I, as a photographer, find my P30 Pro good enough for family, friends and vacation shots and all other casual snapshots.
I use my DSLRs only when I can't afford to lose a shot, when I need top IQ and when I have special lighting needs.
To take a memory of your kid playing, of grandma in the garden, of your cat chasing the dog or some cityscape, the mobile phone is good enough for many people.
Lightfield or similar trickery with maybe 20 regular cameras together with software assemble a larger sensor, could work.
Most smartphones lenses are f1.8 to f2.5 which is plenty fast. The images aren't sharp because the sensor is too damn small, it's more about sensor size than pixel density. That's why a 4x5" sheet of film still blow any digital camera away in term of details, sharpness, tonal range, &c. even when using 80 years old glass in front of it. You realistically can't get more than maybe 12 mega pixels of data from a phone sensor, you can't cheat optic laws by adding pixels.
This is the new 108mpx sensor used on phone, zoom anywhere in the pic and it looks more like a blurry abstract painting than a picture:
https://cdn.dxomark.com/wp-content/uploads/medias/post-41332...
https://cdn.dxomark.com/wp-content/uploads/medias/post-41332...
Compare that to a 6x6 camera (40 years old camera/lens) :
https://i.imgur.com/eEWx7ln.png
100% crop on the left tower: https://i.imgur.com/qy157qy.png
Please demonstrate how it "blows away" any digital camera with a 80 year old lens. This sounds to me like a completely made up claim without any data behind it. (data != one random image)
In what way? Smartphone sensors have the highest pixel density and they can still resolve an amazing amount of detail through their not-so-great tiny lenses. When it comes to ILCs, everytime there is a bump in sensor resolution, it is entirely possible to extract more detail from the same lens (provided its was a good lens to begin with). And ILCs are still not at the level of pixel density as smartphones, so there is quite a bit of room for improvement.
> So if you have several times the size of a "full frame" digital camera, you can achieve better results with 1940s technology
Not exactly because larger sensors mean larger lenses because of the larger image circle, and then you have a manufacturing challenge of grinding and polishing an even larger amount of glass, so it doesn't scale linearly like that.
>. See this stunning 1942 photo from a large format camera: https://upload.wikimedia.org/wikipedia/commons/2/21/Adams_Th....
That image is very very nice from a compositional standpoint. I respect the work of AA, but I don't feel its rude to point out that it is not particularly sharp by modern standards - the point I assume you were trying to demonstrate? Here's one that even an total rank amateur like me can produce.. which is sharper, with an el-cheapo bargain-bin samyang lens on a non-pro camera body (6d) https://imgur.com/a/l4HBWyi
The lenses aren't good enough to resolve details on the new "full frame" 50+ megapixels sensors. The smaller the sensor the more the lens has to resolve. a 4x5" sheet of film is ridiculously big compared to a "full frame" sensor, hence the lens aren't as stressed [0]
> Smartphone sensors have the highest pixel density and they can still resolve an amazing amount of detail through their not-so-great tiny lenses.
No they don't, there are many algorithm tricks to make you think it resolves more details but when you look closely there isn't any details, look at these 64mp samples [1] [2] [3] [4], they don't even look like pictures anymore.
> Not exactly because larger sensors mean larger lenses because of the larger image circle ... larger amount of glass
LF lenses are much simpler than regular lenses, they don't need floating elements, they're not zooms, they don't even need moving parts because you focus through the bellow. There is more glass if you look at the weight, but less glass in term of groups/elements. If anything they're smaller than new "full frame" lenses [5], they only get big if you want something faster than f5.6-4
This is what LF gives you: http://www.drumscanning.co.uk/wp-content/uploads/2015/09/vel... or https://youtu.be/sqN7n9bXgtU?t=288
[0] https://www.quinnimages.com/wp-content/uploads/2017/07/sizec...
[1] https://cdn57.androidauthority.net/wp-content/uploads/2019/0...
[2] https://cdn57.androidauthority.net/wp-content/uploads/2019/0...
[3] https://cdn57.androidauthority.net/wp-content/uploads/2019/0...
[4] https://cdn57.androidauthority.net/wp-content/uploads/2019/0...
[5] https://galerie-photo.com/manuels/nikkor-lenses-for-large-fo...
I can accept that as a general statement, but it depends on many factors. For e.g. The 4x5 example you linked to was scanned at 4000DPI. 4000 DPI across a 24x36mm sensor gives approximately 20 mega pixels - which is entry level full frame. So its not stressing the lens any more than your LF capture.
Now its true that, at 60+ mega pixels you're doing 6500+DPI, and 240MP (pixel shift) even more so, you're stressing the optics, but we've found that the optics is quite capable of delivering.
https://www.dpreview.com/reviews/image-comparison/fullscreen...
https://www.albertdros.com/single-post/2019/09/02/Sony-A7RIV...
>No they don't, there are many algorithm tricks to make you think it resolves more details but when you look closely there isn't any details, look at these 64mp samples [1] [2] [3] [4], they don't even look like pictures anymore.
Those are JPEGS on which noise reduction and sharpening was added by the camera. We'll have to compare RAWs to see the pixel level sharpness.
But again, its not my position that smartphones, which don't have the best lenses, can compete with X,Y or Z format. I'm just pointing out that we're not at the limit of optics, as with every bump in MP we see more details, especially so in professional cameras with professional grade lenses.
>LF lenses are much simpler than regular lenses, they don't need floating elements, they're not zooms, they don't even need moving parts because you focus through the bellow. There is more glass if you look at the weight, but less glass in term of groups/elements. If anything they're smaller than new "full frame" lenses [5], they only get big if you want something faster than f5.6-4
>This is what LF gives you: http://www.drumscanning.co.uk/wp-content/uploads/2015/09/vel.... or https://youtu.be/sqN7n9bXgtU?t=288
We'll have to do a side by side with the same image to really compare. I don't think you'll come away with LF "blows away" 35mm digital as the poster I responded to claimed.
If you take two pictures of the same scene, one with a small sensor, and the other with a larger sensor, the geometric difference is apparent.
Scale a 5 meter tree down to the size of a phone's sensor, and you're having to tightly pack the detail of the tree, to something smaller than your fingernail.
But shrink that same tree onto a bigger sensor, or a sheet of 4x5 film, and you don't compress the tree as much.
We don't notice the shortcomings of phone sensors until we try to print the photos. They don't enlarge very well. Printing an 8x10 from a cellphone is challenging, and the grain and imperfections are magnified.
But you can enlarge that piece of film to 8x10 no problem, and the detail is astounding.
If you only look at phone pictures on a phone, then they look very nice, and a piece of 4x5 film is bigger than some phone screens. There are more silver crystals (pixels) in the film than there are pixels in your phone screen, so you can't appreciate the total detail of film on a small phone either.
The thing to consider is choosing the sensor or the film size based on how you want to display the image. Then you pick the best camera for the situation, considering cost, convenience, usability, etc.
I have worked a lot of different analog film cameras and digital ones, and I've noticed that most cameras can take great pictures. You end up paying more for usability and quality. Pro bodies have dedicated buttons and whatnot, so you can accomplish what you need, very quickly.
A touch-screen camera with pure software controls is never going to be for pros, but, cell phones have their moments of better performance for getting the job done.
A cellphone camera is like a piano that plays chords but not individual notes as far as usability. A manual camera lets you do what you want and make up new chords.
Hope that clarifies the image sensor aspect. Aperture helps, but it can never make up for the constraints of the sensor.
Have you verified this personally or seen any data on this? I find this claim very hard to believe. I make several 24x36 prints in a year (albeit from APS-C and FF sensors, but sometimes heavily cropped) and 8x10 is a tiny size that should not present a challenge for any modern cellphone's sensor.
I have been meaning to shoot a roll of ISO 16 film, the CMS20 II by Adox. They claim it can resolve up to 500MP of details.
Would be great to compare to results to full frame with adapted film lens.
Here's their film page with sample pictures: http://www.adox.de/Photo/adox-films-2/cms-20-ii-adotech-ii/
My canon is currently mid-roll with infrared, so it'll be a while before I get around to using it.
P30 Pro has a Sony camera chip, like almost all other "superphones"
Sony was making more money on smartphone chips than on the regular camera sensors.
Now, Eliot Management probably made them to shoot themselves in the foot:
1. The consumer product unit will now not be able to finance the RnD to push the tech forward. Imagine Sony cameras with Chinese CCDs.
2. On the opposite, without strong incentive to do loss making RnD, the IC unit will become just another commodity component maker
3. Asian companies tend to not to survive "financial engineering" style manoeuvres you see in America
[1]https://www.samsung.com/semiconductor/image-sensor/mobile-im...
Is the ISOCELL any good?
It very much seems to be mostly numbers but it really can't back up the claims in actual visual image quality. The noise and required light to get anything near good shots is very high, the DOF on such a large sensor is downright tight, which is nice for making pretty looking shots..but really bad for just someone trying to capture a photo. Even on the S20 Ultra taking a photo of a person from 5 feet away results in the nose and eyes being in completely different focus.
The fact of the matter is that in normal use, a medium format mobile sensor with a moderate pixel count is preferrable. (I.E. iPhone 11 Pro, Pixel 4) as for 95% of shots you want a wide DOF to put all of the content in focus and great per-pixel aperture to reduce exposure time and ISO so you can reduce camera shake and Noise. Even in the shots you want DOF (Portrait), faking it tends to lead to a more adjustable shot where you can correct after the fact instead of being actually stuck in a inch-wide focus gap where any movement of the camera or subject results in out of focus capture.
Plus most photos are posted online in the 1-2MP size. Now cropping is a great extra 5% item, but for those who want really great photos they would have a lot to gain from a much larger sensor with a larger lens and lower MP count to get way less noise and better exposure times (I.e. The Sony A7R IV) Whereas dedicated full-frame equipment just trashes mobile results.
We went through some “megapixel wars” in the past and it turns out, like you said, there are a lot of tradeoffs here and optimizing for high pixel count means you have to make unacceptable sacrifices elsewhere.
It's strongly suspected, but not outright confirmed last I checked, that they supply some canon compact cameras too.
> If anyone tells you that one camera is better than the other, for any particular activity based on image quality, you should laugh at them. In their face, if possible.
https://m.dpreview.com/articles/4000816220/nikon-d850-vs-son...
But the general consensus is to get rid of your old glass if you're switching to mirrorless because the AF performance with adapters is on the "it barely works enough to not call it unusable" level.
As anyone that isn't living under a rock might know, the point and shoot markets are basically annihilated from existence. The professionals/pro-amateurs who care about image quality etc didn't like the Nikon 1 and EOS M.
Out of nowhere, Sony released the E mount Alpha lineup, which reached the #1 position in mirrorless cameras. They also released the Alpha 7, which was AFAIK the first full frame mirrorless (that means the imaging sensor is the same size as 35mm film. Most DSLRs used APS-C sized sensors, which are 1.5 times smaller than 35mm film). This was unprecedented because everyone thought mirrorless was basically a toy.
Both Nikon and Canon have realized their mistakes and recognized mirrorless is serious business and is the future. Nikon has scrapped the Nikon 1 line and started making Nikon Z, which is a super pro/prosumer oriented mirrorless lineup with a full frame sensor. Canon hasn't scrapped the old mirroless line but is pushing the RF mount lineup which is also full frame.
Sony was able to seize the market by being first to market by a long shot, not to mention the vertical integration from their imaging sensor division to deliver quality products. But now that Nikon and Canon entered the game with guns blazing, we'll see what happens to Sony.
One thing of note is that Sony has opened up the E mount spec and allowed 3rd parties to create lenses. As a result there's a fuck ton of 3rd party lenses (most of them of dubious quality, but some are excellent) with autofocus support, and not just from Tamron and Sigma.
It's a really interesting time to be in this hobby/profession IMO.
IDK what they're like nowadays, but Sigma has in the past had a reputation for a larger quality variance than others.
That might not seem like a big deal but it certainly is when it’s hanging off of you for a full 12 hour wedding.
That said I still have a Sigma 35mm and 20mm in my kit.
Now, this is starting to change somehow but many, me included will wait a few more years because the cost of replacing many pro grade lenses is still very high.
I got sick of the kit lens so I got a cheap used 50mm which can shoot tack sharp and I just unboxed a 90mms macro from the G series and will try it out today.
This is what makes their move to mirrorless so tricky. While adapters for the existing lenses exist and work, in the long term you would want to upgrade many of your lenses to their mirrorless counterpart. Especially, if most new lens designs are made for the mirrorless lines.
The M(240) then added the EVF, which also precedes the A7.
All of this is a bit of trivia, as most people are not aware of the Leica M series, so I thought it would add to the discussion. The A7 certainly was the first EVF-only mirrorless 35mm camera and was sold in far higher numbers than any digital M.
I hope you realize that live-view shooting was essentially turning a dslr into a mirrorless one (caveat obvious OVF/EVF differences). Canon's dual pixel AF tech is still miles ahead of Sony's PDAF when it comes to low light focusing, and video AF. Each pixel on a canon sensor can be used for AF as well as imaging. Sony uses masked pixels which limits them because they canno't mask too many pixels for fear of IQ degradation, and having fewer AF pixels limits them in low-light focusing.
>Both Nikon and Canon have realized their mistakes and recognized mirrorless is serious business and is the future. Nikon has scrapped the Nikon 1 line and started making Nikon Z, which is a super pro/prosumer oriented mirrorless lineup with a full frame sensor. Canon hasn't scrapped the old mirroless line but is pushing the RF mount lineup which is also full frame.
Or rather, like mature businesses, they realized that they could keep making money with DSLRs and slowly milk them dry and then ramp up mirrorless and make money there too.
>One thing of note is that Sony has opened up the E mount spec and allowed 3rd parties to create lenses.
That's because Sony had no expertise in making high quality professional grade lenses and teamed up with Zeiss and others to help them get up to speed initially.
Services, yeah, Canon are miles ahead but one step at a time.
I'm a Nikon shooter, mainly wildlife. If I was starting over today I don't know which system I'd buy into.
I think just about everything else about the Fuji cameras is markedly better than Sony's mirrorless offerings. I occasionally wish I had a full-frame sensor, I suppose, but those are very overrated by prosumers.
Sorry, what is this and why is it revolutionary? (Why did Nikon and Canon miss it?)
The biggest things holding back mirrorless technology previously was battery life and autofocus technology. I am not sure why Nikon and Canon didn't invest more into mirrorless technology sooner, and I'm no photography expert, but I think they were just too entrenched in the DSLR technologies they pioneered.
You nailed it on the head for the most part.
Nikon had a mirrorless camera in the J1 series, which had some very nifty features but was just an absurdly expensive system for what you got (especially since it is basically a dead end, proprietary platform.)
Canon did a bit better with their first MILCs, but they weren't amazing. At the time they were introduced Canon suffered from a lack of on-sensor Phase Detection AF, and Contrast detection systems weren't as fast as they are now.
I mean, a big issue is glass; While there are exceptions to this, you can't just re-use an existing lens design on a mirrorless without making the thing longer than it needs to be (Which looks off on a Mirrorless.) Which means that Nikon/Canon can't just release the 18-55 EFAF Mark 10 lens.
Designing proper lenses takes time and money. I mean, the first few years I had an E-mount, the pickings were slim for lenses, while the a6000 was a good start it wasn't till the A7II and A7RII with their ability to do usable PDAF with adapted lenses that the platform felt safe to pick up.
But also think about it from a marketing perspective - even now most people buying cameras aren’t professionals. A big bulky camera right next to a svelte one in a store will look dated. That said Sony went too far with it, in my opinion. A nice grip is important for professionals.
How can glass lenses be "faster?" Allow you to shoot with a faster shutterspeed?
For many years, basically Nikon and Canon split up the "professional" market between each other. They already had this position with analog SLRs, and continued after the transition to digital. The reason is a mix of having very good cameras, a large lens setup, especially considering the more exotic/expensive lenses mostly used by professionals and most importantly, with the service organizations. At each olympics, they would have large service centers on-site, for repair and exchange. And also outside of these events, they have services networks which enable a professional to get the equipment serviced/exchanged with loaners in minimal time, which is a very obvious benefit. Which basically prevented other brands to make a huge inroads in the professional markets.
So their profits were driven by the established systems. They were milking this cash cow and consequently were late to the mirrorless game. Their recent offerings are nice, especially Canon has some very interesting lenses for their new mirrorless system, but even for those companies it takes a whole lot of time to have a remotely complete lens offering.
It might also be worth while to look beyond the 35mm format. In 2013 Olympus brought the OMD E-M1 to the market, which was possibly the first mirrorless camera, not only matching but in a few aspecs exceeding the performance of comparable DSLRs. This shows, how much Canon and Nikon are lagging behind.
I think that’s it. But it seems they are catching up very quickly now. The upcoming EOS R5 looks like a very good camera.
Mirrorless allows the camera to use the image sensor itself to do all the work. That means autofocus can be much better, using every pixel on the sensor for autofocus. It means you can run computer vision software in real time to do fancy things, like finding out if there are eyes in the image - most likely, you want the eyes to be in focus. You can focus in near darkness, when it's too dark for you to see. And other stuff.
The biggest downsides are 1) the electronic viewfinder, these are not as nice for most people, but with the big caveat that they do allow you to frame an image in near total darkness where an SLR wouldn't allow you to do that and 2) the big one is battery life, it is 5-10X worse with mirrorless than with SLR.
I like using the viewfinder though, and I like digital ones over optical ones.
Those were introduced in 1930s and 1950s respectively. Since the launch of Nikon F in 1959 until some time in 2010s the viewfinder technology didn't change. Means shutter became metal and electronic, autofocus invented, man went to the Moon and back, digital CCD/CMOS photography came around, Sun goes down but mirror and prism SLR technology of 1950s just kept on.
One of the reasons why mirrorless didn't appear sooner is because LCDs or OLEDs viewfinder suck(ed). They used to be in SVGA or XGA resolutions, refresh at 30-60Hz or less, few hundreds ms late, dim, power hungry and whatnot. Wasn't so advantageous in terms of weight or cost as well. Optical quality of SLR viewfinder never disappointed its users on the other hand. It is and always had been a joy to be able to just look into it and see through the lens, especially when films take hours to develop.
Sony, however, knew that no one uses viewfinder in a point and shoot digital cameras, or no one has any problems without SLR viewfinder in video cameras, dad on holidays and professional broadcasters alike. So after they merged camera division from Konica Minolta in 2006, they just walked out and spoke out the obvious. Camera nerds were shocked, I was shocked, so were Nikon and Canon. Leica may have been laughing, Zeiss was happy collecting licensing fees. I was skeptical, and I could verify it feels very disappointing looking into EVF. But Sony knew it ain't a problem and it wasn't.
Nikon and Canon probably thought the quality of EVF never satisfies professional photographers, from their narrow expertise.
As far as I know Nikon started off making Contax clones, not Leica clones. Canon did Leica clones, though. Keep in mind that Leicas up until the M3 didn't have a combined viewfinder/rangefinder. That was an innovation from Zeiss first used in the Contax II.
And there's been quite a few SLRs before the Nikon F, it was just Nikon that made it so solid and efficient that made professional photographers switch from using a rangefinder to SLR.
The big advantage of EVFs is that it can do so much more than an optical viewfinder. You can enlarge it, display focus peaking, but most of all, you get a preview of the image you would take - especially previewing the exposure is very helpful. And of course, if you want to do video, the viewfinder of a SLR wouldn't be usable at all.
My reaction is the opposite. I always felt disappointed by how primitive the optical view finder on modern cameras is versus electronic ones. It doesn't get the light levels right and it can barely show you any relevant information; there needs to be a separate 7-segment display on the body showing pertinent information. How primitive!
EVFs, on the contrast, show you everything you need to know right in the viewfinder (including advanced features like histograms, zebra stripes, etc., that an optical viewfinder could never possibly do).
Nikon and Canon have long used Sony sensors.