CHDKPTP - PC Remote Control Performance Analysis - page 6 - RAW Shooting and Processing - CHDK Forum
« previous next »
Pages: 1 2 3 4 5 [6] 7 8 9 10 11 ... 47   Go Down

CHDKPTP - PC Remote Control Performance Analysis

  • 465 Replies
  • 130417 Views
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #50 on: 02 / September / 2012, 10:44:50 »
Advertisements
Yes that was what I had read a couple of weeks ago: BADPIXEL removes what I understand as hot pixels and BADPIXEL.BIN removes dead pixels.  For my instrument, instead of using BADPIXEL for hot pixels, I do this by Dark Subtract ON because it provides the latest temperature condition of the CCD at the accepted expense of a 2nd exposure.  My question relates not to hot pixels or dead (0) pixels, but rather to marginal unresponsive pixels that may be near 0 but are not.  Could you address this question please?
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #51 on: 02 / September / 2012, 10:53:16 »
FLAG - minor CHDKPTP caveat ... When I press SHUTDOWN button should reflect the disconnection by reverting to CONNECT.  At present it stays as DISCONNECT.
Logged
*

Offline waterwingz

  • ******
  • 12140
    • Waterwingz
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #52 on: 02 / September / 2012, 11:55:42 »
Quote from: SticK on 02 / September / 2012, 10:44:50
My question relates not to hot pixels or dead (0) pixels, but rather to marginal unresponsive pixels that may be near 0 but are not.  Could you address this question please?

In  include/camera.h  there is this :

Code: [Select]
#define DNG_BADPIXEL_VALUE_LIMIT        0   // Max value of 'bad' pixel - this value or lower is considered a defective pixel on the sensor

This gets used in the function  raw_init_badpixel_bin()  in  core/dng.c  as a pixel limit value.

A quick grep of the trunk show that only the ixus220_elph300hs ixus300_sd4000 & ixus115_elph100hs change this in their platform_camera.h configuration file.

Code: [Select]
#undef  DNG_BADPIXEL_VALUE_LIMIT
#define DNG_BADPIXEL_VALUE_LIMIT        16      // Camera sensor apparently has bad pixels with value > 0 & <=16
So if you tweak this value in your build,  you apparently can raise the "dead pixel" threshold.  Unfortunatly that's a build option and not changeable by the end user.  Might be something to add to the already complicated RAW/DNG menu ?

HTH ?

Logged
Ported :   A1200    SD940   G10    Powershot N    G16
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #53 on: 02 / September / 2012, 12:20:10 »
We have a problem.  I don't see the "No Image" corruption yet.  SD passes chkdsk too.

Conditions: DNG 1.1, Man bad px rem OFF, Dark subtract OFF.  BADPIXEL.BIN present.

I am back to the point of yesterday before we started looking into the "No Image."  This *looks* like it's related to the new shoot command resulting in corruption of the DNG file.  Because there are multiple issues the interactions are somewhat confusing and I haven't got a good handle on it yet. 

Here's what I can tell you for now.

==>> There is at least one shoot parameter scenario that produces a corrupt DNG file.  I detect the corruption in RAWTherapee 4.  The thumbnails and EXIF data show up but when I open the file it does the fast demosaic and then freezes completely requiring a kill.

Camera is set to M, 1/15, f2.8, ISO 100.

  a) Press SHOOT BUTTON.  No DNG corruption.

  b) shoot -tv=4 -av=2.8 -sv=200 :  DNG corrupted.

  c) shoot -tv=1/15 -av=2.8 -sv=100 (to mimic camera settings): No DNG corruption.

This is what I found yesterday.

SECONDARY PROBLEM ...

The EXIF (Tv Av ISO) is not completely accurate in any of these, especially ISO // it's really off sometimes.  For example, I've noticed that if I give shoot -sv=100, the EXIF ISO is 164 and the image is indeed brighter that pressing SHOOT with Canon set to 100.  So there's something weird here too.

@waterwingz
Very very nice to know it's possible // thanks // please allow me to look into this later and will make more sense once the new CCD is installed in the instrument.
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #54 on: 02 / September / 2012, 13:16:29 »
DNG Corruption  SX110 test

I set up the SX110 as the S90:

  a) Press SHOOT BUTTON.  No DNG corruption.  EXIF ISO=100 // OK

  b) shoot -tv=4 -av=2.8 -sv=200 :  DNG corrupted, EXIF ISO=329

  c) shoot -tv=1/15 -av=2.8 -sv=100 (to mimic camera settings): No DNG corruption, but ISO is 164.

I think this a very important test to consider: same behavior across both cameras.
Logged
*

Offline waterwingz

  • ******
  • 12140
    • Waterwingz
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #55 on: 02 / September / 2012, 13:46:58 »
Quote from: SticK on 02 / September / 2012, 13:16:29
  c) shoot -tv=1/15 -av=2.8 -sv=100 (to mimic camera settings): No DNG corruption, but ISO is 164.
Sounds like part of this is the "real" vs "market" ISO discussion that has been going on for years on this forum ?

http://chdk.wikia.com/wiki/CHDK_User_Manual#Override_ISO_value
Logged
Ported :   A1200    SD940   G10    Powershot N    G16
*

Offline reyalp

  • ******
  • 14111
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #56 on: 02 / September / 2012, 13:56:09 »
Quote from: SticK on 02 / September / 2012, 12:20:10
==>> There is at least one shoot parameter scenario that produces a corrupt DNG file.  I detect the corruption in RAWTherapee 4.  The thumbnails and EXIF data show up but when I open the file it does the fast demosaic and then freezes completely requiring a kill.
Can upload a sample of the "corrupt DNG" ?
Quote
The EXIF (Tv Av ISO) is not completely accurate in any of these, especially ISO // it's really off sometimes.  For example, I've noticed that if I give shoot -sv=100, the EXIF ISO is 164 and the image is indeed brighter that pressing SHOOT with Canon set to 100.  So there's something weird here too.
ISO difference is probably "real" vs "market", which I explained at least once before. 100 real -> 164 market is a pretty typical delta. For other parameters, some minor variation between Canon display values and actual values is expected. E.g what displays as 1/125th could really be ~1/128th etc and different values could show up in different places.
Logged
Don't forget what the H stands for.
*

Offline reyalp

  • ******
  • 14111
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #57 on: 02 / September / 2012, 13:58:58 »
Quote from: srsa_4c on 02 / September / 2012, 08:19:49
- use CHDK's File Browser to rename a file (you can find it in the "local menu" under the menu item "More"). The local menu can be reached by pressing "left". You may find the rename box a bit complicated to use (I do). After the rename operation, check for corruption.
Alternately, you could use chdkptp
=return os.rename('A/FOO','A/BAR')

Or run llibtst.lua, which exercises all the OS and IO functions.
Logged
Don't forget what the H stands for.
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #58 on: 02 / September / 2012, 14:42:45 »
"Corrupted" DNG from S90
http://www.sendspace.com/file/y3lr7q

Quote "ISO difference is probably "real" vs "market", which I explained at least once before."

I know you have explained it to some degree.  But, there is inconsistency in user usage.  Also I explained before ... differently this time: that I have noticed when I change the "market" value say by 1 f-stop in the shoot command, the "real" value (the one you are referring to in ISO given by CHDK)  remains unchanged.  That's what puzzles me.  I recall talking about this before. 

Next in that respect ... is CHDK doing what I am seeing?  Again ... if I SHOOT at Canon 100, and then want the same exposure with {shoot -sv=100} I see a brighter image, corresponding one could estimate by eye to the EXIF value of 164 ??  I don't object to "market" vs "real", but ... I am having trouble reconciling the usage discrepancy.  Are you saying then that shooting with Canon ISO 100 results in different intensity than shooting Nikon ISO 100?  Do you see my position? 

Useful for me then would a table then of "real" ISO values vs "market" values (that I can feed shoot and test exposures with) in steps of 1-stop that correspond to the VGA 2x gain steps, and validate everything in a consistent manner.  I think I've asked for this too, last week I recall,when you originally explained the issue??  Some (few) of my questions go unanswered (as probably some of yours too).  For example, what is the "real" value of Canon ISO 80 for the S90?  What is the "real" value of Canon ISO 80 for the SX110?  Are they different?  What is the "real" value of Canon ISO 1600?  Are the "real" values linear, ie +1 f-stop is twice the previous value, whatever it is?

I just need something consistent.

Quote " For other parameters, some minor variation between Canon display values and actual values is expected"

Yes that is what I see, so I take it to be normal.
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #59 on: 02 / September / 2012, 14:59:37 »
Re: the "corrupted DNG" I just sent you ... RAWTherapee 4 chokes on it, but, DCRAW -4 -T does appear to convert the file correctly.  I tried two files, so this could be a RAWTherapee problem, not yours, but please check.
Logged
Pages: 1 2 3 4 5 [6] 7 8 9 10 11 ... 47   Go Up
« previous next »
 

Related Topics
SimplePortal 2.3.6 © 2008-2014, SimplePortal