Histogram with no binning for non zero buckets

Submitted by JHeger on

Forums: 

I tried to make histograms from the RAWs of a Fuji T4, a white picture, all pixels overexposed, and a black picture with lens cap.

I would like to have the bucket size to be one for all non zero values. Is there a way to do this?

I tried linear x axix and non-linear x axis. With linear x axis and non auto the minimum bucket size is always reset to 32. Auto also does not always give buckets size one for non zero buckets. With the black picture it did but only about 2 million pixels were counted (sum of R, G,and B channels.)

Is there any setting I have missed so far?

Actually I was surprised to see that the exported CVS values are influenced by the x axis setting. I had expected that only the histogram drawing would affected. But of course this is not a problem.

Is there a reason why in linear and non linear x axis setting the order of column is different?

LinearStart LinearEnd EVStart EVEnd count_R count_G count_ B

vs

Start EVEnd LinearStart LinearEnd count_R count_G count_ B

Anyone any idea?

Best
JHeger

Dear Sir:

Submitted by lexa on

Dear Sir:



Exported CSVs begins with primary (X-Axis) bin range and then secondary/optional bin.



Minimal histogram bin size is calculated based on current screen size: displayed histogram bin size should be 'several whole pixels' (1, or 2, or 3, etc) to avoid display aliasing.



So, if one wish to use single value bins, one needs to select min-max range small enough to fit on screen.

Also, you may export histogram 'raw data' via Save Full data button: it exports 'raw' histogram:  pixel value, 3/4 pixel data countts, so bin size is exactly 1.

"Save Full Data" does the

Submitted by JHeger on

"Save Full Data" does the trick. So now I get the data I want. Thanks for the help.

Best regards
JHeger

Fuji S3

Submitted by Salo Mihai (not verified) on

Hello!
Old fuji S3 demosaicing: S and R pixel problem solved? Thanks!

RawDigger displays S and R

Submitted by lexa on

RawDigger displays S and R frames as separate frames, use frame selector at the program bottom bar.

Thanks! But how to deal with

Submitted by Salo Mihai (not verified) on

Thanks! But how to deal with "prepare data for profiling" with the S3 S+R? Avg. green, white patch etc ( for the 2 frames...?)

1st (normal exposed) frame

Submitted by lexa on

1st (normal exposed) frame data should be enough. Even if your raw processing software merges two sub-images: the normally exposed one is most weighted so color conversion in your software should be based on this frame color.

Thanks! What about the

Submitted by Salo Mihai on

Thanks! What about the "normal" S+R demosaicing (12mp tiff)?

it's definitely not a

Submitted by lexa on

it's definitely not a RawDigger's task: this is a RAW data inspection tool, not a RAW converter.

Yes, of course. But I don't

Submitted by Salo Mihai on

Yes, of course. But I don't believe (with my limited knowledge) that building a profile without using R pixels (brightness related) can make a reliable profile...maybe you can enlighten me...thanks!

Add new comment