mini 0806 (mini 0803 pro) demo video

Sila said:
Dashmellow, explain 18 and the 22nd second of this video where there is no difference
trust the eyes :)
Click to expand...

@Sila, I'm not sure quite what point you are trying to make here or why?

The difference between shutter speed and frame rate is well known and was covered previously in the various links from my previous post.

Regarding the first YouTube video you've posted there is a pronounced and noticeable difference in motion blur between the two frame rates compared with the higher rate looking better.
Even Ivor Feric, the author of the video makes note of it in his comments attached to the clip and apparently his purpose for posting this video was to demonstrate this. For some reason, which I do not understand, you've reached your own opposite conclusions and didn't take the time to read his remarks about his video. While you state that there is "almost no difference at 1/4000th second". Feric states the opposite, "Even at 1/4000th of a second motion blur on 25fps is significant".

Here is a screen shot of Feric's comments about this video with some annotations I've added.

25v50fps.jpg

Regarding the second video, you point out that within the entire 40 second clip there are two anomalous consecutive 1 second segments (glitches?) where there appears to be identical severe motion blur with both 30 fps and 60 fps. For some reason you've decided to ignore the whole rest of the 40 second video which includes 6 excellent examples of an extreme difference in motion blur between 30 fps and 60 fps where the value and performance of the higher frame rate is demonstrated clearly and dramatically.

Once again, you apparently didn't bother to read the descriptive comments posted by the author of the video (iWazzuu) where he explains that his purpose in posting this clip was to show how much better things look at 60 fps.

GoPro-30v60.jpg

Gopro60.png

As for why there is a glitch in the first video, I have no idea but I speculate that it is simply a processing error in the camera's DSP. I think @Nigel's comment that it is a GoPro and not a motion picture camera is correct.
The GoPro and related cameras using Ambarella processors (and others, of course) do a remarkable job for what they are and for what these DSPs cost but they are often being pushed very hard. That is why these cameras get so hot.
While technically, on a basic level a GoPro and a high end motion picture camera are doing essentially the same thing when they capture video you just can't expect a 3 or 4 hundred dollar GoPro to perform as perfectly 100 percent of the time as a digital motion picture camera with multiple powerful processors and tons of memory that costs hundreds of thousands of dollars or even as well as a video-capable DSLR or dedicated video device costing a couple of thousand. I believe there is a good chance the GoPro could experience a hiccup now and again and perhaps drop a few frames here and there. Who knows, perhaps there is some other logical explanation someone else can offer but why dwell on such a thing and ignore the "big" picture which is that 60 fps will definitely give you sharper footage and still frames.
 
Last edited by a moderator:
Dashmellow said:
...
Click to expand...
There is clearly more going on in that last image than simply the difference between 30 and 60 frames per second. The 30fps clearly has 6 superimposed images while the 60fps has a single image. There is no reason the 30fps image could also not be a single clear image but the camera is attempting to blur the motion to make it look smoother when viewed as a movie. Clearly this is unwanted for an accident camera where you may want to study individual frames! I have to wonder, if the camera thought it a good idea to use 6 exposures for the 30fps then why did it not use 3 exposures for the 60fps and mess that up too - probably ran out of processing time.
 
The fact that 6 out of 8 segments of the video show a dramatically better image at 60 fps would seem to make it unlikely that the camera "probably ran out of processing time". Nevertheless, I think you make a good point about the multiple images. I believe the explanation is more likely to do with the nature of how YouTube renders video frames and also may be related to the fact that the video was rendered at 30 fps which the author makes a point of mentioning. It is unfortunate that we don't have the raw footage to view.

As I've said, I don't believe the posting of these videos was particularly helpful to the discussion but as I've also said, it should be noted that the authors of both videos were attempting to demonstrate that higher frame rates provide better results.
 
Actually, it might be interesting to do some experiments with 3D in a dash cam to see it it might offer some sort of benefit for capturing important traffic details in a way that are not captured in 2D.
 
Dashmellow said:
Actually, it might be interesting to do some experiments with 3D in a dash cam to see it it might offer some sort of benefit for capturing important traffic details in a way that are not captured in 2D.
Click to expand...

plenty of resources online, how to set the cameras up, what software to use etc
 
jokiin said:
plenty of resources online, how to set the cameras up, what software to use etc
Click to expand...

Could be an interesting project. I'll have have to add it to my "To Do" list, which seems to be getting longer and longer by the day. :) When summer finally winds down I hope to return to some of my DIY dash cam projects.
 
Dashmellow said:
Actually, it might be interesting to do some experiments with 3D in a dash cam to see it it might offer some sort of benefit for capturing important traffic details in a way that are not captured in 2D.
Click to expand...
That was why I suggested it, NASA use it on the Mars rovers for a good reason and 3D was widely used by the British back in WWII for aerial reconnaissance because they could get a lot more important information using it, and back then it was far from easy to do. I suspect it would be more useful than 60fps. (As long as the camera has a "non-movie" mode which doesn't attempt to artificially blur motion.)
 
Nigel said:
(As long as the camera has a "non-movie" mode which doesn't attempt to artificially blur motion.)
Click to expand...

I'm not aware that video cameras "attempt to artificially blur motion" in "movie mode", whatever that is.
 
Last edited:
JooVuu said:
Personally I think you guys are thinking too small...

Anyone want to help JooVuu develop this camera into a car camera?

http://www.red.com/store/products/e...ide-ssd-carbon-fiber-and-magnesium-lens-mount

Anyone?

But seriously, I would always have fps over mega pixel purely so you can catch more of the 'action'. Action being movement of cars/people/dogs/ghosts...the usual.
Click to expand...

Yeah, that's one of the Red Digital Cinema cameras I mentioned back in post #16 and while it's pretty amazing and cool, the EPIC Red Dragon only can do a measly 300 fps at 2K. So, JooVuu man, I think you too are thinking way too small! We'll need a Vision Research Phantom Flex camera as a starting point if you wanna' think BIG developing the ultimate high fps dash cam. It'll do 2570 fps at 1980 x 1080 2.5K and up to 10750 fps at 640 x 480. There's also the Phantom Flex 4K that will do 2048 x 1080 at 2000 fps.

So count me in if you need help or a beta tester and I assume JooVuu will be funding this venture, no? The Phantom camera line goes for $50,000 to $150,000 depending on specific model and features and of course we'll need two of them to develop this in 3D.... so, Add to Cart? :)

Can't wait!! I can just see it now, "Officer, here's what happened, down to the tiniest detail". "Wait, put on these special glasses".

 
Last edited by a moderator:
Dashmellow said:
Yeah, that's one of the Red Digital Cinema cameras I mentioned back in post #16 and while it's pretty amazing and cool, the EPIC Red Dragon only can do a measly 300 fps at 2K. So, JooVuu man, I think you too are thinking way too small! We'll need a Vision Research Phantom Flex camera as a starting point if you wanna' think BIG developing the ultimate high fps dash cam. It'll do 2570 fps at 1980 x 1080 2.5K and up to 10750 fps at 640 x 480. There's also the Phantom Flex 4K that will do 2048 x 1080 at 2000 fps.

So count me in if you need help or a beta tester and I assume JooVuu will be funding this venture, no? The Phantom camera line from goes for $50,000 to $150,000 depending on specific model and features and of course we'll need two of them to develop this in 3D.... Add to Cart? :) Can't wait!! I can just see it now, "Officer, here's what happened, down to the tiniest detail". "Wait, put on these glasses".

Click to expand...

Oh this is awkward...because we've been and done that 2000fps is now childs play. We're looking at this http://gizmodo.com/5867562/unbelievable-trillion-frames-per-second-camera-captures-light-in-motion ....whatcha got in response to that ;)

PS
We thank you very much for your large donation of roughly £100million to JooVuu Limited for us to develop this...we'll be sure to send you some email updates and a mug.
 
Last edited by a moderator:
You must log in or register to reply here.
Back
Top