I have used Neatimage for quite a while, but I have never known if I was creating my profiles right (every time I updated).
When you bring in the target image, and click "auto profile with calabration target," it brings up a small square sample section, in the middle green color. Are you supposed to just go with that for a sample? If so, why have a multi colored target, and not just a green page? Or, are you supposed to change the size of the sample section to include the whole page?
It seems like if you just use the small sample size it chooses, you could create as good a profile using any picture that has some undefined background.
Thanks,
Les
Creating profiles
You just have to make sure (1) the target shot looks good (the target was shot correctly), (2) Auto Profile with Calibration Target selects a flat featureless part of the target (in the central square). The profiling itself including the noise analysis of the selected part as well as the rest of the target is done by Neat Image fully automatically. It does analyze the whole shot, all parts of the the target, so you don't need to do that manually. Quite easy, isn't it?
Hope this helps.
Vlad
Hope this helps.
Vlad
>>The profiling itself including the noise analysis of the selected part as well as the rest of the target is done by Neat Image fully automatically. It does analyze the whole shot, all parts of the the target, so you don't need to do that manually. <<
Okay, so you are saying that it analyzes the whole image - in this case, the whole test image - not just the little square selection area. So, what is the purpose of the little square selection area? Does it do something special in that area?
>>Quite easy, isn't it? <<
It is as simple as can be. It makes people suspicious because you provide profiles, making it seem that some people might not have the ability to do it. You only have to take pictures of the target using all the different profiles you regularly use once. It is very simple to keep a file of these images, and just run the profile work on them each time you upgrade. We have two cameras, and use about 4 different profiles each. It takes me only a few minutes to do all settings for both cameras.
I could never understand why people would take the time to search this site for someone else's profiles. It is very easy, and takes very little time. How can you be sure they did it right? I just always wondered if you needed to expand the target area for the whole thing. I think I have made my profiles both ways. Not sure I noticed a difference.
So, if it analyzes the whole image, what does it do differently in the little target box?
Thanks for the help,
Les
Okay, so you are saying that it analyzes the whole image - in this case, the whole test image - not just the little square selection area. So, what is the purpose of the little square selection area? Does it do something special in that area?
>>Quite easy, isn't it? <<
It is as simple as can be. It makes people suspicious because you provide profiles, making it seem that some people might not have the ability to do it. You only have to take pictures of the target using all the different profiles you regularly use once. It is very simple to keep a file of these images, and just run the profile work on them each time you upgrade. We have two cameras, and use about 4 different profiles each. It takes me only a few minutes to do all settings for both cameras.
I could never understand why people would take the time to search this site for someone else's profiles. It is very easy, and takes very little time. How can you be sure they did it right? I just always wondered if you needed to expand the target area for the whole thing. I think I have made my profiles both ways. Not sure I noticed a difference.
So, if it analyzes the whole image, what does it do differently in the little target box?
Thanks for the help,
Les
The selection box that you see (and can move) is used to select the area of the primary, rough analysis. After rough analysis, the fine-tuning analysis is applied and it uses the whole image. It is very important to do the primary analysis using a flat featureless area. Since it is not always possible to find such an area automatically (or at all), you are given the control and you can check and adjust the selection using the box. If the automatic selection is not very good, assist NI and select a better area to get a more accurate profile.
If you did noise analysis manually, not using Auto Profile, but using the Rough Analyzer and then Fine-Tuning Analyzer (which is possible when you work in Advanced mode (menu Tools | Advanced mode); the manual profiling procedure is described in this part of the user guide), you would also notice that building a profile is a two-step process:
1) rough analysis (of one manually selected area);
2) several fine-tuning analyses (of several manually selected areas).
The Auto Fine-Tune function automates the step 2 and allows to do fine-tuning in one click. Auto Fine-Tune automatically finds many flat featureless areas in the whole image and analyzes them all thus automating the whole step 2.
The Auto Profile function automates both step 1 and 2 together allowing to do all profiling in one click. It automatically finds an area to be used for rough analysis, analyses it and then applies Auto Fine-Tuning as described above.
So, normally you just click once and get a good noise profile. However, if you go into details (which may be useful when it is difficult to find suitable areas automatically) then you see that there are many small analyses in this process and the whole image is used to build an accurate profile.
I recommend to check the Device Noise Profile section of the user guide for more details on profiles and profiling process. I am sure you will find many useful tips there.
Hope this helps.
Vlad
If you did noise analysis manually, not using Auto Profile, but using the Rough Analyzer and then Fine-Tuning Analyzer (which is possible when you work in Advanced mode (menu Tools | Advanced mode); the manual profiling procedure is described in this part of the user guide), you would also notice that building a profile is a two-step process:
1) rough analysis (of one manually selected area);
2) several fine-tuning analyses (of several manually selected areas).
The Auto Fine-Tune function automates the step 2 and allows to do fine-tuning in one click. Auto Fine-Tune automatically finds many flat featureless areas in the whole image and analyzes them all thus automating the whole step 2.
The Auto Profile function automates both step 1 and 2 together allowing to do all profiling in one click. It automatically finds an area to be used for rough analysis, analyses it and then applies Auto Fine-Tuning as described above.
So, normally you just click once and get a good noise profile. However, if you go into details (which may be useful when it is difficult to find suitable areas automatically) then you see that there are many small analyses in this process and the whole image is used to build an accurate profile.
I recommend to check the Device Noise Profile section of the user guide for more details on profiles and profiling process. I am sure you will find many useful tips there.
Hope this helps.
Vlad