Post added 09-29-21 at 04:04 AM ----------

... what one sees in the viewfinder is a closer approximation of the image that will be captured...

I've always had an issue with the WYSIWYG claim.
Sensors capture data--the closest we get to that is a raw file. But a raw file isn't directly viewable--it has to be interpreted to be displayed on the electronic viewfinder, and that interpretation is susceptible to all the limitations of in-body image conversion. Rather than 'what-you-get', and EVF shows you a fairly 'neutral', speed-optimized rendering of the raw data that might or might not correspond to what the photographer would actually do with the image in post.
So an EVF shows you neither the world actually out there nor the final image, but some intermediate thing that, if the photographer shoots raw and 'processes to taste', might never be seen again.
traveler

[brainwave]

Primarily computational photography (CP), and it's not really here yet. By 'computational photography' I mean the ability of the always-exposed sensor to capture a light information stream that extends both before and after you actually press the shutter, and use all that information to calculate or build up what the scene must have been, through frame stacking and other means.
The primary issues with this are that (1) there's no 'real' camera available today that does is as well as a top-notch cell phone, (2) it's still fairly primitive: 'artificially enhanced' reality isn't good enough yet, and (3) there's no camera available today that's 'future-proof' on this capability: current cameras lack the power to fulfill the promises of CP, even if the necessary software were available.
So future or not, there's no current reason to switch.
traveler

---------- Post added 09-29-21 at 04:04 AM ----------


I've always had an issue with the WYSIWYG claim.
Sensors capture data--the closest we get to that is a raw file. But a raw file isn't directly viewable--it has to be interpreted to be displayed on the electronic viewfinder, and that interpretation is susceptible to all the limitations of in-body image conversion. Rather than 'what-you-get', and EVF shows you a fairly 'neutral', speed-optimized rendering of the raw data that might or might not correspond to what the photographer would actually do with the image in post.
So an EVF shows you neither the world actually out there nor the final image, but some intermediate thing that, if the photographer shoots raw and 'processes to taste', might never be seen again.
traveler

Agreed. I feel that the camera manufacturers dropped the ball yet again by ignoring innovations in the cellphone space and ignoring computational photography as an option in mirrorless. Imagine what could be done with the size of APS-C sensors and the great optics of dedicated lenses. For instance, apple iPhone today in night mode (or pixel for that matter) does way better than ANY dslr or mirrorless system hand-held in low lights. And its not even close, regardless of how much you push ISO up. SImilarly, digital high-res zoom by Pixels could empower shorter / smaller lenses to achieve even sharper images. The image segmentation and selective masking in apple iphones makes the camera content aware, and enables it to expose correctly for each different object such as foilage (where leaf sharpness is prioritized) or faces (where eye-focus is prioritized) or the sky (where underexposure is done intentionally to bring out beautiful colors). Whats more, it even saves this data into the apple proRaw format, so that future image editing softwares can virtually eliminate careful time consuming masks over different subjects when necessary. The raw file already has embedded depth information, and one could "select" trees, humans, faces, sky, etc. and edit accordingly. Even the localized gradient HDR that basically blends the HDR across small chunks of the image to create a sense of "naturality" is way better than 2-5 image HDR on cameras. It just feels inertial to me at this point, where camera manufacturers continue to do what they knew once was the best practice. This is very typical with japanese companies. At least Pentax, continuing the DSLR legacy, feels "classic" done right.


Regarding the other point of WYSIWYG - I agree with you on that count as well. Recently, the camera sensor saw this image :


However, what I could get after post was this :

So how can WYSIWYG a good thing?

[Rondec]

Primarily computational photography (CP), and it's not really here yet. By 'computational photography' I mean the ability of the always-exposed sensor to capture a light information stream that extends both before and after you actually press the shutter, and use all that information to calculate or build up what the scene must have been, through frame stacking and other means.
The primary issues with this are that (1) there's no 'real' camera available today that does is as well as a top-notch cell phone, (2) it's still fairly primitive: 'artificially enhanced' reality isn't good enough yet, and (3) there's no camera available today that's 'future-proof' on this capability: current cameras lack the power to fulfill the promises of CP, even if the necessary software were available.
So future or not, there's no current reason to switch.
traveler

---------- Post added 09-29-21 at 04:04 AM ----------


I've always had an issue with the WYSIWYG claim.
Sensors capture data--the closest we get to that is a raw file. But a raw file isn't directly viewable--it has to be interpreted to be displayed on the electronic viewfinder, and that interpretation is susceptible to all the limitations of in-body image conversion. Rather than 'what-you-get', and EVF shows you a fairly 'neutral', speed-optimized rendering of the raw data that might or might not correspond to what the photographer would actually do with the image in post.
So an EVF shows you neither the world actually out there nor the final image, but some intermediate thing that, if the photographer shoots raw and 'processes to taste', might never be seen again.
traveler

I think a lot of folks are straight out of camera jpeg shooters. For them, an EVF does have what you see is what you get. On the other hand, Forumites tend to be RAW shooters who push things in different directions and may shoot a scene underexposed and push the shadows in post to avoid clipping. For them, an EVF is certainly not going to tell them what the end result image is.

[UncleVanya]

I think a lot of folks are straight out of camera jpeg shooters. For them, an EVF does have what you see is what you get. On the other hand, Forumites tend to be RAW shooters who push things in different directions and may shoot a scene underexposed and push the shadows in post to avoid clipping. For them, an EVF is certainly not going to tell them what the end result image is.

I struggle sometimes with EVF or live view being in that mode - I can’t focus or compose if it’s too dark! I like seeing an approximation- most of the time - but being able to disable it quickly is important.

[c.a.m]

If manufacturers are (mostly) no longer producing DSLRs, the "Era of the DSLR" is over.

One such subject for research would be production and shipment data.

For example, CIPA* reports the following data as of July 2021.

For the period January-July 2021, production of interchangeable lens cameras totaled nearly 3,200,000 units (3.2 M). SLR cameras accounted for 1.36 M (42.5%) and mirrorless 1.84 M (57.5%).

* From CIPA's website: "The Camera & Imaging Products Association (CIPA) is an international industry association consisting of members engaged in the development, production or sale of imaging related devices including digital cameras."


- Craig

[aldo taddia]

One such subject for research would be production and shipment data.

For example, CIPA* reports the following data as of July 2021.

For the period January-July 2021, production of interchangeable lens cameras totaled nearly 3,200,000 units (3.2 M). SLR cameras accounted for 1.36 M (42.5%) and mirrorless 1.84 M (57.5%).

* From CIPA's website: "The Camera & Imaging Products Association (CIPA) is an international industry association consisting of members engaged in the development, production or sale of imaging related devices including digital cameras."


- Craig

This data means that the SLRs are not finished. Surely the mirrorless will increase the percentage share but, for many, photography is still based on the use of the Reflex.

[luftfluss]

One such subject for research would be production and shipment data.

For example, CIPA* reports the following data as of July 2021.

For the period January-July 2021, production of interchangeable lens cameras totaled nearly 3,200,000 units (3.2 M). SLR cameras accounted for 1.36 M (42.5%) and mirrorless 1.84 M (57.5%).

* From CIPA's website: "The Camera & Imaging Products Association (CIPA) is an international industry association consisting of members engaged in the development, production or sale of imaging related devices including digital cameras."


- Craig

These statistics seem to support the premise of the article at Imaging Resource... but I think most of us already intuited the approximate percentages, no?