The No.1 Website for Pro Audio
Why does everyone want 96khz ?
Old 10th December 2019
  #781
Lives for gear
Quote:
Originally Posted by ardis View Post
Over projection systems it absolutely does...... You don't play the Nagra tape projecting film...... Mag stripe or optical sounds better FASTER.....35mm sounds better than 16mm and 24fps is a faster frame rate than 18fps.....

How can you guys not get this? .....
I do get this, someone said they chose 24 fps because it meant better sound quality. I was pointing out that the fps does not effect sound quality.

I am an editor and have spliced both 16mm and 35mm film.
Old 10th December 2019
  #782
Lives for gear
 
cjogo's Avatar
Trying to listen to a specific KHz
4
Share
Old 10th December 2019
  #783
Here for the gear
 

Quote:
Originally Posted by kennybro View Post
Quote:
Originally Posted by TobeNaive View Post
I'll start at digital recording limitations and work out from there.
Seems very backwards to assume that everything you are doing involved with accurately capturing and reproducing a guitar sound (everything from guitar selection to mics to processing to monitor choice and placement to room treatment to etc... etc...) is dead on the money, and that digital recording technology is the element holding you back from sonic guitar nirvana. But whatever...

The logic-conundrum with this is that how could you ever know what is causing the problem if you focus only on the medium of digital recording technology before looking at anything else in the chain? That is the ONE THING in the entire chain that is the least within your control. Do you always focus on element of process outside of yourself first when you are having a problem accomplishing something?
I wouldn't say or assume that, I just happen to have the capability to use 96k right now and have decided I'm going to. I am more than happy to support innovation on all fronts, and this is just a piece of that puzzle. The fact that it looks like the market is finally moving in that direction makes it very easy for me to literally simply click a button and make it happen.

Quote:
Originally Posted by kennybro View Post
I suppose you might claim that if you do everything exactly the same, and record to analog tape, the playback sounds just like you in a room playing your guitar? Again... whatever. Problem solved for you. Why even bother with digital?
The funny thing is a lot of the HD audio tracks available are analog tape transferred to the 48k or 96k or 192k format.

It's interesting that the most revered recordings of all time seem to have been recorded on analog tape.

It would be interesting to see a format perfectly emulate analog tape to the point that analog tape could truly be seen as redundant by its fiercest proponents, but sadly that hasn't happened.

I'm excited at the moment just to see CD quality come back as the base tier HD streaming option.
Old 10th December 2019
  #784
Gear Guru
Quote:
Originally Posted by Mr Funk View Post
I do get this, someone said they chose 24 fps because it meant better sound quality. I was pointing out that the fps does not effect sound quality.

I am an editor and have spliced both 16mm and 35mm film.
Then you should understand that faster projection speed absolutely effects sound playback. You're a film editor and don't get that?!
Old 10th December 2019
  #785
Lives for gear
 
kennybro's Avatar
Quote:
Originally Posted by ardis View Post
Over projection systems it absolutely does...... You don't play the Nagra tape projecting film...... Mag stripe or optical sounds better FASTER.....35mm sounds better than 16mm and 24fps is a faster frame rate than 18fps.....

How can you guys not get this? .....
This is absolutely true. I think the rift comes with the "reason for choosing" the frame rate being because of sound quality. 24 was chosen because that was the rate with the "sound on synced disc," Vitaphone system, not specifically because it sounded better; although there is not doubt that 24 certainly sounds better than 16, with sound-on-film tech, either optical or mag stripe.

Just like a 16 track 2 inch machine sounds better than a 24, and 30ips sounds better than 15ips (although not for lo freq response). More tape real estate used = better sound quality.
1
Share
Old 10th December 2019
  #786
Gear Guru
Quote:
Originally Posted by kennybro View Post
This is absolutely true. I think the rift comes with the "reason for choosing" the frame rate being because of sound quality. 24 was chosen because that was the rate with the "sound on synced disc," Vitaphone system, not specifically because it sounded better; although there is not doubt that 24 certainly sounds better than 16, with sound-on-film tech, either optical or mag stripe.

Just like a 16 track 2 inch machine sounds better than a 24, and 30ips sounds better than 15ips (although not for lo freq response). More tape real estate used = better sound quality.
Respectfully I'm no film historian but you can't say that faster sound playback wasn't a consideration in choosing 24fps over 18fps especially with 35mm film. If sound was synced to disc logically the sound playback system had to travel faster no? I'm certainly no expert on the Vitaphone system but sync speed is sync speed...... Of course that doesn't matter if it was an independent system...

BTW optical tracks also sound better at high speed (more detail) .....
Old 11th December 2019
  #787
Lives for gear
 
norfolk martin's Avatar
 

Quote:
Originally Posted by David Rick View Post
Yo
Everyone agrees that there's still room for improvement in speakers. Someday maybe I'll finally retire my ADAMs and replace them with your yet-to-be-announced product.

David
I hope the new super transducer product causes vibrations to radiate from any number of specific 3D points in the space arouse the listener, replicating the positions of the actual instruments

No more relying on the "stereo effect"!

You'd probably have to have some limiting parameters to stop the sound of an orchestra 30 feet away from you appearing outside the wall of your room and annoying the neighbors, or some idiot producer putting a trumpet 1" from your ear, however.
Old 11th December 2019
  #788
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by TobeNaive View Post
I wouldn't say or assume that, I just happen to have the capability to use 96k right now and have decided I'm going to. I am more than happy to support innovation on all fronts, and this is just a piece of that puzzle. The fact that it looks like the market is finally moving in that direction makes it very easy for me to literally simply click a button and make it happen.



The funny thing is a lot of the HD audio tracks available are analog tape transferred to the 48k or 96k or 192k format.

It's interesting that the most revered recordings of all time seem to have been recorded on analog tape.

It would be interesting to see a format perfectly emulate analog tape to the point that analog tape could truly be seen as redundant by its fiercest proponents, but sadly that hasn't happened.

I'm excited at the moment just to see CD quality come back as the base tier HD streaming option.
Emulating tape perfectly is one thing, but DSD 5.6mhz can pretty much perfectly reproduce recordings off tape and sound at least as good. Unfortunately the workflow is non-starter for most modern music production, and it doesnt add the tape artifacts that some people want.
Old 11th December 2019
  #789
Gear Addict
 

Quote:
Originally Posted by norfolk martin View Post
I hope the new super transducer product causes vibrations to radiate from any number of specific 3D points in the space arouse the listener, replicating the positions of the actual instruments

No more relying on the "stereo effect"!

You'd probably have to have some limiting parameters to stop the sound of an orchestra 30 feet away from you appearing outside the wall of your room and annoying the neighbors, or some idiot producer putting a trumpet 1" from your ear, however.
Funny, but two channels is all you need.. and the right signal.
Two ears, two channels. Binaural!
That simple.
Old 11th December 2019
  #790
Lives for gear
 
Plush's Avatar
96kHz. or die. 192k to feel good. 768k--yeah!! Get there with Internet Company daw.
Old 11th December 2019
  #791
Lives for gear
 
kennybro's Avatar
Quote:
Originally Posted by ardis View Post
Respectfully I'm no film historian but you can't say that faster sound playback wasn't a consideration in choosing 24fps over 18fps especially with 35mm film. If sound was synced to disc logically the sound playback system had to travel faster no? I'm certainly no expert on the Vitaphone system but sync speed is sync speed...... Of course that doesn't matter if it was an independent system...

BTW optical tracks also sound better at high speed (more detail) .....
No debate. Optical as well as mag stripe sounds better at faster speeds. But Vitaphone (as well as other non-file disc systems) had zero connection to film speed other than syncing to it. 24 fps was used for A Plantation Act, (1926) which was actually for first talkie using Vitaphone, not The Jazz Singer (1927).

Just saying that I've never found any evidence that engineers involved with sound-on-film decided on 24 fps specifically because it sounded better. The standard was adopted, according to sources I've been able to discover, because The Jazz Singer was massively successful, and uses of the sound-on-film technology wanted to be technologically "in sync" with the most successful company (Warner) and their most successful film of the day. Warner was producing hundreds of shorts using Vitaphone for projection in thousands of theaters across the US.

Standardization was the reason. Sure, faster film speed sounded better for sound-on-film, but why not choose 30, and make it even better, if that was the core reason?

Please.... research this in the stacks, or just do simple google searches. You will find that every source points to persistence of vision plus film stock consumption as the reasons for 24. 24 was first chosen by the users of Vitaphone, and their sound considerations were totally disconnected from film speed.
Old 11th December 2019
  #792
Lives for gear
 
kennybro's Avatar
Quote:
Originally Posted by Mr Funk View Post
I do get this, someone said they chose 24 fps because it meant better sound quality. I was pointing out that the fps does not effect sound quality.

I am an editor and have spliced both 16mm and 35mm film.
Actually, mag stripe certainty sounds somewhat better at 24 than at 16, just like tape recorder speed/quality connection. I believe optical also better at higher speeds.

But yes you are right... This is not the core reason it was chosen. That's the sticking point.
Old 11th December 2019
  #793
Motown legend
 
Bob Olhsson's Avatar
 

This matches my understanding:

https://www.premiumbeat.com/blog/adv...o-frame-rates/
2
Share
Old 11th December 2019
  #794
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by sax512 View Post
But then again, what SHOULD your mix really sound like, if it were accurately reproduced?
One can't know, because sound capture and reproduction is an unnatural process in almost all of its stages, with the exception of binaural.
Within the context of analog to digital conversion, the recording should sound identical to the input signal chain. It's not more complicated than that, the rest is up to the engineer's discretion.
Old 11th December 2019
  #795
Lives for gear
 
kennybro's Avatar
Quote:
Originally Posted by Bob Olhsson View Post
This matches my understanding:
Thanks Bob. Nice article. Haven't seen this one.
Old 11th December 2019
  #796
Gear Addict
 

Quote:
Originally Posted by IanBSC View Post
Within the context of analog to digital conversion, the recording should sound identical to the input signal chain. It's not more complicated than that, the rest is up to the engineer's discretion.
I already explained why this comparison test is easier said than done. If I explained it again, I expect you to disregard my explanation once again.
Go back and read my answer, if you really want to know it. Otherwise, I'll continue to ignore you, as you add no value to this conversation, in my opinion.
Old 11th December 2019
  #797
Lives for gear
Quote:
Originally Posted by ardis View Post
Then you should understand that faster projection speed absolutely effects sound playback. You're a film editor and don't get that?!
I’m having one last attempt then I’m giving up trying to explain my point. A Nagra or DAT or DIVA or whatever, records at a set speed/resolution, irrespective of what frame rate the camera is shooting. So the quality of the sound has been set at the capture stage.
Old 11th December 2019
  #798
Lives for gear
Quote:
Originally Posted by kennybro View Post
Actually, mag stripe certainty sounds somewhat better at 24 than at 16, just like tape recorder speed/quality connection. I believe optical also better at higher speeds.

But yes you are right... This is not the core reason it was chosen. That's the sticking point.
Yes, of course once you transfer the captured audio to another medium, the medium will make a difference on how much it degrades that audio
1
Share
Old 11th December 2019
  #799
Lives for gear
 
monomer's Avatar
 

Quote:
Originally Posted by David Rick View Post
For localization, a minimum-phase filter would be much better, but then we have phase distortion.
Why would you ever consider a minimum phase filter in relation to localization, when there are linear phase filters?
Wouldn't you want a filter that leaves all the time relations between frequencies alone?
Old 11th December 2019
  #800
Gear Addict
 

Quote:
Originally Posted by monomer View Post
Why would you ever consider a minimum phase filter in relation to localization, when there are linear phase filters?
Wouldn't you want a filter that leaves all the time relations between frequencies alone?
If I understood him correctly (David, please forgive me if I didn't), his point is that the part of a signal that is limited to the audio band, with all ultrasonics removed, is audibly different from the whole original signal, because some audible 'time smear' happens, somehow.
Old 11th December 2019
  #801
Quote:
Originally Posted by RightOnRome View Post
I'm rooting for team David!
The difference between Gearslutz and politics is that Marco (sax512) and I are actually on the same team. We both want to make the best recordings we can; we simply have somewhat different ideas about what that actually entails. I believe the previous statement is true of most folks posting here, even though some have gotten a bit out of their technical depth. It's important for recording engineers to understand the scientific underpinnings of our profession and to pursue continuous improvement in technical practice, but it's just as crucial to keep in mind the overall goal of serving music and musicians.

David
4
Share
Old 11th December 2019
  #802
Lives for gear
 
12tone's Avatar
 

Quote:
Originally Posted by RightOnRome View Post
I'm rooting for team David!
So which one is Edward, and which one is Jacob?
Old 11th December 2019
  #803
Gear Addict
 

Quote:
Originally Posted by David Rick View Post
The difference between Gearslutz and politics is that Marco (sax512) and I are actually on the same team. We both want to make the best recordings we can; we simply have somewhat different ideas about what that actually entails. I believe the previous statement is true of most folks posting here, even though some have gotten a bit out of their technical depth. It's important for recording engineers to understand the scientific underpinnings of our profession and to pursue continuous improvement in technical practice, but it's just as crucial to keep in mind the overall goal of serving music and musicians.

David
Right on.
Old 11th December 2019
  #804
Lives for gear
 
juiseman's Avatar
Pretty good stuff here; I will have to agree with Divid. Sorry @ sax512 .
Both of you have more knowledge on this subject than I can hope to understand.

My only addition to this is that a higher sample rate sounds more accurate to me
on my setup. is it my Converters?, Plugins?, DAW DAE in 32 or 64 bit mode or
my speakers? or the 3 beers I had before making this assessment? I'll let
others argue about that. I bounce back between 48K (because of my ADAT preamp) and 96k quite often. And like I said before; I actually like my gtr sims at 48K vs 96k; sounds a little softer (and warm? I hate using warm...) but the FX plugs (reverb, delay) sound cleaner to me at 96k...Plus the RTL is a tad lower
at 96k..so I use that when I can.
Old 11th December 2019
  #805
Gear Addict
 

As I said before, the beauty of physics is that it doesn't care about anybody's opinion, just facts.

Personal preference is a whole different discussion, and I would never tell anybody what they should or should not do.
If you like 96 kHz, there's no mathematical demonstration I can produce to change what is a personal choice. Kind of like I would never stop a smoker from lighting up.
2
Share
Old 11th December 2019
  #806
Deleted 2ef94c5
Guest
Quote:
Originally Posted by David Rick View Post
We both want to make the best recordings we can; we simply have somewhat different ideas about what that actually entails.
Well I certainly hope it entails using your ears!

The best recordings I have ever heard were done at 44khz by a guy with skills that you, myself and most anyone here will likely never surpass in this lifetime. And he does it time and time again in multiple studios with different musicians using different gear. He listens closely and reacts accordingly. All thought is secondary and supportive to what he hears as the primary indicator of reality.

Do we care more about making the best recordings possible or being right scientifically? These are mutually exclusive brain activities as you cannot think and listen at the same time to the best of your abilities (good luck trying!) and the sooner we realize this, the sooner we will improve our recording game in ways that go far beyond what gear or sample rate we use.
2
Share
Old 11th December 2019
  #807
Lives for gear
 
monomer's Avatar
 

Quote:
Originally Posted by sax512 View Post
If I understood him correctly (David, please forgive me if I didn't), his point is that the part of a signal that is limited to the audio band, with all ultrasonics removed, is audibly different from the whole original signal, because some audible 'time smear' happens, somehow.
Well, there is a paradox here. Mathematically a linear phase filter does seem to give more 'proper' results. However, if its workings is transposed into the hearable range there is obvious ripple (noted as 'time smear').

I found this site of someone who discusses some of this stuff and who actually did some tests. Maybe the pictures are good as a reference in this discussion.
http://archimago.blogspot.com/2018/0...e-digital.html
(seems there are lots of interesting things to be found on this blog concerning converter filters so maybe it's worth looking around)

From those graphs it does seem that linear phase gives the least errors, tho i think that that particular sample contains little (visible) frequencies close to Nyquist and so may not be completely honest.
But you do see more error in the resulting waveform from the minimum phase filter compared to the linear phase.
Note that an intermediate filter is also discussed, one that tries to minimize the perceived bad effects of both linear and minimum phase filters.

Moreover, this test is strictly about upsampling interpolation. There may be a difference between removing frequencies instead of representing already existing frequencies at a higher samplerate.

So maybe when removing frequencies a different kind of filter is more optimal than the linear phase. This may have to do with physiology and the general causality of sound in the natural world.
What i mean is that technically a linear phase filter is not causal. It only becomes causal if you introduce a window and a delay.
But this is not how things work in the real world. There is no easy way to dampen, say, a snare drum, so that it exhibits pre-ripple. The impulse will be coherent, even when the high frequencies are severely dampened. You do indeed get that low-passed 'tump'.
Yet when i filter a snare with a linear phase LPF i will hear this ringing or 'smearing'. So a linear phase filter operates on the signal in a way that does not match the way we find dampened impulses to sound in nature.

BTW, thinking about this, is this pre-ripple not simply a consequence of the filters steepness?
What shape would a sinc wave take on when it would result in a gentler slope of low pass? Would that be like a summation of sinc waves over the transition band?
1
Share
Old 11th December 2019
  #808
Motown legend
 
Bob Olhsson's Avatar
 

Nobody actually records at 44.1 or 96k. A modern converter samples much higher than both and then the audio is converted down to the chosen output sample rate. This is really about when, where and how frequently (i.e. oversampled processing) to make that sample rate conversion which will end up in the delivery format.
3
Share
Old 11th December 2019
  #809
Gear Nut
 

Quote:
Originally Posted by sax512 View Post
As I said before, the beauty of physics is that it doesn't care about anybody's opinion, just facts.

Personal preference is a whole different discussion, and I would never tell anybody what they should or should not do.
If you like 96 kHz, there's no mathematical demonstration I can produce to change what is a personal choice. Kind of like I would never stop a smoker from lighting up.
I still do not believe it is the converter's or, in general, the signal source's job to present an easy signal to reproduce to the loudspeakers. Those who actually generate the signal cannot know in advance on what kind of transducer the signal will be reproduced, nor what the requirements are (except for when standards come into play). If a speaker behaves badly beyond a certain frequency range, I would argue that it must be the speaker's designer's job to ensure that the speaker can conform to the design's intentions as close as possible and anything else is a serious failure on their part (or deliberate disregard, when IMD due to ultrasonic components is negligible in the context of the speaker's performance).
Old 11th December 2019
  #810
Gear Nut
 

For instance, if we were to consider practical speaker limitations it would be unwise to allow much bass frequency to feature in the mastered signals. Very few speakers can reproduce those frequencies anyway and there is a very real risk of damaging the speakers outright due to overexcursion. Mastering engineers ought to distort anything below 80Hz to smithereens, after all most bass we hear in real life is crassly distorted. Instead, this would be viewed as bad practice (except when deliberate, of course) and we actually expect speakers to be designed to reproduce what they can. All those tiny Bluetooth speakers (or even smartphone speakers) don't blow up (usually) because the signal is treated (filtered and compressed) in order to prevent damage to the transducers and maximise perceived volume. The same sort of applies to usual consumer speakers. One exception to the rule is for guitar and bass speakers that, in fact, should not be used for full range music (especially guitar speakers).
1
Share
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump