Home Products Download Order Contacts


Subject: Re: ACR Calibration

Compare two Canon 1Ds Mark II cameras.

The images were shot by John Cornicello, Adobe Systems Inc. using two different cameras, both Canon 1Ds Mark II. The images are HN1X9386.CR2 and LQ5E9384.CR2. The lighting was studio strobes. Thank you John.

The first two test shots have a very small difference in the exposure settings, so John was kind enough to shoot two more with synchronized settings.

I ran a calibration script against both images. The results were very similar but I believe the exposure was about 2/3 of a stop from where I think they should have been. I base this on the fact that the resulting brightness setting (91) is higher than I expected for this camera. I have access to other 1Ds Mark II daylight images where it comes out to about 40.

That said, the calibration results were very similar, but there were some differences in the color settings. Overall, the results correlated very closely. All patch errors and the summaries are very similar. The reports can be seen here: HN1X9386
and LQ5E9384 .

The next step was to synchronize the ACR settings for both images and run two more reports. The control image was LQ5E9384 and the settings were modified in HN1X9386. The first report simply shows the new delta values. This time the correlation is even better. The second report was obtained by modifying my ReadMacbethPatches script to use the other image as the target values. The result was fascinating. No patch value differs by more than 0.7%. This is within the range of errors introduced by having to use an 8-bit histogram to read the patch values. In other words, no discernable difference at all. I have never seen a calibration attempt come anywhere near this close.

As I mentioned the first two images had an exposure difference of 1/3 stop. The shutter speeds were 1/125 and 1/160 respectively. This test produced considerably more image differences than the second test. Some will argue that with strobes this will make no difference at all. But the typical strobe duration is between 100 microseconds and one millisecond. The longer duration is usually only at very low power settings. Thus with slower shutter speeds, there can be considerable ambient light included.

So my conclusion is that there is no discernable difference in the sensors in these two units. The differences introduced by lighting, camera settings, and ACR settings are far greater than any differences in these sensors.

Your mileage may vary.
Cheers, Rags :-)


View All Messages in adobe.photoshop.camera.raw

ACR Calibration =>


Copyright 2006 WatermarkFactory.com. All Rights Reserved.