18
\$\begingroup\$

The other day I went hiking through a mountain in west Tokyo to take wilderness pictures for reference for the new videogame I'm working on.

For the past few years, I've only relied on my smartphone for taking pictures, but I thought this would be a good moment for undusting my old Canon PowerShot S5 IS, which I think was released in 2007, but since it has a bigger sensor, it should take better (technically, not artistically) pictures than a smartphone, even on auto, right?

So I took most of the pictures in the trip with the Canon, and just a few to share on social media with my Samsung Galaxy S8+ smartphone, which is a bit more than a year old.

Most locations were relatively low-light, but since the smartphone was taking decent pictures, I went and took all the Canon pictures on auto. And when I got home, to my surprise, pretty much every single picture with the Canon was blurry.

For example, these are two pictures of the same waterfall, taken less than a minute apart, with no noticeable changes in environment lighting between both:

Samsung Galaxy S8+ (Not bad, and actually usable):

Samsung Galaxy S8+ picture

Canon PowerShot S5 IS (pretty much unusable):

Canon PowerShot S5 IS picture

| Camera        | Samsung SCV35 (S8+)     | Canon PowerShot S5 IS   |
|---------------+-------------------------+-------------------------|
| Exposure time | 0.030 s (1/33)          | 0.077 s (1/13)          |
| Aperture      | f/1.7                   | f/2.7                   |
| Focal length  | 4.2mm (35mm equiv 26mm) | 6.0mm (35mm equiv 38mm) |
| Metering Mode | center weight           | pattern                 |
| Resolution    | 4032 x 3024             | 1600 x 1200             |
|---------------+-------------------------+-------------------------|
| Flash used    | No                      | No                      |
| ISO equiv.    | 200                     | 200                     |
| Whitebalance  | Auto                    | Auto                    |
| Date/Time     | 2018/06/09 15:37:44     | 1980/01/01 00:00:38     |
|---------------+-------------------------+-------------------------|
| Exposure      | program (auto)          |                         |
| Orientation   | rotate 90               |                         |
| CCD width     |                         | 5.72mm                  |
| Focus dist.   |                         | 65.53m                  |

I know that I have more control with the Canon, and I could have used manual settings, and a tripod to take much better pictures, but with a larger sensor, I was expecting to get better pictures from the Canon on auto than the Samsung.

So why is the smartphone picture better than the dedicated camera one? Has technology evolved so much that a 1 year old smartphone takes better pictures than a 10 year old high level point-and-shoot? Would getting a more recent camera help me take better point-and-shoot pictures on auto?

Edit: I added the EXIF metadata, and it is obvious that the Samsung used a much faster shutter speed which led to the blurring. However, those were the settings both cameras chose for Auto. I had supposed that a larger sensor in the camera meant the camera can choose more aggressive settings for Auto and eventually getting a better picture on Auto.

\$\endgroup\$
4
  • 1
    \$\begingroup\$ I would have a look at the Canons settings. It appears that auto favors light instead of speed. You may want to turn to the "I want to decide the shutter speed, auto for rest including ISO" and experiment what happens when you go to 1/50 and 1/100 second speed. You might be surprised. \$\endgroup\$ Commented Jun 23, 2018 at 9:51
  • \$\begingroup\$ @ThorbjørnRavnAndersen: And it seems like the main reason why auto worked better on the Samsung is that, time and ISO being equal, the Samsung can have a larger aperture than the Canon, which definitely helps in low light situations. It would have been nice if the Canon would have compensated for the low light and aperture with a larger ISO instead of longer time, but it clearly didn't. \$\endgroup\$ Commented Jun 23, 2018 at 10:57
  • \$\begingroup\$ @ThorbjørnRavnAndersen: Interesting that you mention brand, but I don't doubt that the ability of the Samsung to go to f/1.8 gives it more flexibility when choosing settings for full Auto. How would you rank brand, age (sensor quality) and the f number range in importance when taking full auto pictures? \$\endgroup\$ Commented Jun 23, 2018 at 11:02
  • \$\begingroup\$ @PandaPajama I do not know, and I do not expect my google-fu to be better than yours. The exact answer to your question can only be answered by Canon (or someone who reverse engineered the firmware of the camera). Everyone else is just guessing. Next time learn to use what you have next time before going on such a trip, and consider if you should get a newer camera if your current one is not good enough for you. \$\endgroup\$ Commented Jun 23, 2018 at 22:13

7 Answers 7

30
\$\begingroup\$

No, your camera sensor is not bigger than your smartphone sensor, they are both about 1/2,5".

The difference is that your Canon camera uses ancient sensor technology (more noise), smaller aperture (less light) and no sophisticated image processing, so no wonder it will produce inferior images in nearly all conditions, except when zooming in, thanks to its 12x optical zoom, not available in your smartphone.

Typical fate of compact cameras from 10+ years ago...

\$\endgroup\$
10
  • 1
    \$\begingroup\$ The Canon EXIF metadata says the sensor width is 5.72mm, while the S8+ specs say it's 1/3.6", which if my calculations are correct, that is about 4mm wide. Both are small, but the Canon has more than twice the surface area. Are advances in CCD technology so large that a new camera with half the sensor surface area takes noticeably better pictures than an older camera with twice the sensor surface area? \$\endgroup\$ Commented Jun 23, 2018 at 9:55
  • 6
    \$\begingroup\$ 1/3.6" is the front selfie cam and the main camera of S8+ is 1/2.55" - samsung.com/global/galaxy/galaxy-s8/specs \$\endgroup\$
    – szulat
    Commented Jun 23, 2018 at 10:04
  • 1
    \$\begingroup\$ Same size sensor, but the new one packs 12.9 megapixels & the old one a mere 2MP. Sensors have come a long way in 10 years. Even 'budget' DSLRs these days have 24MP on crop frame, high end go to 50 on full frame. \$\endgroup\$
    – Tetsujin
    Commented Jun 23, 2018 at 10:25
  • 1
    \$\begingroup\$ Canon PowerShot S5 IS has 8mp sensor according to Canon website. \$\endgroup\$
    – xiota
    Commented Jun 23, 2018 at 10:28
  • 1
    \$\begingroup\$ The EXIF data says it's 1600 x 1200 - so it's either been trimmed, or was taken at some lower-res setting. I don't know that camera at all, so don't know if it has that option. \$\endgroup\$
    – Tetsujin
    Commented Jun 23, 2018 at 10:32
23
\$\begingroup\$

The primary reasons for the differences in image quality between the two cameras are the settings and technique used to capture the image, not sensor technology.

To get sharper images, increase ISO, increase shutter speed, and use shorter focal length. Choose appropriate aperture and exposure compensation. Use shutter half-press, and hold steady while releasing shutter.

Sensor Technology

It is impossible to say what role advances in sensor technologies might have played in the quality of the images because neither image takes full advantage of the sensor technology available.

Differences in settings and technique used to capture the images accounts for the vast majority of differences between them. Except for resolution, either image could conceivably have been taken with the other camera had similar settings been used. 

Common Reasons for Blurry Photos

  • ISO and Aperture. Modern cameras, including smart phones, adjust ISO and aperture more responsively than older cameras. It's also possible the camera was set to a fixed ISO or aperture that was unsuitable for the lighting conditions.

  • Shutter Speed. If ISO was set too low or aperture closed too much, shutter speed would have to decrease to compensate. Slower shutter speed allows more time for movement to affect the image.

  • Exposure Compensation. Exposure compensation could have been set accidentally, which could result in slower shutter speed. Large areas of the waterfall and some of the leaves have blown out highlights.

  • Focal Length. The Canon S5 IS has a 36-432mm equiv focal length. If you were shooting at the long end, camera movement would have been magnified.

  • Image Stabilization. Since it seems the camera might not have been used in a while, image stabilization may have been turned off. Also, older IS technologies were not as effective as modern ones. A dozen years ago, I would have been lucky to get an additional stop with IS. More recently, I have gotten up to 5 stops. The phone does not have IS, but this is where the camera could have helped, but might not have.

  • Depth of Field. Phone cameras tend to have extremely wide depth of field, so minor changes in focus can be more difficult to see in the final image. Dedicated cameras potentially have shallower depth of field, so incorrect focus may be more obvious.

Comments based on Exif

The Canon camera was set to ISO 200. Recommend increasing ISO or setting to Auto. If already on Auto, change to fixed ISO at highest setting that with acceptable noise. Although this is the same ISO as the phone, the phone aperture was about 1.3 stops faster (F1.7 vs F2.7). To compensate, camera ISO would need to be 600+. However, in older cameras, ISO usually jumped from 400 to 800. You would need to decide whether ISO 800 has an acceptable noise profile.

Your phone shutter speed was 1/33s vs 1/13s on Canon. Since this is about 1.3 stops difference, this is where the camera chose to compensate for the smaller aperture. The camera may also be over-estimating its ability to stabilize the image. Personally, I would rather max out ISO than drop to speeds below 1/40s, even with image stabilization. Many modern cameras have settings to indicate preferred max ISO and slowest shutter speed.

The ISO, aperture, and shutter speed on both the phone and camera were set to give equivalent exposures, so exposure compensation did not play a role in this case. However, Canon cameras I've used in the past had a tendency to overexpose, so -1/3 EV might be beneficial.

The Canon camera was not set at its maximum resolution (8 mp).

The hazy appearance in the phone image is caused by light reflecting or scattering between the lens and sensor (glare/flare). Shading the lens with your hand or a hat would have further improved the image.

\$\endgroup\$
4
  • \$\begingroup\$ I have added the Exif metadata in the original post. However, isn't the entire point of Auto for the camera to choose "good enough" settings for me? It seems that the Samsung's choices for Auto were better than those of the Canon. Is this because it's an old camera? or because it's a dedicated camera? \$\endgroup\$ Commented Jun 23, 2018 at 9:15
  • 2
    \$\begingroup\$ Looking at the specs, the Canon only goes to f/2.7, while the Samsung took that picture at f/1.8. That is a large deficiency of the Canon compared to the Samsung... \$\endgroup\$ Commented Jun 23, 2018 at 10:37
  • 1
    \$\begingroup\$ @PandaPajama: Auto just has some heuristics that try to work for all situations. The heuristics might not be perfect, or they might not match up well with the right choices for that shot. It would be interesting to know whether manual settings could have gotten a much better picture from the Canon. (Remember the phone has a much more powerful CPU to run software to decide on settings, so I wouldn't be surprised if it's Auto made better decisions more of the time. And the benefit of more years of software development, probably.) \$\endgroup\$ Commented Jun 23, 2018 at 17:59
  • \$\begingroup\$ @PandaPajama you know about Use shutter half-press, and hold steady while releasing shutter part, right? \$\endgroup\$ Commented Jun 24, 2018 at 17:34
11
\$\begingroup\$

Yes, the technology of your Samsung smartphone is the reason it gave you a better photo than your ancient Canon S5is. The S5is photo is unusable because of motion blur. You got motion blur because of the slow shutter speed which the camera chose because when the light is so low, that was the only parameter it could adjust.

The other options were already maxed out.

The lens aperture was already as open as possible. This is the negative side of the hi-zoom cameras: to maintain an affordable cost and weight, they tend to capture less light than a fixed lens. The Samsung's aperture was 1.7, the Canon 2.8. Quite a difference.

The only parameter left is ISO. And that's where the technology really makes a difference. When dpreview tested the S5is, they found excessive noise above ISO 100. And Canon knows the limits of their sensors too and that's why auto mode did not go above ISO 200.

I own and used the predecessor, the S3is and also found its low light performance really lacking. But the good news is, if you buy a current camera, it will likely perform much better than your S5is. A Canon PowerShot G7 X Mark II, for example, does offer a 1.8 aperture as well as much higher usable ISO.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Thanks. This pretty much sums up the conclusions we've reached with everybody else \$\endgroup\$ Commented Jun 24, 2018 at 0:00
  • \$\begingroup\$ Rubbish. The Powershot image is bad because of bad technique. Inferior technology had nothing to do with it. \$\endgroup\$
    – user4792
    Commented Jul 9, 2019 at 14:46
3
\$\begingroup\$

Everything xiota said, and a few more thoughts.

It's the word "control" that's important here. A camera will allow you to take better pictures than a smartphone, but it will also allow you to make mistakes. Smartphones are set up so anyone can take pictures - the camera is more reliant on decisions made by the photographer.

It also depends how you're composing the shot. I'm old school myself, and while I've used the screen to compose pictures, I'm much happier using a viewfinder - I find holding my arms close to my head gives a much more stable stance, which is particularly noticeable with slower shutter speeds.

It looks like that's what's going on with this picture - the "milky" appearance of the waterfall suggests a slower shutter speed on the Canon than the Samsung. Depth of field is less (suggesting a wider aperture), but to me much of the blur looks motion related (camera shake) rather than focus related.

\$\endgroup\$
5
  • \$\begingroup\$ You say that it allows me to make mistakes, but isn't the entire point of Auto to let the camera choose some "good enough" settings for me? Are you saying that for point-and-shoot, I should completely forget about a dedicated camera and stick to smartphones? \$\endgroup\$ Commented Jun 23, 2018 at 9:12
  • 3
    \$\begingroup\$ @PandaPajama - If you want something that will do everything for you, you might get better pictures on a smartphone, but it sounds like you care enough about photography to have bought a camera and to be asking questions on this site - I think you're close to finding the limitations of a smartphone. Interestingly, full Auto is the only camera setting I don't use - if you have a play with aperture priority, shutter priority and Program, you'll soon see the differences aperture and shutter speed make, and will be able to use them creatively. There will be mistakes, but that's half the fun. \$\endgroup\$ Commented Jun 23, 2018 at 9:29
  • \$\begingroup\$ in this particular case, my purpose was to take reference pictures, not artistically pleasing pictures, and since travel was involved, mistakes were definitely not fun in this context. I chose to leave less to chance, so I chose the Canon expecting for it to give me better pictures in Auto, especially knowing I'd be going to low-light locations. I was clearly wrong, and I want to know if my mistake was on choosing an old camera, or a dedicated camera. Alternatively if I should stick to smartphones or get a newer camera when shooting for this particular purpose. \$\endgroup\$ Commented Jun 23, 2018 at 10:14
  • 1
    \$\begingroup\$ I think the 'mistake' if anything was to choose an old camera. A new one would still leave room for error, same as anything, but they do tend to be 'smarter' in full auto now [not that I ever use full auto even for point & shoot occasions. I always like to choose aperture & ISO to force a given shutter speed 'bracket'.] You can do similar with the old Canon, as has been pointed out in other answers, but tbh the phone will still probably do a better job at guessing. \$\endgroup\$
    – Tetsujin
    Commented Jun 23, 2018 at 10:30
  • \$\begingroup\$ @Tetsujin: I've learned a bit more about cameras in the past few hours. It seems like the mistake was choosing this particular camera for no-zoom low light photography. This is a high zoom camera, and as such has a small sensor to fit it in a point-and-shoot package. I should probably get a camera with a larger sensor, sacrificing on zoom, for purposes like this one. I still wonder how the Samsung takes such good pictures with a small sensor though... \$\endgroup\$ Commented Jun 23, 2018 at 13:50
1
\$\begingroup\$

The Canon is shooting with an aperture a stop or so smaller than you used for the smartphone, and I believe that you held your phone far steadier than you held the camera, which would lead to more blur regardless of the different exposure times.

The rotten shot from the Canon has nothing to do with inferior technology and everything to do with inadequate technique. If you can't keep a camera steady at t 1/13 sec then put it on a tripod. Alternatively open up the lens or raise the ISO to get a faster shutter speed.

Having said that you must have been REALLY shaky when using the Powershot if its IS was unable to cope.

Why is it that when photos don't turn out well people blame their cameras?

Put your Powershot in my hands and I'll get you an image every bit as good as the one you got with your smartphone, at least at the sort of resolution you have displayed here. Of course the smartphone will make bigger prints or display better on a high-res monitor, as it has more pixels.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ "Why is it that when photos don't turn out well people blame their cameras?" — likely because camera manufacturers sell their products as magical devices that turn the money you spend into better photographs. \$\endgroup\$
    – mattdm
    Commented Jul 9, 2019 at 15:24
0
\$\begingroup\$

The smartphone used a faster shutter speed. You need a faster shutter speed or use a tripod. The green foliage looks like camera movement. The falling water looks like the effect you aim for with a ND filter.

\$\endgroup\$
-3
\$\begingroup\$

2007 Linux vs. 2017 Mac

As others have said: technology marches on. Optics and sensors have progressed. But the phone also has a little advantage that your old camera does not have: it is a computer, with an application helping it along.

For example: where you previously had to take two/three/more pictures manually, and then import them into an image editing software in order to get an HDR picture, the phone now does that automatically for you.

You say you took the picture with a Samsung Galaxy. Personally, I was enamoured by the Galaxy's camera and associated app, especially with its ability to perform in low-light and extreme back-lit conditions.

enter image description here

enter image description here

enter image description here

Unfiltered, full auto, HDR on, taken by a Galaxy S6

Now granted, cameras give you extremely fine-grained control, if you want that. And if you know what you are doing, and put in the effort, with that you can do some amazing things.

In essence, what you are describing, where you have an old camera vs a new phone, is kind of like a 2007 Linux vs a 2017 Apple computer.

A 2007 Linux computer — especially if it was purchased at the equivalent price of a Mac — is by no means a bad computer today. In the right hands, and with the right effort, the Linux computer can still perform miracles.

Nevertheless, a modern Mac will do lots of grunt-work for you, and will do so much more streamlined than a Linux computer. Backed by hardware that is more than ten times faster and with its modern operating system it will outperform the older computer in many aspects, especially when doing common, everyday tasks.

And such is the situation with modern phones vs. old cameras. The phone may not have the full optic and photographic capability that an old camera has, and with a skilled and dedicated photographer/editor the camera is can still achieve great things. But for simple, everyday photo-tasks, the modern phone has reached such a level that it competes with — and often outperforms — what cameras did 10 years ago, because the phone can use the on-board CPU — and sometimes even cloud-based automation — to do mundane tasks that make the image better.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ It seems that some people have taken offence at this post. You would do well to explain your down-votes instead of just using the votes to express petty dislike. \$\endgroup\$
    – MichaelK
    Commented Jul 12, 2019 at 8:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.