<blockquote id="quote"><font size="2" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by starjim</i>
<br />In reading this thread I see a couple of issues that really keep our feet in the mud and it is fast turning hard........
If I process an image on my own is it OK? How do I avoid the complaint of you introduced those artifacts........
We need to agree to agree somewhere.........
<br /> Lets set the parameter of just whose images are acceptable and leave it at that, Laney, Malin, NASA????? If I process an image on my own is it OK? How do I avoid the complaint of you introduced those artifacts...
<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">
starjim, (this is a belated response to a previous message)
Those are fair questions on a subject that we have been bantering back and forth for awhile.
I think there is a fairly straightforward answer to it, although I don't presume to know everything there is to know about the subject. Far from it.
Here are some of the criteria that I think should apply. I don't pretend this is complete or absolutely 100% right, it's just what I think about the subject.:
1. We have to start with the raw data, from some reliable source. In this case we're mostly dealing with MOC data from the MSSS website. If we don't start from the raw data, we have to be able to reproduce the starting image by some know set of steps. As PheoniX_VII said in a previous message, that could mean using a specific program and version. But the key is reproducibility. Plus, we should not add or remove data indiscriminately (that's pretty obvious), except when producing a "key" to be used as a reference to the real data. Or, as Tom has explained many times, we should not introduce "bias".
2. It has long been accepted that contrast and brightness adjustments are acceptable, because they are just making the composition easier to see, and not changing it's basic makeup. Saturation must be avoided in the process, though, because a saturated pixel contains no data.
3. Smoothing and averaging are useful, for the same reason as they are when charting something. They can help in seeing the big picture, without drastically altering the image. But they do need to be explained, because there is always a trade-off when doing this type of filtering. Detail is lost. Some of it may have been noise, and not really there, but some of it is real detail that is lost.
4. As we showed in the Nefertiti topic (and Tom explained) Histogram Equalize is a useful tool for spreading the data over the range of the eye, without introducing bias.
5. I tend to agree with Neil, that "negatives" or "brightness reversal" should not be used as the final product, anymore than you would sit around Thanksgiving Dinner with your family looking at negatives. They may serve some intermediary purpose, though, when trying to "eke" out detail in the image, as would certain more sophisticated techniques used when making measurements, such as Fourier transformations.
6. In general, I think "effects" should be avoided, for anything other than artwork or photography. There are so many wild effects available today, where would one draw the line? That's my personal opinion, there could be exceptions.
As to whether or not "you" did it, or "I" did it, and whether or not we introduced those artifacts, and how to defend ourselves, other than what I said above, I'd have to say that the starting point would be that we have to be able to explain what we did, and take it from there. That's easier said than done sometimes, but we have just as much right to try it as anyone.
rd
Topic Author
]