SG9665GC firmware updates and pre release access

I know how to do the calibration in my sleep at this point.

You did in fact tell me almost a year ago that the calibration was permanently secured at the factory. Now you are retracting that in the PM you just sent me. As a customer who has been repeatedly assured of this or that about this camera only to later learn that statements you made are no longer actually true, what am I to think?

I do appreciate the offer of a refund. I will consider it but you are missing or more likely ignoring the point. Your main goal always seems to be to marginalize any problems that come up and sweep them under the rug regardless of how well documented they may be or how long they have been going on. God forbid that anyone might discover that the Emperor isn't wearing all of his clothes.

As I mentioned, the camera otherwise has a lot going for it and after a long wait there has been a concerted and much appreciated effort to deal with the issues I first raised 18 months ago via new the beta firmware. It does strike me as rather unusual for a camera listed on DCT as the "best camera on the market today for video quality" to still be mucking around with beta firmware that seeks to address serious exposure issues after two years on the market in my opinion but for now, as I said, I am still hanging in there. Nevertheless, I will continue to report my experiences both good and bad.
 
Last edited:
Dashmellow said:
I was advised that the calibration was now permanently secured and quarantined at the factory and I wouldn't need to concern myself with that any further especially with the latest V2 unit that was shipped to me.
Click to expand...

Novatek engineers told us automatic calibration was possible awhile ago, but when we enable automatic calibration the image quality goes to crap so we were not able to go with auto calibration. :( sorry (we tried/really wanted to!)
Manual Initial and or Re-Calibration should be a 1 time thing, but it seems maybe latest Beta messed it up for some. If this is the case, a quick re-calibration should get things back on track. We need to make sure this last calibration issue doesn't happen for non-beta/release versions of firmware for sure.
 
Street Guardian USA said:
Novatek engineers told us automatic calibration was possible awhile ago, but when we enable automatic calibration the image quality goes to crap so we were not able to go with auto calibration. :( sorry (we tried/really wanted to!)
Manual Initial and or Re-Calibration should be a 1 time thing, but it seems maybe latest Beta messed it up for some. If this is the case, a quick re-calibration should get things back on track. We need to make sure this last calibration issue doesn't happen for non-beta/release versions of firmware for sure.
Click to expand...

To tell you the truth I could give a damn about some hot pixels. What I really want is reliable exposure and contrast. As I mentioned, the hot pixel thing only showed up yesterday and I haven't had time to fully explore beta 27 as yet.
 
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
 
jokiin 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
Click to expand...

Thanks. As I've said, although I reported it I don't see some hot pixels as being that big a deal. I was just surprised that it popped up all of a sudden. I would much prefer to have reliable good exposure across a range of lighting conditions.
 
Dashmellow said:
Thanks. As I've said, although I reported it I don't see some hot pixels as being that big a deal. I was just surprised that it popped up all of a sudden. I would much prefer to have reliable good exposure across a range of lighting conditions.
Click to expand...

yourself and @harsh have mentioned some hot pixels coming back, from memory though harsh said it showed up in one AE setting but not the other, not sure if the AE setting used when calibrating has any impact on calibration results, as you know calibration is just a process to map to memory any pixels which are hot and disable them at appropriate times so I'm not sure if which table is used at the time of calibration has any impact on the calibration results or not and if this would cause differing results depending on which table was in use at the time of calibration compared to what is used for recording, that will require some checking and probably some experimentation to see if that's even a possible cause/effect, could be something else totally, not really sure as yet but will be looking into it, I just spent 7 hours with the engineers yesterday going through upcoming changes with the engineers so will add this to the list
 
FYI - The hot pixels are more prominent with low-cut, the lower hot-pixel isn't as visible with centre and full-frame.
 
I'm getting some hot pixels on my 18 month old SG9665GC, as seen here:


Can someone inform me how to re-calibrate to fix this?
 
Curious, but why do calibration instructions need to be private? Do we need like a special code for each device?
 
ChibiSqueeze said:
Curious, but why do calibration instructions need to be private? Do we need like a special code for each device?
Click to expand...

no, we just like to make sure that anyone that does it is stepped through the process so that it is done right, also if it was just public we'd have no feedback on how many people have needed to do this
 
I've noticed some hot pixels on my 1 week old GC that came with beta 27 installed.

I'd prefer that the calibration was checked before it was shipped. I don't mind too much, assuming that the recalibration works when I get a chance to do it.

EDIT: It's just one hot pixel, and some dust on my monitor! :oops: I can live with that - not worth recalibrating IMO.
 
Last edited:
TonyM said:
I've noticed some hot pixels on my 1 week old GC that came with beta 27 installed.

I'd prefer that the calibration was checked before it was shipped. I don't mind too much, assuming that the recalibration works when I get a chance to do it.
Click to expand...

Instructions sent. Check PM inbox
 
jokiin said:
no, we just like to make sure that anyone that does it is stepped through the process so that it is done right, also if it was just public we'd have no feedback on how many people have needed to do this
Click to expand...

How many have needed it?

I too have wondered why you don't just post a thread about the issue and provide the calibration instructions for this very well known well documented long term problem.

What exactly do you mean "we'd have no feedback on how many people have needed to do this"?.

There have been many, many reports of hot pixels with the V2 camera and the truth of the matter is that you and @Street Guardian always try to act like you really don't know or don't want to admit that this is an ongoing issue that has been happening for many many months for a wide swath of owners. We were clearly told repeatedly that the V2 version of the camera included a cure for the hot pixel issue that manifested with the V1 and yet there has been an apparent concerted effort to downplay, and marginalize any and all reports that the problem persists, perhaps even worse than it did with the V1.

To quote what I said in this thread back on October 18th:

"Your main goal always seems to be to marginalize any problems that come up and sweep them under the rug regardless of how well documented they may be or how long they have been going on."

This is essentially the same approach that happened with the High Contrast/Dynamic range issue where nothing was actually acknowledged or addressed until a year after it was reported to you and thoroughly documented.

So now here we have @Street Guardian USA saying above, "but it seems maybe latest V2 Beta messed it up 'for some'. (Yup it must be just with the "latest beta?)

And you saying above, "I don't know if there has been any mistake that has overwritten the calibration info". (A recent mistake? with beta 27?)

These comments clearly imply that this is some recent manifestation that you have been previously unaware of that "may" have been introduced by the recent beta firmware updates. You really would have readers of this thread believe this might be a new problem mysteriously introduced by the recent beta?

But the facts about the hot pixel issue apparently just don't support this. One has merely to look at the numerous posts about hot pixels on the V2 version of the SG9665GC going back over half a year or more to see that you have been well aware of the pervasiveness and longevity of the problem.

So again, "What do you mean "we'd have no feedback on how many people have needed to do this"?. Of course you know there is a significant problem and you have for quite a long time.

As I've said repeatedly, the SG9665GC is basically a great camera with many good qualities but one with some significant long term issues that have badly needed resolution. Attempts to downplay and deny problems for fear of jeopardizing sales is a strategy that can only work for just so long and when the Emperor turns out not to be wearing all of his clothes, he loses credibility. That is the risk here for the long term.

I do apologize, jokiin as I don't wish to be adversarial but it is hard to ignore the truth of all this.

The truth of the matter is that when I got home this evening and reviewed some footage and I couldn't believe how bad the hot pixels were. It seems as if they are getting worse but I can't tell for sure. I will try yet another calibration.
Anyway, I got curious about this and decided to scan the forums for reports of hot pixels and here are only "some" of the results I found which I have listed in the links below. I was actually rather surprised to discover just how many complaints of hot pixels there are and for just how long they have been reported.

Then when I saw the most recent post about, "if it was just public we'd have no feedback on how many people have needed to do this", as if there is some sort of mysterious obscure problem you are trying to get to the bottom of I felt compelled to post this response.

Hot pixel reports:

New SG9665GC V2 has Hot Pixels

Hot Pixel Recalibration Help
https://dashcamtalk.com/forum/threads/hot-pixel-recalibration-help.19744/#post-257547
SG9665GC v2 Hot/stuck pixels + video quality issue

My SG9665GC has about 10 hot pixels of varying brightness that are visible in moderate to low light conditions

SG9665GC V2 got hot pixel
https://dashcamtalk.com/forum/threads/sg9665gc-v2-got-hot-pixel.20720/
Hot Pixels - Related to F/W Version?

Hot pixels














 
Last edited:
you can take your tin foil hat off, in total we've given out the instructions around 30 times, there are probably people out there that have hot pixels that aren't bothered about it so likely there are more that, if we just posted the instructions publicly we'd have no idea how many have found this problematic enough to want to adjust it

as an FYI all sensors have this, Sony acknowledge that it is impossible to avoid based on current manufacturing technology and consider it to be normal, there are other variables as well, see below

the calibration process (calibration is probably an inaccurate description but that's how it is termed) does not actually calibrate the sensor but masks the effect by mapping which pixels show this issue and when the pixels around the affected pixel are dark or light (whichever the case may be) turn off the affected pixel, we have tried an automated process to dynamically mask the pixels as that is an available option in the SDK but we found that when we used it there was a negative impact on overall image quality, it does work fine on some other model sensors so maybe it's something that can be improved, I did question the engineer about this 2 weeks ago and the answer was still to do it manually as there hadn't been any progress made on the automated method

Sony doc.jpg
 
Tin Foil Hat? Facts are facts.

As @ChibiSqueeze wisely asked, why not just admit there's been an issue going back to the beginning and just post the calibration instructions for anyone who may need them rather than be coy about how prevalent the problem has been and for how long? That is the issue, not whether hot pixels are "always there".

 
Dashmellow said:
As @ChibiSqueeze wisely asked, why not just admit there's been an issue going back to the beginning and just post the calibration instructions for anyone who may need them rather than be coy about how prevalent the problem has been and for how long? That is the issue, not whether hot pixels are "always there".
Click to expand...


As I've mentioned already we want to keep tabs on how prevalent it is or isn't, how many people have issues with it etc, we also want to make sure it's done right and people are stepped through the process, the way we do it now we have a good idea of how often it comes up, how best to deal with any questions about the process etc, if we posted the details the first time it came up we would have no idea how many people had found issue with it, regardless of how you try and present it this has nothing do with being coy, trying to hide something or whatever, it's a public form, everyone can see anyway, you're making out like it's some big secret, even Sony acknowledge that it's a reality

However your 'issue' that you have a problem with is not that hot pixels happen it's that we didn't just post the details of how to calibrate publicly, really? If we had this info but didn't help people then I could appreciate your point but really I don't understand where you're coming from and why you think it's an issue that we prefer to know how many times something comes up, we can't improve things or know where we're at based on rumour and innuendo

A bit of fact checking (you're the 'facts are facts' guy here remember) but the question was

ChibiSqueeze said:
Curious, but why do calibration instructions need to be private? Do we need like a special code for each device?
Click to expand...

In case you missed it to which I answered

jokiin said:
no, we just like to make sure that anyone that does it is stepped through the process so that it is done right, also if it was just public we'd have no feedback on how many people have needed to do this
Click to expand...

He liked that post so I'm going to take a stab that he read and understood my response


Now when I read this question which I already answered I seem to have missed the bit where (according to you, the 'facts are facts' guy from earlier on) he quote "wisely asked, why not just admit there's been an issue going back to the beginning and just post the calibration instructions for anyone who may need them rather than be coy about how prevalent the problem has been and for how long?" Perhaps those are your words rather than his?

This is why I have a problem basing what we do on rumour and innuendo, even when things are written in black and white for all the world to see there are those that see what they choose to see
 
You must log in or register to reply here.
Back
Top