10. Pulse Code Modulation - Digital Audio Fundamentals

Pulse Code Modulation is an encoding mechanism, a way of representing digital data for the purposes of transmission and storage. But what form does this encoding take, and why is it needed? In this video, we'll explore the the concept of encoding and modulation as a corner stone of digital audio. We'll start from first principles and discover different modulation strategies and their advantages, and naturally arrive at why PCM is so resilient and has survived the test of time.
Find the full playlist here: • Digital Audio Fundamen...
Content:
0:00 Encoding
1:21 Frequency Modulation
4:28 Pulses - Digital encoding
6:06 Pulse Width Modulation
6:25 Pulse Position Modulation
7:59 Pulse Amplitude Modulation
9:05 Pulse Code Modulation
9:50 Bandwidth of PCM
11:30 Overview of ADC
This video series explains the fundamentals of digital audio, how audio signals are expressed in the digital domain, how they're converted and transformed and the advantages of working with digital signals. Find the full playlist here: • Digital Audio Fundamen...
If you've got any questions, suggestions or recommendations, type them out here, or send me a message on any of my social channels mentioned below.
A lot of time was spent on creating this series, and I plan to do more. So please consider subscribing if you wish to be notified about more releases in the future. And if you feel generous: / akashmurthy
Check out the audio track in the intro and outro here: / tinderbox
Check out my music here: farfetchd-official.bandcamp.com/
Give me a shout here:
Facebook - / akash.murthy.319
Instagram - / thrifleganger

[@binarywoif2852]

I've already seen quite a few videos and am really impressed, both by the content explained and the quality of the explanation + graphics quality. I really hope you'll get the recognition you deserve for all the efforts you put into this!

[@akashmurthy]

Thank you very much for taking the time off to write me a comment! These keep me going! :)

[@alfredbasi5937]

I've seen quite a few videos on PCM but was unable to wrap my head around all these concepts until now. Your explanation was really informative and easy to understand, thanks!

[@snowboarder12653]

Absolutely incredible. Such a good explanation. Thank you!

[@akashmurthy]

Thank you so much! :)

[@marktrivette4272]

This dude does excellent videos, breaking down complex EE that a layman can understand. Nice work.

[@Sams911]

wow, I'm up late watching videos and this gem shows up! Still surprised you don't have more subscribers!

[@akashmurthy]

Haha, thanks! All in good time.

[@deltakid0]

@@akashmurthy you don't have too much subscribers, but you will have them because yeah, this is a gem. By the way, this is somewhat 3 Blue 1 Brown style which is kinda classy. Thank you.

[@akashmurthy]

@@deltakid0 thanks for the motivation! Im quite a fan of 3b1b!

[@blakemartin5296]

this playlist is incredible. Thank you for your contributions to the distribution and improved communication of knowledge!

[@akashmurthy]

Cheers mate! My pleasure!

[@dadominicanstyl]

Amazing explanation, the clarity of how you present the subject it is the best I have even seen! Thank you so much.

[@akashmurthy]

Thank you for checking it out mate!

[@Apoorvpandey]

phenomenal explanation, the touch of real example values made this even better!

[@akashmurthy]

Glad you think so! :)

[@hawthornrabbit]

Fantastic video, this really helped me to understand frequency modulation and PCM like nothing else has.

[@mingjunyu1127]

Much better than my professor.

[@nelizevanschalkwyk408]

This video's illustrations are so satisfying to watch

[@paveldroo]

This is simply excellent content, thanks for your work!

[@yayayaiamlordeyayaya]

Wow ❤, this is just awesome ! Thank you ! Best video about PCM

[@chefatchangs4837]

These videos are fantastic, thank you sir

[@zheyuanzhang8926]

Amazing video! Thank you!

[@francescom.1558]

This is absolutely amazing! I am currently studying this stuff and your explanations are really helping me a lot

[@akashmurthy]

That's brilliant! All the best

[@ralphr.schmidt716]

Very good explanations, with a clear voice and at the right pace. I would give 5 our of five stars if I could (for every video that I have watched from your channel that is).

[@akashmurthy]

Thank you for the feedback! I'll take the verbal stars, thanks :)

[@brydon8359]

By far the best video I have seen on the subject. Thanks a lot!

[@akashmurthy]

Thanks so much!

[@rogo7330]

Such a simple explanation, you can just take it and code for something. Thank you!

[@akashmurthy]

Thanks for that! And you're welcome.

[@DanJanTube]

Really well explained, thank you.

[@akashmurthy]

Thank you! :)

[@esmaelkaronjr.6346]

Excellent quality!

[@akashmurthy]

Thanks mate!

[@kaushikgupta1410]

This video is a masterpiece!!

[@hendrikoffermans]

I agree, I am watching all videos on this channel and the level of explanation is so high (I am a teacher myself).. really amazing that this channel has so little subscribers.. excellent work! Thank you very much for your great videos.

[@akashmurthy]

Brilliant! I appreciate the comment, thank you. It would be stellar if you could share this as reference material with your students.

[@yakovsushenok4009]

Thank you for the series. I'm doing a master thesis and these really helped out to understand some of the stuff :)

[@akashmurthy]

Great stuff! I'm glad it's helping you out for your thesis..

[@mehmetcaloglu9409]

bro , amazing job, thank you.

[@ranjitakhadka7839]

wow such a good explanation.....very few people are so good at explaining things this clearly.

[@akashmurthy]

Thanks so much! Glad you think so.

[@madhavrabinbipin2116]

this is exactly what i wanted. thanks a lot

[@akashmurthy]

Glad this helps! Thanks!

[@alexanderquilty5705]

This is literal gold content.

[@akashmurthy]

Ah thanks mate!

[@alexanderquilty5705]

I was thinking about doing similar videos related to teaching electrical and computer engineering topics, but I haven’t really committed. I have no experience with video editing either. 😅

[@akashmurthy]

@@alexanderquilty5705 if you have the time, go for it! Learning editing and animation is really just skills that you learn progressively as you make videos, and get better and faster at it as you make more..

[@pranavgupta27]

Thank you so much for such informative videos. Keep it up. God Bless.

[@akashmurthy]

You're welcome!

[@HakunaMatata0707]

Great explanation.

[@nelizevanschalkwyk408]

Thank you for creating such good content!

[@akashmurthy]

Thanks very much for checking it out!

[@bugss_]

i've been searching for a video like this in german but still ended up watching this. This was a great help though, thank you!

[@akashmurthy]

Thanks very much! I'm glad you found the channel!

[@javiergonzalezsantamaria9302]

Thank yoouuuu, video edit is amazing!!! and it's such a good explanation!!!!

[@akashmurthy]

You're welcome mate!

[@rheamaebibay7650]

Omg! The layout is so nice 💓

[@akashmurthy]

Aww..thanks!

[@subhajitdebnath9085]

Your vedios are great and helpful to understand deep learning of audio coding..

[@akashmurthy]

Glad it helps!

[@michaeltina3866]

GREAT JOB... MAY GOD BLESS YOU

[@akashmurthy]

Thanks so much!

[@thopmemestelugu5260]

Excellent explaination and information sir

[@akashmurthy]

Thank you! :)

[@michaeldausmann6066]

fantastic explanation

[@akashmurthy]

Thank you :)

[@abhishekpatil2444]

nice explanation

[@rahatpreo3209]

best video on this topic on YT

[@akashmurthy]

Thanks mate!

[@soyusunk]

Thank You very much.

[@andytwgss]

can't wait to see PDM (ie DSD) / D-S modulation !

[@MrJef06]

Awesome video series, thanks! One small mistake though: around 4:00 the Phase Modulation graph actually depicts Frequency Modulation.

[@akashmurthy]

Thanks for the feedback! I have to admit that I was having a hard time depicting phase modulation. I hadnt encountered phase modulation before, and the math for the animations didn't work out. But if you Google image search for phase modulation, there's quite a strong similarity to frequency modulation. So, I used that instead.

[@MrJef06]

@@akashmurthy you are right, the difference between FM and PM modulation is not abvious at all when the input signal is a sine wave. It would look much different with a square wave for example. I just noticed you re-used the exact same animation ;-) Thanks for replying!

[@shaunfossett]

You explained this very well in very simple terms. I just wish people who "produce" music understood this versus relying solely upon what software like FL Studio does. The old school method is better

[@MahmudurRahman-em3mn]

OMG what was that! This channel will take 'youtube study tutorials' to a whole new level. I mean Its been 3 years I am totally dependent on youtube for my engineering degree 😅and have watched thousands of tutorials, some of them were best, but today wasn't having a good one on PCM and found this channel and I am SPEECHLESS! Bro thanks for providing these for free! Dhonnobad❤‍🔥 just saying, I think you need to work on the thumbnails for these awesome tutorials otherwise it looks just like a boring ppt slides!

[@akashmurthy]

Thanks man, glad you found it useful!

[@stephanethomas2186]

I totally agree. The animations are very good. I can’t find any downside to those videos. That’s clear and right to the point. I started this set of videos at number four or such because I wanted to have a better understanding of the pros and cons of using a high sample rate for computer music, now I’ll have a short night cause I’m watching an explanation of what’s PCM and I just can’t quit watching! ^^

[@nguyenhuyhoang2545]

Great. Thank you

[@Frame_Fusion]

Okay so, for some reason I entered this channel, thinking it would be another PPT/paper explanation But damn, this is so so underrated, the graphics, the explanation. Everything is amazing! I wish you reach great heights! Subbing for your true efforts :)

[@akashmurthy]

Thanks very much for the sub mate! :)

[@arunlinus007]

Just be consistent bro. You will definitely shine on youtube

[@akashmurthy]

Cheers!

[@chikkivenu6632]

Hi Akash, Great work. I have not seen video which is better than yours. for 2 reasons - knowledge on the subject. 2- explanation which is supported by GREAT Graphics.(wondering how did you do it.). I would request you to put some video on the different codec which are there are in market. starting from adpcm, celp codes, svc, silk. I am network voice engineer in mnc these learning are adding more confidence. Also can put some video on commanding ulaw and alaw. Thanks you looking forward to more videos from you.

[@christianreyalison8082]

Thank you :)

[@sloblo7032]

Thank you for the clarification you provide. I have a question about the windoing process in the case of audio processing and not speech. What is the size of the window? We know that in the case of speech it is 10 milliseconds.

[@antoinefortin1386]

I cannot understand why these are not more popular than this

[@akashmurthy]

All in good time my friend!

[@mariammohamedsaad6513]

THANK YOU

[@elijahjflowers]

Thank you.

[@akashmurthy]

You're welcome!

[@chikkivenu6632]

Thanks

[@nLakers1290]

so in PAM, only amplitude varies and width a position are constant right ?

[@jamilajamila3658]

Sir what will be the maximum transmission bandwidth in PCM system?

[@kalakxfif9473]

really good explanation. also, how do you create these animations?

[@akashmurthy]

Thanks a lot! I use Adobe after effects for the animations.

[@mohsens22]

Thanks for this video :) I learned a lot and I searched a lot about wave modulations and so on. But I have some questions. Why 20-20khz doesn't have enough power? What makes 100Mhz more power? If more frequency means more power, why Wifi (2.5Ghz) doesn't spread as much as FM? And isn't power all about amplitude? Why in normal transmission we have more noise? comparing to 100MHz transmission Why noise only effect the amplitude? I assume that because mixing is basically adding amplitude of waves, the noise which is mixed, actually interferes with the amplitide and not the frequesncy directly. But hey, we still can create new frequencies via mixing amplitides, right? So noise can interfere with the frequency or at least data it's carrying How is FM Quantized and encode/decoded real time? Assuming that FM is an Amplitude to Frequenct converter, In a given time we can measure a point and see what Amplitude that has, but in a given time, we can't have understand the frequency without having a range of time and mesuring cycle. The circuit only can see one number per time, right?

[@akashmurthy]

Hey mate, thanks for the question. 1) 20kHz sound is different from a 20kHz signal. Sound is a mechanical, longitudinal wave. A signal, in this case, is an electromagnetic wave. Totally different concepts right? From Google, "The energy carried by an electromagnetic wave is proportional to the frequency of the wave." So yes, your 2.5GHz signal contains more energy, thus potentially data carrying capacity than a 100MHz FM signal. Why doesn't it "spread much"? You can draw parallels from sound waves here. Though they are not the same, you can make comparisons. What do you hear outside a nightclub? You hear the music, but only the low thumping bass sounds. Lower frequencies can pass through barriers like walls and sound proofing, but the higher frequencies are absorbed and are blocked by barriers like walls. Similarly, FM signals is lower frequency compared to WiFi signals. FM signals can penetrate through walls and barriers much easier than Wifi signals. 2) Noise only affects amplitude. If you add noise to a sound file, what do you hear? You hear noise interleaved with the sound. You don't hear the underlying sound being different in anyway, you don't hear the frequency response of the sound any different with or without the noise. Noise is by definition a signal with randomized characteristic. You are talking about specific cases where 2 pure sine tones for example, when mixed, gives some sort of beating effect, where the frequencies is different. Like I said, we are dealing with random characteristics here. 3) I think it's gone a little off topic here, this video isn't meant to be anything about FM. You can find better resources for this elsewhere. I've already talked a little about frequency modulation (encoding). Your question is related to frequency demodulation (decoding - not the right term for it, but oh well). Electronic circuits need to detect changes in frequency, and not the actual frequency itself. Electronic components do this readily. Read up about it by googling FM Demodulation circuits. The change in frequency is mapped to change in voltage which drives the speaker to produces the underlying sound. Again, google has all the answers if you want to go deep.

[@karanamraghavendrarao5042]

Voice is so good

[@jesusalonsocano7245]

Hola, podrias ayudarme con una práctica de digitalización PCM?

[@dr.fritzprengel2378]

*Great explanation!* If an old CD-player from the 80s send PCM Audio to a new DAC/-Amplifier - then only the pure CD data is send to the (good new) DAC and is converted there with modern methods?

[@akashmurthy]

Thank you! I'm not sure I understand the question. I don't know how an old CD player is capable of sending digital data out. Usually, it reads a cd, which is a digital medium, it has a small DAC which converts to digital data extracted from the cd to analog outputs (earphones). There is no digital to digital transmission of data in this workflow. Only modern amplifiers support SPDIF or optical out, which can transmit digital data.

[@dr.fritzprengel2378]

@@akashmurthy Thx! I meant: Old CD Player with _Cinch Digital Output_ ↓↓↓Digital Signal↓↓↓ New Amp with modern DAC = sounding like a new CD Player instead of the old one?

[@akashmurthy]

@@dr.fritzprengel2378 Right, I think if a modern amplifier recieves a faithful digital signal, then it will definitely have a better output than old CD players, since Digital to analog (DAC) conversion technology has massively improved over the years, and the individual electronic components have reduced in price over the years. But reading the CD to produce the digital signal in the first place is also technology that's improved, in terms of error detection and correction. So that matter as well I suppose.

[@dr.fritzprengel2378]

@@akashmurthy OK, thank u 👍

[@rajendrabiswas]

Is dts and Dolby superior to pcm audio ? I am wanting to buy a DAC amp for headphones

[@akashmurthy]

The encoding formats make no difference. It's lossless audio both cases. The hardware that you can build compatible with certain encoding formats can vary quite a lot, and that's where the advantage comes from. The hardware manufacturer may choose to support one encoding mechanism over the other while choosing components. Some components are more capable of taking advantage of the sampling and quantisation process underneath the hood. Is one subjectively/objectively better than the other is not a discussion I'm willing to have.

[@artpinsof5836]

What are your thoughts on DSD and sigma-delta encoding? There is one famous engineer of speakers on KZbin, that if you, Google DSD, you’ll see 1 million videos of him saying that it is better somehow than PCM… Most people in forms disagree, but he has many fans who agree with him as well… I would love to hear your thoughts on this, or if you have any links to describing the benefits of this form of encoding in comparison with PCM, that would be much appreciated! Thank you greatly for the incredible Educational content!

[@akashmurthy]

Hey there, to be honest, I don't know enough about Delta sigma modulation to comment on the fact that one is better than the other. DSD is just a brand name for Delta Sigma modulation. It's trade off between word length (bit depth) and sample rate. Delta sigma modulated signal is represented as a single bit stream, but requires super high sample rate (in the order of mega hertz) for it to be effective. In a sense, this is easier to implement since modern chips have very high speed and precise clocks. A lot of the times, modern ADCs and DACs use delta sigma modulation to convert from analog to digital and vise versa, but covert to PCM internally to be represented in applications. This is because manipulating digital audio represented as a DSM stream is quite complicated. Even applying gain is so much hassle. Whereas in PCM, it's quite trivial, since the data is represented as you'd expect in the analog domain, and the same principles apply.

[@Hi-kj8pg]

Mention when signal rise and fall for pwm, ppm pcm , give example with exact values

[@Nerthexx]

I see, so there is no episode 9. Also, could be useful to hear about resampling techniques in maybe some of the next videos ;)

[@akashmurthy]

Yea! I was working on both 9 and 10 at the same time, ended up finishing 10 first. Sure, the digital to analog process is on the story board. I'll have to get to it at some point of time! The trouble is, these videos take a lot of time to make. Cheers.

[@bflmpsvz870]

Can you make one on Pulse Density Modulation please?

[@akashmurthy]

Sorry, this series is mostly concerned with audio. PDM is rarely used in the domain of audio.

[@carlosa.chacon985]

Now we need DSD jeje!

[@akashmurthy]

Some day!

[@may8049]

do we need to calculate the mathematically unique continuous solution for the samples when outputting sound? maybe we don't need them at all? only need to send the discrete data to the speaker and apply the voltage that's enough?

[@akashmurthy]

Discrete data is just information. Bits and bytes are abstract information that represents the state of a system. To make any use of it in the real world (like driving the speakers), you need to convert discrete data into continuous electrical signals.

[@may8049]

@@akashmurthy but at which step we convert to continuous signal, i'm still confused at the process, is there a formula implemented in DAC that converts data to its original sin wave? how does it input data, output sin wave.

[@akashmurthy]

@@may8049 I'd like you to check out video number 2 (Sampling Theorem) from 12:28 onwards. There is no formula involved. And yes, the step where discrete data points are converted into continuous analog signals is the DAC. Simplifying the process quite a lot: the data points are connected, or interpolated by electrical circuitry to create a continuous signal. Then this signal is filtered with a low pass filter with its cut off as the nyquist. This is the critical step, which eliminates all the interpolation errors and produces the original signal back.

[@may8049]

@@akashmurthy OH, that's it! the low pass filter, so after applying the low pass filter,the square wave turned out to be as close as the original sin wave because all the square part the high frequency part gets filtered off,right? i finally understand, bro, you are awesome!

[@akashmurthy]

That's it! You got it!

[@subhajitdebnath9085]

Sir, please make vedio on AC-3 coding system and dts coherent acoustic coding system..because I can't found in youtube..

[@akashmurthy]

Sorry, I can only talk about what I know. Also, I want to talk about general concepts here, and not patented, product specific topics.

[@subhajitdebnath9085]

Yes sir I understand then please make vedio of digital filters MDCT, and ADPCM what is LPC etc..it will help us to understand coding technology

[@akashmurthy]

@@subhajitdebnath9085 that's a reasonable request! Ok, I'll keep it in mind and work on a video for that. Thanks for the suggestion.

[@subhajitdebnath9085]

Thank you sir for accepting my request.

[@danadam0]

nitpick: uppercase 'K' usually means binary prefix, i.e., x1024, so 705.6 should rather use kbps instead of Kbps.

[@akashmurthy]

Really? 0x is hexadecimal, 0o is octal, I thought binary is 0b. I mean it depends on the context right, different languages choose to support different prefixes. But in most of the cases I've seen, the metric multipliers like kilo and mega and giga are always abbreviated with capitals, like K, M, G. But I was careful not to capitalize 'b' in 'Kbps', since 'KBps' is reserved for kilo Bytes per second. Let me know if you disagree.

[@danadam0]

@@akashmurthy I didn't mean binary representation of a value, for which indeed 0b is often used. I meant a binary prefix of a unit (of a multiple of units, to be precise). Lowercase 'k' usually means a multiple of 1000x and uppercase 'K' a multiple of 1024x. 1024 is 2^10, hence binary. Wikipedia has a lot more about that: en.wikipedia.org/wiki/Binary_prefix en.wikipedia.org/wiki/Byte#Multiple-byte_units en.wikipedia.org/wiki/Kilobit

[@akashmurthy]

@@danadam0 interesting! Thanks for the insight!

[@ahmednor5806]

🙏🌹🌹🌹🌹

[@Thoracius]

Not clear from this video how PCM identifies the start of a new 16-bit sample.

[@akashmurthy]

Yea good call. Most of the videos in the series so far has been concentrating on the analog to digital side of the conversion. I haven't really talked much about the other way round, the digital to analog conversion. There is much to talk about there, including in this case - clock synchronization, jitter, and all of that good stuff. I guess I'm holding that off for later.

[@akashmurthy]

Whops, sorry, I didn't really comprehend the question correctly. Yea, most of the time, in RAW PCM, there are no identifying features to mark the beginning of a new sample. Computers are quite accurate to synchronize with clock to count 16 bits, mark that as one sample, and carry on with the next sample. Sometimes, when PCM data is fed into DAC for playback, channel codes can be added to signify the start of a frame of samples for multichannel streams, so that the DAC could be more accurate and rely on these additional codes for synchronization.

[@Thoracius]

Look forward to hearing about that side of things, jitter, etc. Indeed I was curious whether after a long string of silence (all zeroes) whether two devices would maintain perfect synchronization. Or if a decoder starts receiving a streaming PCM signal mid-sample how it would know to wait for the beginning of the next 16-bit series of 1s and 0s and how to know when that occurs. With the pulse width examples that was clear but not with PCM.

[@lordpatil7716]

Is this manim?

[@akashmurthy]

No, this is After Effects and expression scripting.

[@ECE_Third]

I wished you have been my teacher 😂😂