The paragraphs below describe the process I went through to create Spectrograms for Speech. Read on to learn about what started the idea, what I learned from my literature search, and how I chose a computer program that makes using spectrograms affordable and accessible in the clinical setting.
I first became interested in spectrograms in the Acoustics & Phonetics course I took as a post-baccalaureate student at Portland State University. I was a piano teacher by trade before I came to speech-language pathology, so I have always been interested in music and sound. When the acoustics unit was introduced, I was instantly “converted” to speech science. I was fascinated that sound could be translated into a consistent visual signal. My professor told us about PRAAT, a free downloadable program that can be configured for spectrograms. I went home, installed the program, and began recording and visualizing any and all sounds I could think of. I made a spectrogram of middle C on my piano, I pet my cat until he meowed loud enough for the microphone to pick up a signal, I subjected my husband to hours of weird vocal sounds I could create to see how the spectrogram was affected. I was hooked.
About a year after that Acoustics & Phonetics class, I got the opportunity to work with a middle school girl in my first clinical practicum. One of her goals was to work on vocalic /r/. She had been in therapy for six years prior to our meeting and had already tried all the “tricks” in the bag. But she had not worked with spectrograms. So we set out to visualize that magical third formant. Spectrograms, at this point, became more than making funny black lines run across the computer screen. This project is the product of my search for a practical tool that marries acoustic science with clinical application.I needed to relate the purpose of spectrograms–an abstract and foreign concept–to her in terms and ways that she would understand. We colored third formants and made imaginary pirates jump down the formant “plank.” It was a learning process for both of us. While our time together was limited and I was not able to fix /r/ for her permanently, spectrograms did, however, elicit a correct vocalic /r/ (albeit inconsistently) in more than one context. The next term, I was given yet another opportunity to use spectrograms to remediate vocalic /r/ with a middle school student. This time, I was able to develop a hierarchy of nonsense words that shaped /er/ from “crate.” He was highly successful and came to enjoy “computer time” during our therapy sessions. Again, our time was short, and the term ended with me feeling like so much more could have been done.
I decided that a practical clinical tool was needed if I was going to go on and use spectrograms in my clinical practice after I graduated. Up until now, I had been using my professor’s equipment and lab, which were no doubt excellent resources (albeit expensive and unportable). But I knew someday I would be out in the real world, and like many practicing clinicians today, I probably would not have the budget to purchase equipment that is used in university and research settings.
This project is the product of my search for a practical tool that marries acoustic science with clinical application. At this point, it is by no means complete as I anticipate making changes and adding information to this project in the months and possibly years to come. I welcome you to read on to learn more about the detailed process I worked through to bring you Spectrograms For Speech.
Before I could go out and develop a clinical tool that applies spectrograms for remediating speech, I first needed to make sure my idea was founded in strong evidence and theory. It was important to me that clinicians trust that using spectrograms for articulation therapy is backed by the best available evidence, or at the very least, grounded in solid theoretical foundation. You can read more about what I learned with regards to the history of the clinical use of spectrograms, empirical studies that demonstrated the effectiveness of spectrograms, and the theoretical underpinnings of spectrograms by clicking on the links below.
Once I learned that using spectrograms for articulation therapy was grounded empirically and theoretically, I set out to find a useful computer program that would suit the “everyday” clinician.
The two most important features of a spectrogram software program that directed my selection were 1.) real-time analysis, and 2.) relatively inexpensive (e.g., under $50) or free. It is important that visual feedback for articulation therapy is in real-time (i.e., scrolls across the screen simultaneously with the speaker’s productions) so that students can make modifications to their productions in the moment rather than wait to see the feedback and then make modifications accordingly albeit late when the sensations of the production have passed. It was also important that the program be affordable as school clinicians’ budgets are very tight. I, myself, also did not have an unlimited budget to try out every software program under the sun. For the most part, I considered only those programs that were free or were under $50.00 that could be trialed for free. Lastly, it was important that the software program be compatible on a PC computer. For one, my own home computer is a PC, so it would need to work on my computer first before I could test it out and recommend it. Secondly, most clinicians in the work setting have access to PCs; it makes sense to cater to the larger audience.
The flowchart below explains the process I went through to select a spectrogram program for this project. You may visit the links to learn more about each program that I have listed below:
No Real-Time Analysis