Quantcast
Once and for all...44.1khz or 48khz? - Gearspace.com
The No.1 Website for Pro Audio
Once and for all...44.1khz or 48khz?
Old 1 week ago
  #1
Here for the gear
 
ToastMalone's Avatar
 
Once and for all...44.1khz or 48khz?

I’d like to get this discussion off on the right foot as I’d really like to get some clear answers/advice on this topic.

I use Ableton 11, a UA Apollo x8p interface and Macbook Pro (Intel). I record vocals, guitars and bass. I use a lot of virtual instruments. I also use some samples.

I used 44.1/24 all of 2020. In 2021 I switched to 48/24. I am still unsure which I should default to and I would ideally like to come to a final decision on this topic.

Here are some highlights of things I’ve considered:

- When I sent my last song away to be mastered by one of the top mastering engineers in the industry, they requested 48/24 as the delivery format.

- 48khz has the smoother rolloff above 20khz and into the audible range which is primarily why I switched to 48khz.

- If most of the samples I use within a session are 44.1khz...would my track overall sound better if there was no need for sample rate conversion within the project? (Ableton itself recommends NOT to mix sample rates within a single project).

- If the delivery format for streaming/CD is almost always 44.1khz...wouldn’t it just make sense to work at 44.1khz?

- There is slightly less latency at 48khz, but it uses more CPU...

- The more I work at 48khz the more I am creating a mismatch of stuff. Now many of my own personal samples are 48khz because I have been working at that sample rate. So now, even if I go back to 44.1khz I still have a bunch of random things and unfinished projects at 48khz...

I would really like to come to a definitive decision on this. Somehow I feel slightly unprofessional working at 44.1khz, it seems like a sample rate that has been well superseded, and most would recommend working at 48khz minimum. But as you can see, I still have some queries.

Please let me know any advice you may have for me on this topic!

Thank you.

P.S. I know some may see this as splitting hairs, but I’m just trying to understand this better. Don’t worry, it’s not preventing me from making music, but it is bugging me enough to start this thread
Old 1 week ago
  #2
Lives for gear
 
🎧 10 years
44.1khz was created as the CD specification for the most arbitrary (read: ridiculous) of reasons:

a person with power and influence wanted to make sure Beethoven's 9th symphony could fit on a single CD. that is literally the reason 44.1khz was chosen. nothing to do with sound quality. the guy wanted to hear Beethoven on one CD.

PCM digital stuffs its chaos in an attic just above the standard hearing range. with 44.1, that is a very tight attic. fabfilter (one of the best designers out there) can go on for hours about how that tight attic creates problems during mixing and signal processing.

48k is only a slightly higher tax on your system resources, but provides meaningfully more space to stuff the digital garbage. 48k has been a standard in video for a long time. 44.1 is a relic from 1980.

just keep working in 48k and moving over to 48k, and things will be fine.
Old 1 week ago
  #3
Gear Maniac
I went on that rabbit hole some time ago, at the end i've read a lot of discussion, and different point of view, tecnical reason vs ideas, at the end one guy told me, to do not care about it, just to try it myself, and feel if there is any audible differences.

At the end i've choose to work at 44khz going with OS on every plugins, the difference it's very very small and you can hear it only if you are looking for something, otherwise no differences, and even if listened with attention, the difference on some songs it's not worse neither better just slightly different.

Last but not least most of the music now (unfortunatly) it's listened on smartphones and earbuds, 44vs48 isn't a real problem imho.

I feel more difference between 32 and 16bit, this is what i really care
Old 1 week ago
  #4
Lives for gear
 
Rick Dalton's Avatar
Quote:
Originally Posted by ToastMalone ➡️
I’d like to get this discussion off on the right foot as I’d really like to get some clear answers/advice on this topic.

I use Ableton 11, a UA Apollo x8p interface and Macbook Pro (Intel). I record vocals, guitars and bass. I use a lot of virtual instruments. I also use some samples.

I used 44.1/24 all of 2020. In 2021 I switched to 48/24. I am still unsure which I should default to and I would ideally like to come to a final decision on this topic.

Here are some highlights of things I’ve considered:

- When I sent my last song away to be mastered by one of the top mastering engineers in the industry, they requested 48/24 as the delivery format.

- 48khz has the smoother rolloff above 20khz and into the audible range which is primarily why I switched to 48khz.

- If most of the samples I use within a session are 44.1khz...would my track overall sound better if there was no need for sample rate conversion within the project? (Ableton itself recommends NOT to mix sample rates within a single project).

- If the delivery format for streaming/CD is almost always 44.1khz...wouldn’t it just make sense to work at 44.1khz?

- There is slightly less latency at 48khz, but it uses more CPU...

- The more I work at 48khz the more I am creating a mismatch of stuff. Now many of my own personal samples are 48khz because I have been working at that sample rate. So now, even if I go back to 44.1khz I still have a bunch of random things and unfinished projects at 48khz...

I would really like to come to a definitive decision on this. Somehow I feel slightly unprofessional working at 44.1khz, it seems like a sample rate that has been well superseded, and most would recommend working at 48khz minimum. But as you can see, I still have some queries.

Please let me know any advice you may have for me on this topic!

Thank you.

P.S. I know some may see this as splitting hairs, but I’m just trying to understand this better. Don’t worry, it’s not preventing me from making music, but it is bugging me enough to start this thread
Heres something you may find interesting, you may need to expand comments.
https://www.quora.com/Which-audio-qu...48KHz-320-kbps
Old 1 week ago
  #5
Lives for gear
 
🎧 10 years
It's very simple.

For storing files it doesn't matter, but you ad/da may work better at a certain frequency. But last 1%...

Where is can matter is when processing with plugins (oversampling, quality of downsampling filters, aliasing). Depending on the plugin you can have better results with higher sampling rates, but this is just due to processing headroom.

For the SRC with different samples, you have to test it. When it's well made there is minimal or no difference. Theoretically you should have less room for error when using 48khz and upsample 44khz material, because the SRC filter works at a higher frequency this way than the other way around.

I like 48khz for the bit more headroom for aliasing filters. But I would like to work even higher because of even more plugin processing headroom.

I think only very highend recordings of natural things should care about if resampling does make a bit difference (again it's mainly the effect of a SRC filter, not that 44.1 sounds worse).

Most other stuff propably benefits more from having higher processing headroom for plugins, where you have SRC all the time!

Last edited by AreYouHuman; 6 days ago at 06:28 PM..
Old 6 days ago
  #6
Lives for gear
 
No reason not to go 48K particularly as we are now a multimedia world and CD is a dying format.
Old 6 days ago
  #7
Lives for gear
 
Analogue Mastering's Avatar
 
🎧 10 years
48 has become the new 44.1 indeed no reason not to move.
Old 6 days ago
  #8
Lives for gear
 
mbvoxx's Avatar
 
2 Reviews written
🎧 10 years
Everyone I send tracks to requests 24/48 so that's how I leave my system configured. Whether it's preferable for one reason or another doesn't factor in.
Old 6 days ago
  #9
Lives for gear
 
🎧 15 years
Quote:
Originally Posted by ToastMalone ➡️
I’d like to get this discussion off on the right foot as I’d really like to get some clear answers/advice on this topic.

I use Ableton 11, a UA Apollo x8p interface and Macbook Pro (Intel). I record vocals, guitars and bass. I use a lot of virtual instruments. I also use some samples.

I used 44.1/24 all of 2020. In 2021 I switched to 48/24. I am still unsure which I should default to and I would ideally like to come to a final decision on this topic.

Here are some highlights of things I’ve considered:

- When I sent my last song away to be mastered by one of the top mastering engineers in the industry, they requested 48/24 as the delivery format.

- 48khz has the smoother rolloff above 20khz and into the audible range which is primarily why I switched to 48khz.

- If most of the samples I use within a session are 44.1khz...would my track overall sound better if there was no need for sample rate conversion within the project? (Ableton itself recommends NOT to mix sample rates within a single project).

- If the delivery format for streaming/CD is almost always 44.1khz...wouldn’t it just make sense to work at 44.1khz?

- There is slightly less latency at 48khz, but it uses more CPU...

- The more I work at 48khz the more I am creating a mismatch of stuff. Now many of my own personal samples are 48khz because I have been working at that sample rate. So now, even if I go back to 44.1khz I still have a bunch of random things and unfinished projects at 48khz...

I would really like to come to a definitive decision on this. Somehow I feel slightly unprofessional working at 44.1khz, it seems like a sample rate that has been well superseded, and most would recommend working at 48khz minimum. But as you can see, I still have some queries.

Please let me know any advice you may have for me on this topic!

Thank you.

P.S. I know some may see this as splitting hairs, but I’m just trying to understand this better. Don’t worry, it’s not preventing me from making music, but it is bugging me enough to start this thread
Well, you've thorougly read the other three gigantic sample rate threads....and then you come here expecting to get different answers than on the other existing threads?

By the way, I have no idea if your Apollo itself does better at 48k vs other rates. I'm certainly not gonna buy one to shoot out its best sounding conversion rates to my ears.
Old 6 days ago | Show parent
  #10
Here for the gear
 
ToastMalone's Avatar
 
Quote:
Originally Posted by thenoodle ➡️
Well, you've thorougly read the other three gigantic sample rate threads....and then you come here expecting to get different answers than on the other existing threads?

By the way, I have no idea if your Apollo itself does better at 48k vs other rates. I'm certainly not gonna buy one to shoot out its best sounding conversion rates to my ears.
I listed my specs in the hope that perhaps there’s something I don’t know about the DAW or interface I use that might be optimised to favor 44.1khz or 48khz. If such a thing is true, if I am lucky perhaps someone who already knows that information will see this thread.

I don’t know what I don’t know. From what I’ve read so far I wasn’t able to make a definitive conclusion (not to mention I still had further questions that went beyond those discussed in the info I was able to find!).
Old 6 days ago | Show parent
  #11
Gear Guru
 
🎧 15 years
Quote:
Originally Posted by thenoodle ➡️
Well, you've thorougly read the other three gigantic sample rate threads....and then you come here expecting to get different answers than on the other existing threads?
That's because those other threads did not include the phrase: "Once and for all" in their title!
Old 6 days ago
  #12
Lives for gear
 
Monotremata's Avatar
All of the previous "Is 44.1 vs whatever" threads are FULL of information and reasons on everyone's personal preference of one or the other. If you cant make an informed decision based on the YEARS of comments already said and done, I don't think the sample rate is your big issue.
Old 6 days ago
  #13
Lives for gear
 
telecode's Avatar
 
1 Review written
🎧 5 years
I use 24/48. My understanding is the plugins work better at that project rate. (if you are using newer plugins).
Old 6 days ago | Show parent
  #14
Here for the gear
 
ToastMalone's Avatar
 
Quote:
Originally Posted by Monotremata ➡️
All of the previous "Is 44.1 vs whatever" threads are FULL of information and reasons on everyone's personal preference of one or the other. If you cant make an informed decision based on the YEARS of comments already said and done, I don't think the sample rate is your big issue.
I couldn’t. Could you? If so, what were your reasons and what would you suggest to me?
Old 6 days ago | Show parent
  #15
Here for the gear
 
ToastMalone's Avatar
 
Quote:
Originally Posted by telecode ➡️
I use 24/48. My understanding is the plugins work better at that project rate. (if you are using newer plugins).
I definitely like the idea of this. If this is the case then that would definitely be a bonus for working at 48khz. As yet I haven’t come across any method of testing this for myself. Does anybody know if there is a way to determine this? I’ve read a few manuals of new plugins lately and I don’t think I’ve seen it mentioned in any of them.
Old 6 days ago | Show parent
  #16
Here for the gear
 
ToastMalone's Avatar
 
Quote:
Originally Posted by joeq ➡️
That's because those other threads did not include the phrase: "Once and for all" in their title!
Damn right!
Old 6 days ago
  #17
I'm also on the 48/24 bandwagon. Minimal system performance and storage impact compared to 44.1/16 but moves the nyquist frequency a bit away from the audible band, gives me a little spectral wiggle room for time-stretching where needed, offers more dynamic range than my preamps, allows for additional dynamic range for tracking/mixing/gain-staging/headroom/compression/expansion while working.

You could use 44.1 and/or 16 bit, but you have to be careful and conscious of its limits whereas with 48/24 you can just do what you need to do and not worry about it. Whereas going to higher rates and depths doesn't really gain me anything (assuming nonlinear plugins offer oversampling) and just consumes proportionally more resources. Anyway, that's my reasoning.

44.1/16 is fine as an output and playback format now that even cheap converters are well up to the task of getting it right.
Old 6 days ago | Show parent
  #18
Lives for gear
 
telecode's Avatar
 
1 Review written
🎧 5 years
Quote:
Originally Posted by ToastMalone ➡️
I definitely like the idea of this. If this is the case then that would definitely be a bonus for working at 48khz. As yet I haven’t come across any method of testing this for myself. Does anybody know if there is a way to determine this? I’ve read a few manuals of new plugins lately and I don’t think I’ve seen it mentioned in any of them.
From what I have seen on some discussions, it seems when the developers develop plugins, they develop them at a native resolution. Going down to a lower resolution does not automatically mean the plugin will run faster or better and might actually run worse. You might also want to check some of the threads on aliasing and plugins. I am primarily a musician and not a developer or engineer. So maybe others can chime in. I just run my projects at 24/48. It seems the best trade off between file size and performance. My interface can apparently run at 192. I tried that once. It did not work out well.
Old 6 days ago
  #19
Lives for gear
 
Sharp11's Avatar
 
🎧 15 years
I track, mix and deliver mixes and stems at 24/48, it’s what my clients expect and it sounds great.
Old 6 days ago | Show parent
  #20
Here for the gear
 
ToastMalone's Avatar
 
Quote:
Originally Posted by telecode ➡️
From what I have seen on some discussions, it seems when the developers develop plugins, they develop them at a native resolution. Going down to a lower resolution does not automatically mean the plugin will run faster or better and might actually run worse. You might also want to check some of the threads on aliasing and plugins. I am primarily a musician and not a developer or engineer. So maybe others can chime in. I just run my projects at 24/48. It seems the best trade off between file size and performance. My interface can apparently run at 192. I tried that once. It did not work out well.
Interesting. Hopefully I can find more info on this to help me understand it better (and know what to test/look for). I tried running at 192khz the other day as well...my CPU basically maxed out immediately, so that’s gotta be a hard no for me
Old 6 days ago
  #21
Old 6 days ago
  #22
Lives for gear
 
🎧 5 years
IMO this is like asking "should I have 10 potato chips with my lunch or 11?"

I agree you've got basically nothing to lose by going to 48, but the odds of you or anyone else hearing a difference are slim to none.
Old 6 days ago
  #23
Lives for gear
 
LASTLAVGH's Avatar
It's 24/48.


You can close the thread now.

Edit: has no one linked the Dan Worrall thing?
https://www.youtube.com/watch?v=-jCwIsT0X8M
Old 6 days ago | Show parent
  #24
Lives for gear
 
🎧 15 years
Quote:
Originally Posted by joeq ➡️
That's because those other threads did not include the phrase: "Once and for all" in their title!
I got mixed up with the other sample rate thread that's labeled "a thread to end them all".

So this one is ...once and for all a thread to end them all.

How many pages before people start linking to the goatee guy with the dot-connect video?
Old 6 days ago
  #25
Gear Maniac
 
🎧 5 years
Quote:
Originally Posted by ToastMalone ➡️
- If most of the samples I use within a session are 44.1khz...would my track overall sound better if there was no need for sample rate conversion within the project? (Ableton itself recommends NOT to mix sample rates within a single project).
Yes! Well, it depends on the samples and maybe there won't be any audible difference.
But Ableton Live's resampling quality is awful. See this thread.
Until that quality improves, I recommend using some other SRC for the files before importing them into Ableton. (But note that if you'll be making pitch changes in Ableton they'll get resampled again anyway.)


As for getting the definite answer: I don't think you'll get one, because it just doesn't really matter that much and there isn't a big difference between those two sample rates.
I like 48k, mostly for practical reasons (it's standard for video/broadcasting/Android...) and partly because it might offer a sonic advantage, but it's not overkill (like 96k). 48k makes using 16bit for delivery a bit easier (there's less need for 24bit), because you can noise shape the dither further up. It's a very small difference, though.

Plugins working better at higher rates? Yeah, but by now plugins that benefit from it already use oversampling internally (or have that option), so you don't need to run the whole project at a higher rate.

In short: I wouldn't lose any sleep over it. If someone or something demands a certain rate you can always do the SRC in the end.
Old 6 days ago | Show parent
  #26
Lives for gear
 
jaddie's Avatar
 
🎧 10 years
Quote:
Originally Posted by gearstudent ➡️
44.1khz was created as the CD specification for the most arbitrary (read: ridiculous) of reasons:

a person with power and influence wanted to make sure Beethoven's 9th symphony could fit on a single CD. that is literally the reason 44.1khz was chosen. nothing to do with sound quality. the guy wanted to hear Beethoven on one CD.
Nope, that's not why they picked 44.1.

The big question was how do we record the bit stream resulting from PCM quantization of two-channel stereo with adequate audio bandwidth? If the sampling frequency was of no concern, the only hardware of the time that could be used would be some rather special and expensive digital instrumentation/data recorders, and there were systems that did that (Soundstream used a custom modified Honeywell 18 track instrumentation drive running at 30-35ips).

But the desire was for a practical and affordable system. There were already some very common and low-cost devices that had the bandwidth: video recorders, specifically the U-Matic format. And if the correct sampling frequency could be chosen, the data could be easily formatted to fit within a standard video frame. However to do that, the sampling frequency would need to work mathematiclly with the number of video lines per field and frame, and the frame rate. If you consider NTSC monochrome video, 44.1/16 two channels plus a block of error correction code, slices up perfectly into an NTSC frame of 30 frames/sec and 525 lines. 48kHz would not. That's also why EIAJ Consumer PCM recording systems ran at 44.056kHz...because cosumer VCRs were pre-set to record only NTSC color, which has a frame rate of 29.97 frames/sec, the sampling frequency had to be changed slightly to fit.

The entire CD recording, editing, and mastering system was running on U-Matic monochrome machines. 44.1 was the only choice.

The story your refer to above, though since shown to be myth, was that Norio Ohga, a Sony exec, wanted the entire Beethoven's 9th Symphony to be recorded on a single CD. In reality, the EFM code they were working on might have been able to store over 90 minutes on a 120mm disc, but the packing density was lowered to 74 minutes, to permit more reliable disc reading.
Quote:
Originally Posted by gearstudent ➡️
PCM digital stuffs its chaos in an attic just above the standard hearing range. with 44.1, that is a very tight attic. fabfilter (one of the best designers out there) can go on for hours about how that tight attic creates problems during mixing and signal processing.
Chaos? In an attic? Wow. Yeah, no actual science. There. The one problem that exists is the anti-aliasing filter needed for 44.1kHz vs 48kHz. There is a slight difference IF you maintaint the same actual -3dB frequency. Otherwise, if you try to actually use that extra 2kHz of bandwidth, the filters are identical in slope.
Quote:
Originally Posted by gearstudent ➡️
48k is only a slightly higher tax on your system resources, but provides meaningfully more space to stuff the digital garbage. 48k has been a standard in video for a long time. 44.1 is a relic from 1980.

just keep working in 48k and moving over to 48k, and things will be fine.
And if you do, you will ALWAYS face resampling for release. What resampling does is calculate new sample data based on old samples by applying a calculated FIR filter to perform interpolation between the new desired samples that don't exist yet because of the difference in sample rate. While the process has been somewhat refined by now, it should be pointed out (to trigger those with OCD) that your original 48kHz data must be altered to create new samples at 44.1kHz. In absolute terms, this is a lossy process. In practical terms, it's pretty transparent, if you don't keep resampling over and over.
Old 6 days ago
  #27
Lives for gear
 
Analogue Mastering's Avatar
 
🎧 10 years
Release is no longer an issue as CD is dead, all major aggregators work 48/24 native rate.
Old 6 days ago | Show parent
  #28
Lives for gear
 
telecode's Avatar
 
1 Review written
🎧 5 years
Quote:
Originally Posted by Analogue Mastering ➡️
Release is no longer an issue as CD is dead, all major aggregators work 48/24 native rate.
i just read an article claiming CD sales are higher than ever.
Old 6 days ago | Show parent
  #29
Lives for gear
 
jaddie's Avatar
 
🎧 10 years
Quote:
Originally Posted by Analogue Mastering ➡️
Release is no longer an issue as CD is dead, all major aggregators work 48/24 native rate.
You mean like iTunes and Spotify and YouTube? All distribute at 44.1KHz, and if content is loaded at anything else, THEY downsample it.

Several will accept 16 or 24 bit files, but 24 bit for distribution is pointless. If there's a bitrate trade-off, take the bitrate and give up bit depth. Always.
Old 6 days ago | Show parent
  #30
Lives for gear
 
Analogue Mastering's Avatar
 
🎧 10 years
Quote:
Originally Posted by jaddie ➡️
You mean like iTunes and Spotify and YouTube? All distribute at 44.1KHz, and if content is loaded at anything else, THEY downsample it.

Several will accept 16 or 24 bit files, but 24 bit for distribution is pointless. If there's a bitrate trade-off, take the bitrate and give up bit depth. Always.
No I don't mean platforms I mean aggregators, distributing to platforms
but also on the platforms there is a bit more to it https://help.apple.com/itc/videoaudi...l#itced32b09c6
📝 Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 1655 views: 178721
Avatar for b0se
b0se 29th December 2020
replies: 342 views: 49872
Avatar for szyam
szyam 6 days ago
replies: 3306 views: 70381
Avatar for Synth Guru
Synth Guru 21st August 2021
Post Reply

Welcome to the Gearspace Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…

Forum Jump
Forum Jump