Thank you Very good explanation , just wanted to know how can utilise this Spectrum analysis for automotive noise source analysis ? ( Mean from the noise data need to find the problematic frequency thus improving part quality) thanks in advance.
Hi, at 1.12 you say that you will select the stuff in waveform view. I couldn't see how you did that. I couldn't see which buttons you were using. Can you please explain again? Cheers :)
Waveform view is the default view that is displayed when you first load the audio file into audacity. He basically duplicated the track, and then switched the duplicated track (the bottom one) to spectrogram view. Both tracks are the initially the same (until he edits one or the other). In this case he simply used the original track (the top one that was still in waveform view), zoomed into the area he wanted to analyze, and then dragged left to right to select the area. Hope that makes sense!:)
is there a way to make audacity remember the plot spectrum window size? every time i open audacity and open plot spectrum I have to stretch the window out every time. it only remembers the spectrum window size if its the same session
Hi, I need your help. I read some physics paper and get a question. If i plucked the rubber band and record its sound to find the frequency that the rubber band vibrated. Why the paper choose the peak of plot spectrum and tell that is the frequency of the vibrated? Sorry for bad english
I like the black box explanation of the analysis algorithms. Completely useless except maybe to inspire some future audio software engineers but yet cute and fun.
Hello! I need help. So, I am not anywhere related to the field of physics or sound engineering. But I'm looking into audios and frequencies for a research. So for this I am trying to allot a particular frequency (probably something like fundamental or prominent frequency) to a portion of sound track (max 8-10 seconds). First of all, I am not sure if that is a correct thing and if we can actually allot one frequency for a 10secs audio track. Secondly, if something like that exists, then can I assume it is the tallest or highest peak seen in such 'plot spectrum' graph in Audacity? My terminologies might be wrong, I apologise.
Hi! I hope you can help me, I've started using Audacity because we want to record son ultrasonic vocalizations (of rats) so we want to count how many times do they emit certain frequencies. We read in some articles they use Audacity to analyze it, but we are clueless, is there any chance you know how?
Either show a spectrogram (with appropriate settings) and count the number of times the power enhances at the frequency you're interested in or filter the data to the frequency band and count the number of individual wave packets?
Thanks for the video! Do I understand correctly: the spectrum analyzer shows, how to say, what frequencies have what level? I came up with a problem and thought this analyzer could help me localize the problem, but actually did not fully understand what you were saying. The problem is with my microphone: I try to overcome the keyboard been heard when recording videos as much as I possibly can. If you don't mind I will describe what I've done so far to better understand what I want :) I've covered the backside of my micro (which is facing the keyboard) with some kind of soundproofing material. The sound of keyboard changed significantly (bass disappeared), but typing is still heard, though quieter. Then I fully covered the micro with that material and nothing changed compared to the previous shenanigan, after which I came to conslusion, that not the sound goes to micro through the open area, but this material still passes through some frequencies of the keys been pressed. So, I want to find out what frequencies still pass through this material for better understanding how to block them, preferrably by searching for another material. Can spectrum analyzer help me in this question? P.S. Btw, I have one wish for your video: could you please increase the frame rate of screen-recording? It really hurts to see the slide-show :) Anyway, everything else is OK to me :) Thanks!
AmegaPlay yes that’s right, that’s what a spectrum shows. If you had audio with and without the soundproofing of nearly identical keystrokes you could calculate a transfer function - how the soundproofing suppresses the levels as a function of frequency. Sorry about the frame rate. I did set it to be 25fps but for some reason it turned out much lower and I didn’t have the time to redo these guides. My channel usually does videos of a different style but these were to help students on a project I run.
does the Audacity s/w have a mixer function ( non-linear kind ) , like radios use ? and can the s/w operate on a signal that is coming in real-time and not a recorded file ?
I would like to use the plotted spectrum and use it as a sort of frequency filter to apply it to another sound file. To put it another way, I want two audio clips to share the same characteristics without having to do the work manually with Audacity’s Filter Curve EQ effect. It’s probably impossible, but can anyone point me in the right direction?
Hi Martin, Hope you're doing well & safe! Thanks for the detailed explanation (this is going to be a long read, apologies) If I must use this data and improve my recorded audio (in this context, I'm referring to my voice modulation); how do I focus on certain frequency groups and modify them to sound more natural? My voice naturally is quite balanced as it is, with an inclination towards base frequencies. If I want to explore this further and improve my voice modulation, how do I read and make necessary modifications to my audio file, please? Also, is it possible to save certain settings as a preset so that I could remember those values and apply in whichever DAW software I might be choosing to work with? I know I have posted a lot of questions; however, I'm very much thankful to you and the others who have been posting such detailed tutorials online! Cheers from Bangalore, India!
@@DrMartinArcher thanks so much for your quick response! Any particular frequencies I must be focused on, if the voice must sound real? How do I apply those settings during a podcast or maybe while streaming live? :) Sorry if my questions are too basic/stupid.. Cheers!
If I want to see the individual note to, let's say, the opening chord to The Beatles "A Hard Day's Night" can you do that in Audacity? If you have any videos demonstrating this principle, can you direct me?
You can use the same tool, just switch over to the Pitch Correlation and the biggest peak will be at the fundamental frequency of the note. Audacity I don’t think translates that into a note (ie a letter) like Audition does, but such lists of notes and their frequencies can be found online.
Dear Martin, I need your help. I construct string instruments and I use audacity to define the frequency of the wood (soundboard alone) and in different steps till the instrument is completed (without armed with strings). The procedure is the following (i.e for the free soundboard) : I keep the soundboard from an antinode point and I tap it with my figure. The signal is recorded through a microphone to audacity and the analysis of spectrum, gives several peaks of different size of dB in different frequences. By tapping the soundboard, I define with my ear the corresponding note (A ,or D, or E,..- according to the thickness of soundboard and some other parameters). I expect that the first (higher dB value) peak of the spectrum will be the same note of what I hear. But it doesn't happen every time. Can you please give me an explanation? Many thanks.
The largest peak in a standard spectrum won't always be the fundamental frequency since higher harmonics will be present in different amounts depending on the instrument. It's best to use the autocorrelation methods to mitigate this. Hopefully they work better for you!
while trying to code something (alike to birdgenie) that would take an audio file and identify bird species by sound, I ran into the fourier transform. My maths not being great, which is an understatement, my attempts stranded. Would any of the methods you show in your video have the same accuracy? Could they be an easier or even better alternative to the fourier transform?
Hi Martin, I have a recording of someone's speaking voice. The sound is relatively stable (no shouting, whispering, background music etc just 'normal' talking). Is the peak of the highest wave (let's say 130 Hz) = Main frequency of the voice?
Be careful with your terminology. Pitch is the same as frequency or musical note, has nothing to do with volume - you can shout, whisper, or talk with roughly the same pitch. Also when you’re looking at the spectrum, what’s plotted is the power (same as intensity or amplitude squared) at each frequency when you consider the overall waveform as a sum of pure tones. So there is no “highest wave” but you can have a peak in power. The highest peak in power will be the dominant frequency or pitch of the audio, yes.
@@DrMartinArcher By 'highest wave', I mean the highest point on those purple graph (idk the correct word lol). Anyway, thanks for the clear explanation :))
@@DrMartinArcher Martin, the said audio has the highest power in 130 Hz, but also has the second highest power in 230 Hz (the purple graph has almost the same height as the 130 Hz one), what's that mean?
Hi can someone tell me how I can get the spectrum (Hz) from a track, for example from a meditation track. I need to know how many hertz this track is, is there any program that calculates it?
This video shows you how to get the spectrum. A spectrum does t give you just one frequency value, it tells you how much power (amplitude squared) there is at all frequencies. Perhaps you’re looking to get the tempo of the track out instead?
528Hz music (and some other purported special frequencies) is a weird myth and the number is not used in any proper scientific way / definition www.medson.net/528hz-myth.html
@@DrMartinArcher I like gtoroja am looking to analyze a song in order to find out if it indeed is tuned to the common 440hz or something else like say 432hz. I'm changing the hz and playing w the speed and want to confirm once I'm done everything that the song is tuned to 432hz. Thanks for any help you might be able to provide. Have a great day
Thank you ,i will use it to analize the frecuencies that make up a drop of water sound
Thank you Very good explanation , just wanted to know how can utilise this Spectrum analysis for automotive noise source analysis ? ( Mean from the noise data need to find the problematic frequency thus improving part quality) thanks in advance.
Hi, at 1.12 you say that you will select the stuff in waveform view. I couldn't see how you did that. I couldn't see which buttons you were using. Can you please explain again? Cheers :)
Waveform view is the default view that is displayed when you first load the audio file into audacity. He basically duplicated the track, and then switched the duplicated track (the bottom one) to spectrogram view. Both tracks are the initially the same (until he edits one or the other). In this case he simply used the original track (the top one that was still in waveform view), zoomed into the area he wanted to analyze, and then dragged left to right to select the area. Hope that makes sense!:)
@@bullpuppy7455 Thanks! :)
Can I see the spectrum plot for just the left or right channel of a stereo recording rather than a single, summed plot?
I think if you just select only one channel then it will plot the spectrum for just the channel and times selected
@@DrMartinArcher Ah, ok I will split the tracks and see if that works. Thanks for the super fast answer!
is there a way to make audacity remember the plot spectrum window size? every time i open audacity and open plot spectrum I have to stretch the window out every time. it only remembers the spectrum window size if its the same session
Hi, I need your help. I read some physics paper and get a question.
If i plucked the rubber band and record its sound to find the frequency that the rubber band vibrated.
Why the paper choose the peak of plot spectrum and tell that is the frequency of the vibrated?
Sorry for bad english
😭 I am struggling just a little bit with this subject, not gonna lie! 😭
Thanks sir
how to remove noise from videos without losing sound quality
I like the black box explanation of the analysis algorithms. Completely useless except maybe to inspire some future audio software engineers but yet cute and fun.
Very good explanation. I think this is just the thing I am looking for to analyse some audio signals. Thanks 😀 👍
Hello! I need help. So, I am not anywhere related to the field of physics or sound engineering. But I'm looking into audios and frequencies for a research. So for this I am trying to allot a particular frequency (probably something like fundamental or prominent frequency) to a portion of sound track (max 8-10 seconds). First of all, I am not sure if that is a correct thing and if we can actually allot one frequency for a 10secs audio track. Secondly, if something like that exists, then can I assume it is the tallest or highest peak seen in such 'plot spectrum' graph in Audacity?
My terminologies might be wrong, I apologise.
Hi! I hope you can help me, I've started using Audacity because we want to record son ultrasonic vocalizations (of rats) so we want to count how many times do they emit certain frequencies. We read in some articles they use Audacity to analyze it, but we are clueless, is there any chance you know how?
Either show a spectrogram (with appropriate settings) and count the number of times the power enhances at the frequency you're interested in or filter the data to the frequency band and count the number of individual wave packets?
Thanks for the video! Do I understand correctly: the spectrum analyzer shows, how to say, what frequencies have what level? I came up with a problem and thought this analyzer could help me localize the problem, but actually did not fully understand what you were saying. The problem is with my microphone: I try to overcome the keyboard been heard when recording videos as much as I possibly can. If you don't mind I will describe what I've done so far to better understand what I want :)
I've covered the backside of my micro (which is facing the keyboard) with some kind of soundproofing material. The sound of keyboard changed significantly (bass disappeared), but typing is still heard, though quieter. Then I fully covered the micro with that material and nothing changed compared to the previous shenanigan, after which I came to conslusion, that not the sound goes to micro through the open area, but this material still passes through some frequencies of the keys been pressed.
So, I want to find out what frequencies still pass through this material for better understanding how to block them, preferrably by searching for another material. Can spectrum analyzer help me in this question?
P.S. Btw, I have one wish for your video: could you please increase the frame rate of screen-recording? It really hurts to see the slide-show :) Anyway, everything else is OK to me :) Thanks!
AmegaPlay yes that’s right, that’s what a spectrum shows. If you had audio with and without the soundproofing of nearly identical keystrokes you could calculate a transfer function - how the soundproofing suppresses the levels as a function of frequency.
Sorry about the frame rate. I did set it to be 25fps but for some reason it turned out much lower and I didn’t have the time to redo these guides. My channel usually does videos of a different style but these were to help students on a project I run.
does the Audacity s/w have a mixer function ( non-linear kind ) , like radios use ?
and
can the s/w operate on a signal that is coming in real-time and not a recorded file ?
I would like to use the plotted spectrum and use it as a sort of frequency filter to apply it to another sound file. To put it another way, I want two audio clips to share the same characteristics without having to do the work manually with Audacity’s Filter Curve EQ effect. It’s probably impossible, but can anyone point me in the right direction?
Hi Martin,
Hope you're doing well & safe!
Thanks for the detailed explanation (this is going to be a long read, apologies)
If I must use this data and improve my recorded audio (in this context, I'm referring to my voice modulation); how do I focus on certain frequency groups and modify them to sound more natural?
My voice naturally is quite balanced as it is, with an inclination towards base frequencies.
If I want to explore this further and improve my voice modulation, how do I read and make necessary modifications to my audio file, please?
Also, is it possible to save certain settings as a preset so that I could remember those values and apply in whichever DAW software I might be choosing to work with?
I know I have posted a lot of questions; however, I'm very much thankful to you and the others who have been posting such detailed tutorials online!
Cheers from Bangalore, India!
Prabhath MP sounds like you want to do some EQ at the very least, perhaps also multiband compression. Have a play!
@@DrMartinArcher thanks so much for your quick response! Any particular frequencies I must be focused on, if the voice must sound real? How do I apply those settings during a podcast or maybe while streaming live? :) Sorry if my questions are too basic/stupid..
Cheers!
How do I find the bass frequency in a song I Rebassed?
When you upload a file onto audacity, how do you find the frequency of that audio file.
is it audacity only can read freq 8 khz
what does this jargo mean ? we need lament terms
If I want to see the individual note to, let's say, the opening chord to The Beatles "A Hard Day's Night" can you do that in Audacity? If you have any videos demonstrating this principle, can you direct me?
You can use the same tool, just switch over to the Pitch Correlation and the biggest peak will be at the fundamental frequency of the note. Audacity I don’t think translates that into a note (ie a letter) like Audition does, but such lists of notes and their frequencies can be found online.
do it slow...
Dear Martin, I need your help.
I construct string instruments and I use audacity to define the frequency of the wood (soundboard alone) and in different steps till the instrument is completed (without armed with strings). The procedure is the following (i.e for the free soundboard) : I keep the soundboard from an antinode point and I tap it with my figure. The signal is recorded through a microphone to audacity and the analysis of spectrum, gives several peaks of different size of dB in different frequences. By tapping the soundboard, I define with my ear the corresponding note (A ,or D, or E,..- according to the thickness of soundboard and some other parameters).
I expect that the first (higher dB value) peak of the spectrum will be the same note of what I hear.
But it doesn't happen every time. Can you please give me an explanation?
Many thanks.
The largest peak in a standard spectrum won't always be the fundamental frequency since higher harmonics will be present in different amounts depending on the instrument. It's best to use the autocorrelation methods to mitigate this. Hopefully they work better for you!
@@DrMartinArcher Thaks again.
while trying to code something (alike to birdgenie) that would take an audio file and identify bird species by sound, I ran into the fourier transform. My maths not being great, which is an understatement, my attempts stranded. Would any of the methods you show in your video have the same accuracy? Could they be an easier or even better alternative to the fourier transform?
The methods I show are all based on the Fourier transform and extensions of it.
@@DrMartinArcher Watched the video again and now see my understanding was incorrect. Sorry for wasting your time.
Hi Martin, I have a recording of someone's speaking voice. The sound is relatively stable (no shouting, whispering, background music etc just 'normal' talking). Is the peak of the highest wave (let's say 130 Hz) = Main frequency of the voice?
Be careful with your terminology. Pitch is the same as frequency or musical note, has nothing to do with volume - you can shout, whisper, or talk with roughly the same pitch. Also when you’re looking at the spectrum, what’s plotted is the power (same as intensity or amplitude squared) at each frequency when you consider the overall waveform as a sum of pure tones. So there is no “highest wave” but you can have a peak in power. The highest peak in power will be the dominant frequency or pitch of the audio, yes.
@@DrMartinArcher By 'highest wave', I mean the highest point on those purple graph (idk the correct word lol). Anyway, thanks for the clear explanation :))
@@DrMartinArcher Martin, the said audio has the highest power in 130 Hz, but also has the second highest power in 230 Hz (the purple graph has almost the same height as the 130 Hz one), what's that mean?
Hi can someone tell me how I can get the spectrum (Hz) from a track, for example from a meditation track. I need to know how many hertz this track is, is there any program that calculates it?
This video shows you how to get the spectrum. A spectrum does t give you just one frequency value, it tells you how much power (amplitude squared) there is at all frequencies. Perhaps you’re looking to get the tempo of the track out instead?
@@DrMartinArcher So a meditation track that tells me its description that it is at 528 Hz, how is it defined?, thanks.
528Hz music (and some other purported special frequencies) is a weird myth and the number is not used in any proper scientific way / definition www.medson.net/528hz-myth.html
@@DrMartinArcher I like gtoroja am looking to analyze a song in order to find out if it indeed is tuned to the common 440hz or something else like say 432hz. I'm changing the hz and playing w the speed and want to confirm once I'm done everything that the song is tuned to 432hz. Thanks for any help you might be able to provide. Have a great day
A little technical for me but helpful.
That’s the point of tutorials