The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
48kHz
Old 25th November 2014
  #31
Lives for gear
 
PatrickFaith's Avatar
 

I was at home consumer electronic store today buying xmas stuff, ton's of 7.1 surround sound for consumers there (i'd say 4k and 7.1 was the big push in the store). I saw ton's of 7.1 focused on gaming and object oriented sound also. Anyway thats 48/16 minimum and often 96/24.

I'm doing everything now at 96/24, to me it's not so much the frequency resolution (which seems to always be the big debate on gearslutz, but i think is off point in the discussion) but the dynamic range for object oriented sound and the need to resolve phase relationships in rooms.
Old 25th November 2014
  #32
Lives for gear
 
Dpro's Avatar
Quote:
Originally Posted by drz View Post
This is Rupert Neves opinion about the over-20kHz-frequency-capturing theory in this video at 41:00, actually 44:00, but youve got to bring a little time, he is developping it slowly.
short version:
some mixer asked Rupert Neve to look at some channels at a console, because they didnt sound as good as the others, and they measured a difference between those channels that peaked at 54 kHz.
that had an audible (good) effect that is audible in the frequency range we humans are able to hear.
He is also referring to a study of a japanese scientist in the manual of my Portico II channel

But he is talking also about recording devices that capture at sampling rates in the MHz range, if I understood correctly..
He is most likely citing Ohashi who's theories on high freqs and the test he did to prove them have never been repeatable. In the Scientific world if a persons theory and test to prove it cannot be repeated then it remains theoretical at best.
Rupert Neve was a good electrical engineer who designed some great sounding boards but that does not make him knowledgable in the world of digital audio or its intricacies.

Like others have stated the highest freq the human ear can possibly hear tops out around 22kz and that is usually a baby . Most Golden Ears hear around I would say 19k-20k with a lot of,the great engineers probably peeking in the 16-17k range.
Given that a sample rate of double 22k ala Nyquist is going to give us all we can possibly hear. Anything above pretty much can lower the noise floor but the freqs up there we wil not hear. Nor do I believe they have any cause on the freq we do hear. In fact I don't think anyone has proven scientifically that it does . Even what Lavry siad about 60k was theoretical.
You can read up on what JJ Johnston has to say about this as he is the father of several codecs and knows this stuff better than the revered Mr Neve.
Old 25th November 2014
  #33
Lives for gear
 
Grant Ransom's Avatar
 

Quote:
Originally Posted by PatrickFaith View Post
I was at home consumer electronic store today buying xmas stuff, ton's of 7.1 surround sound for consumers there (i'd say 4k and 7.1 was the big push in the store). I saw ton's of 7.1 focused on gaming and object oriented sound also. Anyway thats 48/16 minimum and often 96/24.

I'm doing everything now at 96/24, to me it's not so much the frequency resolution (which seems to always be the big debate on gearslutz, but i think is off point in the discussion) but the dynamic range for object oriented sound and the need to resolve phase relationships in rooms.
Going over 44.1kHz SR does one thing: It moves the aliasing filter upwards. These filters have been under development for decades. You will probably have to be told that something is at 96kHz - as the differences are subjective and depend on the hardware. Many systems themselves sound worse at Hi-SR.
Higher SR itself does not do anything to stereo imaging.

As people have pointed out time after time here: Once you can describe perfectly a frequency waveform at 20k (and therefore everything below) placing more sample points on the waveform achieves nothing in the human hearing bandwidth.
You CAN however lose accuracy as you drive sampling systems faster - especially consumer ones.

"Object oriented sound"?
"resolve phase relationships in rooms"? How would one do that with the audio system sample rate and bit depth?
Old 25th November 2014
  #34
Gear Head
 

I love my high-end CD player. I personally find original CD's more appealing than downloaded music.

My high-end CD player also has the ability to load up to 192khz WAV files if I wish.

With SRC technology today, is there really any audible differences when converting the final product down from 48khz to 44.1khz?

Despite my recently acquired collection of CDs, I myself work in 48khz. Recording musicians to 96khz and SRC'ing the individual tracks down to 48khz seems to be the norm nowadays.
Old 26th November 2014
  #35
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by Grant Ransom View Post
Going over 44.1kHz SR does one thing: It moves the aliasing filter upwards. These filters have been under development for decades. You will probably have to be told that something is at 96kHz - as the differences are subjective and depend on the hardware. Many systems themselves sound worse at Hi-SR.
Higher SR itself does not do anything to stereo imaging.

As people have pointed out time after time here: Once you can describe perfectly a frequency waveform at 20k (and therefore everything below) placing more sample points on the waveform achieves nothing in the human hearing bandwidth.
You CAN however lose accuracy as you drive sampling systems faster - especially consumer ones.

"Object oriented sound"?
"resolve phase relationships in rooms"? How would one do that with the audio system sample rate and bit depth?
To my ears, these filters are a big deal, even those in the nicest newest converters effect the sound a lot. Comparing an analog or DSD recording (which lack the filters) to a PCM conversion of the same thing, the sound definitely changes. In a general sense, sound quality (notice I didn't say resolution) increases with greater filter bandwidth and gentler slope even beyond 384khz, well up into the mhz range. I know there are associated problems with quantization noise as a tradeoff, however.

To my understanding, for both the filter and analog circuit of the ADC, higher bandwidth equals faster slew rate, which I think is a hugely overlooked element of sound quality. This is what Mr. Neve was talking about, I'm assuming.

I hear it talked about a lot, I haven't personally heard a system that performed worse at 192k, 96k, 2.8 or 5.6mhz. Some have sounded different or equivalent, but I haven't heard one that sounded worse.
Old 26th November 2014
  #36
Lives for gear
 

Quote:
Originally Posted by IanBSC View Post
To my ears, these filters are a big deal, even those in the nicest newest converters effect the sound a lot. Comparing an analog or DSD recording (which lack the filters) to a PCM conversion of the same thing, the sound definitely changes. In a general sense, sound quality (notice I didn't say resolution) increases with greater filter bandwidth and gentler slope even beyond 384khz, well up into the mhz range. I know there are associated problems with quantization noise as a tradeoff, however.
The modern converters now all over-sample and so the aliasing filters are the same for DSD and PCM. Lower sample rate PCM output is then achieved with very high quality SRC in the digital domain.

In fact, this converter has exactly the same A/D front end for DSD and PCM. It is used in the Merging Horus.
Arda Technologies - AT1201 Flagship Audio A/D Converter
Old 27th November 2014
  #37
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by David Spearritt View Post
The modern converters now all over-sample and so the aliasing filters are the same for DSD and PCM. Lower sample rate PCM output is then achieved with very high quality SRC in the digital domain.

In fact, this converter has exactly the same A/D front end for DSD and PCM. It is used in the Merging Horus.
Arda Technologies - AT1201 Flagship Audio A/D Converter
I'm pretty sure oversampling ADCs still have to use a digital decimation filter, often with a very sharp slope, for downsampling.

I've actually encountered the Horus before, our mastering engineer uses one, and there is still a huge difference between standard sample rate PCM audio and DSD.
Old 22nd January 2015
  #38
Gear Addict
I've been helping out a band I recorded with a music video using Final Cut Pro X. The audio was mastered at 44.1 but Final Cut Pro sample rate converts it to output @ 48kHz (there is no other option) when it renders the video. Then, when the video is uploaded to Youtube, it will be transcoded and audio sample rate converted back to 44.1. In other words the audio is converted unnecessarily from 44.1 > 48 > 44.1 and there is a noticeable hit in sound quality. I've also realized through this process that most video editors (even the pros) don't know anything about audio.

Wanted to bump this thread as a real example of why 48kHz should become the standard audio sampling rate. Yes, I realize it's not likely going to happen though :>)

Last edited by captainj; 22nd January 2015 at 06:56 AM..
Old 22nd January 2015
  #39
Gear Guru
 
Brent Hahn's Avatar
 

Quote:
Originally Posted by captainj View Post
... the audio is converted unnecessarily from 44.1 > 48 > 44.1 and there is a noticeable hit in sound quality. <> ... a real example of why 48kHz should become the standard audio sampling rate.
Made me curious. Here are two short 44.1 WAV's, taken from a live radio show we do here. One has been converted from 44.1 to 48 and then back to 44.1, the other hasn't. See if you can tell the difference.
Attached Files

Sample_Rate_Test_A.wav (3.64 MB, 438 views)

Sample_Rate_Test_B.wav (3.64 MB, 648 views)

Old 23rd January 2015
  #40
Quote:
Originally Posted by captainj View Post
I've been helping out a band I recorded with a music video using Final Cut Pro X. The audio was mastered at 44.1 but Final Cut Pro sample rate converts it to output @ 48kHz (there is no other option) when it renders the video. Then, when the video is uploaded to Youtube, it will be transcoded and audio sample rate converted back to 44.1. In other words the audio is converted unnecessarily from 44.1 > 48 > 44.1 and there is a noticeable hit in sound quality. I've also realized through this process that most video editors (even the pros) don't know anything about audio.

Wanted to bump this thread as a real example of why 48kHz should become the standard audio sampling rate. Yes, I realize it's not likely going to happen though :>)
What makes you think YouTube is down sampling to 44.1? You can upload videos at 48k sample rate...And download them off at the same rate if you have the tools.
Old 23rd January 2015
  #41
Lives for gear
 

Quote:
Originally Posted by captainj View Post
The standard sample rate for music releases should be changed from X to X kHz.

-The video standard is already 48kHz......

CD is the only thing that has been tying us to 44.1 ........

isn't it time to standardize?..........

-Many people (myself included) believe that 48kHz sounds .........

I'd like to hear people's thoughts on this............
What do you care? You track music, right? Do whatever you want for tracking. There are no "standards". There are only a multitude of "versions" your clients will pay you for......

They want a 44.1 mix? Fine, do it.
They want a 48k mix? Fine, do it.
They want a 96k mix? Fine do it... but if you want to collect the dough for that one, you're gonna have to track in 96k.... which maybe you don't want to... which is okay... you just lose the extra $ for the project. And don't think you can just upsample.... the audiophiles will crucify you when they find out.
Neil Young and Pono wants your 192k mixes?.... Fine... do it.... but you'll have to track in 192. And remember, in Neil's head, 192k IS the standard now.

Again... there are no real standards. The last one I had to really live with was cutting to vinyl. Now, there's a headache standard.

What you have now is an ala carte situation where you can mix to any old rate that people care to pay for...... evergreen income for you when you think about it.

Now, if you really like 48k as your self-imposed standard..... Neil Young isn't going to like it much.

If I were going to be picky about standards, I'd vote that Les Pauls should once again be the standard. But it's shaky... we now seem to be so tied to the Stratocaster standard.
Old 24th January 2015
  #42
Gear Guru
 
Brent Hahn's Avatar
 

Quote:
Originally Posted by thenoodle View Post
If I were going to be picky about standards, I'd vote that Les Pauls should once again be the standard. But it's shaky... we now seem to be so tied to the Stratocaster standard.
Neil probably doesn't much like that, either.
Old 24th January 2015
  #43
Lives for gear
 
GP_Hawk's Avatar
Quote:
Originally Posted by Brent Hahn View Post
Neil probably doesn't much like that, either.
lol
Old 25th January 2015
  #44
Here for the gear
 

Re: the CD vs stream debate. I thought Taylor Swift was quite savvy in releasing first on physical format CD/vinyl and then later on to streaming sites. I agree though that there doesn't seem a huge difference in fidelity between 44.1 and 48k.
88.2 and 96k does sound much better to my ears (in my humble studio).
Old 25th January 2015
  #45
Lives for gear
 
matucha's Avatar
I hate 48k for music when the target is CD. Handling different sample rates makes things messy and it's easier to make some kind of sample rate related mistake (you can't really hear something is wrong right away when you don't know the songs, not in some genres at least).

I also had one ridiculous request from a label to supply wavs as 24/96 even though the recording was 24/44.1. There was no way in the world for me to explain them how wrong it is on every level. Customer is always right.

When something goes from 88.2/96 to 48, I don't feel as I'm loosing the most important info from the sound. 44.1 crosses that line. On the other hand I'm perfecly comfortable with 44.1 all the way through the process when working with pop/rock/electronic music. Perhaps the most frustrating thing is when you loose something you really liked after the SRC and the compensation via EQ doesn't sound as good.
Old 27th January 2015
  #46
Quote:
Originally Posted by matucha View Post
I hate 48k for music when the target is CD. Handling different sample rates makes things messy and it's easier to make some kind of sample rate related mistake (you can't really hear something is wrong right away when you don't know the songs, not in some genres at least).
If you're externally clocking, you should always be extra careful!

Quote:
Originally Posted by matucha View Post
I also had one ridiculous request from a label to supply wavs as 24/96 even though the recording was 24/44.1. There was no way in the world for me to explain them how wrong it is on every level. Customer is always right.
We've all had them...

Quote:
Originally Posted by matucha View Post
When something goes from 88.2/96 to 48, I don't feel as I'm loosing the most important info from the sound. 44.1 crosses that line. On the other hand I'm perfecly comfortable with 44.1 all the way through the process when working with pop/rock/electronic music. Perhaps the most frustrating thing is when you loose something you really liked after the SRC and the compensation via EQ doesn't sound as good.
With most mastering engineers at a certain level doing some form of OTB processing, this is less of an issue. Replay @ 48k , re-record in @44.1 No SRC required.
Post Reply

Welcome to the Gearslutz 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…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump