Signal Processing vs DSP System Toolbox -- Which One?

I think I might have the answer to my question, but I still want to put it out here in the community and get some experienced feedback.

I am a non-engineering type who is interested in signal processing functionalities from a holistic learning standpoint. Specifically, I am an finance professional who suspects that there may be some interesting tie-ins between signal processing and financial time series data analysis. It would be very easy for me to go for the financial suite of toolboxes and stay within that space. But I have a deep suspicion that the frequency domain may have myriad stories to tell about financial time series, and I am intrigued as such.

In short, my goal is to augment my finance intuition with a rich exploration of the frequency domain space -- signal processing and analysis, so to speak.

Additionally, I am a relative novice user of Matlab, and I thought that exploring the more engineering side of the product suites (e.g. signal processing) would deepen my learning of the Matlab environment and thus make my user experience much richer once I work with the financial suite of products.

Also, I am a hobbyist in photography and Argentine tango music, and thought doing signal processing and analysis of my photographic works and tango music library would be a fun engagement.

With the aforementioned as background, please share your thoughts about which toolbox would be the best fit and why. That said, please refrain from asking my any further questions of clarification, as everything that needs to be known is stated herein. If there is uncertainty and ambiguity, just caveat your response accordingly.

All well intentioned thoughts and feedback are deeply appreciated. Thanks :-)

Answer 1

[Star Strider]

The Signal Processing Toolbox is intended primarily for offline use, i.e. signals already acquired. The DSP Toolbox is intended for the design and realisation of real-time signal processing, eventually in hardware. For what I understand your purposes to be, I would suggest the Signal Processing Toolbox.

If you are a photographer, you will likely want the Image Processing Toolbox. I’m adding it to your associated products list.

I suggest you peruse the Academia — Student section of the MathWorks site. There is a link to books that you will probably find helpful in making full use of the Toolboxes you’re interested in.

You can download and experiment with trial versions of MATLAB software, so you can see what Toolboxes are most appropriate.

Answer 2

[Puneet Rana]

To add to Star Strider's answer above:

• DSP System Toolbox (DST) has more specialized filter design algorithms (e.g., multirate, multistage, adaptive filters etc) than Signal Processing Toolbox (SPT). In fact, SPT is a required product in order to use DST.
• DST has blocks that can be used to build systems in Simulink as well - SPT works in MATLAB only.
• You mentioned analyzing signals in frequency domain. SPT has functions to do this for offline data. DST has equivalent System objects that show you the plot, with options for measurement data from the plot, while streaming the data so you can observe the trend in time. In general, DST is more powerful for modeling and simulating systems for streaming applications.
• Although it may not be useful to you based on what you have mentioned about your application, SPT works with floating-point signals, and DST has algorithms for both floating and fixed-point.
• One more point since you mentioned your interest in audio. DST has features to read/write and play/record audio, as well as algorithms for audio signal processing.

To summarize, your choice should be based on the type of modeling, the type of data, and the type of algorithms you wish you use. As Star Strider mentioned, trial versions are a great way to explore this.