IT-8 Calibration for current version of DCPro
IT-8 Calibration for current version of DCPro
With the new camera workflows, is it still preferred to create profiles by shooting the target and calibrating or is it preferred to use the embedded profile? Currently the profile I had built with an earlier version doesn't seem to be as accurate as using the embedded profile, I was planning on redoing the profile by reshooting the target and more carefully lighting it. If it is still preferred to create a custom profile, are there any guidelines on how to light the target, I'm planning on using a single flash directly behind the target as far away as I can get it to make the light as even as possible. Should I be measuring any point in the target with the densitometer to insure I have properly exposed it to insure I have create the optimal target for your profiler?
Charles Bush
-
LSI_Rossee
- LSI Staff
- Posts: 303
- Joined: Mon Oct 13, 2003 12:14 pm
- Location: Kiel, Germany
- Contact:
Dear bushfoto.
A camera is really just a scanner up on end with a lens right? Oh, but lighting is different isn't it? That and other factors can make camera profiling a much different experience than scanner profiling.
Early camera profiling adventurers found that simply placing the profiling target in the scene can create a profile for use in that lighting condition.
But any other lighting condition, even as much as moving a light, can cause enough variance that a new profile is required.
A properly built camera profile, in combination with correct gray balancing, characterizes the camera system effectively for a wide range of lighting conditions. Occasionally a new profile is required when lighting is drastically different than the profiled lighting or a camera is particularly sensitive to infrared wavelengths.
When it comes to building quality camera profiles, you can never spend too much time setting up the lighting and target for capture. This is the process that makes or breaks the quality of the profile. Absolutely even lighting from a single light source is required. Using two light sources is murder for camera profiling. Any deviation in color temperature between the two lights and you will have a color bias across the target that will kill your profile.
A digital camera can pickup as little as 1/10th stop variation, so be very careful. Gray balance is the secret to using profiles in a wide range of circumstances. It effectively calibrates your camera for each lighting situation and is the key to profiles remaining valid. Gray balancing a camera can be done in several ways including automatically, in camera, in the camera software or in SilverFast DC Pro using the embedded profiles for your specific camera.
SilverFast DC Pro offers a single interface for a wide range of cameras and also supplies some "tweaking" tools for adjusting white balance and other image characteristics as the file is opened. What SilverFast DC Pro brings to the table is a great replacement for the bewildering and problematic software typically included with cameras.
But the secret behind which profiles to use really is which mode you are shooting in. If using the RAW mode of your camera an ICC profile is not needed as source gammut since the file contains the naked, raw, bare, unlogarithmitised data from your camera?s Chip, including Bayer matrix. The embedded camera specific profiles in SilverFast DC Pro assure that the grey balance is calibrated and accurate after the interpolation of the bayer matrix. That means you might as well convert directly into your RGB-workspace i.e. AdobeRGB or ECI_RGB.
When shooting JPEG or TIFF the camera takes care of the grey balance for you before writing the file. In that case I recommend creating an ICC-Profile with SilverFast for your Camera and lighting situation and assign this to your file as the input profile.
In a digital workflow situation I recommend to shoot raw with one image containing the IT8 chart, preferrably a test shot. Later when converting from Raw you can attach the ICC-Profile, created from the Raw image in SilverFast DC Pro for your lighting situation, to your 8-bit Tiffs or Jpegs as source profile.
I hope my "little" excerpt could clarify your issue.
Kind regards.
A camera is really just a scanner up on end with a lens right? Oh, but lighting is different isn't it? That and other factors can make camera profiling a much different experience than scanner profiling.
Early camera profiling adventurers found that simply placing the profiling target in the scene can create a profile for use in that lighting condition.
But any other lighting condition, even as much as moving a light, can cause enough variance that a new profile is required.
A properly built camera profile, in combination with correct gray balancing, characterizes the camera system effectively for a wide range of lighting conditions. Occasionally a new profile is required when lighting is drastically different than the profiled lighting or a camera is particularly sensitive to infrared wavelengths.
When it comes to building quality camera profiles, you can never spend too much time setting up the lighting and target for capture. This is the process that makes or breaks the quality of the profile. Absolutely even lighting from a single light source is required. Using two light sources is murder for camera profiling. Any deviation in color temperature between the two lights and you will have a color bias across the target that will kill your profile.
A digital camera can pickup as little as 1/10th stop variation, so be very careful. Gray balance is the secret to using profiles in a wide range of circumstances. It effectively calibrates your camera for each lighting situation and is the key to profiles remaining valid. Gray balancing a camera can be done in several ways including automatically, in camera, in the camera software or in SilverFast DC Pro using the embedded profiles for your specific camera.
SilverFast DC Pro offers a single interface for a wide range of cameras and also supplies some "tweaking" tools for adjusting white balance and other image characteristics as the file is opened. What SilverFast DC Pro brings to the table is a great replacement for the bewildering and problematic software typically included with cameras.
But the secret behind which profiles to use really is which mode you are shooting in. If using the RAW mode of your camera an ICC profile is not needed as source gammut since the file contains the naked, raw, bare, unlogarithmitised data from your camera?s Chip, including Bayer matrix. The embedded camera specific profiles in SilverFast DC Pro assure that the grey balance is calibrated and accurate after the interpolation of the bayer matrix. That means you might as well convert directly into your RGB-workspace i.e. AdobeRGB or ECI_RGB.
When shooting JPEG or TIFF the camera takes care of the grey balance for you before writing the file. In that case I recommend creating an ICC-Profile with SilverFast for your Camera and lighting situation and assign this to your file as the input profile.
In a digital workflow situation I recommend to shoot raw with one image containing the IT8 chart, preferrably a test shot. Later when converting from Raw you can attach the ICC-Profile, created from the Raw image in SilverFast DC Pro for your lighting situation, to your 8-bit Tiffs or Jpegs as source profile.
I hope my "little" excerpt could clarify your issue.
Kind regards.
IT-8 Calibration for current version of DCPro
Well it helps a bit but I'm still perplexed by a few things.
First, are the built in profiles within DC Pro for raw more accurate than what could be buiilt from your IT-8 target and software even considering the variations from camera to camera? Seems to me that there is probably enough variation between cameras that would make a well done custom profile worthwhile for raw camera files, unless I'm missing something. If I'm using a custom built profile that has been buiild using a properly calibrated white balance, will the white balance in DC Pro still be accurate or is it treated differenctly with the profiles you have embedded. Are there other things that are embedded with your current workflow that won't be available using a custom profile?
I understand the critical need for accurate white balance prior to shooting the target, the plan is to create a custom white balance using a grey card in place of the target prior to shooting the target. I also understand the need to very evenly light the target to insure building the best profile. If I had studio lights with a soft box I'd be using that, I don't so using a flash at the greatest distance is probably the best I can do. Like the last time I shoot the target and adjust the lights so there is very little difference in the rgb levles in the corners of the target. I also suspect the exposue levels are very critical to getting it right, particularly dealing with the highlight areas of the image. That's where I'm struggling, if I shoot the target what spot on the target should I be gauging the exposure by and what RGB levels should I be expecting when it's properly exposed so I can make several test shots bring it into silverfast, measure the levels and then adjust the exposure by moving the flash to just the right spot as measured in silverfast. Seems to me that this is very critical for getting it right and right now it is simply a trial and error thing with no known value to be shooting for. As you said a digital camera can differentiate up to 1/10th difference in exposure so it seems to me getting just the right amount of light is critical. A point on the target or a value for one of the squares of a Macbeth color checker would be most helpful, or some other means of accurately measuring the input values.
Finally while I rarely shoot JPG or TIF, it seems to me that by shooting the target in Raw mode and building the profile from that, the results could be very different due to the way the camera makes the conversion to TIF or JPG versus how Silverfast interprets the raw image. Seems that you would need to measure the target through the conversion not as a Raw file to take into account the characteristics of the camera.
It's probably not something you really want to get into, but a paper on how to best create a profile for a digital camera using your software down to the level of suggested equipment, shootig diagrams, target measurement, evaluation of the resulting profile, etc. would be very helpful.
First, are the built in profiles within DC Pro for raw more accurate than what could be buiilt from your IT-8 target and software even considering the variations from camera to camera? Seems to me that there is probably enough variation between cameras that would make a well done custom profile worthwhile for raw camera files, unless I'm missing something. If I'm using a custom built profile that has been buiild using a properly calibrated white balance, will the white balance in DC Pro still be accurate or is it treated differenctly with the profiles you have embedded. Are there other things that are embedded with your current workflow that won't be available using a custom profile?
I understand the critical need for accurate white balance prior to shooting the target, the plan is to create a custom white balance using a grey card in place of the target prior to shooting the target. I also understand the need to very evenly light the target to insure building the best profile. If I had studio lights with a soft box I'd be using that, I don't so using a flash at the greatest distance is probably the best I can do. Like the last time I shoot the target and adjust the lights so there is very little difference in the rgb levles in the corners of the target. I also suspect the exposue levels are very critical to getting it right, particularly dealing with the highlight areas of the image. That's where I'm struggling, if I shoot the target what spot on the target should I be gauging the exposure by and what RGB levels should I be expecting when it's properly exposed so I can make several test shots bring it into silverfast, measure the levels and then adjust the exposure by moving the flash to just the right spot as measured in silverfast. Seems to me that this is very critical for getting it right and right now it is simply a trial and error thing with no known value to be shooting for. As you said a digital camera can differentiate up to 1/10th difference in exposure so it seems to me getting just the right amount of light is critical. A point on the target or a value for one of the squares of a Macbeth color checker would be most helpful, or some other means of accurately measuring the input values.
Finally while I rarely shoot JPG or TIF, it seems to me that by shooting the target in Raw mode and building the profile from that, the results could be very different due to the way the camera makes the conversion to TIF or JPG versus how Silverfast interprets the raw image. Seems that you would need to measure the target through the conversion not as a Raw file to take into account the characteristics of the camera.
It's probably not something you really want to get into, but a paper on how to best create a profile for a digital camera using your software down to the level of suggested equipment, shootig diagrams, target measurement, evaluation of the resulting profile, etc. would be very helpful.
Charles Bush
Return to “DCPro (digital camera)”
Who is online
Users browsing this forum: No registered users and 1 guest