Mix to 96KHz = "less bass" than 48KHz???

[cameron]

To me this doesn't 'compute'...

I was told recently by a pro that to mixdown off an analog console and capture this at 96KHz will 'sacrifice' a nice fat bass response, and was therefore told that a final two-bus capture at 48KHz (before mastering) was much better for rock.

Now technically, the difference from 48 to 96 should only be in representation of frequencies at 24KHz and above - so how can using 96KHz affect the bass waaay down the other end of the spectrum??

Perhaps I'm missing something about destructive interference, harmonics, bandwidth capabilities of equipment, mojo, etc. I repect this guy's opinion given his work, so there's gotta be something behind his recommendation.

CLA apparently mixes 16bit/48KHz though his SSL console onto 1/2" tape, and then captures off that tape at 24bit/96KHz via Protools for delivery to mastering.. and his stuff is bassy!

c

[jupiter8]

There is nothing in the Nyquist-Shannon theorem to support that (as you already know). So i see two possible reasons for this. Either he has a worked with an ADA converter that has those characteristics and believes that is applies to all ADA converters or he's just plain confused with tape speed on tape machines.

You can be pretty much oblivious to all tech stuff and still do great work.
I remember reading an article on one of the great producers (don't remember who. That guy that did the Norah Jones stuff i believe) He had no clue on what even the boxes were called. He knew what they did though and that was enough.

"Give more of that blue one"
"You mean the Focusrite?"
"I don't know.The big blue one over there"
"That's a Focusrite"
"Whatever."

Though he was a bit more diplomatic than that but you get the picture.

[Farview]

Originally Posted by jupiter8

There is nothing in the Nyquist-Shannon theorem to support that (as you already know). So i see two possible reasons for this. Either he has a worked with an ADA converter that has those characteristics and believes that is applies to all ADA converters or he's just plain confused with tape speed on tape machines.

You can be pretty much oblivious to all tech stuff and still do great work.
I remember reading an article on one of the great producers (don't remember who. That guy that did the Norah Jones stuff i believe) He had no clue on what even the boxes were called. He knew what they did though and that was enough.

"Give more of that blue one"
"You mean the Focusrite?"
"I don't know.The big blue one over there"
"That's a Focusrite"
"Whatever."

Though he was a bit more diplomatic than that but you get the picture.

+1

I forgot which ADA designer said it (I'm thinking Lavry) "If you can hear a difference between sample rates, your converter is broken".

[Bob Olhsson]

More aliasing in his converter at 48kthan 96k can cause that illusion.

His mixes most likely need more bass!

[Ethan Winer]

Originally Posted by Farview

"If you can hear a difference between sample rates, your converter is broken".

Exactly. Also, human perception is very fragile, and it's very difficult to know what changed or even if something changed unless the difference is really blatant. Likewise, most project studio control rooms are not well treated acoustically, and even tiny changes in head position can have a large effect on what is heard. See this article for more on that phenomenon:

Why We Believe

--Ethan

[807Recordings]

Originally Posted by Ethan Winer

Exactly. Also, human perception is very fragile, and it's very difficult to know what changed or even if something changed unless the difference is really blatant. Likewise, most project studio control rooms are not well treated acoustically, and even tiny changes in head position can have a large effect on what is heard. See this article for more on that phenomenon:

Why We Believe

--Ethan

I like how you ended the article on how both people where right. That has been my arguement all along in that we might not have been measuring the right things. Tuned room is critical and one of the biggest things I have been wanting to get for the reviews I do is better test equipment. You can't trust anything 100%.

[fenderbender.]

Originally Posted by Farview

I forgot which ADA designer said it (I'm thinking Lavry) "If you can hear a difference between sample rates, your converter is broken".

My Lavry Blue is broken? Hmmmmmmm.....

[bob katz]

Originally Posted by Bob Olhsson

More aliasing in his converter at 48kthan 96k can cause that illusion.

His mixes most likely need more bass!

I've been seeing some interesting phenomenon when downsampling 96 to 44 or 88 to 44 especially after dithering to 16 bit (and perhaps ONLY after dithering to 16) that if I'm not careful either the bass seems to come up a hair or perhaps it's because the high end is not as clear. But going down to 48, I haven't seen any significant tonal shifts. The first differences to my ears as the sample rate goes down is a slight loss of depth and width and/or space. The second difference is a loss of "openness" but that may be very well tied in with the first. The net result, though, is that translation of the bottom end, you have to be careful, going down to 44. But I know, the question was at 48K, and not to my experience. 48K is an "excellent" medium.

BK

[bob katz]

Originally Posted by fenderbender.

My Lavry Blue is broken? Hmmmmmmm.....

There's considerable truth in Lavry's remark, "if you can hear a difference between sample rates, your converter is broken". Some of his principles are based on the fact that the main components of the converter are the same and only the decimator changes. However, in my opinion, Dan L. is not COMPLETELY right. I (we) do hear differences. It's just that with a really well designed converter, the differences among sample rates are far less radical than with a cheaper or poorer converter. Especially on the D/A side, where the DAC is always upsampling to a very high rate.

When I got the latest rev. of the Avocet DAC I found that the differences between a 96K source, and that source downsampled to 4424 via the Weiss were MUCH less than ever before. Where previously I had been blaming the Weiss (or the sample rate itself) as the major cause of losses at 44.1 kHz now I had to reassess this and say that the DAC itself is a significant contributor.

[arf]

Actually, dcs had a white paper a few years ago where they stated that bass instruments were less defined at 192k sampling. That may no longer be true with 5.6MHz few-bit modulators, but there may be something to high-speed conversions causing trade-offs in bass instrument definition due to incorrect implementation.

[bgrotto]

Originally Posted by cameron

Now technically, the difference from 48 to 96 should only be in representation of frequencies at 24KHz and above - so how can using 96KHz affect the bass waaay down the other end of the spectrum??

c

It can't. The guy that told you that is confusing Inches/sec with Khz/sec.

[arf]

If the ADC produces higher jitter at the higher sample rate, bass instrument definition will suffer.