30fps v 60fps

[Kremmen]

I have all versions from 1.03 'in stock' so I can backtrack if I want to.

Next weekend when I do the M4 again at 50mph and 60mph in the roadworks zones I'll be able to compare like for like.

From what I've read here though HDR/WDR inserts an artificial frame that can introduce motion blur under some lighting conditions and best avoided if possible, is that still the case ?

 

[jokiin]

From what I've read here though HDR/WDR inserts an artificial frame that can introduce motion blur under some lighting conditions and best avoided if possible, is that still the case ?

in simple terms

WDR is just wide dynamic range, nothing to do with frame rate

Multiple exposure HDR though uses two different exposures to create one frame, so the camera uses two different exposure settings at 30fps each in a 60fps format, where 60fps is normally just continuous recording frame by frame this (HDR) is split into two different exposures and the result is output as the one frame, as there's a 30th of a second difference between each frame used to make the one frame you will often notice ghosting in the image (most noticeable during the day) it's still there at night but less noticeable and due to the dual exposures used things like license plates, street signs etc are generally able to be read where they wouldn't normally be able to due to the motion blur you get in low light

the dual exposure method used in HDR is great for still photography, I don't think it will ever be perfect for video though as movement always adds some amount of difference between the consecutive frames used, the faster the movement the more noticeable it becomes, for something like a security camera it's a lot more successful as the movement is much slower and the bulk of the scene is generally a static environment

 

[Kremmen]

I've just checked my Menu and I have HDR not WDR so 30fps HDR off, my current setting, is my best option.

 

[Nigel]

Next weekend when I do the M4 again at 50mph and 60mph in the roadworks zones I'll be able to compare like for like.

In daylight there should normally be no difference between 30 and 60fps, however there should be big differences between days, and even bigger differences between months due to changes in light levels.

This makes your experiment likely to give the wrong result unless you have two cameras recording side by side at the same time, configured the same except for FPS.

I’m not sure that there really is no difference though, it is possible, and I think likely even though there is no good reason, that the camera has been configured to use different exposure settings for the different frame rates.

There is an additional problem in that the 60fps generally doesn’t have sufficient extra bitrate to be able to record the same amount of detail per frame, so especially under trees and at higher speeds, the image deteriorates and it can be difficult to work out the cause(s).

 

[SWoMP]

Would the 30fps setting generate less unit heat than 60fps?

 

[jokiin]

Would the 30fps setting generate less unit heat than 60fps?

On our SGGCX2PRO+ we do 30fps at 21mbit/sec and 60fps at 42mbit/sec and there is no measurable difference in the temps when recording

 

[Nigel]

On our SGGCX2PRO+ we do 30fps at 21mbit/sec and 60fps at 42mbit/sec and there is no measurable difference in the temps when recording

Is there a measurable difference in power consumption?

Generally more FPS = more power = more temperature. It can be hard to see the difference using a thermometer though.

 

[Nigel]

60fps at 42mbit/sec

That is high for 1080 resolution, should be good image quality

 

[Kremmen]

Here's Windows data for the 2 sets I have :

All rates are higher under 30fps.

 

[Nigel]

All rates are higher under 30fps.

Yes, negative extra bitrate doesn’t make for the best high speed 60fps image quality! But it does make for approximately the same power use and thus high temperature resilience for both rates.

 

[SawMaster]

On our SGGCX2PRO+ we do 30fps at 21mbit/sec and 60fps at 42mbit/sec and there is no measurable difference in the temps when recording

From anyone else I might question this but you know what you're doing so I'm sure it's right. Yet still it seems counterintuitive, as more data and more processing should equal more heat. Are you just not pushing the processor hard? That could explain the temps I think.

Phil

 

[jokiin]

Is there a measurable difference in power consumption?

Generally more FPS = more power = more temperature. It can be hard to see the difference using a thermometer though.

I'd have to double check the power consumption, can't recall as this is something we changed about 18 months ago, running them side by side on the bench 24/7 there was no measurable difference in the temps even after multiple days

 

[jokiin]

That is high for 1080 resolution, should be good image quality

It's selectable 30/36/42, some users just stick with 36 saying that the quality is fine and they just want the extra record time

 

[Kremmen]

I've just offloaded the latest trip and I still have motion blur at 30fps. Not as bad but not as good as previous.

So ........ I've gone back through my archive and I've found some 60fps with quite sharp freeze frame and re checked my old 30fps as well, all good.

Then ..... I only just spotted that I must have knocked the lens during one of my weekly mount/unmount because over the last few months I'm recording less sky and more dash. Not massively so, just one click, hence why I initially missed it. That angle change seems to coincide with the motion blur.

Re-angled and ready for next week.

 

[VIOFO-Support]

All the 60fps freeze frames have motion blur but going back to when I ran 30fps which would have been pre v2 firmware all the freeze frames are a lot sharper with no blur.

Could you please share two videos of these two frames you mentioned?

 

[Kremmen]

Thanks

All these clips are at 50mph for consistency.

This shows a history of previous good freeze frame at both 30fps and 60fps:

--------------------------------------------------------------------------------------------------

3rd Sep 2019, 30fps, that would have been firmware 1.03 (or maybe 1.02):

--------------------------------------------------------------------------------------------------

27th Nov 2020, 60fps, firmware 2.00:

--------------------------------------------------------------------------------------------------

25th Jul 2021, 30fps - Latest - but notice how much dashboard I'm showing and the motion blur, which is constant throughout all the journey:

As well as the camera angle changing I also note that the first bitrate is lower, not sure what 1.03 allowed.

Basically what is the optimum fps and bitrate for a good freeze frame as the difference between 30fps and 60fps is minimal.

 

[Kremmen]

Update:

After having raised the lens to show more sky and less dash the freeze frame is sharper.

I wonder if the logic was looking at the stationary dash taking 1/4 to 1/3 of the frame and under compensating ?

 

[Nigel]

If we ignore very dark conditions without headlights, then 30 and 60fps should have the same motion blur. More dash in the image can change the exposure and thus change the motion blur. Different weather does the same, I suspect most of the differences you are seeing are due to the weather.

If you want to keep the dash in the image then you can set a little -EV instead of adjusting the angle

Reducing bitrate will lose detail, generally it is background details such as blades of grass, then tree leaves. Writing is generally kept sharp until bitrate is far too low.

Since the issue is motion blur, more speed makes it worse, as does getting closer to things.

 

[Kremmen]

Thanks for the input.

This is one of those situations where I'm fortunate to have Tb's worth of old data so I can run comparisons. As you say, I've had to ignore headlight hours as they are unrealistic.

Again, as you say, speed plays an important part and if you look at all my above examples they are all 50mph because I use cruise control through the current M4 average speed controlled roadworks. Because these roadworks have been in place for a few years now I can go back quite a way.

It's an interesting investigation and I will continue to monitor it.

 

[Nigel]

Very difficult to work it all out since there is no way of knowing the light levels, and a 10x increase in light can look very similar on a video, a little extra cloud can make a big difference, even a bit of haze can halve light levels while looking identical to our eyes.

Only way to be certain about the differences when comparing two cameras or two firmwares, or two settings, etc., is to do a side by side comparison with two cameras. Don't necessarily need to be the same camera, just need a reference, if the test camera improves while the reference stays the same then you have a genuine improvement.