2
$\begingroup$

What I have got:

Scan227darkExportedGraph.png

Scan227ExportedGraph.png

(Partially removed the “demo” words. I’ve got only evaluation license for OriginPro, planning to buy it…)

  • An image, exported from a slide of a ppt presentation: BaselineCorrected1.png

(I removed non-English commentary from the slide.)

Notes:

  • The slide of the ppt presentation is not mine and currently I’m unable to contact the creator.
  • On the original ppt slide of BaselineCorrected1.png if I double click those embedded pics containing the red label “- Session001:Scan227 – matched TATP (tr…”, following message pops up: “The server application, source file, or item can’t be found, or returned an unknown error. You may need to reinstall the server application.”
  • Opening with notepad the .spc files from the First Defender, you can read at the beginning: “Session001:Scan227 - matched TATP (triacetone triperoxide) (0.000000 single)?”
  • I’ve been told that the black lines on these graphs on ppt slides – if the picture is containing the label “- Session001:Scan227 – matched TATP (tr…” - are the baseline corrected version of the red one, but I couldn’t get any information about the applied method, function, or computational way of that baseline correction. Here are my questions:
  1. Is it possible anyway, that the plots on the GraphOfScan227dark.oggu and GraphOfScan227.oggu correspond to the Session001-Scan227.jpg’s black line? I mean, if the actual scan is that black line, is it alright, that we see such (blue) graphs, after importing them to OriginPro, and making the plot? (For my eyes, they appear to be non-consistent in peak positions.)
  2. If the first answer is yes, please describe to me, what is the computational relationship between them? How can I “infer” the blue graphs from that black plot in the jpg file, and vice versa (what is it? A baseline correction, peak fitting, signal processing or something like that?)? Or am I maybe completely misunderstanding the process, and overlooked a basic step?
  3. Concerning the picture from ppt slide: can I assume, that the red line plots on the graphs labeled “- Session001:Scan227 – matched TATP (tr…”, are gained from the above provided .spc files or is it false? Are these graphs created by OriginPro? Finally: How could I reproduce them? I’ve got ready the above linked two .spc files: what kind of operation on these files would conclude to the same shape and sight of graph, what we see on BaselineCorrected1.png. My reason for that is that I’d like to do myself that baseline correction, depicted on those ppt slides.

Please, I’d highly appreciate any helpful answer and comment from You! Thanks a lot, in advance.

Response to Jon Custer's comment:

They sent me multiple files, reportedly from the same Raman scan conducted with First Defender. My question is: if you consider the black line of the top picture, and the blue “line” of the Scan227-dark graph; how is it possible, that these two graphs represent the same scan of the same material sample. And I doubt it, because of the obvious differences between these two plots. E. g. The top peak of the blue line at approximately 2500, where can You find that peak on the top pics black line, if they should be the plots of the same sample? Of course, maybe there is a standard operation I'm missing, but I tried many built in functions of OriginPro, and I never could convert the blue graph, to make it look even similar to the black graph. (Sorry if my language is confusing I hope I could clarify my doubts.)

$\endgroup$
6
  • 1
    $\begingroup$ Please use the built-in image uploader, not Google Drive links. Also, one question per post is optimal. Though I don't really see how this is a chemistry question, to begin with. $\endgroup$
    – andselisk
    Commented May 16 at 12:03
  • $\begingroup$ I've posted here because I was thinking this is analytical question related to Raman spectroscopy. Could You please advise me a more suitable place to this question? $\endgroup$
    – AbelBp
    Commented May 16 at 12:10
  • $\begingroup$ There is a lot going on, and mostly in terms of software-related issues and data interpretation. That's not exactly chemistry, but I'm afraid on other SE sites the chance of getting a response will be even lower. You applied a lot of effort writing this question, so I'd leave it as is (just fix the images, please). But ideally you might want not the answer, rather a discussion on issues like this, and some forum, or Reddit would be better than a Q&A site (which Stack Exchange sites are) since there is very little room for back-and-forth discussion and one question per post is advisable. $\endgroup$
    – andselisk
    Commented May 16 at 12:16
  • 2
    $\begingroup$ Thank You for yout patience, and really , I worked hard on this question. I'd be glad with any answer which brings me closer to the final understanding of this. I can provide more information/detail if needed. $\endgroup$
    – AbelBp
    Commented May 16 at 12:30
  • 1
    $\begingroup$ What is your question here? Is it whether or not the top picture, showing a match to a particular compound, is correct or not? $\endgroup$
    – Jon Custer
    Commented May 16 at 16:18

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.