Search
  • Kurt McManus

IS IT TIME TO DITCH MY DSLR?

Updated: Apr 19

I've been lugging a heavy camera bag around with me for many years now and I'm starting to realise that I might not need to (most of the time). Whenever I head out with photography in mind I always take my camera bag which usually contains the gear below:


  • Canon 6D DSLR

  • Canon 17-40mm f/4 L lens

  • Canon 50mm f/1.4 USM lens

  • Tamron 70-200 f/2.8 DI VC USD lens

  • Samyang 12 mm f/2.8 ED AS NCS Fish-eye lens

  • Filters and holders

  • 3x Batteries, head torch, cable release, backup SD card


and the list goes on....








This gear is also what I'd carry on my back whenever I went travelling and you only start to realise how much of a burden it is when you walk around London for 6 hours and fail to capture a single image, meanwhile you start to realise your knees have grown a voice and they're starting to swear in your general direction.


Don't get me wrong, I love my photography.... And I love my 6D and the lenses I've chosen for it. But with the advances in mobile phone cameras over the past few years, I've found myself increasingly reaching for my Samsung mobile instead of my DSLR. And with the advancement of the software that these phones are using to stitch multiple images together for 'night mode' shots and the introduction of large megapixel sensors, I've started to think that there's a real possibility I may not be using my expensive camera gear in a few years time... So I really wanted to do a comparison of my phone and my DSLR, and just see how close these phone cameras are actually getting.


When I came up with the idea, I was using a Samsung S10, which I knew was probably an unfair fight, but with my fiancee picking up the new Samsung S20 with a 64MP sensor, the comparison seemed oh so more attractive...


Samsung S20 and multi-camera array (yeah I know it's pink....)


So I was interested in a number of different comparisons...


  • My DSLR vs Samsung's latest mobile (standard 12MP and 64MP)

  • Samsung S20 vs last year's S10

  • All of the above compared to the GoPro Hero 8 (for interest sake)


This is the camera specs for the Samsung phones (source: Samsung)



Device Comparisons


It's obviously difficult to compare devices like this because they are completely different kettles of fish with sensor size, megapixel count etc and that isn't even considering the glass that's packed in front of them. But I'm interested in whether despite all of this, can the new sensors packed into mobile phones coupled with the improvements in image processing negate some of these glaringly obvious differences? And is the quality 'good enough' for what I'm after as a landscape photographer?


So let's start by looking at a quick comparison of the image sizes these devices produce in megapixels:


  • Canon 6D = 20MP

  • Samsung S10 = 12MP (wide and telephoto), 16MP (ultra-wide)

  • Samsung S20 = 12MP (ultrawide, wide and telephoto), 64MP (wide and telephoto)


Therefore, when I compare the S20, it'll be with the standard 12MP sensor and the 64MP sensor.


Note: This is achieved by selecting the aspect ratio in the camera app and selecting the 64MP 3:4 option as below:


Source: Phonearena

The images produced by these devices will all contain different levels of detail, and you can see a comparison of the different image sizes below:


Comparison of the image sizes between the five devices


So I'm interested in seeing how these levels of detail compare side by side when we get in a lot closer. But I also want to see how they differ at the different focal lengths that are on offer in phones these days with multi-lens arrays. The focal lengths that I'll be comparing between the devices are therefore as below:


Samsung S10 and S20


  • Ultra wide = 13mm (35mm equivalent)

  • Wide angle = 26mm (35mm equivalent)

  • Telephoto = 3x and 10x


Comparative range with Canon 6D and various lenses


  • Ultra wide @ 17mm

  • Wide angle @ 26mm

  • Telephoto @ 82mm and 200mm (to match the 3x and 10x on phone)


GoPro Hero 8


  • 'Linear' = ~ 24mm

  • 'Wide' = ~17mm

  • No Telephoto compared

OK that's enough of the technical stuff. Let's get into the image comparisons!

Now just a heads up that my day job may be in forensic science, so I have a little experience with controls and comparisons etc, but I'm by no means a 'pro' when it comes to comparing image quality between different sized sensors etc, so I apologise if it's not perfect, but I'll do my best.



Test 1 - Image Quality in Daylight


The first test is just a straight out image quality test in daylight utilising the different cameras in the Samsung phones. I needed something with a lot of detail and a cityscape seemed the best approach. So we got up early one morning and headed over to a spot I knew had a great view of Auckland city.


For standardisation I used a tripod and phone mount (not used for GoPro) and then snapped an image at the various camera zoom ranges and then matched these as close as I could with the Canon (GoPro was snapped handheld with 'wide' or 'linear' mode).


We'll start off at the 'ultra-wide' end, which is a zoom range I love for landscape photography (I typically shoot my Canon @ 17mm).


Samsung S20 (Ultra-wide @13mm)


All images were edited as JPG in a similar fashion using Adobe Camera Raw and then cropped to a similar point in the landscape. Just note that I captured my Canon shot using RAW (as this is what I always shoot with) and I didn't use RAW on the phones. This is because I find the RAW photos on the Samsung phones is pretty poor and in-fact the software processing of the standard 'photo' mode tends to bring out better detail etc than this 'pro' mode.


To ensure the differences in detail were more obvious, the images were all sized to match the Canon 6D image. This meant enlarging the GoPro and Samsung (@12MP or 16MP) images and reducing the S20 64MP image.



'Ultra-wide' test



The difference in detail is not completely obvious at this crop (especially if you haven't zoomed in), but the Canon certainly appears to be more detailed while I think the others seem quite similar.


But what if this was to be viewed on a larger scale (for large scale printing etc)? Well if we crop in further then we can start to see the differences become more apparent.



From this we can definitely see the improvement in the Canon over the other devices. But what surprises me is the GoPro appears to be more detailed (or at least finer details) than the two Samsung cameras when using this 'ultra-wide' camera. The S10 may be a smidge better than the S20 but that's easily explained by the reduction in image size from 16MP (S10) to 12MP on the S20 with this mode.



'Wide' test


The next comparison is the 'wide' angle on the phone which is known as '1.0x' on the Samsung camera app. This camera would allow the use of the 64MP sensor, which is where my most interest lay in this test. Could this bump in pixels be enough to take out the Canon?




Again at this point you can't really see a great deal of difference, even if you zoom in on your browser. So we're going to have to get in a bit closer....



At this point, we're definitely seeing the GoPro with the least detail and the Samsung 12MP images are starting to fall behind the Canon. This can however be expected, as we're talking 12MP vs 20MP (not considering anything else about the DSLR and lenses over a phone). But the Samsung 64MP image appears to not only be on par with the Canon, but perhaps a lick ahead...But we might just crop in even further and see how much it's nudged ahead.


Well this confirms it... the Samsung S20 64MP image has more detail then the Canon and it's impressively clean! This slim mobile phone with it's tiny sensor (in comparison) is out-performing my $3500 DSLR and lens combo... I will come back to this discovery later in this piece.


But I know what you're all thinking, 'why don't you just zoom in if you want more data'. So then lets take a look at that then....



'Telephoto' test (3.0X)




Note that the Samsung S10 has an optical zoom of 2.0x, while the S20 has an optical zoom of 3.0x, so we're adding a small amount of digital zoom to the S10 in this test. But... I'm mostly interested in the S20 at this point anyway. From the wider crop of the cityscape above, it's hard to see a difference but they all look pretty nice. The S20 has a smidge more warmth over the S10 which I like. So let's crop in and see the changes in detail:



Now these results really surprised me... the S20 64MP looks more mushy, but let's just crop in one step further to confirm it.


Yup, the S20 64MP is actually lower in detail for some reason. Sure we've got a bigger image to print if we want, but it has less detail than the 12MP image when all things are equal (resized). So I'm not sure what to make of this. Perhaps this is a software issue? Perhaps some blur of some degree at this zoom and sensor combo? I know the Samsung S20 Ultra and the regular Samsung's have issues with their large MP cameras focusing. So perhaps this may be fixed in the future?


I will keep an eye on it and add to this blog if anything comes to light....



'Telephoto' test (10.0X)


Now this test will undoubtedly be the most unfair comparison, but I just wanted to see how the zooms on the S20 stood up to my $1500 telephoto lens on my DSR.


The comparison below is my Canon at it's maximum of 200mm, the Samsung S20 12MP @ 10x zoom and the Samsung S20 64MP @ 6x zoom (and cropped to match the 10x).


The telephoto Samsung images are surprisingly good with all things considered and again, as with the 3.0x test, the S20 12MP image is actually sharper than the 64MP one, so I'll only crop in further and compare the two below:




And as expected, this is where the phones really fall down. BUT not a lot of people are zooming in to scenes at 10x with their phone and expecting to print them big enough to see this kind of detail loss. So for the most part, the zoom on the mobile phones is pretty impressive and appears to be improving over the years.


Test 1 - Conclusion


This test certainly produced a few surprising results, and a few expected ones.


The GoPro (added to this test for interest) is certainly the worst performer in this test, but I don't think it'll be disappointed in itself over this result. The GoPro is primarily a video device and you can take it into places that you'd never dare to with these other devices. And for those reasons alone, it's a great camera. But on a side of that, it can hold it's own with 'ultra-wide' images compared to the Samsung phones and you can even go wider than the others if you really want to. Nice...


GoPro Hero 8 'Superwide' mode vs Samsung S20 'Ultrawide'


The Samsung S20 doesn't provide too much improvement over the S10 when using the 12MP sensor for both the 'ultra-wide' and 'wide' cameras, and in fact loses some room to the S10 on ultra-wide due to the reduction from 16MP to 12MP. But it does provide improved resolution with the 3x optical zoom compared to the 2x on the S10.


But it's when you flick the S20 onto the 64MP sensor where things get really interesting... This is specifically for the standard 'wide' camera where the improvement in resolution far surpasses the standard 12MP images, but also beats out my Canon 6D DSLR and a 17-40mm L lens. Meaning, if I'm walking about in decent light and shooting scenes without too much dynamic range, the S20 could offer a better solution for capturing images at this zoom range (~26mm). However, these mobile phones certainly won't be able to replace your telephoto images if you're looking for as much detail as possible. And I never expected they would (yet.....)


But of course, this has only been tested above in decent morning light, without too much dynamic range... so I can't speak for the comparison between these devices in scenes with low light or higher dynamic range. We're all well aware that low light is where smaller sensors fall down (despite their high pixel counts), but from what I've seen with the various 'night' modes on these latest mobile phones, the image stacking and software corrections are creating some really impressive images at night without the need for a tripod. So I think this is something that will catch up to larger sensors pretty quickly...


I'm also well aware I've compared all of these images with a distant cityscapes and all in the approximate middle of the frame, so I haven't considered how they perform in the corners or with other elements closer to the camera. So I'm not done with this comparison yet... I'll be adding a few more images to compare them in different situations and with different elements. But for now I'm a little stuck inside with this COVID-19 situation. So watch this space...



So is it time to ditch my DSLR?


From what I've seen, I'm incredibly impressed by what Samsung have managed to do with it's mobile phone cameras in these types of conditions, but I certainly haven't completed a full test of the phone's ability, especially in the more difficult lighting conditions.


So I don't think I'll be ditching my camera gear quite yet, but potentially in a few years I think I may need to ask myself this question again... With improvements in the phone sensors and the software that's processing the images, I think mobile phones are closing the gap quicker than I previously thought.


But what this has shown me so far, is that if you're heading out for some photos shortly after sunrise and shortly before sunset, and you don't want to lug your heavy DSLR around, then you will be perfectly fine grabbing your new mobile phone (like the Samsung S20) and your tripod and you'll be saving your knees or back a trip to pain-town. And you'll probably appreciate the ability to edit the images from your phone instead of needing a laptop and a bunch of other programs to get to the final result. I think convenience and 'good enough' will be fine for most people at some point.


So few photographers print their work large nowadays and even if they did, 12MP is more than fine in most circumstances. And as a recent example of this, below is an image I took of my fiancee doing yoga in Bali on my DSLR:



The issue here was, I didn't take my laptop with me to Bali, so I didn't get to edit this photo til around 2 weeks later. However, I also took an (almost) identical photo on my Samsung S10 and edited it on my phone straight afterwards:



Besides it being a little cooler (was obviously my editing mood at the time), there's not a great deal of difference and I was equally happy with both results. Sure when I zoom in to 100%, there's less detail, and maybe some less colour in the sky from reduced dynamic range. But I'd happily get this image printed large for our wall at home and a more significant difference between the two was the convenience of it all. The mobile image was shot and edited within 5 minutes while the DSLR shot came about 2 weeks later (or at least a good 30 minutes later and a whole more fuss if I'd taken my laptop to Bali). So is the difference in detail actually really that important?!


Ultimately I think the software they're packing into these phones nowadays is getting smart enough to counteract what was previously only possible with large sensors and quality optics. Of course a wedding or sports photographer will laugh at these sentences, but I'm not suggesting phones for that type of work. That's mad..... But I'm a landscape photographer and my subjects are a lot more forgiving, and I've got my knees to think about.... :)


Anyway, I hope you guys get something out of this comparison. And stay tuned for an update to this piece when I can get back out and about.


K

0 views

 Kurt McManus Photography 2019