I'm currently working in an area related to speakers. Particularly on ways to improve a speaker through some tabulated parameters. If say, i compare 2 speakers(one that is better than the other) and use an audio software to take the measurement. Having the frequency response of both and compare them in the time,frequency,time-frequency domain. Is there any way, with these parameters, gives me insight on how i can improve my speaker performance that was benchmarked against the better one?
Better? define, please. Do you have an anachoic space? Or are you dealing with in-room response? Real music or white noise or some other test tone situation?
Are you asking how you can make your existing speakers sound "better"? If the question is how would you modify an existing speaker, there are no answers to that question. Do you intend to rebuild the driver bit by bit, changing out the voice coil, spider, cone and suspension? Is this just one driver you intend to tweak or a speaker system with multiple drivers and crossover networks? Would you adjust the enclosure to make the system sound better?
I think the answer would be to remove your wallet from your pocket and buy the "better" speakers.
Hi guys, thanks for your input. I am actually conducting some measurements tests to look into the possibility of improving an existing speaker which was benchmarked against a commercialized speaker. I conduct the experiment in an audio room using a test tone(MLS signal) to get the impulse response of the 2 speakers under test. And from there, compare their responses such as frequency response, comparison in time,frequency and time-frequency domain.
I'm not sure if just by comparing their plots in the different domains and with the magnitudes obtained, can i make some sense into the correlation between the data and possible improvements schemes to the poorer speakers.
Or maybe in other words, by comparing the responses, will it gives me insights on where i should work on to improve the poorer speaker's performance.
Using an impulse response is simply one tool that shows a small portion of the speaker's response. With this tool alone you cannot do more than make subjective guesses at the whole story of the speaker's performance. What you are essentially asking us to do is provide you with an entire schooling in speaker design. This is outside the scope of this forum. You should investigate speaker building text books and web sites. There are forums better suited to answering your specific questions not the generalities you offer here. But keep in mind, 1) your tools are limited and 2) people spend their life learning how to subjectively appraise what they see on objective tests. This is not information that is easily translated to "do 'X' and 'Y' will automatically follow". Even with the sophisticated test equipment available, most of a speaker builder's success is still tied to their intuitive talent and sense of what sounds right and how to get there.