[yurko_yr]

Update: Ricoh Japan requested today the combined darkframe collection.

That is good news on my opinion. They might finally take it seriously.
Yuriy

---------- Post added 11-23-2016 at 07:56 AM ----------

Hi,


So you think that "darker" frame just hide white dots? I am also thinking about the same thing, but I'm not so sure...Even some frames are a bit darker, the white dots are very bright, I think that they should be visible even if darker frame ...Maybe that sensor calibration makes "more" than just affects to native sensitivity (and frame darkness) ?


regards,


Ari

---------- Post added 11-23-16 at 03:36 AM ----------

...For example, look at this thread:
K1 and white dot issues (Same as the D810)


There are some white dot noise samples, and these white dots are very dominant, at worst. ..Then Nikon made some kind of software/firmware trick, and white dot noise almost completely disappeared, it was not calibration, nor that sensitivity would be now "darker" ...


...just my thoughts ;-)


Ari

I think it is something about difference in intensity between some pixels producing white dots and general signal. In some bodies this difference is smaller than in other resulting in normal behavior and sensors producing white dots.
Just my thoughts.
Yuriy

[MJKoski]

Those are my thoughts as well.

[JBox]

If to take JBox pictures from the first upload with ISO100, LENR off, 30 sec, boost exposure 4 stops up and convert to JPEG, it is possible to see some faint dots (200% crop).

Just an FYI, that ISO 100 dark-frame is from before my K-1 was recalibrated, so it is likely there is even less noise at ISO 100 now. I can do another ISO 100 dark-frame if you are interested, but I will probably not be able to before next week.

[MJKoski]

The White dots are not actually white. In my unit there were lots pixels which were fully saturated in red channel. OOC JPG files produced heavy dot show or Adobe converter which transformed those bright pixels into white pixels.

[Medex]

OK. I follow discussion in this thread. And this is my comparison of 4 different Pentax K-1 camera. The picture is below. The base picture is Medex-25600-30s-??*C done 10 days ago and I dont know the exact ambient temperature. But today I shot 4 more pictures with lens cap on, f22, 30 seconds at ISO 25600 in the dark room with OVF closed by hand. Other 3 samples are according the name. All samples where imported to the Adobe LR, WB Temp was set to 6000K, Tint to 0, Sharpening is Off, Color Noise Reduction is Off. Then all pictures were opened in PS as layers and different layers were exposed on the base picture using layer masks. Then the file was saved as 16-bit TIFF to Adobe LR, and Exposure was set to +3 eV. Then TIFF was exported to JPG and opened in Windows Picture viewer. Picture below is a snap (using Snipping tool).
Any thoughts?
P.S. I'm not satisfied with high-ISO pictures in my camera do to white dots. Maybe I expect too much?
P.P.S. Manufacture date of the camera is 2016-04-18. Production code is 2.2. Internal serial number is 2009913

[aikaarska]

OK. I follow discussion in this thread. And this is my comparison of 4 different Pentax K-1 camera. The picture is below. The base picture is Medex-25600-30s-??*C done 10 days ago and I dont know the exact ambient temperature. But today I shot 4 more pictures with lens cap on, f22, 30 seconds at ISO 25600 in the dark room with OVF closed by hand. Other 3 samples are according the name. All samples where imported to the Adobe LR, WB Temp was set to 6000K, Tint to 0, Sharpening is Off, Color Noise Reduction is Off. Then all pictures were opened in PS as layers and different layers were exposed on the base picture using layer masks. Then the file was saved as 16-bit TIFF to Adobe LR, and Exposure was set to +3 eV. Then TIFF was exported to JPG and opened in Windows Picture viewer. Picture below is a snap (using Snipping tool).
Any thoughts?
P.S. I'm not satisfied with high-ISO pictures in my camera do to white dots. Maybe I expect too much?
P.P.S. Manufacture date of the camera is 2016-04-18. Production code is 2.2. Internal serial number is 2009913

Hi,


I don't know for sure, but somehow I feel that we have gone too far - with these conclusions. Okey, I understand white dot noise issue. It's nasty for them whose K-1 sensor produces these white dots too much. Especially if you make a lot of long exposures. BUT these "unscientific" sensitivity level tests... I'm not sure how accurate these test even are, and how significant... C'mon: ISO 26500, 30sec and then push up +3 eV in PP...That's quite extreme, or what you think? What is meaning of these so-called differences in real life shootings? ...Now we just look at these ISO25600, 30sec+3eV noise level collages. And we wonder - we don't even know for sure what we are looking at.


Okey, in real life, if we're shooting the starry sky at ISO 25600, 30sec. Then we process RAW file in PP, and push up eV a bit...If a little bit more sensitive sensor has given a little bit more "brighter" image, then we don't have to raise eV so much - and that's it. Or what do you say, whether this make any sense?
However, as I wrote, the white dot noise is a different issue, isn't different issue ?

Ari


P.S. Another thought: we don't have any references to other cameras. We do not know how others cameras differ from K-1, at same exposure values - whether this is usual, maybe even a normal variation?

[JBox]

If to take JBox pictures from the first upload with ISO100, LENR off, 30 sec, boost exposure 4 stops up and convert to JPEG, it is possible to see some faint dots (200% crop).

I just went back and found that I have not posted an ISO 100 dark-frame here (I had assumed it was one of the initial ones I had posed). Is that one of awscreo's dark-frames?

[awscreo]

I just went back and found that I have not posted an ISO 100 dark-frame here (I had assumed it was one of the initial ones I had posed). Is that one of awscreo's dark-frames?

My files are marked TDK14667 to 4670. Two were iso 100, two iso 25600.

[Photator]

If you mean leaving Pentax forums completely you could just add the emails to spam, most modern systems should filter out all related emails.

Thanks, but I'd prefer to go leaving no trace.

---------- Post added 11-23-16 at 09:58 PM ----------

...but at least if I stay and there is no ban/block policy I now know I can say what I think.

[Medex]

My files are marked TDK14667 to 4670. Two were iso 100, two iso 25600.

I used for comparison one of two ISO 25600 files LENR off

---------- Post added 11-24-16 at 09:48 AM ----------

I just went back and found that I have not posted an ISO 100 dark-frame here (I had assumed it was one of the initial ones I had posed). Is that one of awscreo's dark-frames?

all frames are ISO 25600 30s dark frames acording to info in LR

---------- Post added 11-24-16 at 09:53 AM ----------

Hi,


I don't know for sure, but somehow I feel that we have gone too far - with these conclusions. Okey, I understand white dot noise issue. It's nasty for them whose K-1 sensor produces these white dots too much. Especially if you make a lot of long exposures. BUT these "unscientific" sensitivity level tests... I'm not sure how accurate these test even are, and how significant... C'mon: ISO 26500, 30sec and then push up +3 eV in PP...That's quite extreme, or what you think? What is meaning of these so-called differences in real life shootings? ...Now we just look at these ISO25600, 30sec+3eV noise level collages. And we wonder - we don't even know for sure what we are looking at.


Okey, in real life, if we're shooting the starry sky at ISO 25600, 30sec. Then we process RAW file in PP, and push up eV a bit...If a little bit more sensitive sensor has given a little bit more "brighter" image, then we don't have to raise eV so much - and that's it. Or what do you say, whether this make any sense?
However, as I wrote, the white dot noise is a different issue, isn't different issue ?

Ari


P.S. Another thought: we don't have any references to other cameras. We do not know how others cameras differ from K-1, at same exposure values - whether this is usual, maybe even a normal variation?

exactly, we have no reference. Maybe this is normal behavior of sensor. In real life many of us are using ISO 800-3200-6400 for celestian pictures. Maybe it would by reasonable to compare white dots at these ISOs?
Anyway I expected much clearer dark frames at ISO 1600-3200.
By the way, I have one frame at ISO 1600 with darker band at the bottom, but this band is absent in another frame at ISO 6400.

[aikaarska]

I used for comparison one of two ISO 25600 files LENR off

---------- Post added 11-24-16 at 09:48 AM ----------


all frames are ISO 25600 30s dark frames acording to info in LR

---------- Post added 11-24-16 at 09:53 AM ----------


exactly, we have no reference. Maybe this is normal behavior of sensor. In real life many of us are using ISO 800-3200-6400 for celestian pictures. Maybe it would by reasonable to compare white dots at these ISOs?
Anyway I expected much clearer dark frames at ISO 1600-3200.
By the way, I have one frame at ISO 1600 with darker band at the bottom, but this band is absent in another frame at ISO 6400.

Hi,


What do you mean exactly that you want clearer dark frames at ISO 1600-3200? Do you have white dot noise then - or what issues?


I ask this because some of us have also done ISO 3200, 120sec dark frame RAW tests. Have you done similar, too? It might be useful to see your ISO 3200, 120sec RAW samples also...One could compare then ;-)


regards,


Ari

[Medex]

Hi,


What do you mean exactly that you want clearer dark frames at ISO 1600-3200? Do you have white dot noise then - or what issues?


I ask this because some of us have also done ISO 3200, 120sec dark frame RAW tests. Have you done similar, too? It might be useful to see your ISO 3200, 120sec RAW samples also...One could compare then ;-)


regards,


Ari

Yes, I have 1 frame at ISO 3200 and 120 s, ant 1 frame at ISO 3200 and 182 s. Now I try to inspect those frames and what I'm thinking is that the white dots problem is overestimated. In normal conditions, when I inspect completely unprocessed RAW image in Adobe LR (Exposure 0 level, Clarity 0 level, No sharpening, No Color (chroma) noise reduction), there are maybe small amount of lighter (white) dots in the frame visible at 100 percent magnification, all other "dots" are colorful (red, green, blue) - it is actually chroma noise. But if I try to use standard setting of Lightroom (25 percent Color noise reduction), then chroma noise turns into Luminance noise. And after that if I try to make some Exposure or Clarity adjustment (increase), the problem with this Luminance noise is worsening. In case I postprocess the image of night sky a lot of false "stars" occur.
So, maybe in that situation we should use another postprocessing techniques (e.g. Deep Sky Stacker or similar) and not blame sensor for white dots problem?

[Medex]

here is the frame at ISO 3200 and 120 s. Left standard LR settings, right - Color Noise Reduction off, +3eV

[yurko_yr]

I just went back and found that I have not posted an ISO 100 dark-frame here (I had assumed it was one of the initial ones I had posed). Is that one of awscreo's dark-frames?

Sorry, yes, that was awscreo's file.
Yuriy

[aikaarska]

here is the frame at ISO 3200 and 120 s. Left standard LR settings, right - Color Noise Reduction off, +3eV

My similar sample looks like this: