ID Snapshot Tools 2 Spectra Author
4475 none Snowy Sky cfastie
4476 none Snowy Sky vertical blur (calibrate to display) cfastie

Notifications

Description  

cfastie vertically averaged a spectrum to smooth noise in the image, and re-plotted it. There is some noise reduction. I think this might work even better for a live capture, where the Y-axis represents time. This is a good candidate for a new macro... hopefully soon!

about 8 years ago, cfastie wrote:

I guess this wasn't the best test because I treated the image differently when I did the vertical blur. First, I cropped it to the colorful part (the black part to the left on the original is for calibration). Second, I had adjusted the levels on this image to brighten it and to bring down the blue channel a bit. The original image was not manipulated before it was analyzed (except cropping top, bottom, and right).

about 8 years ago, warren wrote:

Testing set comment email notifications... hold onto your pants!

about 8 years ago, warren wrote:

Also, i misread - here, the noise seems to be greater in the blurred one. But you're right; we could do a more controlled test. I do wonder if there is systematic noise that is common along the vertical axis of a photo which might not be in a time-averaged spectrum.

about 8 years ago, cfastie wrote:

The email notification worked. However, the link to the set in the email was http://spectralworkbench.org/set/show/172 instead of http://spectralworkbench.org/sets/show/172

about 8 years ago, cfastie wrote:

Both the noise and the signal are probably greater in the blurred image because I also brightened the image. So the intensity is greater and everything gets multiplied. I also don't really know what Photoshop is doing when it blurs the image.

Leave a comment

You must be logged in to comment.