I just bought a DJI Osmo Pocket 2 camera to use as a cab ride camera in videos.
The depth of focus is somewhat shallower than I’d hoped (to say the least). The f-stop is fixed, so all I can do is play with the point of focus.
Here’s a short video of some testing I’ve done using four different points of focus - one about 40 scale feet in front of the lens, one at 80, one at 120, and one at 160 scale feet.
I’d sure like some feedback on this. IS the camera suitable at the farther points of focus, or is the depth of focus just too shallow for the application?
One thing I do like is the low camera angle - it scales out to about 6 feet above the ties!
So what do you think? Return and buy a cube camera? Or keep and use for future videos?
It is not great, however, if it is relatively inexpensive it might be worth keeping to shoot back towards a following train or something along those lines.
I have several micro cams, all cheapos. I haven’t done much with them as my arthritis has kept me from working on my layout.
Mine are all pinhole cameras so no focusing available. I like your test methods.
Just when I was about to do my thing on my layout several years ago we had a break-in and they stole my 7” tablet (viewfinder) so I backed off, I still haven’t ordered a replacement. 70 HO locomotives in the garage and thousands of dollars of power tools and all they took was a $40 tablet.
The SQ series cameras have WiFi as well as on board recording, I use the WiFi as a viewfinder.
I found 80’ to be a distinct improvement over 40’ and quite liked it. My next choice would be the next highest focal point, but I would favour the 160’ the least. I need to see details the way I would be ‘seeing’ and looking if I actually were in the cab. To me, the 80’ is about perfect, followed reasonably closely by the 120’.
I’m no camera expert by any means, but I do know what I don’t like. In all four settings there were still some things that were out of focus. The 120’ setting was the best but I still found the out-of-focus elements to be distracting. Perhaps when you have more scenery and detail in the foreground the effect might be less noticable.
I think I can safely predict that you will be taking a fair number of operational videos so if I were you I would invest in a better camera.
Thanks everyone - your feedback is extremely helpful.
I did some additional testing last night, including at some lower light levels. Here’s that video:
This is not an inexpensive camera. Total cost was $350 (not including tax!).
I really like the eye-level view the low lens position allows, and the gimbal-stabilized image (the reason the camera costs what it does) prevents most wobbling. Then again, my video editor has stabilization routines that might eliminate the wobble. So I’m still uncertain of what to do. I have about ten days to decide, so…
Mark, about ten years ago (…already!!!..[:|] ) I uploaded a video of my BlueLine Class A 2-6-6-4 for someone who wanted to see what it looked like and sounded like with a Revolution in it, or was it a Titan…can’t recall, a QSI new variant (oooh, must not use that word. [:#]). I was surprised when youtube prompted me to use their new stab tool. I accepted, being the curious guy and because I had hand-held whatever it was, I think my Powershot A710is.
Let’s just say that it was a huuuuuuuuuge mistake. The image was rendered much worse, essentially spoiled, by that app. So, my advice would be to find contentment with the gimbal method resident in your appliance. If youtube’s app has been improved, maybe what I say is moot ten years later, but I’m not banking on it.
My video editor of choice is Davinci Resolve. I’ve messed with the stabilization processor in it - it’s very good. It doesn’t degrade the video at all. Basically it crops the edges of the frames just a bit as needed to move them around to stabilize the video, then enlarges the entire video clip back to full frame. Very effective.
Dave all my cameras are cheapos. I started out with a wireless analog pinhole camera about 25 years ago, TV Channel 4 about $40. I installed it in the nose of an Athearn PA1. I bought a second analog camera about 10 years later and installed it in a shortened Athearn PA1 frame with a Proto E7 shell.
About 5 or 6 years ago I bought my first micro digital recorder, MD80, much better resolution over the analog cameras but not very reliable. A couple of years later I bought a SQ7, $18. A bit larger than the MD80 but not very reliable either. I moved on to a SQ8, much better camera and half the size of the SQ7.
The SQ8 has been very reliable and has exceptional video quality. About a year ago I liked the reviews of the SQ23 and bought one, about $20. The SQ23 has been very good, all the features work as advertised, very good video (for a pinhole camera).
I haven’t been able to run the SQ23 on my layout do to many things, arthritis pain and I tore up my control panel before the arthritis problem and haven’t been able to run any trains since last December.
Mark
If you decide to return the OSMO Pocket 2 give the SQ23 a shot. It has worked out very good on my bench tests. The focus isn’t real sharp at all distances but it’s the best pinhole camera I’ve ever had. The only thing that isn’t real good is like all pinhole cameras it has a slight fisheye look but it’s the best camera I’ve ever tried for my layout.
The flat car has Mel wheel wipers for track power and a DC to DC converter to supply 5 volts to the micro USB connector on the SQ23. With the WiFi on and in record mode the battery life is about 30 minutes, it takes me that much time to get it ready to take
Have you seen the Choo Choo Vision cameras? They are micro-sized 2MP WiFi cameras that run from track power. GHR line in Ontario is using them for Remote Operations. There’s also a PAN-CAM accessory that allows you to pan the camera.
After reading through everyone’s comments and thinking it over, I returned the Pocket 2 to the store and ordered one of the first models - the Osmo Pocket. It has a slighly smaller aperture (2.0 compared to 1.8) and a somewhat more restricted field of view. I thought these might combine to provide better results for cab ride videos.
I got the Pocket today and tested it out. Here’s a short video of those tests, and the video includes side-by-side comparisons of the recordings from the Pocket 2 and the Pocket.
I prefer the Pocket 1 results overall, but I have a question. One thing that bothers me at the higher ranges is that the foreground track is noticably out of focus with both cameras. That track takes up a lot of space in the frame so, to me, it is very obvious. I know you said that you like the ability to mount the camera quite low, but I wonder if the blurry track would be less obvious if the camera was mounted higher, say at the engineer’s level of sight?
