Audio quality is reduced when converting

[ginlane]

Hi there, I’m Gin. Please, if possible, could I get some simple, non-jargon help. My mum died in May, she was a singer. We have many cassette tapes of her singing, which many in our local area would love to hear, so I came up with a plan. I would digitize the cassettes using Audacity audio software and then convert the audio file into a video file with an added photo or 2 for a bit of colour and then upload them to a registered channel on Youtube where they would be available for all to access. The trouble is, when I convert the digitized audio file to a video file, the audio quality became horrible, much less clear and bright, in fact, almost a bit muffled. I used Powerpoint for the conversion, from an MP3 in to an MP4 file. I feel there is no point in doing this lengthy project if the beautiful music is ruined at the end, and I won’t upload if my mum doesn’t sound her best. I need a bit of friendly help from someone to make sure the converted video files are as clear as the original audio files. I am posting these files on to Youtube rather than any other platform because they will be accessed by primarily older people with limited knowledge of the internet, but everyone knows and understands Youtube. I am happy to take advice on your recommended video editing software and which is best audio file format to use for the process.
I look forward to your help kind regards Ginlane

[super8rescue]

Record the cassette in to audacity, export as a wav file.
Mp3 is not the way forward as the audio will get compressed again in the video edit.
Before you export the audio, highlight the track with a double click, and use Normalise (defualt settings) in the effects menu.
Edit out any blank bits too before exporting the wav file
While recording the audio make sure the levels don't clip (red in the recording meter)
Then use a free editor like shotcut to create the final video. Import the audio, and images, export as h264 video in something like 1280x720p (even windows movie maker would do a better job than powerpoint to make a video in mp4 as you are more interested in the sound elements)

Use Youtube for tutorials on recording in audacity, shotcut edits and exports, and windows movie maker if you can't use shotcut.

[babygdav]

https://m.youtube.com/watch?v=XEyn2NYVZqw
https://m.youtube.com/watch?v=ns0JPLpN8t8

If you're running a pc on Windows 10, try the super simple Windows Photos app that includes video editing, titles, sound, etc.

It ought to retain the quality of your recordings (mp3? Did you do them already?). Do listen to your mp3 audio first - they're ok, right?

(If not, record as noted to wav files - 44.1,khz, 16-bit will do - cd quality.)

[babygdav]

Also, no idea how good your recordings are.
Post a 10 second clip captured as a Wav file format 44.1khz, 16bit.

You may need to reduce common tape background noise.
https://manual.audacityteam.org/man/noise_reduction.html

Even use a more advanced correction tool. E.g.
https://www.izotope.com/en/products/rx.html

[hello_hello]

As others have said, if possible, record as a wave file and keep it that way until the final conversion.

There shouldn't be a need to convert the MP3 audio though. MP3 can be stored in an MP4 container.
https://www.coolutils.com/Formats/MP4
If the program you're using is re-encoding the MP3 audio to another format, use a different one.

If you're recording directly to MP3 and need to edit the audio, try Mp3DirectCut. It can losslessly edit MP3 and AAC audio, and for MP3 it can also adjust the volume and apply fade-ins and fade-outs, without having to re-encode it.

[gdgsdg123]

...convert the audio file into a video file with an added photo or 2 for a bit of colour and then upload them to a registered channel on Youtube where they would be available for all to access.

Quite similar to what I did...

Check this. (the "best command for preparing music uploading for YouTube" part)


Also note:

Do realize: the audio(s) and the video(s) are separate (independent) streams in essence.

So are the subtitles, miscellaneous, etc.

Source

Anything shown on Twitch (or YouTube, whatsoever) usually won't be what you uploaded.

On the Lossy Encoding

...export as a wav file.
Mp3 is not the way forward...

Why WAV?.. For it doesn't introduce unnecessary losses during the processing.

(If not, record as noted to wav files - 44.1,khz, 16-bit will do - cd quality.)

For digital processing, 48000 Hz (48 kHz) is preferred.

Why?.. For it's dividable by 8 *, 12, 15, 24, 25, 30, 50, 60, etc. * And most digital audio systems expect the input to be of 48000 Hz. (all exceptions are converted to be of the expectation in runtime *)

* 1 Byte = 8 bits, which is the smallest unit of the CPU's handling. (complying to this aids in the processing efficiency * Note)
* Those numbers are typical video frame rates. (for the ease on the audio/video synchronization)
* To avoid unnecessary artifacts caused by improper resampling implementation (...though very unlikely unless you were using some piece of crap), it's desirable to perform the resampling in advance in a controlled environment (such practice may also marginally aid in the audio quality).

...try Mp3DirectCut. It can losslessly edit MP3 and AAC audio, and for MP3 it can also adjust the volume and apply fade-ins and fade-outs, without having to re-encode it.

There are limitations...

To achieve compression, such codecs use an approach similar to the Inter interpolation in video encoding. Which inevitably limits the available frames can be used for the "Cut & Join" without re-encoding.

[babygdav]

There's a difference between what's on the tape and the minimum acceptable digital format to capture in versus perhaps an ideal music format to edit in (should the audio editor not be smart enough).

https://www.nationalaudiocompany.com/wp-content/uploads/2017/08/NAC_DuplicationGuideli...s_wLINKS-1.pdf

Assuming you're mastering copies of cassette tapes, 44.1k 16-bit is usually more than sufficient..why? The tapes themselves can't handle more physically. (Why 44.1 if tapes only can handle up to 20? Nyquest theorem, use double) You can try to record 96khz 128 bit quality to a tape, but it physically can't reproduce such nice sound.

When recording the tapes to digital, sure, you can go as high as your equipment allows - 48khz, 96khz, etc 24-bit, 32-bit, etc - there's nothing "wrong" with that. But for any miniscule incremental audio improvement you may get, it's typically recording the "noise" in the system. Ie. You're equipment only handles up to 16khz and you recorded voice to tape up to that frequency. Anything you pickup beyond that likely wasn't part of the original voice - just noise you'll pickup in any analog system (E.g. Humans don't sing at 48khz frequency, so what are you picking up using 48khz to record above the typical 20-20khz of recordings?)

....

Frequency side, bits can help.

More bits allow finer adjustments in the sound level and processing. 2 bits is like painting her voice with 2 crayons - Not looking good. 16-bits is fine for CDs, but like many have noticed with hires audio, 24 to 32-bits can be better for pulling out nuances during editing.

44.1 32-bit can be better than 48+ 16-bit for recordings pulled off cassette tapes, especially if they were not made to metal tapes with professional equipment, mics, etc.

But, honestly, listen first. Try 44.1 16 bit, 96 24 bit and see if you can even hear a difference.

...

As for mp3/aac or any other LOSSY MUSIC format, these will never be as good as a NON-LOSSY format wav/flac/etc.
It's good to keep a recording of the tapes in NON-LOSSY format so you retain the most quality for later editing and listening.

[ginlane]

Additional: Hi again, it’s amateur hour over hear, I’m a complete beginner, I’ve watched a few audacity tutorials to help me improve some 40 year old amateur audio singing recordings which were originally on cassette. Specifically I took the knowledge from this YouTube video: https://www.youtube.com/watch?v=7_vvfMOqHiw
So this is what I’ve been doing. I used, in order:
Noise Reduction, (If you have a particularly noisy piece of audio you can change the value in the noise reduction DB box to between 12 to 18) then hit okay)
then Compressor, then Normalise, then Equalisation, Select curve, Bass boost,
then Equalisation, Select curve, Treble boost,
then Bass and treble, (Adjust the sliders as desired, you can preview if needed)
then Limiter, limit to -4 dB, and finally, Amplify
I find that after applying all of these settings and then playing my track through my widescreen TV speakers, I have a kind of crackly distorting buzz at the high end of the audio. (I have sent a short WAV clip with this post for your perusal) Now when I just play the audio through my computer, well then the buzz is not there and the same with headphones, it’s only through my widescreen TVs speakers that the distortion happens. I’m afraid my audio will not bear being played through big speakers which sucks! it’s a classic case of a little bit of knowledge is a dangerous thing. Please can someone help me to find a solution to this problem, maybe giving me an idea which of those settings it was which caused this problem in the first place. Ongoing, I am digitising more old cassette tapes and attempting to improve the sound quality. I don’t want to make things worse during restoration so I’d like to figure out a non-distorting way of improving my recordings!

Kind Regards Gin

[ginlane]

Additional:

Hi again, it’s amateur hour over hear, I’m a complete beginner, I’ve watched a few audacity tutorials to help me improve some 40 year old amateur audio singing recordings which were originally on cassette. Specifically I took the knowledge from this YouTube video: https://www.youtube.com/watch?v=7_vvfMOqHiw

So this is what I’ve been doing. I used, in order:

Noise Reduction, (If you have a particularly noisy piece of audio you can change the value in the noise reduction DB box to between 12 to 18) then hit okay)
then Compressor, then Normalise, then Equalisation, Select curve, Bass boost,
then Equalisation, Select curve, Treble boost,
then Bass and treble, (Adjust the sliders as desired, you can preview if needed)
then Limiter, limit to -4 dB, and finally, Amplify

I find that after applying all of these settings and then playing my track through my widescreen TV speakers, I have a kind of crackly distorting buzz at the high end of the audio. (I have sent a short WAV clip with this post for your perusal) Now when I just play the audio through my computer, well then the buzz is not there and the same with headphones, it’s only through my widescreen TVs speakers that the distortion happens. I’m afraid my audio will not bear being played through big speakers which sucks! it’s a classic case of a little bit of knowledge is a dangerous thing. Please can someone help me to find a solution to this problem, maybe giving me an idea which of those settings it was which caused this problem in the first place. Ongoing, I am digitising more old cassette tapes and attempting to improve the sound quality. I don’t want to make things worse during restoration so I’d like to figure out a non-distorting way of improving my recordings!

Kind Regards Gin

[aedipuss]

it's usually called hiss. i used a de-hisser plugin on it.

maybe upload a sample of the original before you adjusted it and i'll see if there's an easier way for you to make them ok.

[babygdav]

Is the sample you attached the original recording? Or processed?
(Sounds like a fair recording in a non-professional setup.)
Send an original clip if you didn't already post it.

What's your setup - how is it played, equipment, what software, record settings?

[gdgsdg123]

...when I just play the audio through my computer, well then the buzz is not there and the same with headphones, it’s only through my widescreen TVs speakers that the distortion happens. I’m afraid my audio will not bear being played through big speakers which sucks!

Ensuring the consistency of the playback experience among random devices?..

"No man ever steps in the same river twice, for it's not the same river and he's not the same man."

[babygdav]

[Attachment 51478 - Click to enlarge]
[Attachment 51479 - Click to enlarge]
[Attachment 51480 - Click to enlarge]
[Attachment 51481 - Click to enlarge]

Anyways, the file doesn't seem to be fully corrected.
I can hear tape hiss in the background throughout.

Using Adobe Audition (only to save time to show as an example, you can likely do the same in Audacity).
There are two filters of note - Noise Reduction and Hiss Reduction. Both can work on tape hiss.
Photo W1 shows the spectral image of the recording - nothing of note able 16Khz, so probably regular cassette tape and equipment, not metal/dolby/etc.
(No real reason to go above 44.1khz for recordings)

Hiss reduction is an art because every time you reduce hiss, you also take some quality away from the original recording.
There are more sophisticated tools like Izotope R you can use.
https://www.youtube.com/watch?v=n2SfxMpQSwQ

Compare the two - I did a Noise Reduction (W2) and Hiss Reduction (W3) on the same original.
You can listen and hear the reduced hiss, as well as the slightly rounded off voices.
More advanced spectral filters, cutoff filters, Izotope, etc. can all be used to "clean up" the recording, but that will take a lot of trial and error and testing to see which methods work the best for you.

The key here is reducing the hiss improves the audibility of the voices. You can hear them clearer with less hiss.

....

Typically, for audio, there's a LOT of standards as to how loud the music/sound track should be set to.
(eg https://www.gearslutz.com/board/post-production-forum/229741-standard-mixing-levels-mo...cials-etc.html)

Now, for home users, you don't need to understand all that.

Notice W4. The volume goes all the way to 0dB = 100% volume.
Great for PCs because they can handle that perfectly fine, probably not for TV home systems where the audio is typically mastered with a LoWER dB maximum.

Instead of normalizing to 0dB, try -6dB, -9dB, even -12dB and play the audio on your home system to see if it still overloads your speakers.


...

Tiny/Minor things that can help squeeze more quality out of these tapes:

1. clean and demagnetize the tape heads.
http://soundfirst.com/cleandemag.html

2. Try recordings with Dolby/Dbx turned on and off.
While you SHOULD have recorded with such turned on in the first place, in some cases, the audio compressor aids in making the tapes a tiny bit clearer even if you didn't.

3. Use a professional cassette deck.
These usually have a higher quality signal path, so less noise in the audio output.

4. Use a higher quality sound card for the computer to capture, along with better cables.
Most are good enough for cassette audio, but you can still use a high quality sound card to get the most out of the tapes.

But I'd try Izotope first. It is designed to do well all the cleanup you'll want to make to your analog recordings.
https://www.youtube.com/watch?v=n2SfxMpQSwQ

[ginlane]

it's usually called hiss. i used a de-hisser plugin on it.

maybe upload a sample of the original before you adjusted it and i'll see if there's an easier way for you to make them ok.

Hi aedipuss, thanks for your reply,

I am attaching a new sample of the original recording

The sample I sent you was of the recording after the changes where made.

So I also listened to the original recording through the TV speakers, and there is still a little bit of high-end distortion crackle, but not as much as after all the adjustments I made. I’m thinking it’s one of two things, 1. maybe my TV has poor speakers, (but then it doesn’t usually crackle playing other stuff) or 2. it something to do with how the audio was originally recorded, this was done about 30 odd years ago.
See I think that, seeing as the distortion is also a little bit on the original recording, then that makes me think that, this issue stems from how the recording was originally made? Is that possible? I’ve played it on all the devices I have, (admittedly, only tablets and phones, I don’t have a decent hi-fi)

[ginlane]

Leyburn Ladies Choir

[ginlane]

[Attachment 51485 - Click to enlarge]

Is the sample you attached the original recording? Or processed?
(Sounds like a fair recording in a non-professional setup.)
Send an original clip if you didn't already post it.

What's your setup - how is it played, equipment, what software, record settings?

Hi babygdav,

It was originally recorded on an old-fashioned hand-held cassette recorder, looking similar to the one in the photo I have sent as an attachment, obviously a model from about 35 years ago.
I have a Microsoft surface laptop 2 which I’m currently doing all the audio clean up work on. I did the original digitization transfer about six months ago using my dads high end cassette deck to a Samsung laptop with separate jack holes for headphones and mike.
The connecting cable had dual RCA red and white connections at one end and a jack connection at the other end. I plugged the RCA end of the cable into the cassette deck and the jack end into the headphones hole on the Samsung laptop.
I used Audacity for the digitisation. I didn’t mess about with any settings as far as I remember, I’m a beginner, (I will have to do some more digitisation soon, any suggestions on what settings I should be using for that?) Anyway…. I hit play on the take deck, record in Audacity, at the end I saved the output as an Audacity file to work on later.

[aedipuss]

attachment didn't upload???

[aedipuss]

since the tape was recorded without dolby noise reduction the best you can do is de-hiss it first. i added a bit of reverb to make it sound large and normalized it. how much better it can be made is iffy.

[babygdav]

1. The majority of the audio quality comes from the way it was recorded - on that portable cassette recorder.
Nothing you can do about the "sounds like a home recording" audio. No way to go back in time and use a better mic/etc.

2. You pulled a decent recording off the tapes (keeping in mind the tapes are probably regular, not metal, so you see a 16Khz rolloff).
No need to go back and try to get a better capture given how it was recorded and the general decent quality of the audio recordings.
Nothing objectionable about the capture given the age, etc, so you've got the transfer process down.

Yes, if you must improve, you can always use much higher quality equipment (http://indigitization-toolkit.sites.olt.ubc.ca/files/2018/09/Buying-Recommendations_Aug_2018.pdf).
But realistically, it isn't going to sound much better given how it was recorded.

3. Now, as for the recording itself.
a. It's got a nice -6dB maximum. That's perfectly fine and no need to normalize or amplify the recording.
The reason is when you do, you introduce more artifacts that come from raising the noise floor at the same time.
b. Also, the hand-held recorder looks like one of those auto-volume recorders, so that explains the higher noise levels in the very beginning, and lower when the singing starts. If you try to normalize such a recording, you'll get a mess. At most, a manual amplification by a set dB, but again, not at all needed.
c. Generally, a good transfer you can do a bit to to clean it up.

4. In Audacity, you'll want to noise reduce, then adjust the various parts (treble, bass, mids) using a EQ.
Don't go crazy with the adjustments, a light to medium adjustment is all you can do. Heavy and it'll distort badly.

No need to normalize (that'll make things worse).
You have 6dB or so to work with, so you shouldn't need a limiter or anything while EQ'ing - else, you're overdoing it.

Beyond that, it's finessing. How much work do you want to do?
It depends on how much clean up you want to do - you can spend hours cleaning up a single song.
Given the age, it's expected there will be a bit of noise and such after you've gone through it.

If you want the best for a reasonable amount of time and money spent, Izotope, BUT you will never get a CD quality recording out of it.
https://www.izotope.com/en/products/rx.html
https://www.tacsystem.com/upload/izotope/iZotope_RX_Restoration_Guide_v_1.pdf

If it becomes too much, you can hire an audio engineer who knows and does this sort of work.

Besides Audacity, there's other "simple" audio cleanup tools, but don't expect a miracle.
https://www.antarestech.com/product/soundsoap-5/

5. As for the TV, I wouldn't worry about it. If the original and edited recording sounds fine on the PC with headphones and has a decent -6dB headroom, it is NOT the problem.
It could be how you play it on the TV (eg. transfer to CD/DVD at levels too high) or something else.

[babygdav]

since the tape was recorded without dolby noise reduction the best you can do is de-hiss it first. i added a bit of reverb to make it sound large and normalized it. how much better it can be made is iffy.

All that processing is a touch heavy handed - you can hear the clipped high-frequency versus a simple, lighter noise-reduction (done in Audition).
It's an art, so how much of the full range of the voice you keep versus filter out through noise-reduction/de-hissing is a balance.

Best is to listen to the original, then the filtered version to compare. At some point, you'll subjectively feel it's too much, and you'll have to back off.
Otherwise, over filtering can reduce the audio quality to the point where the audibility is worse than desired.

[gdgsdg123]

The volume goes all the way to 0dB = 100% volume.
Great for PCs because they can handle that perfectly fine, probably not for TV home systems where the audio is typically mastered with a LoWER dB maximum.

Instead of normalizing to 0dB, try -6dB, -9dB, even -12dB and play the audio on your home system to see if it still overloads your speakers.

If such "overloads" do happen... the equalizer settings of the playback device were usually being messed with.

1. maybe my TV has poor speakers

Isn't it clear enough?..

...when I just play the audio through my computer, well then the buzz is not there and the same with headphones, it’s only through my widescreen TVs speakers that the distortion happens. I’m afraid my audio will not bear being played through big speakers which sucks!

[babygdav]

Buzz. Could be a low or high level repeating electric noise embedded in the audio. At namm right now, but you can use cut off filters to test. Everything below 1000 hz, above 8000 hz, just reduce it to 0. Play.

[hello_hello]

...try Mp3DirectCut. It can losslessly edit MP3 and AAC audio, and for MP3 it can also adjust the volume and apply fade-ins and fade-outs, without having to re-encode it.

There are limitations...

To achieve compression, such codecs use an approach similar to the Inter interpolation in video encoding. Which inevitably limits the available frames can be used for the "Cut & Join" without re-encoding.

It's not really like that.
http://www.mp3-converter.com/mp3codec/frames.htm
MP3 yses something called a byte reservoir, which I don't fully understand, but I think it means if the bitrate required for a frame exceeds maximum, the extra bits can be stored in another frame, or something like that. I don't know how much it effects cutting, but in my experience the distance between points you can cut for a typical MP3 wouldn't be more than 40ms.

[gdgsdg123]

It's not really like that.
http://www.mp3-converter.com/mp3codec/frames.htm

It has to act like that...

The situation quite resembles encoding videos of 1x1 resolution (only 1 single pixel in each frame).


Caveat: the "frame" I mean is not the "frame" mentioned in that link... in which, it's called "sample".

[hello_hello]

I'll confess I don't follow. MP3 frames have a fixed duration. They hold 26ms worth of samples. Byte reservoir aside, each frame is independent. The higher the bitrate, the more bits each frame stores, but the sample rate doesn't change.

Anyway, while I'm arguing.....

(If not, record as noted to wav files - 44.1,khz, 16-bit will do - cd quality.)

For digital processing, 48000 Hz (48 kHz) is preferred.

Why?.. For it's dividable by 8 *, 12, 15, 24, 25, 30, 50, 60, etc. * And most digital audio systems expect the input to be of 48000 Hz. (all exceptions are converted to be of the expectation in runtime *)

* 1 Byte = 8 bits, which is the smallest unit of the CPU's handling. (complying to this aids in the processing efficiency)
* Those numbers are typical video frame rates. (for the ease on the audio/video synchronization)
* To avoid unnecessary artifacts caused by improper resampling implementation (...though very unlikely unless you were using some piece of crap), it's desirable to perform the resampling in advance in a controlled environment (such practice may also marginally aid in the audio quality).

The reason why everyone in the world uses 48k rather than the 44.1 used by CDs, is because of the Nyquist sampling theorem.

A frequency can only be sampled if the sample rate is at least double that frequency. So if you're using a 44.1k sample rate, it means you need a filter that starts rolling off at 20kHz and removes everything above 22.05khz.
If you sample at 48k you can use a filter that starts to roll of at 20kHz, but has to until 24kHz to remove everything. A low pass filter with a gentler slope can potentially sound better and cost less, at least in the early days of digital. As far as I know, CDs use 44.1k because Sony already had a digital format and a vested interest.

The sample rate doesn't effect the number of bits per sample. 8 bit audio uses 8 bits per sample so each sample can be assigned one of 256 different values. 16 bit audio can be have one of 65,536 different values per sample, and 24 bit audio can have 16,777,216... but in each case the number of bits per sample is divisible by 8, even if on playback they're whizzing by at a rate of 22.05k, 44.1k, 48k or 96k per second. And..... not that I'm sure it means anything in particular..... but.....

CDs have a rate of 44100 samples per second, at 16 bits per sample, which is 705,600 bits per second, or 88,200 bytes per second, and both are evenly divisible by 12, 15, 24, 25, 30, 50, and 60, etc

[gdgsdg123]

And..... not that I'm sure it means anything in particular..... but.....

CDs have a rate of 44100 samples per second, at 16 bits per sample, which is 705,600 bits per second, or 88,200 bytes per second, and both are evenly divisible by 12, 15, 24, 25, 30, 50, and 60, etc

...That's not how things work.

* 1 Byte = 8 bits, which is the smallest unit of the CPU's handling. (complying to this aids in the processing efficiency)

The processing efficiency here, is mostly about the compression efficiency. (speed, quality (if lossy), compression ratio)

If the content is being used (for playback only) in uncompressed form, there'll be little difference on the efficiency.

[hello_hello]

And..... not that I'm sure it means anything in particular..... but.....

CDs have a rate of 44100 samples per second, at 16 bits per sample, which is 705,600 bits per second, or 88,200 bytes per second, and both are evenly divisible by 12, 15, 24, 25, 30, 50, and 60, etc

...That's not how things work.

I did a little research and it seems it has nothing to do with being "divisible by 8", and everything to do with an integer number of samples at a given frame rate, at least in the analogue days, so you're correct in that respect.

From Wikipedia: https://en.wikipedia.org/wiki/Sampling_%28signal_processing%29#Sampling_rate

44,100 Hz.
Audio CD, also most commonly used with MPEG-1 audio (VCD, SVCD, MP3). Originally chosen by Sony because it could be recorded on modified video equipment running at either 25 frames per second (PAL) or 30 frame/s (using an NTSC monochrome video recorder) and cover the 20 kHz bandwidth thought necessary to match professional analog recording equipment of the time. A PCM adaptor would fit digital audio samples into the analog video channel of, for example, PAL video tapes using 3 samples per line, 588 lines per frame, 25 frames per second.

48k sampling was chosen because it also matched the NTSC 29.97 frame rate, in respect to an integer number of samples.

48,000 Hz
The standard audio sampling rate used by professional digital video equipment such as tape recorders, video servers, vision mixers and so on. This rate was chosen because it could reconstruct frequencies up to 22 kHz and work with 29.97 frames per second NTSC video – as well as 25 frame/s, 30 frame/s and 24 frame/s systems. With 29.97 frame/s systems it is necessary to handle 1601.6 audio samples per frame delivering an integer number of audio samples only every fifth video frame.

48000 / (30000 / 1001) * 5 = 8008

In a digital world I think all that matters is the audio clock and the video clock stay in sync. In the real world I doubt many players sync to a TV at exactly the correct rate, so they'd probably adjust their audio clock to match (I'm theorising a little), but it'd be about the video and audio playing at the correct speed to stay in sync and not about the audio rate being divisible by eight, or an integer number of samples matching the frame rate.
ReClock does a similar thing for PCs, where the video and audio are often independent. My video card connects to both my TV and PC monitor at 60.002Hz, The displays no doubt adjust themselves to stay in sync with the card, but it'd mean for the frame rate to match the audio rate, the video frame rate won't necessarily match the display's refresh rate and you can end up with the occasional dropped frame or tearing in the picture.
I'm pretty sure by default ReClock adjusts the PC's audio clock so it can feed the sound-card samples at a rate that matches the video card's refresh rate. This is what it's doing for the video I'm watching at the moment. The audio clock is adjusted to 48062 Hz. See the pic below.

Anyway, the point I'm trying to make is the audio sample rate and frame rate probably don't need to have a special relationship in the digital world, or be even multiples of each other as such, as long as the audio and video clocks have a way to stay in sync. Mind you I'm no expert. That's why I don't mind a bit of a debate in forums. I often end up learning something.

* 1 Byte = 8 bits, which is the smallest unit of the CPU's handling. (complying to this aids in the processing efficiency)
The processing efficiency here, is mostly about the compression efficiency. (speed, quality (if lossy), compression ratio)

If the content is being used (for playback only) in uncompressed form, there'll be little difference on the efficiency.

Then why is uncompressed audio such as PCM sampled using multiple of 8 bit-depths? 8, 16, 24, 32 etc?

[gdgsdg123]

Anyway, the point I'm trying to make is the audio sample rate and frame rate probably don't need to have a special relationship in the digital world, or be even multiples of each other as such, as long as the audio and video clocks have a way to stay in sync.

Indeed not necessary but preferable. (for editorial purposes)

...a bit of a debate in forums.

We debate... for the good of the community.

* 1 Byte = 8 bits, which is the smallest unit of the CPU's handling. (complying to this aids in the processing efficiency)

The processing efficiency here, is mostly about the compression efficiency. (speed, quality (if lossy), compression ratio)

If the content is being used (for playback only) in uncompressed form, there'll be little difference on the efficiency.

Then why is uncompressed audio such as PCM sampled using multiple of 8 bit-depths? 8, 16, 24, 32 etc?

Mind the context of the quote, it's about the sample rate.

[gdgsdg123]

Note on the Lossy Encoding

The reason I used FLAC as a container was to avoid double-compression (once from the game and another time from my recording).

With appropriate settings using an adequate lossy encoding method (namely, Vorbis, in this scenario).

Double, triple, quad, penta... such compression won't make much difference. (the spectrogram of the output remains in extreme similarity)



The quality of the material itself plays a much bigger role. (i.e. a properly processed 128 kbps * MP3 can be of much higher quality than a faulty FLAC)

The encoding method is merely a container, it does not guarantee quality. It only guarantees the upper limit of the quality.







2 post-processed version on YouTube based on your source:
(higher quality than the FLAC there, and likely higher quality than what's in the game... even again lossy compressed)

Lineage - Kenji Kato
The Pass - Naoki "naotyu-" Chiba



And another source of "Lineage - Kenji Kato" on YouTube. (presumably recorded with a phone... or microphone at best)

Serves as a good example for comparison.

Source from a dead forum (FFShrine), accidentally found it in my local post archive...


* When the number of audio channels is not clearly mentioned we can simply assume it's 2. (quite a number of programs actually behave like this way...)

But also note: many audio encoders (especially the advanced ones) also exploit the redundancy between the audio channels.