Kamkar beating on B1W

[estore009]

@kamkar1 and @dash riposki and @estore009 plus anyone else experiencing audio problems with the BiW

I have just noticed that the mic is right next to the cam's push-button, and in my cam that button is loose. This could produce a rattle which could resonate at a low frequency which would be easily picked up by the adjacent mic . If those experiencing this issue could place a piece of tape on the button to the case preventing rattle, we can then see if the loose button is the problem.

At worst if that does nothing we will know one place the problem isn't at which can help guide us further. I'll join the testing this afternoon or tomorrow when it's warn enough outside to be playing with cams. It's -6C (20F) and dark here so non-critical things can wait a bit

Phil

Thanks for your thought,we will have a test

@kamkar1 and @dash riposki and @estore009 plus anyone else experiencing audio problems with the BiW

I have just noticed that the mic is right next to the cam's push-button, and in my cam that button is loose. This could produce a rattle which could resonate at a low frequency which would be easily picked up by the adjacent mic . If those experiencing this issue could place a piece of tape on the button to the case preventing rattle, we can then see if the loose button is the problem.

At worst if that does nothing we will know one place the problem isn't at which can help guide us further. I'll join the testing this afternoon or tomorrow when it's warn enough outside to be playing with cams. It's -6C (20F) and dark here so non-critical things can wait a bit

Phil

Not the problem form its button,we still on the way to test more and try to figure it out!

 

[SawMaster]

Well that's one less thing to consider so it is progress. I'm going to try disassembling mine today for a better look.- maybe that will get me some new ideas. Hope I don't break the nice little beastie I noticed quite a difference between @dash riposki road-tractor and pick-up truck recordings which leads me to believe this is vibration-related. The big truck will have far more strong low-frequency vibrations versus the smaller one. The problem was still there, but lessened overall with the smaller truck which is the clue

Phil

 

[Module 79L]

Yeah i dont think its a rattle picked up, i am more inclined to say it is resonance that's picked up.

I dont think its the microphone that's bad, it may be the microphone used are not optimal for this kind of application, but its not bad i think, maybe applied bad or installed bad.

I also think resonance...or something.
If I thought I could take it apart, and get it back together again I'd maybe try a few dots of hot glue inside the case in empty spaces.

Trust me, I know what I'm talking about.

 

[SawMaster]

OK, just a quickie before I get to my transmission work. I got it apart and as I expected the mic is fixed to the case in a hole and held with a dab of hot-melt glue.. This is typical of most (if not all) dashcams, so in itself shouldn't be a problem. But with my example, the glue dab was very small and attached to one side only, and then just barely holding the mic. This allowed the mic to move in the hole on the free side, and a gentle pull on the wires released the glue and the mic came loose. I see a distinct possibility here that in other units the mic could be totally loose inside the cam which would amplify vibrations if it were just laying against the case banging around.

In the other cam's I've been inside there was more glue better distributed here. I re-glued it and reassembled the cam. I will do a road-test later today and post a series of pics on the disassembly process later tonight.

@Module 79L you could be right, but given your audio experience I think you'll agree that a loose mounting could play a role (rumble is the audio term used to describe that over here). There's enough space inside so later on I am going to foam-mount the mic element and if that does nothing then I have a mic from a G1W-C which had good audio that I will swap into here. With some luck it will be compatible and show whether you are right.

Time for me to get greasy again

Phil

 

[Module 79L]

@Module 79L you could be right, but given your audio experience I think you'll agree that a loose mounting could play a role (rumble is the audio term used to describe that over here).

Phil

What we hear in the video samples is not the sound of a loose mic, it's low frequency distortion. Those mics have to be very sensitive to be able to pick up low volume external sounds through the tiny holes in the casing, so when they're not good or aren't well calibrated, any high volume sounds, especially the low frequency ones, tend to overwhelm the mic and cause distortion.

If you decide to swap mics, bear in mind that they have polarity. Make sure you mark the positive pin of the G1W-C's mic before removing it from the PCB. I'm assuming the B1W's mic has red and black wires, so in principle the red should be the positive.

 

[kamkar]

Reporting back.

My testing yesterday did not give me any results i was hoping for, upon reviewing the footage i noticed that the distortion are indeed strongest when i driver over a manhole cover and that do give some low frequency noise in the car.
So if it is possible in FW to cut off the lower part of the frequency range that would be a option.

I have yet to see if the change to 60 Hz alleviate the dropped frame issue, when i came home yesterday from helping my friend i was too tired and had a headache so after eating a little and digesting for a hour or so i went to bed.

 

[kamkar]

Nope !

I have just looked at my footage from yesterday recorded with a 60 Hz setting for the light, and the duplicate frame every 5 frames are still there in daytime footage.

Visually looking the the footage recorded with the 60 Hz setting i dont see any flickering in the lights i pass buy ( street lights - shop window lights - gas station billboards - LED billboards )
Everything look normal, so maybe the 50/60 Hz option are not working and just go with 50 Hz no matter what is selected ??

I am not sure what i could do / use to test that 50/60 Hz setting, i am thinking maybe my computer monitor that i run at 144 Hz now maybe i could change it to some other refresh rate that would witk with 50 or 60 Hz but would then not work with the other setting.

 

[SawMaster]

Just got time to check my clips and wow- the noise is awful My minivan is relatively quiet and during my testing there was little difference between driving and stopped by ear, but in the clip linked below the difference is like night and day. One minute clip with me stopped near the end, the rattles when I went onto the bridge are tools in the back and not the cam This is a stretch of road near my home so when I get the mic swapped out we can directly compare the before-and-after. Not sure when I can get the mic swap done, I have to take advantage of the mild weather and get my transmission work done right now.
B1W Audio Test-1

Phil

 

[TonyM]

Thinking back to my early tests with this camera, I recall the amplitude of the recorded audio would vary even whilst the car was stationary at traffic lights. I'll do another check to see if it's still doing that.

 

[Mtz]

50Hz or 60Hz are not about video fps, but about electricity. I know that some companies are advertising their video resolutions for example ast 1080p50Hz, 1080p60Hz to be 1080p 25fps and 1080p 30fps, but this is just stupid.

enjoy,
Mtz

 

[kamkar]

Yes i know, but jokiin suggested that changing it to 60 Hz might alleviate the dropped frames issues, as it seem some cameras might have this issue.
But i notices it dont matter if i have it set to 50 or 60 Hz, neither of them give me any flashing/flickering light in my apartment light or any other light along the roads, and we are 50 Hz here in Denmark.
I also tried to set my computer monitor to a range of Hz from 60 on up to 144 Hz but that was all also butter smooth footage i captured on the B1W

 

[dash riposki]

Maybe the problem is we're not driving chinese made cars?

 

[SawMaster]

Maybe the problem is we're not driving chinese made cars?

I would not be surprised in the least if it turned out that the "Engineers" hardly ever or never tested the product outside of the office. Almost nobody likes to take their work home after a long day, and nobody does something which they feel is unnecessary. This is universal human nature not limited to any place in particular And indeed we do not know which roads, vehicles, and what speeds were used if they did do any real-world testing. That could make a huge difference as it does with Korean cams which fare poorly at the faster speeds the rest of the world drives Even Blueskysea may not know these things if they are not working from the same office.

But rather than questioning or bashing, let's keep our focus on solving this one problem for now, as the world needs a cam like this if we can get the audio sorted out We can bash to our heart's content afterward

Coffee break over, back under the big old van I go
Phil

 

[Leonauto]

...
But rather than questioning or bashing, let's keep our focus on solving this one problem for now, as the world needs a cam like this if we can get the audio sorted out We can bash to our heart's content afterward

It seems to me that I have become a bit closer to understanding the solution and direction of our actions to help engineers Blueskysea improve the audio recording quality in B1W and remove distortion.
What will be useful now:
In order to analyze the noise spectrum, two records are needed:
# 1 - recording sound B1W with distortions, and
# 2 - recording sound (side-by-side) on any other device (dashcam, video camera, voice recorder, etc.) without distortion.
That is, we need this record as a source of sound, which engineers can reproduce and test how this sound will record the B1W and at what frequencies and under what conditions distortions appear.
I hope this is the right way and I hope that this should help improve the quality of audio recording

 

[kamkar]

Here a lot of people take their work with them home, but its not us regular hands on workers.
But people pushing papers check their work mail very often also out of the work place and even during holidays, but i assume its just another excuse to have a phone in the hand.
Personally i never minded working and put the hours in, and if push have come to shove i have been known to jump out of bed and help where i can, off course mostly paid work, but for friends i dont get payment.

And really friends should read friend, cuz outside of family i just do stuff like that for 1 person.

 

[jokiin]

50Hz or 60Hz are not about video fps, but about electricity.

correct, but not not always, it's a common tactic on some of the lower end chipsets to tie frame rate to the frequency setting to avoid flicker and in doing so they either add 25fps and 50fps options, or they run everything at 30fps or 60fps and inject the missing 5 or 10 frames for each setting

I know that some companies are advertising their video resolutions for example ast 1080p50Hz, 1080p60Hz to be 1080p 25fps and 1080p 30fps, but this is just stupid.

as above it's because they don't have proper 50Hz/60Hz anti flicker capability that they do this

 

[jokiin]

I would not be surprised in the least if it turned out that the "Engineers" hardly ever or never tested the product outside of the office.

quite a lot of firmware engineers don't own cars, and don't drive anyway, I have known of some engineers that when they need to test something if the boss isn't there with his car they call a taxi and go for a drive with the cameras

 

[Module 79L]

It seems to me that I have become a bit closer to understanding the solution and direction of our actions to help engineers Blueskysea improve the audio recording quality in B1W and remove distortion.
What will be useful now:
In order to analyze the noise spectrum, two records are needed:
# 1 - recording sound B1W with distortions, and
# 2 - recording sound (side-by-side) on any other device (dashcam, video camera, voice recorder, etc.) without distortion.
That is, we need this record as a source of sound, which engineers can reproduce and test how this sound will record the B1W and at what frequencies and under what conditions distortions appear.
I hope this is the right way and I hope that this should help improve the quality of audio recording

@kamkar1 already posted a video with that comparison in this thread (B1W vs. SG GC).

 

[Module 79L]

Just got time to check my clips and wow- the noise is awful My minivan is relatively quiet and during my testing there was little difference between driving and stopped by ear, but in the clip linked below the difference is like night and day. One minute clip with me stopped near the end, the rattles when I went onto the bridge are tools in the back and not the cam This is a stretch of road near my home so when I get the mic swapped out we can directly compare the before-and-after. Not sure when I can get the mic swap done, I have to take advantage of the mild weather and get my transmission work done right now.
B1W Audio Test-1

Phil

One other thing I've noticed before in @kamkar1 samples that I haven't mentioned yet is how muffled the normal audio (if you can call it that) sounds. The mic swap will tell us where the problem is: in the hardware or in the firmware.

 

[kamkar]

I would assume if the microphone pick up too much low frequency sound, then when driving while holding it in the hand would mean distorted sound, on the other hand if the audio are good with a handheld camera in a driving car the issue must be picked up via the camera mount and housing.
I am going for a little drive tomorrow to visit my dear old mother, will bring powerbank and micro USB wire to do a handheld test too.