Why do I have a feeling that you are trying to sell something here?EdgarRothermich wrote: ↑Wed Sep 29, 2021 10:53 pm The new standard is 48kHz
It is always better to convert down than to convert up
Don't assume that technology/standards/requirement doesn't move forward.
Often, they use oversampling internally and don't care if you are stuck in the ancient 44.1kHz world.
What sample rate you're running your DAW(s) on?
-
- KVRAF
- 35410 posts since 11 Apr, 2010 from Germany
-
- Banned
- 380 posts since 13 Jul, 2021
That's my core audience.
-
- KVRAF
- 8802 posts since 7 Oct, 2005
Although I can use 96kHz with my hardware, I use 44.1 kHz most of the times. Sometimes 48kHz but not more.
Most of the plugs I use are oversampled, so why bother with higher resolution that just takes more space and CPU with no noticeable difference in sound?!
Most of the plugs I use are oversampled, so why bother with higher resolution that just takes more space and CPU with no noticeable difference in sound?!
-
- KVRAF
- 2834 posts since 8 Jun, 2018
always make this difference: the sample rate of your recording and the sample rate at which people and yourself, problably, listen.
what do i mean? 96Khz isn't better than 44.100. it hasn't nothing to do with higher frequences..
hi-res audio? those claims, are very doubtful, although there seem to be studies, that give a small difference.
why use higher sample rate: aliasing... Nyquist.
but don't bother, use oversampling of the plugin itself, or it already does it... or use a wrapper, like kushview element ($2), and it oversamples (does not work always, i believe).
i use 48.000, i am used to that, not my ears, but that setting, and... 32 float nowadays.. also not for better sound, but 'treatments', if you do them a lot, o well 24 bits is also enough, or 16, but for CPU or diskspace it does not matter (for me), and some standalone programs, do export 32 float files, and i use those recordings, for further treatment.
but this is about sample rate.
the discussion about hi-res audio is cluttered (??) with a lot of statements, that are most of the times incorrect, and if you call your cables interconnects... well... and buy a powerstrip for $7000, and you think wow it sounds great now, you hear the sound only of the money...
so, perhaps hi-res audio can be something, or not. i doubt it.
but we are talking about sample rates in DAW's, in studio's (with cables and power strips, that will cost less than one when walks into a hifi shop (they still exist!), and buy 'special cables', much less... if they see a studio, they think.... and that i play with my $7000 power strip????).
of course do want you want, i don't have to tell you, who am i, nobody (odysseia..).
but i think sample rate in a studio is mainly for counteract digital artefacts caused by Nyquist, the mirror! the mirro!
and digital artefacts can also be from another source, not only sample rate.
so, in the end sample rate isn't that important.
but why do i use 32 float, 48.000Hz????????
disclaimer: my speakers, for listening to music, come from a so called hifi-shop, where you only can buy stuff that costs as much as big second car, or a bentley, but people dump there old speakers, i take those dumps... o well: one set, of course rejected the cables they wanted to sell me, that costed almost as much as the speakers themselves... and that was the entry level product...
the speakers i bought cost less than Reason..... but sound as good, when they were new, and costed an arm and a leg.
what do i mean? 96Khz isn't better than 44.100. it hasn't nothing to do with higher frequences..
hi-res audio? those claims, are very doubtful, although there seem to be studies, that give a small difference.
why use higher sample rate: aliasing... Nyquist.
but don't bother, use oversampling of the plugin itself, or it already does it... or use a wrapper, like kushview element ($2), and it oversamples (does not work always, i believe).
i use 48.000, i am used to that, not my ears, but that setting, and... 32 float nowadays.. also not for better sound, but 'treatments', if you do them a lot, o well 24 bits is also enough, or 16, but for CPU or diskspace it does not matter (for me), and some standalone programs, do export 32 float files, and i use those recordings, for further treatment.
but this is about sample rate.
the discussion about hi-res audio is cluttered (??) with a lot of statements, that are most of the times incorrect, and if you call your cables interconnects... well... and buy a powerstrip for $7000, and you think wow it sounds great now, you hear the sound only of the money...
so, perhaps hi-res audio can be something, or not. i doubt it.
but we are talking about sample rates in DAW's, in studio's (with cables and power strips, that will cost less than one when walks into a hifi shop (they still exist!), and buy 'special cables', much less... if they see a studio, they think.... and that i play with my $7000 power strip????).
of course do want you want, i don't have to tell you, who am i, nobody (odysseia..).
but i think sample rate in a studio is mainly for counteract digital artefacts caused by Nyquist, the mirror! the mirro!
and digital artefacts can also be from another source, not only sample rate.
so, in the end sample rate isn't that important.
but why do i use 32 float, 48.000Hz????????
disclaimer: my speakers, for listening to music, come from a so called hifi-shop, where you only can buy stuff that costs as much as big second car, or a bentley, but people dump there old speakers, i take those dumps... o well: one set, of course rejected the cables they wanted to sell me, that costed almost as much as the speakers themselves... and that was the entry level product...
the speakers i bought cost less than Reason..... but sound as good, when they were new, and costed an arm and a leg.
Primoridal Music: sadà\exposadà - Indusrial & Expanding Your Mind Hurts: Sound Brut
-
- KVRAF
- 35271 posts since 14 Sep, 2002 from In teh net
-
RugerioDelStereo RugerioDelStereo https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=466409
- KVRist
- 150 posts since 23 May, 2020
44.1. I use a lot of effects plugins to mangle sounds, so need the processing power for that. Also I am making industrial/electro which has not really a hi-fi sound.
-
- KVRian
- 1244 posts since 14 Apr, 2008 from /* whitenoise */
I started working ITB around 2005 and got educated by a lot of experience people during my time working in the digital realms. Which sample rate to use depends highly on your specific setup and style of music. Do you use midi software synthesizer only? Do you record a lot of material? How do you process your signals? And so forth.
I learned that the integrity of the audio signal is the most important thing. That means to change (regarding sample rate and bit depths) the signal as less as possible.
Dan Worral made a really good video about this topic.
https://m.youtube.com/watch?v=-jCwIsT0X8M
For my own stuff I produce and mix at my home studio I use 24bit 48khz. I produce using some hardware synthesizer and sampler which I record and from there I'm ITB only. Effects and mixing is done ITB and I use oversampling carefully and mostly just for saturation, distortion, wave shaping and stuff. I use TDR plugins on my master chain in HQ mode as well as apShaper in HQ mode before export.
I learned that the integrity of the audio signal is the most important thing. That means to change (regarding sample rate and bit depths) the signal as less as possible.
Dan Worral made a really good video about this topic.
https://m.youtube.com/watch?v=-jCwIsT0X8M
For my own stuff I produce and mix at my home studio I use 24bit 48khz. I produce using some hardware synthesizer and sampler which I record and from there I'm ITB only. Effects and mixing is done ITB and I use oversampling carefully and mostly just for saturation, distortion, wave shaping and stuff. I use TDR plugins on my master chain in HQ mode as well as apShaper in HQ mode before export.
Last edited by noiseresearch on Thu Sep 30, 2021 11:05 am, edited 1 time in total.
/* whitenoise */ /* abandon */ /* reincarnated */
-
- KVRist
- 126 posts since 13 Aug, 2017 from Gothenburg
48kHz, mainly so that I can use the "class compliant" driver on mac and don't have to install anything (Line6 Helix).
-
- KVRist
- 279 posts since 27 Aug, 2015 from Paris, France
Good ol' 44.1kHz here.
There's an article here that I wouldn't pretend to fully understand, but that seems to indicate this is the case : https://www.psaudio.com/copper/article/ ... onversion/
Take my words as a question more than a truth
Doesn't it depend on the target audience ? I mean, if you're working for TV, you'll need a 48kHz, so that's ok. But if your work is for an audio CD or streaming service, you will have to provide a 44.1kHz file, which means your 48kHz will have to be converted to 44.1, and I've read that converting to a non-integer factor will lead to some losses/artifacts. Which would mean that it might not be such an improvement.LeVzi wrote: ↑Wed Sep 29, 2021 2:32 pmYeah the improvements are tiny but any are good right ?FapFilter wrote: ↑Wed Sep 29, 2021 1:10 pmI'm not sure it will bring much of quality improvements. 48k just happens to be a standard for TV and cinema, as far as i'm concerned.
88k and up should bring you the benefits you are getting from oversampled plugins without their drawbacks of additional sample rate conversions.
With the big drawback of course that you will use a lot more CPU and HDD space
There's an article here that I wouldn't pretend to fully understand, but that seems to indicate this is the case : https://www.psaudio.com/copper/article/ ... onversion/
Take my words as a question more than a truth
Soundcloud - Synthwave & More https://soundcloud.com/canapelee
-
- KVRAF
- 1863 posts since 11 Apr, 2008
96k for music and sound design. I don't need to deal with aliasing, beg devs to add OS to plugins etc.
For many plugins, I can hear the difference with the benefit of a higher sample rate. Not all of them benefit from a higher sample rate tho. It's simply more convenient to me than switching OS in plugins - I don't need to care about it anymore.
Also, I do sound design and a higher sample rate has another benefit: I can slow down samples (recorded in 96k or higher) without affecting its quality (to some extend of course). Development of sounds like dragons and monsters... ultrasonic microphones rules but I can't handle 384kHz sessions yet so it's converted to 192k or 96k... sad face!
It also applies to recoding analogue synths for music. I can slow down track by half and have no audible degradation (high-frequency loss).
When I do work that has to be in 44.1k or lower (mobile devices), then I work in 44.1k. At that point, there are no benefits in working at a higher sample rate if the final format will not even have anything above 12kHz, will be compressed sometimes even to ADPCM and will be listened through the phone speaker
The evolution of technology is great and I can't wait to switch to 192k at some point for good
Ps. I don't care that others can't hear any difference in higher sample rates or they don't even understand how digital audio works. I just came here to give an answer to OT, not to argue or tell others what SR they should use.
For many plugins, I can hear the difference with the benefit of a higher sample rate. Not all of them benefit from a higher sample rate tho. It's simply more convenient to me than switching OS in plugins - I don't need to care about it anymore.
Also, I do sound design and a higher sample rate has another benefit: I can slow down samples (recorded in 96k or higher) without affecting its quality (to some extend of course). Development of sounds like dragons and monsters... ultrasonic microphones rules but I can't handle 384kHz sessions yet so it's converted to 192k or 96k... sad face!
It also applies to recoding analogue synths for music. I can slow down track by half and have no audible degradation (high-frequency loss).
When I do work that has to be in 44.1k or lower (mobile devices), then I work in 44.1k. At that point, there are no benefits in working at a higher sample rate if the final format will not even have anything above 12kHz, will be compressed sometimes even to ADPCM and will be listened through the phone speaker
The evolution of technology is great and I can't wait to switch to 192k at some point for good
Ps. I don't care that others can't hear any difference in higher sample rates or they don't even understand how digital audio works. I just came here to give an answer to OT, not to argue or tell others what SR they should use.
-
- KVRAF
- 6458 posts since 17 Dec, 2009
Standard where tho? Its not standard anymore i thinkDJ Warmonger wrote: ↑Wed Sep 29, 2021 2:42 pm 88.2 KHz. Which is double oversampling of standard frequency.
There are many benefits of it, especially reduced aliasing but not only.
Using Ryzen 3900 for the task, but older 1700 also handled it nicely.
-
- KVRian
- 1000 posts since 25 Feb, 2008 from Sydney, Australia
44.1kHz
What gets me, is that if someone requests a 24bit 48k file, I can just work at 44.1, export to 44.1, and simply upsample the mixdown, and the client won't, or can't, know.
If they can't know, without being specifically told, then something very nebulous is afoot.
What gets me, is that if someone requests a 24bit 48k file, I can just work at 44.1, export to 44.1, and simply upsample the mixdown, and the client won't, or can't, know.
If they can't know, without being specifically told, then something very nebulous is afoot.
Prestissimo in Moto Perpetuo
-
- KVRAF
- 8823 posts since 6 Jan, 2017 from Outer Space
Anything above 20 kHz of your sound is for dogs and bats!
Even the 2 kHz difference between 44.1 and 48 kHz do make a difference. Its the analog filters you need before your DAC for recording. On 44100 hz the filter needs to cut between 20 and 22.05 kHz by maybe 50dB, which creates all kinds of ripple in the audible area. On 48 kHz it needs to be only half as steep. On 96 kHz a simple filter would be sufficient already… (a 24 dB per octave filter would cut it already by 48 dB, as it may alias into the inaudible area between 20 kHz and nyquist. That gives you about 2 octaves range for the filter. Whereas the range between 20 and Nyquist at 48 kHz is a much smaller intervall…
I would always prefer to record on 96 kHz, the difference is pretty audible even on cheap audio interfaces. Later for processing and mixing it depends on your non-linear fx like saturation. If they have internal oversampling you are fine to run your session in 48 kHz like I do mostly.
Sample rate conversion is pretty good nowadays, and as Edgar said downsampling is better than the other way around…
