Why don't you use GoPro or Garmin?

So I've had a number of people ask me why I don't use GoPro or Garmin action cameras, and stick to my Sony.

GoPro may have the biggest name in the market, but it seems to be mainly through marketing, as opposed to any actual feature or price differentiation.  

Garmin does have additional advanced GPS and Gyroscope features related to cycling, etc., but like Sony, hasn't released a new model in a few years.  

Of course there is no "best", just best for your requirements - so let's cover mine:

1. I want to take videos with a camera mounted on my handlebar.  These videos are primarily for safety / evidence purposes, so I don't want to bother with gimbals, helmet cams, chest harnesses, or selfie sticks most of the time.  I want to stick it on my bike and ride, and it should work.  

2. The battery should last at least an hour, and preferably it can be externally powered.  The battery must be removable it the camera can't be externally powered.  

3. It should be rain proof because, it rains outside sometimes.

4. It should work reasonably well in the evening.  A bit of image noise is okay, turing into a blurry mess is not.  

5. It should have image stabilization that works well with the vibrations from a bike without jolting the video, and this should work even at night.  

6. It should have enough resolution to make out license places, etc.  (4K at a decent data rate / shutter speed should be plenty).  

7. GPS Data recording is a nice to have.  

8. It shouldn't require special proprietary software (which the vendor will then inevitably stop supporting in a few years).  

That's it.  Easy, right?

More specifically, It doesn't need any of the following:

1. A fancy onboard color LCD screen.

2. A "Selfie" screen (Why?!)

3. 360 video

4. Any "Filters", WiFi, mobile phone apps, etc.  

5. Web Cam features

6. 8k or anything like that 

7. Live streaming or anything like that

I want to take out the MicroSD card when I get home, copy it to my computer, and then edit or archive it.  WiFi cloud uploads are okay, I guess.  Having USB3 on the camera with a USB-C cable so I can transfer files to the computer at high speed without removing the MicroSD card would be fantastic.  

Now it's story time, let's talk about the different cameras I have owned and tried.

1. I had a Sony point-and-shoot in round 2012.  I tried mounting this on the bike.  It looked like blair witch project.

2. I bought a Sony ActionCam, which was black, and had only electronic image stabilization.  (I think it was model AS50).  

It was okay, but the night video was terrible.  I sold it to Janpara.  This camera is waterproof without a case, but needs a dive case for any deep water immersion.  This camera was released after USB C existed, but for some reason they chose to use Micro USB.  

3. I bought the Sony ActionCam FDR-X3000.   (Which I still have, and use for most of my videos).  

Same waterproof situation as #2 above.  

Again with the MicroUSB.  I will say quickly that the Sony software is absolutely aweful, and hasn't been updated in a while, and the live streaming feature seems to be discontinued now.  The camera has a basic black and white LCD built-in, and if you really want a color screen to watch previews, then Sony sells a separate screem that you cat attach to your wrist, handlebar, etc.  I didn't bother with this.  More about this camera later.  

4. I bought the Garmin Virb Elite.  

This camera is huge and heavy, but it has some fancy GPS and Gyro features, and works very reliably.  The battery lasts up to 3 hours, which is amazing, and it can connect to ANT+ and bluetooth sensors to record cycling data.  It is very sturdy and waterproof without any case.  A bit too clunky for my bike, and the video is only 1080p@30FPS.  Interestingly, this camera also had a front-lit display on the top.  I really liked the cycling specific features, like the ability to show the exact speed, G-force, RPM, etc., on the video, but the night performance and resolution were not good compared with the Sony X3000.  Mini USB.  Electronic stabilization only, and poor night performance.  

5.  Next I got the Garmin Virb Ultra 30.  

This camera is in the size and form factor of a typical GoPro, and has specs similar to the Sony - that is, it could do 1080p at 60fps or 4k at 30fps.  This camera was smaller than the Virb Elite, while still retaining the awesome G-Metrix features.  This camera has a touch screen on the back, similar to recent GoPro models.  It has no mounting hole or fingers, so you must put it in the dive case (or some other harness) in order to mount it to anything.  Again with the Mini USB.  

This camera can be powered by USB, but there is no way to connect the USB while it is in the dive case, and no way to mount it when it's not in the dive case.  

There are three solutions to this connundrum:

a. Buy the special dive case with the power cord. (Powered mount bundle).  

b. Buy the non-waterproof frame.

c. Drill a hole in the dive case so you can feed the USB cable in.   

 Option A costs $100 (!), and I couldn't find the powered mount anywhere in Japan anyway.  Option B I also couldn't find anywhere, so option C it was.  That means I can't use it for diving anymore - which isn't a big deal to me until I want to sell it, and the next person wonders why there is a hole in the formerly waterproof case.  

Conclusion: Fancy GPS features, but only electronic stabilization, and poor night performance.  

6. Finally, I wanted to try a 3D camera, and it just so happened that Garmen came out with the Virb 360.  

This camera was fun to play with, but it only worked well in very bright sunlight, and I quickly found out that 5.7k is not nearly as high resolution as it sounds when you are taking a 360 degree video.  For example, if you record the video in 360 mode, and then later want to create a 2D video with a relatively normal view port, then you "reframe" the 360 video into something like 1080p if you are very lucky.  So while yes, you can do things that you wouldn't be able to do on a "normal" 2D camera, using a 3D camers for 2D video means you will end up with low res footage, but need a lot of storage and processing power to edit it.  The Virb 360 also has the awesome G-Metrix features and fantastic stabilization, but again very poor low light performance.  and... again with the Micro-USB.  Micro USB implies USB 2, which means you don't want to use the USB cable to transfer your absolutely huge 360 videos.  The camera could do stitching and stabilization internally for 4k videos, but if you want to use the 5.7k resolution, then you need to let your computer handle those features.  The software was very good, but had gotten buggy with recent operating system releases, and it seems Garmin hasn't updated it.  I enjored playing with this camera, but the technology just isn't there yet.  I checked with other people about 360 cameras from GoPro, Insta360, and others.  None of them are reasonable to use to capture high res 2D video or use on cloudy days.  

The Virb360 also had the limitation that the stabilization would only work if you captured 360 video, it didn't work  at all with 180 video.  This is a shame, because otherwise I might have used it more for capturing 2D video.  

At the end of the day, the low light performance on the virb 360 was the worse of all the cameras so far.  I enjoyed using the 360 to play around with, and take on a few trips hiking and skiing, but since that meant it would only be used a few times per year, I sold it.  

So what am I left with?

Well, Basically I still use my Sony X3000!  It has the best image quality, the best stabilization, and the best night quality.  Granted the night quality is poor, but it's still better than the others.  

This makes sense for two reasons:

1. With electronic image stabilization, you are essentially just taking a big video and then cropping out part of the video.  When the camera shakes, you can compensate for that by moving the section you are cropping, so that it looks like the camera hasn't moved.  This software trickery is indeed fancy, but it can't fix everything.  Immagine at night you are riding down the street and there is a bright street light, which leaves a streak on the video.  It leaves this streak because it actually leaves a streak on the sensor from maxxing it out "Blow out".  Moving the crop area is not going to get rid of the streak, just make it look strange.

The X3000 has optical stabilzation, which means it can simply prevent the image that hits the sensor from shaking to begin with.  If the image reaching the sensor is stable, then there is no streak caused by the light.  This alone means that no electronic stabilization system can compete with an optical stabilization system.  

2. In order to cope with low light situations, cameras usually compensate by increasing the shutter time.  The longer the shutter is open, the more light can get in, and the less noisy the image will be.  The problem is that if the camera is moving during this time, then the image will be blurry.  Blurriness will be more evident on high contrast sections of the image (f.e. lights and signs), again adjusting the cropping won't solve this problem.  

Optical stabilization can't completely solve this issue.  A faster exposure will mean more noise (grain), and a slower one will mean more blur.  The only real way to solve this is to have a larger sensor, or some more sensitive sensor technology.  Still, it seems that electronic stabilization makes the issue seem worse, whereas optical stabilization doesn't.  I have my theories as to why this might be the case.  

Still, what about the latest cameras, surely they are better?  GoPro was maybe on the Hero.. 7? when the Sony X3000 came out, so I figured it might be fair to compare a newer generation  GoPro or Insta360 against the X3000.  

I only have access to a Hero 9 Black at the moment - but from what I have seen of reviews, the Hero 10 and Hero 11 have basically the same low light performance as the Hero 9.

Interestingly, I noticed that even during day time shots, there are market differences between the GoPro and the Sony - and not flattering differences.  

Before I did any night videos, I wanted to do some day time videos so I could get the settings correct and establish a baseline.

Before we go any further, yes I am lazy and so I just took screen shots of my entire desktop.  Deal with it.  

I rode the same route twice, once with the GoPro Hero 9 Black, and once with the Sony X3000.  I used the highest frame rate available at 4K, with both set to "Wide" angle, and the standard bitrate.  

The screen shots are obviously much smaller than the original files, but even so you can see a huge difference!  (You can click on the screen shots to view them at a larger size).  

The texture of the road is washed out on the image in the right, and the white building in the upper right quadrant looks... faded and mushy.  (Sony is on the left here, GoPro on the right)

Look at the difference in the gas station sign in the image above (Sony is on the right)

Look at the definition on the leaves of the bushes in the image above.

And.. one more for good measure.  The contrast on the sony seems much better.  
(Looks at the cracks in the paint on the crosswalk)

It was at this point that I realized that the angles are very very different between the two cameras.  In this case, the Sony is the one on the right, and the traffic light dead ahead looks much farther away on the Sony, while the sides of the image (including the blue animal hospical building on the left) are cut off on the GoPro.  

The Sony has only "Narrow" and "Wide" options for angle, while the GoPro actually has three options.  It seems the GoPro has a "SuperView" option that more closely corrosponds to the Sony's "Wide" option.  I decided to change this before testing night time shooting.  

Another change that I made was to the data rate.  For the day time I had been using 50Mbps on the Sony and "Standard" on the GoPro.  Since I had no idea what actual bitrate the GoPro was using, I decided to change it to "High", and change the Sony to 100Mbps.

Finally, the Hero 9 can handle 4k @ 60 fps, while the X3000 can only do 4k @ 30 fps.  As mentioned above, I I was comparing both at their maximum frame rates in order to give the GoPro the best change of success - but it occured to me that this decision might have been misguided.  All things being equal, a higher framerate will require more data, which means if both cameras were given the same bandwidth quota then the camera with the higher frame rate could potentially end up with worse detail.  Also, for night, forcing 60 fps might create more grain.  To make things fair, I set the GoPro to 4k @ 30fps instead for the night shooting.  

Before we move onto the night section, I should say I was quite surprised.  I fully expected the GoPro to win the daytime shootout since it is 4 years newer.  The way I figured it, the night time performance might be impossible to beat because of pesky things like... physics, but the advances in software stabilization may on average beat the optical stabilization during the day.  Plus, the Hero being 4 years newer probably has a more advanced sensor, etc.  - and yet somehow the Sony beat it at every turn!

The night time tests.

Before we start, let me be clear, neither of the cameras produces what I would call great night shots.  Something like my Alpha camera can absolutely crush either of these cameras - but it's not an action camera I could mount on a bike.

So the criteria here isn't so much "which one looks good", as "which one looks least bad".  

Again, I ran the same course twice, just before midnight, once with the Sony, and once with the GoPro.  

As mentioned above, both cameras were set to 4K at 30FPS, with the highest possible bit rate, and the widest possible angle, and maximum stabilization.  

Right off the bat, this crosswalk is a blurry mess on the GoPro!

This one is interesting, as the brick with the holes in it on the left actually looks better on the GoPro to me, but the mailboxes look way better on the Sony.  

This is a good shot to show that the angle is now at least closer to being the same.  

Again with the blurriness on the road, but also the difference in the road sign and the yellow shop sign is a bit unexpected.  On the original, I can read the writing on the blue street sign on the Sony, but on the GoPro it's illegible.  

This one looks pretty bad on both cameras to be honest, but it looks even worse on the GoPro.  

It's hard to even make out the word "Dominos" on the GoPro.  

It's hard to see at this reduced size, but the GoPro is much noisier.  

The building is more defined on the Sony, and the image is noisier on the GoPro.  

To be sure, the GoPro has some advangates:

1. A larger accessory ecosystem

2. WebCam functions

3. Streaming functions that actually work

4. Touch screen

5. Cloud service with automatic WiFi uploads

6. Horizon lock

All great stuff, but kind of meaningless if the image quality is sub-par.  

I wasn't surprised that the Sony won for nighttime shooting - but for day time, I absolutely was.

Also, I kind-of wanted to be proven wrong because at the moment there is no upgrade path for me.  Someone mentioned to me that the Insta360 RS has a 1 inch Leica sensor, but the demo videos I have seen of that used in the night time don't necessarily give me confidence.  Perhaps the noise will be less, but the stabilization still looks sub par.  

Also, given that the screen shots above are small and perhaps difficult to see, and just so people don't think I am cherry picking, I have uploaded the original videos, unedited to YouTube.  (Of course, YouTube compression makes everything look worse, but I am not sure if that will accentualte the differences, or hide them).  

Another reason to view the videos is that it is a more fair comparison.  Unforunately, comparing the videos sway things even more in favor of the X3000.  

You'll want to click "Youtube" to view the videos on the YouTube site itself, and then set the resolution to 4k.  

Sony X3000 Night Cycling Test

GoPro Hero 9 Night Cycling test

I would be interested to test with the Hero Black 11 or the Insta360 RS with the 4K boost or 1 inch Leica lense, but sometimes tells me to be suspicious.  If 4 years of progress hasn't already beaten the  X3000, then 6 years might not make much of a difference.  I certainly wouldn't buy a new camera without testing it out first.  

I understand why the stabilization on the Sony is better in some ways, especially at night - but I don't understand why the Daytime video from the GoPro was so bad.  At first I thought the lense must be dirty.  

I thought about buying an extra X3000 as a back-up unit, but they are now going for almost twice their original retail price, even used!  It looks like other people like them too!  

One thing I did notice is that the GoPro has an advanced settings screen, which is awesome.  If you have any recommended settings for me to try on the GoPro, let me know, and I will borrow it again to give it another shot.  

Final Disclaimer:

1. The videos and photos are note edited in any way.

2. The X3000 isn't for everyone.  If you need 4k@60fps or something like 5.3k, then the X3000 won't cut it.  The X3000 can't be powered from USB, only battery, and  it uses microUSB (which is only USB2 data transfer).  There is no webcam function, and the streaming function is no longer supported.