# Converting from 48 kHz to 44 kHz looses the highest frequencies \[with various software\]

**URL:** <https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971>\
**Category:** Off-Topic\
**Created:** [May 16, 2020, 5:11pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971 "2020-05-16T17:11:27Z")\
**Posts on this page:** 10\
**Page:** 1

<div class="post-metadata">

**Author:** ![Zerow](https://community.mp3tag.de/user_avatar/community.mp3tag.de/zerow/32/11379_2.png) [@Zerow](https://community.mp3tag.de/u/Zerow)\
**Post date:** [May 16, 2020, 5:11pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/1 "2020-05-16T17:11:27Z")

</div>

I had this problem ever since, but it did not bother me in the past as I did not knew what exactly it meant. It goes like this

[Below I am referring to my latest case of this problem - if particular files for testing will be needed I can provide them; but I think this canbe discussed without even seeing the data]  
` `  
` `  
I use DVD Audio Extractor 7.1.1 [and before that version 6.3.0] to extract stereo PCM WAV files. I set them to be saved as 48 kHz. As a result the files I get have all the frequencies visible on the spectrogram, going almost to the very top [~23.5 kHz]. But if I extract the same tracksto a 44 kHz WAV file, then the frequencies in them will go only up to ~20.5 kHz. So this is a problem of DVD Audio Extractor? I wish, it was so

Because I tried converting those better ones WAVs [48kHz] into FLACs and WVs, with both NCH Switch Sound File Converter Plus 4.35 and AIMP Converter 4.51 2080. The result was more or less the same, with minimal variations- i.e. a profound lost of high frequency data. I also tried converting those 48 kHz WAVs in Audacity 2.3.2 and Sony Sound Forge 7.0 to 44 kHz- as a result the new WAVs also lost the frequencies above the 20 kHz. And all those losses do not look like a ordinary plain cut-off line- instead of that they look "natural", as if the audio was recorder / released in that way [as opposed to being crassly compressed somewhere further down the line]  
` `  
` `  
So how can this be happening? I've seen this through years over and over with all kinds of files from various sources: spectrogram shows "everything" when the file is 48 kHz, but after conversion / re-sampling it becomes more inferior than it should [loosing not only that what was above the 22 kHz level]. And yes: I double checked the settings of the output formats - it is not a case of wrong settings in multiple pieces of software; to which attests also the fact that I do not have no such problems when converting 44 kHz files that have full frequencies [they retain frequencies up to 22 kHz]

---

<div class="post-metadata">

**Author:** ![TAC109](https://community.mp3tag.de/letter_avatar_proxy/v4/letter/t/848f3c/32.png) [@TAC109](https://community.mp3tag.de/u/TAC109)\
**Post date:** [May 17, 2020, 6:01am UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/2 "2020-05-17T06:01:07Z")

</div>

Read up on the Nyquist Limit.

---

<div class="post-metadata">

**Author:** ![Zerow](https://community.mp3tag.de/user_avatar/community.mp3tag.de/zerow/32/11379_2.png) [@Zerow](https://community.mp3tag.de/u/Zerow)\
**Post date:** [July 13, 2020, 8:26pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/3 "2020-07-13T20:26:53Z")

</div>

So what I understood is this

It is normal that this happens? And that all of the software that I happen to use apparently have such filters that they automatically wipe those frequencies?  
` `  
` `  
And so the new question wold be: is there a sense in looking for some other converters, because they might preserve those frequencies? Or should I work around that issue by using what-u-hear method for creating a new file with desired parameters?

---

<div class="post-metadata">

**Author:** ![TAC109](https://community.mp3tag.de/letter_avatar_proxy/v4/letter/t/848f3c/32.png) [@TAC109](https://community.mp3tag.de/u/TAC109)\
**Post date:** [July 13, 2020, 11:49pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/4 "2020-07-13T23:49:07Z")

</div>

You will know of Scotty's (of Star Trek fame) quote - "You cannot change the laws of physics, Captain!" That’s what the Nyquist Limit is like - a law of physics. If you want to keep the frequencies above 20k, you’ll have to keep your wav files at 48kHz.

---

<div class="post-metadata">

**Author:** ![Zerow](https://community.mp3tag.de/user_avatar/community.mp3tag.de/zerow/32/11379_2.png) [@Zerow](https://community.mp3tag.de/u/Zerow)\
**Post date:** [July 14, 2020, 9:51am UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/5 "2020-07-14T09:51:55Z")

</div>

> [@TAC109](#):
>
> [...]  
> If you want to keep the frequencies above 20k, you’ll have to keep your wav files at 48kHz.

Or... push the record button in an editor set up to create a 44 kHz file and then push the play button of a the 48 kHz source file?  
` `  
` `  
With one song it is not a problem. With an entire album it is a drag. And there is of course the issue of possible glitches in playback and of addition of extra sounds

---

<div class="post-metadata">

**Author:** ![ohrenkino](https://community.mp3tag.de/user_avatar/community.mp3tag.de/ohrenkino/32/4843_2.png) [@ohrenkino](https://community.mp3tag.de/u/ohrenkino)\
**Post date:** [July 14, 2020, 10:50am UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/6 "2020-07-14T10:50:34Z")

</div>

Hi @zerow, old house!  
Did you have a look at the facts of the

> [@TAC109](#):
>
> Nyquist Limit

?  
If not, then we can stop this discussion straight away as it will not come to any result.  
And if you had a look at it, then the discussion will also come to an end as  
@TAC109 already said:

> [@TAC109](#):
>
> You cannot change the laws of physics

---

<div class="post-metadata">

**Author:** ![Zerow](https://community.mp3tag.de/user_avatar/community.mp3tag.de/zerow/32/11379_2.png) [@Zerow](https://community.mp3tag.de/u/Zerow)\
**Post date:** [April 22, 2022, 12:14pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/8 "2022-04-22T12:14:10Z")

</div>

> [@ohrenkino](#):
>
> Hi @zerow, old house!  
> Did you have a look at the facts of the
> 
> > [@TAC109](#):
> >
> > Nyquist Limit
> 
> ?  
> If not, then we can stop this discussion straight away as it will not come to any result.  
> And if you had a look at it, then the discussion will also come to an end as  
> @TAC109 already said:  
> [...]

> [@TAC109](#):
>
> You will know of Scotty's (of Star Trek fame) quote - "You cannot change the laws of physics, Captain!"  
> [...]

If this is true, then how did **I finally managed to convert 48kHz stereo FLACs to both FLACs and WAVs that are 44kHz in format and have frequencies above 20 kHz**?

This time I was using dBpoweramp Music Converter 17.2 and it has no problem with such conversions. And apparently this even works when the source file is a 96 kHz 24-bit one - as I converted such high-end FLAC to WAV, loaded it to that outdated Sony Sound Forge 7.0 audio editor, saved it as [44kHz 16-bit](https://community.mp3tag.de/uploads/short-url/62oykbV6z2SYfXEtTKV5dkgMaf2.zip) and then loaded it once more and save it as [48kHz 16-bit](https://community.mp3tag.de/uploads/short-url/3DtfVYLUCSFaKYuBkZOUEMPzEbU.zip). [Unfortunately Sound Forge does is not capable of saving WAVs with higher values of frequencies and bits - but if it would be needed I can provide that source file in some other format or try to edit it in some other audio editor. And the reason I used an editor to cut out a small piece of music was so that this forum would accept for storage the relatively small size of files] Just use any spectrogram software and compare these two files and see fr yourself, that the 44 kHz version does not have a cut-off line at ~20.5 kHz

Did dBpoweramp Music Converter cancelled that Nyquist Limit? Was Scotty wrong? Or is it me that is committing some sort of fallacy?

---

<div class="post-metadata">

**Author:** ![ohrenkino](https://community.mp3tag.de/user_avatar/community.mp3tag.de/ohrenkino/32/4843_2.png) [@ohrenkino](https://community.mp3tag.de/u/ohrenkino)\
**Post date:** [April 22, 2022, 12:40pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/9 "2022-04-22T12:40:33Z")

</div>

> [@Zerow](#):
>
> If this is true,

It is true.  
The fact is being discussed (no not really, it is a fact) in [Wikipedia](https://en.wikipedia.org/wiki/Nyquist_frequency) at length.  
So if you have any further problems understanding this phenomenon, try to consult training pages for physics.  
Or perhaps a little easier mathematics: you need at least twice the sampling frequency to detect the real frequency. So 44/2=22 plus minus some artefacts =\> 20 something, which is what you got.

---

<div class="post-metadata">

**Author:** ![Zerow](https://community.mp3tag.de/user_avatar/community.mp3tag.de/zerow/32/11379_2.png) [@Zerow](https://community.mp3tag.de/u/Zerow)\
**Post date:** [April 22, 2022, 7:28pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/10 "2022-04-22T19:28:31Z")

</div>

ERGO: I have hallucinations that I see data above ~20.5 kHz in files converted from 48 kHz to 44 kHz with the usage of dBpoweramp Music Converter?

---

<div class="post-metadata">

**Author:** ![ohrenkino](https://community.mp3tag.de/user_avatar/community.mp3tag.de/ohrenkino/32/4843_2.png) [@ohrenkino](https://community.mp3tag.de/u/ohrenkino)\
**Post date:** [April 22, 2022, 7:38pm UTC](https://community.mp3tag.de/t/converting-from-48-khz-to-44-khz-looses-the-highest-frequencies-with-various-software/48971/11 "2022-04-22T19:38:04Z")

</div>

> [@Zerow](#):
>
> ERGO: I have hallucinations ...?

Could you please research the answer to that question yourself?  
You noted that frequencies are lost when you convert to a lower sample rate - that is quite obviously not surprising to anyone who can interpret the Nyquist rule.  
IMHO it will not lead anywhere to discuss one kilohertz more or less
