Measuring true value of shutter-speed overrides from average pixel luminance - page 3 - General Discussion and Assistance - CHDK Forum

Measuring true value of shutter-speed overrides from average pixel luminance

  • 27 Replies
  • 6153 Views
Advertisements
I have given-up with CHDK DNG and ImageMagick.

I am using Reyalp's tool and looking at the data with ImageJ's histogram function.
It indicates the median and other values and automatically only uses the area of the histogram that contains pixels to draw the screen image.
If CS4 can do that I have not found how.

Now, I need to make the histogram as narrow as possible by using a steady, monochromatic, very diffuse light source.

Problem is, even if I find the true shutter values for my A620 the same factors may not apply to other cameras if it depends on how quickly the mechanical shutter can close.

I have given-up with CHDK DNG and ImageMagick.

I am using Reyalp's tool and looking at the data with ImageJ's histogram function.
It indicates the median and other values and automatically only uses the area of the histogram that contains pixels to draw the screen image.
If CS4 can do that I have not found how.
As I said above, I'm mainly using DNG because I'm testing on the SX10.  Whatever you can get to work, as long as it is giving you the linear response expected.

Now, I need to make the histogram as narrow as possible by using a steady, monochromatic, very diffuse light source.
The histogram doesn't necessarily need to be very narrow, you just need to minimize over/under exposure.  But a narrow histogram does help to give a wider testing range.

If you can't diffuse evenly over the entire image, you can always crop a central smaller portion from the images that is within the even illumination area and get proper results (as long as the cropping is equal from image to image).

Problem is, even if I find the true shutter values for my A620 the same factors may not apply to other cameras if it depends on how quickly the mechanical shutter can close.

This is true, absolute shutter (and aperture) values achieved are really only valid for the model tested.

Interestingly, I have also found that in really fast shutter settings, there is some variable functionality:  For example, when I had previously done some shutter tests, I could reliably get exposure variation down to about 1/32000 (whether it is actually 1/32000 is yet to be determined) and typically beyond that, the exposure stays somewhat constant.  But occassionally at faster shutter speeds, a single shot will show up that does have a lower exposure.  I find this can be sometimes found if you set a static shutter speed and then take a series in burst mode.  This may mean that occasionally, the camera will actually achieve a faster speed, but this is unreliable and can't be depended upon to be consistent.

*

Offline reyalp

  • ******
  • 12151
But occassionally at faster shutter speeds, a single shot will show up that does have a lower exposure.  I find this can be sometimes found if you set a static shutter speed and then take a series in burst mode.  This may mean that occasionally, the camera will actually achieve a faster speed, but this is unreliable and can't be depended upon to be consistent.
This has been noted before with extreme high speed overrides. Another thing that has been noted ISTR are shadows, where parts of the image are exposed different amounts. Av may have some influence on this.

One can reasonably assume that Canon puts in their limits for a reason.
Don't forget what the H stands for.

Measuring by pixel values won't work. Due to the wide range of sensitivity due to ambient temperature of the sensor while it is being used from shot to shot. This is the same problem people ran into when trying to figure the top ISO range of their cameras. The first shot after powering-up the camera might look like ISO1000, but by the 20th shot it might look like it was shot at ISO1800, though the ISO override value was never changed. The only reliable method would be by photographing something at a known rate of motion and then figuring out how much angular distance it has moved during the exposure. Whoever first proposed this pixel value method doesn't know much about these cameras.


*

Offline reyalp

  • ******
  • 12151
Whoever first proposed this pixel value method doesn't know much about these cameras.
CCD cameras are routinely used to make extremely precise measurements of absolute luminosity. See the page I linked earlier.

I'm not sure how big the effect of temperature is on CHDK cameras. barberofcivils results suggest that it isn't significant for a series of a few exposures, since the response over his tests almost precisely linear, but in any case we can monitor sensor temperature, and if necessary calibrate it by taking series of photos with identical settings. Or just wait for it to stabilize before you start testing.

If you want to post actual data, feel free ;)
Don't forget what the H stands for.

CCD cameras are routinely used to make extremely precise measurements of absolute luminosity. See the page I linked earlier.
I agree with that (I use them professionally for this purpose), but there is a difference between a camera that is made for such a task (a scientific CCD camera) and a consumer camera. Even if the sensor is identical, we can not be sure that there has been no pixel-processing in the raw-files.

I would also prefer to use a more direct measurement of exposure time (like taking a frame of something with a fixed and known velocity).

Measuring by pixel values won't work. Due to the wide range of sensitivity due to ambient temperature of the sensor while it is being used from shot to shot. This is the same problem people ran into when trying to figure the top ISO range of their cameras. The first shot after powering-up the camera might look like ISO1000, but by the 20th shot it might look like it was shot at ISO1800, though the ISO override value was never changed. The only reliable method would be by photographing something at a known rate of motion and then figuring out how much angular distance it has moved during the exposure. Whoever first proposed this pixel value method doesn't know much about these cameras.

Thanks for the input.  This is not meant to be a precise measuring method, but rather a way of verifying and quntifying the general magnitude of the higher shutter speeds. Just in the same way that when you set your shutter speed to 1/1000s, it is actually corresponds to 1/1024s, and the actual achieved speed may be different, but we at least know its in the general 1/1000s range and not 1/800s or 1/1200s.

Of course other methods would be more accurate, but this is also meant to be a quick and dirty method that can be done quickly and easily without special equipment, just approriate software for analysis.

Previously I had always tested shutters and apertures in this manner visually just to verify that an exposure difference was present, and that the override was working, and would just assume the speed/aperture corresponded to the setting.

While I do understand what you are trying to point out, I can point out that in the proper tests that I have done for both shutter and aperture, the results have shown a more-or-less linear response as expected.


*

Offline reyalp

  • ******
  • 12151
I would also prefer to use a more direct measurement of exposure time (like taking a frame of something with a fixed and known velocity).
For really accurate measurements I'd agree. At the very least, you'd have to do very precise characterization. However, what we are looking for is to tell whether one stop of commanded shutter speed results in roughly one stop change of raw values. I'm no CCD expert but it's very difficult to believe a few C of temperature variation would prevent us from doing this. From what I understand, the major effects of CCD temperature are dark current and noise, not actual sensitivity, and these are very minor contributors at the kind of shutter speeds we are talking about (it would be a serious issue if we were investigating the low end) There's no way a minor change in sensor temp results in a stop worth of sensitivity change. If it did, metering wouldn't work.

The raw values are what we are really concerned with anyway: that's what the camera itself uses to create the jpeg. If say 1/4000th->1/8000th is reliably one stop, that's all we need to know.
Don't forget what the H stands for.


 

Related Topics