SG9665GC firmware updates and pre release access

Just went out and checked, both cameras were set to "low", which explains the behavior. I just set them both to "off", problem solved! Apologies jokiin, no need to investigate this PEBKAC error (or maybe it's PEBDCAC -- Problem Exists Between Dash Cam And Chair ;) )

Sutton said:
(similar to a smoke detector in your kitchen going off if you burn your toast, at least you know its standing by)
Click to expand...
Honestly, I wouldn't mind if the older "event" footage was cycled at the same time as the regular stuff. It does seem to redirect the footage to the "RO" folder on an event, rather than simply recording the same file twice like my Thinkware F750 did, which is nice. Good lord I hated that thing...

Anyways, reviewing my 128 GB Adata cards (I swapped them on 10/27 in favor of 200GB PNY cards), the oldest drive I have a full record of is 10/12, which has 17 "uneventful" videos, and 51 "event" videos, each lasting about a minute. The "RO" folder has files from 10/5 to 10/27, but the main folder only goes as far back as 10/5, right around when I upgraded the firmware and formatted the card. In the card with a capacity of 119 GB, a full 37.9 GB (31.8%) is considered "eventful". 30 GB even (25.2%) if you only count from 10/11 and more recent.

The part I don't like is that once "event" footage is stored, it is kept, even if that means a full third of the card is nothing but event footage. That's why I'm turning it off -- I still have the screen come on for three minutes at the start of each drive so I can make sure it powers up properly and isn't experiencing issues, but I don't need to be storing pseudo-eventful drives in a manner that minimizes the space available for actually eventful drives that may not involve an impact. I'd prefer to make it to a computer and simply pull the most recent footage.
 
jokiin said:
neither of us said anything of the sort, once again you're reading things that haven't been written, Jon said "but it seems maybe latest Beta messed it up for some", I said "I don't know if there has been any mistake that has overwritten the calibration info, I said I would check, could be something unrelated, give us a chance to at least investigate and see if any connection", you just turned that into "we clearly stated the beta messed it up" both responses were directed at yourself so I'm not sure if you don't read things or what, anyone is welcome to go back a page and read that for themselves, it's there in black and white, how can you possibly see it as anything other than what is there, I don't get it
Click to expand...

Well, yes...for "some". Either way, both you and Jon were definitely suggesting the possibility or even likelihood that the new beta somehow wiped the calibration data. Since several people so far have reported hot pixels only since installing the recent beta 27 it seems likely that something must be happening to the "locked" calibration.

Then again, according to the numerous threads about hot pixels with the V2 going back half a year or more, there must still be a hot pixels calibration problem long after we were told in no uncertain terms a year ago by Street Guardian USA, They locked the sensor calibration data into a protected memory area so it will no longer get over written during the next firmware update. Once the sensor is calibrated perfect at the factory, it will stay that way for the life of the product? (10/15/15)

Like I said earlier, I don't see some hot pixels as being a major problem. It has more to do with how this particular issue, like certain other issues with the GC are being presented. It feels as if forcing individual purchasers of the GC to formally request the calibration fix-it is a way of compartmentalizing the issue. I started this discussion because I agree with another member who asked the entirely logical question about just posting the fix-it instructions here on DCT. (Which happens to be publicly available on YouTube and has been since May, 2015 anyway!) It strikes me as interesting that most of the people in the links I posted who were reporting hot pixels where not aware that this has been a long time problem with this particular camera. Previously, when people reported hot pixels, a link would be publicly posted on the forum to niko's "Defect Pixel Calibration on Street Guardian SG9665G YouTube video tutorial". It's odd that now it is a link that requires a private communication to obtain.





 
Last edited:
our engineer checked and we have made no changes to any code in regard to where the calibration information is stored, it's still in the protected area of memory, they are checking with Novatek in case there's something that has changed that they are not aware of, Novatek hasn't come back to us on this as yet (I checked yesterday for any update), if and when we get any news from Novatek that there are changes we will address them accordingly if that's what is needed
 
jokiin said:
our engineer checked and we have made no changes to any code in regard to where the calibration information is stored, it's still in the protected area of memory, they are checking with Novatek in case there's something that has changed that they are not aware of, Novatek hasn't come back to us on this as yet (I checked yesterday for any update), if and when we get any news from Novatek that there are changes we will address them accordingly if that's what is needed
Click to expand...

Thanks for reporting back about that. Since people have been reporting hot pixels for at least the last six months or so there must be some explanation other than "locked calibration" in a protected area of memory. Could there perhaps be some calibration error at the factory in a certain percentage of the units going out the door since re-calibration by the user seems to help?
 
Dashmellow said:
Thanks for reporting back about that. Since people have been reporting hot pixels for at least the last six months or so there must be some explanation other than "locked calibration" in a protected area of memory. Could there perhaps be some calibration error at the factory in a certain percentage of the units going out the door since re-calibration by the user seems to help?
Click to expand...

Straight out calibration errors could happen, that is always a possibility as they are done manually, can never rule out human error

There are other variables though, as I mentioned previously there is an automated function which can dynamically turn white pixels off, the current version of this however has a very negative impact on overall image quality so we choose not to use it, with some sensors like the AR0330 this function can work quite well (and are less noticeable due to smaller individual pixels) but unfortunately not so with the IMX322, now for sensors were this does work well what you don't notice is that hot pixels aren't a one time situation, they can change based on a number of factors such as geographic location, altitude, time, temperature, accumulated hours of use etc etc so it's not as cut and dried as just saying lets calibrate this and that's the end of it, when you have a dynamic software solution that does this on the fly you're blissfully unaware that these variables exist but when white pixel info is mapped and stored in memory it's only based on what pixels were showing up at that particular point in time when the calibration was done, another time, another location, the information can change so recalibration is available to address that, it would be nicer not to have to do this but the reality is that the downsides of the automated process (impact on image quality) far outweigh the annoyance of either having a couple of bright pixels or needing to recalibrate, recalibration is definitely the lesser of two evils

As to why a few of these came up after updating the firmware we still have to wait and see what Novatek say to see if something has been changed that we're not aware of, perhaps it's something that is more noticeable due to different metering tables and the differing emphasis on parts of the image compared to the prior version, not sure, something we're still looking at, as with any issue that crops up you can't address it unless you can identify the cause, until such time there's no point speculating, it's just a tedious process to work things out
 
With all the discussions here about white pixels, and the others about night-time quality improvements, at this end-of-year season, I'm not surprised more people check their videos for white dots and do find them.
Much more than when we were discussing about whitewashed-out skies in high noon summer shots.:D
 
Is there any update on being able to adjust the bitrate in menu? Would be good for those where duration of capured footage is more important than having the best possible image quality such as commercial use, car going into the garage for a few days etc. Thanks.
 
That's something I want to look at but not until the other IQ related adjustments are sorted, they still have priority
 
"IQ" = Image Quality

Just clarifying before anyone misreads that as an unintended insult.
 
daniel_bel said:
With all the discussions here about white pixels, and the others about night-time quality improvements, at this end-of-year season, I'm not surprised more people check their videos for white dots and do find them.
Click to expand...

never checked mine for them, nothing stands out, that's all I'm worried about, I'm not into cinematic productions using dashcams.

Shep said:
"IQ" = Image Quality
Just clarifying before anyone misreads that as an unintended insult.
Click to expand...

and all this time I thought jokiin was insulting my lack of intelligence all the time.
 
Sutton said:
Where is the cops when we need them?
Click to expand...
Good question, where is @DashCamMan ? Seems the lot of you have forgotten rule #1:
1. Treat other members with respect. While debating and heated discussion is fine, we will not tolerate personal attacks, rudeness or inflammatory posts.
Click to expand...
Funny, I thought cops were here to enforce the rules, not be your own personal servant for whatever flawed agenda you have :rolleyes:
 
I've already tagged @DashCamMan , I'll let him decide what to do here. Have fun with your ignorance.
 
I deleted posts from both Sutton and Shep and provided warnings. I believe both played a responsibility in the escalation. Please avoid insults going forward.
 
DashCamMan said:
I deleted posts from both Sutton and Shep and provided warnings. I believe both played a responsibility in the escalation. Please avoid insults going forward.
Click to expand...
@DashCamMan Fair decision and will abide by your advice.
 
arcticfire said:
Is there any update on being able to adjust the bitrate in menu? Would be good for those where duration of capured footage is more important than having the best possible image quality such as commercial use, car going into the garage for a few days etc. Thanks.
Click to expand...

jokiin said:
That's something I want to look at but not until the other IQ related adjustments are sorted, they still have priority
Click to expand...
Going in the other direction, is the hardware currently running at its maximum bitrate?
 
That would decrease the blocks! But how far can you take that?
 
the recording continues in the absence of the driver in the car

aowNs.jpg


 
So the bitrate could be higher? What's the limiting factor? Is it something like temperature, reliability, card compatibility etc., or is there no visible image quality improvement?
 
You must log in or register to reply here.
Back
Top