avatar

morespaceecho

Platinum Blonde

Posts: 2,304 Member Since:29/01/2011

#21 [url]

Apr 16 17 7:18 PM

silvertone wrote:

One of the ways I intend to use the Langevin tube console and Presto 3 track is to mix down to DSD for my stereo format.    From there go analog out of the DSD player and into the analog side of my mastering chain.

Also I plan to be able to, at the push of a button, send the whole tube recording chain into the mastering room to monitor mixes pre and post mastering.  I had the Electrodyne setup this way and it was a great way to be able to mix. It truly spoiled me.

I'm also going to be able to multitrack in Pro Tools and send stems out to mixdown though the tube console to the 3 track to capture its "sound" and can go full round Robin back out, through the mastering chain and print that back into the mix.


that is going to be a really sweet setup!

88.2 is as high as i go as well. lately i've just been going with good old fashioned 44.1...
 

www.oldcolonymastering.com

morespaceecho.bandcamp.com

Quote    Reply   
avatar

waltzingbear

Platinum Blonde

Posts: 1,240 Member Since:16/07/2012

#22 [url]

Apr 16 17 9:22 PM

I'll put my 2 cents in with going to 96k rather than multiples of 44.1 The CD is going away, no reason to keep that compromise sample rate, 48k is much better all around. And by going to multiples of 48k, we simplify life by being the same sample rate that is associated with visual media.

So, quit using multiples of 44.1 and use multiples of 48k, simplify life!

In the long run its all going to be associated with imagery anyway ;-)

None of this is a technical argument other than simplifying life and getting rid of 44.1k
remember the original research showed the low 50s being a good target.

Cheers
Alan

Alan Garren
Waltzing Bear Audio

Quote    Reply   
avatar

zmix

Aqua Marine

Posts: 4,069 Member Since:20/01/2011

#23 [url]

Apr 16 17 9:35 PM

gtoledo3 wrote:
For sure Chuck, not being literal at all, hence the analogy.

It is very different than perceptual encoding, but I think "perceptual encoding on steroids" might be a fair expression.

There is a decision made that human perception is such that pushing all of the noise to HF is the best judgement call, and as a side effect of this, certain frequency ranges have less resolution. In a very general sense, this is a similar logic process to perceptual encoding; human perception is more attuned to some frequencies, so more of the data should go towards representing them.

Of course the actual methodologies between the two worlds are vastly different, and there are also many spins on perceptual encoding.

With PCM there is obviously the similar judgment that a given frequency range is most important, but there isn't the next step of making resolution effectively differ across that frequency range.

I see what you mean..   Maybe more like "sweep it under the rug" or "kick the can down the road"   Maybe it's the HF contamination that I find objectionable in the sound, even though it's usually LP filtered at 50Khz it's possible that it can cause increased distortion in the speakers if all that energy is causing the voice coils to heat up.


I recall having similar discussions about noise shaped dither.  I cannot stand the sound of it.  I never could, but several mastering engineers I know would *read* the theory behind it and dogmatically apply it to everything. 


PCM actually works really well...





Quote    Reply   
avatar

Mike Rivers

Aqua Marine

Posts: 2,514 Member Since:13/10/2012

#24 [url]

Apr 17 17 9:00 AM

waltzingbear wrote:
I'll put my 2 cents in with going to 96k rather than multiples of 44.1 The CD is going away, no reason to keep that compromise sample rate, 48k is much better all around. And by going to multiples of 48k, we simplify life by being the same sample rate that is associated with visual media.

So, quit using multiples of 44.1 and use multiples of 48k, simplify life!
 


While it's been quite some time since any of my recordings have been replicated as pressed CDs, I typically will make an audio CD for the client. I remember all the warnings about the dangers of sample rate conversion and truncation and used to heed them. But once I was working on a 48 kHz 24-bit file in Sound Forge and, without thinking, used Sound Forge's tools to split the file into CD tracks and burn a CD, and the CD sounded just fine - played at the right speed, and didn't sound any worse, to my ear, than the file playback. I don't know what Sound Forge did internally to get to the CD sample rate and word length, nor do I know that if I had a really pristine recording with things that decayed to studio silence rather than audience noise. Maybe I'd notice some conversion artifacts on a different kind of project.

But, also, with what I record and with the gear I have to record it with, I don't notice any difference between recording at 44.1 and 48 kHz. So I'm not surprised that live DSD recording didn't really wow me.

 



For a good time, call mikeriversaudio.wordpress.com

Quote    Reply   
avatar

maarvold

Aqua Marine

Posts: 3,106 Member Since:23/01/2011

#25 [url]

Apr 17 17 9:04 AM

waltzingbear wrote:
I'll put my 2 cents in with going to 96k rather than multiples of 44.1 The CD is going away, no reason to keep that compromise sample rate, 48k is much better all around. And by going to multiples of 48k, we simplify life by being the same sample rate that is associated with visual media.

So, quit using multiples of 44.1 and use multiples of 48k, simplify life!

In the long run its all going to be associated with imagery anyway ;-)

None of this is a technical argument other than simplifying life and getting rid of 44.1k
remember the original research showed the low 50s being a good target.

Cheers
Alan

 
Since I read Dan Lavry's point-of-view that 60k would be the ideal sampling frequency, I have been happily working at 48k almost exclusively.  Occasionally I will be asked to do a project at 96k--I'm doing 2 now--and I think the top end can be more prone to be a bit more 'bothersome' if I'm not careful.  

Quote    Reply   
avatar

chetdelcampo

Silverado

Posts: 101 Member Since:27/03/2013

#26 [url]

Apr 17 17 9:31 AM

compasspnt wrote:
I have found the playback of DSD 5.6 mixes to be the most accurate to the desk output of anything I've tried.


But I prefer mixing back into PT for rock...DSD for jazz, classical, soft acoustic music...


I can understand this.  Even though I find there's generally a significant compromise to DAW capture compared to what's coming off the desk in the room.

I'm still trying to sort out conclusions here with recently mixing to a Tascam DA3000 DSD compared to, mixing to it PCM vrs back into the box.
If I do use DSD, I tend to run out of it into a good AD and back to DAW for any pre-mastering instead of using DSD > PCM conversion SW. 

One thing I don't see much discussed regarding sample rates is that at least "to me" 44.1 retains more of what I like in the low end / low mids.  
Which again, for me, leads me to just stick with it. 

 

Quote    Reply   
avatar

zmix

Aqua Marine

Posts: 4,069 Member Since:20/01/2011

#27 [url]

Apr 17 17 10:16 AM

chetdelcampo wrote:
compasspnt wrote:
I have found the playback of DSD 5.6 mixes to be the most accurate to the desk output of anything I've tried.


But I prefer mixing back into PT for rock...DSD for jazz, classical, soft acoustic music...


I can understand this.  Even though I find there's generally a significant compromise to DAW capture compared to what's coming off the desk in the room.

I'm still trying to sort out conclusions here with recently mixing to a Tascam DA3000 DSD compared to, mixing to it PCM vrs back into the box.
If I do use DSD, I tend to run out of it into a good AD and back to DAW for any pre-mastering instead of using DSD > PCM conversion SW. 

One thing I don't see much discussed regarding sample rates is that at least "to me" 44.1 retains more of what I like in the low end / low mids.  
Which again, for me, leads me to just stick with it. 

 
I'm sure that Terry meant that he was recording the output of his desk back into PT.

The "low end / low mids" issue you mention is probably related to the phenomenon we've all experienced in mixing and mastering:

If you get the HF right the lows sound bigger.  

I've observed ultrasonic contamination and strange aliases in tracks I've received at 96kHz. All of the spec sheets I've seen show that converters perform much better at their lowest sample rates in terms of distortion, noise and aliasing (suggestions for data sheets wanted here!)

96kHz has not yet sounded particularly "good" to me, even when I'm making a choice to begin a session.

Quote    Reply   
avatar

morespaceecho

Platinum Blonde

Posts: 2,304 Member Since:29/01/2011

#28 [url]

Apr 17 17 10:56 AM

i've done a few records at 88.2 and i dunno if they really sound any better than the ones i've done at 44.1.

i'd like to prove it to myself one way or the other. it's kind of a hard thing to really test, no? the best test scenario i've come up with is:

write a really generic song
record a verse and chorus of guitar/bass/drums at 44.1. then do the same at 88.2.
apply the same plugins to both sessions, make sure all fader levels match exactly, render mixes, compare. 

there will of course be playing discrepancies between the two, but i think if the song is boring enough and i play as straightforward and simply as possible, that would mostly remove enthusiasm as a variable:)
 

www.oldcolonymastering.com

morespaceecho.bandcamp.com

Quote    Reply   
avatar

soapfoot

Ruby Baby

Posts: 7,323 Member Since:04/02/2011

#29 [url]

Apr 17 17 11:12 AM

why not just get two of the exact same make and model of converter, set one at 44.1 the other at 88.2 and mult all inputs to both at the same time?

brad allen williams

Quote    Reply   
avatar

jesse decarlo

Platinum Blonde

Posts: 1,515 Member Since:24/03/2013

#30 [url]

Apr 17 17 11:14 AM

morespaceecho wrote:
i've done a few records at 88.2 and i dunno if they really sound any better than the ones i've done at 44.1.

i'd like to prove it to myself one way or the other. it's kind of a hard thing to really test, no? the best test scenario i've come up with is:

write a really generic song
record a verse and chorus of guitar/bass/drums at 44.1. then do the same at 88.2.
apply the same plugins to both sessions, make sure all fader levels match exactly, render mixes, compare. 

there will of course be playing discrepancies between the two, but i think if the song is boring enough and i play as straightforward and simply as possible, that would mostly remove enthusiasm as a variable:)


 

I think the only way to be sure is to have two of the same model AD/interface each connected to its own computer, and running at two different sample rates. Record the same performance into each, using mic splitters so each sees the exact same input. Even then it only answers the question for that particular AD. 

edit: Brad beat me to it. 

Quote    Reply   
avatar

morespaceecho

Platinum Blonde

Posts: 2,304 Member Since:29/01/2011

#31 [url]

Apr 17 17 11:27 AM

soapfoot wrote:
why not just get two of the exact same make and model of converter, set one at 44.1 the other at 88.2 and mult all inputs to both at the same time?


hahahahaha "just".

right, no problem, i'll peel off a few thousand from that pile of money i sleep on:)

www.oldcolonymastering.com

morespaceecho.bandcamp.com

Quote    Reply   
avatar

soapfoot

Ruby Baby

Posts: 7,323 Member Since:04/02/2011

#32 [url]

Apr 17 17 11:40 AM

morespaceecho wrote:

soapfoot wrote:
why not just get two of the exact same make and model of converter, set one at 44.1 the other at 88.2 and mult all inputs to both at the same time?


hahahahaha "just".

right, no problem, i'll peel off a few thousand from that pile of money i sleep on:)

No need for that! You could find a friend to loan you two of the same model, or another of the model you already have, or try it next time you're in a studio that's thusly equipped!

brad allen williams

Quote    Reply   
avatar

morespaceecho

Platinum Blonde

Posts: 2,304 Member Since:29/01/2011

#33 [url]

Apr 17 17 12:14 PM

who's got 6 channels of lavry blues and a focusrite isa428 i can borrow?

a better idea is how about someone who's already set up in a legit recording studio do this test? could you guys do this at the bunker?

i agree that multed inputs is a much better idea than my plan!

www.oldcolonymastering.com

morespaceecho.bandcamp.com

Quote    Reply   
avatar

zmix

Aqua Marine

Posts: 4,069 Member Since:20/01/2011

#34 [url]

Apr 17 17 12:17 PM

I have done the following:

Received a track recorded at 96kHz.

Mixed it at 96kHz

Noted an incredible amount of ultrasonic interference and anomalies.

SRC'ed down all of the individual audio files to 44.1kHz

Replaced those audio regions in the very same "96kHz" mix after changing the session rate to 44.1

Everything in the mix was identical, every automation move, every insert, etc.

I sent the client a 320kbps MP3 of each mix, but did not explain what the difference was.

They vastly preferred the 44.1 mix, as did I.

The SRC lopped off all the unwanted RF and "ultrasonic" information, the mix sounded more open, more musical , way less "annoying"...

PCM audio works really well these days, people, especially at lower sample rates.

Quote    Reply   
avatar

soapfoot

Ruby Baby

Posts: 7,323 Member Since:04/02/2011

#35 [url]

Apr 17 17 12:19 PM

We could manage that, but we'd have to set up a second DAW. Maybe one day if we've got a tech day going and nothing else going on it would be a fun experiment!

brad allen williams

Quote    Reply   
avatar

morespaceecho

Platinum Blonde

Posts: 2,304 Member Since:29/01/2011

#36 [url]

Apr 17 17 12:41 PM

i'd certainly be interested to hear the results!

in the meanwhile, chuck's test above is easy enough to recreate, if i have time later i'll try and find something reasonably hifi to work with and do mixes at 88.2 and 44.1.

www.oldcolonymastering.com

morespaceecho.bandcamp.com

Quote    Reply   
avatar

silvertone

Aqua Marine

Posts: 2,743 Member Since:26/01/2011

#37 [url]

Apr 17 17 1:50 PM

I have done enough tests here at SIlvertone to know that with my system 88.2 and 96k definitely sounds the best for tracking with the converters I use. (Dangerous, Lynx Aurora, HEDD and Troisi).

I first heard it when I was transferring tapes from my 1" 8 track analog machine while trying to figure out what sample rate would sound best. Set the session up for 88.2 and transferred one song in, then changed the sample rate to 44.1 and transferred the same song in the same file. Both sounded very good. The higher sample rate one sounded closer to what the output of the console sounded like. Checked at 96k and 48k as well. Did the same test when I transferred all my 2" 16 track tapes. Went with what my ears told me sounded best.

Easy enough to do with an Ampex ATR102 and a great stereo mix. Just print the same mix at all different sample and bit depths and compare the results back to the original source. What sounds closest? What changed the most? Go with what your ears prefer.

When comparing DSD I'm always playing back through the built in DA coming into the mastering console analog. The high bit rate PCM from the daw is coming in digital. Both go out though the same Dangerous converters for monitoring. Even with the extra conversion step the DSD always sounds like the output of the analog console (which is also patched into the mastering console through an analog input and is monitored out the same Dangerous DAC). It's easy to hear as you can level match and A/B/C in an instance.

Silvertone Mastering, celebrating 28 years in business.

www.silvertonemastering.com

Quote    Reply   
avatar

zmix

Aqua Marine

Posts: 4,069 Member Since:20/01/2011

#38 [url]

Apr 17 17 2:13 PM

I am the sort of person who has infinite curiosity and patience, and I always like to know "why" something seems "different", so I endeavor to learn the cause.

In my test, it turned out that there was RF and other ultrasonic interference within the individual tracks which were causing artifacts in the *audible* spectrum, which went away when I downsampled to 44.1, and these were simply files within a mix, and no difference in "converters" except for my monitoring DAC, which upsamples everything to the same rate.

I wouldn't just assume that listening to 2 channels at 44.1 or 96kHz is going to make the same difference as listening to 50-60 channels at 44.1 or 96kHz.

Last Edited By: zmix Apr 17 17 2:17 PM. Edited 1 time.

Quote    Reply   
avatar

soapfoot

Ruby Baby

Posts: 7,323 Member Since:04/02/2011

#39 [url]

Apr 17 17 3:38 PM

"sounds better" is also subjective. One person's "bright and detailed" is another person's "etched and hard". So it's hard to be precise if our question is "which sounds better?"

Once at our studio we set up a blind test with playback of a capture of a mix into the Lynx Aurora at 96k as choice "A", and the Studer 1/4" at 15 IPS as choice "b." We didn't tell the listeners what was different.

Three people preferred the Studer, one person preferred the digital capture, and everyone had their reasons why (before the sources were revealed. All four seemed to hear the same sorts of things; they just had different preferences/priorities/opinions of which one was "better" or "more hi-fi."

brad allen williams

Quote    Reply   
avatar

zmix

Aqua Marine

Posts: 4,069 Member Since:20/01/2011

#40 [url]

Apr 17 17 3:45 PM

soapfoot wrote:
"sounds better" is also subjective. One person's "bright and detailed" is another person's "etched and hard". So it's hard to be precise if our question is "which sounds better?"

Once at our studio we set up a blind test with playback of a capture of a mix into the Lynx Aurora at 96k as choice "A", and the Studer 1/4" at 15 IPS as choice "b." We didn't tell the listeners what was different.

Three people preferred the Studer, one person preferred the digital capture, and everyone had their reasons why (before the sources were revealed. All four seemed to hear the same sorts of things; they just had different preferences/priorities/opinions of which one was "better" or "more hi-fi."
Absolutely, and declaration of sound preferences are as meaningless as someone describing their girlfriend's breasts, the only reasonable reply is "how nice for you"...

In my case, when I hear a difference, I want to know if I can understand it and use it to my advantage or in a way that suits my tastes... or in the case of DSD, I'r rather use a format that I prefer and I can actually edit, send to a client, etc...




Quote    Reply   
Add Reply

Quick Reply

bbcode help