I'd Like To Try The Sheffield Experiment. Advice requested!

Archived from groups: rec.audio.pro (More info?)

I have a copy of the KJV along with several other direkt2disk
recordings from the 70's and 80's. They are in, semi decent shape but
good enough for some basic testing as the ticks and pops will be on the
digital copy as well. Unfortunately the systems I was using when I
first purchased those records were less than great,
Dual/Empire/Pickering/Stanton681EEE and so forth in addition to the
fact that I played them over and over again

I have a Thorens TD318 with the TP-21 arm and a Shure V15mr Type V
cartridge which is a couple of months old.
I am using a Creek preamp in the studio and I have Dynaaudio BM15a
speakers in my studio along with an Allen Heath MixWiz desk.

My casual listening room has an Apt-Holman preamp going through Maggie
MG12 speakers driven by a Rotel RB-1090. I also use the Creek in this
room as well, especially when the switches on the Apt go dodgy on me.
That's my weekend project. A couple of hours with Caig Labs fluid
should help things.

I have a Rotel RCD-1072 CD player. So anyhow, I can do level matching
and rapid switching only in the studio. I can however pipe up either
signal from the studio to my home system, but only one at a time and
their will be a lag between them. So any true A+B comparisons must be
done in my studio.

I do plan to listen to both on my home system taking notes of my
observations, just for curiosity sake.

I plan to transfer into Sound forge 8 via the Creek and do both a 16/44
and 24/96 transfer. I will then burn 2 audio CD's, one direct 16/44 and
the other down sampled from the 24/96 using Sound forge.

I will use no noise reduction, normalizing etc.

QUESTION: Should I do the bit depth conversion before or after the
sample rate conversion?

QUESTION: Should I use anti-aliasing?

I plan to do a level matched listening comparison between vinyl
playing(pre converter), the 16/44 and 24/96 audio in my studio.

Due to logistics, I plan to do a casual listening of all 3 on my home
system at various levels.


What I hope to discover is if I can:

1. Hear any difference between 16/44 and 24/96.

2. If the translation of vinyl to digital loses anything or leaves
audible artifacts.

This is NOT a "which format is better sounding",nor is it a blind test
to discover which is which. It is an experiment to listen and as
honestly as possible, understanding the obvious biases that are
interjected by a non blind test, to see if I can hear any differences.

I am a degreed EE but have never worked as one because my love of music
has lead me into other fields as a musician, jazz pianist to be exact.

Obviously not quantifying measurements and relying on just my ears is
going against my grain as an engineer, but as a musician, my ears and
sense of musicality is a better tool for this type of highly subjective
experiment. I am 45 years young and have a decent set of ears, although
not as good as when I was 18 and wild!
I have a bias toward CD and while vinyl has a special place in my
heart, I often wonder what some of my favorite vinyl recordings would
have sounded like had they been digitally recorded.

Again, I am not trying to determine whether vinyl sounds better or
worse but if the digital transfer and or down sampling has an audible
effect.

So, can anyone offer any advice for me to try and make this experiment
at least technically valid as far as the transfer, down sampling and
such is concerned?

ChowForNow

Allen Catalano

Archived from groups: rec.audio.pro (More info?)

On 30 Jun 2005 14:26:46 -0700, "Allen Catalano"
<allen_catalano@yahoo.com> wrote:

>So, can anyone offer any advice for me to try and make this experiment
>at least technically valid as far as the transfer, down sampling and
>such is concerned?

Anybody in your region do record cleaning (alcohol bath + vacuum)?
Try it once and you'll buy your own machine. The're essential.

Good fortune; sounds like a good project for the Fourth.

Chris Hornbeck
"Chief Assistant to the Assistant Chief" -F&S

Archived from groups: rec.audio.pro (More info?)

On 30 Jun 2005 14:26:46 -0700, "Allen Catalano"
<allen_catalano@yahoo.com> wrote:
-------------------------8<----------------------------
>I plan to transfer into Sound forge 8 via the Creek and do both a 16/44
>and 24/96 transfer. I will then burn 2 audio CD's, one direct 16/44 and
>the other down sampled from the 24/96 using Sound forge.
>
>I will use no noise reduction, normalizing etc.
>
>QUESTION: Should I do the bit depth conversion before or after the
>sample rate conversion?

Yes, you do it after the sample rate conversion.

>QUESTION: Should I use anti-aliasing?

In Sound Forge you have the anti-aliasing option, use it at
downsampling. It filters out the artifacts. While at it, do the DC
offset correction and as the last step before saving, do dithering by
triangular highpass and high pass contour noise shaping.
>
>I plan to do a level matched listening comparison between vinyl
>playing(pre converter), the 16/44 and 24/96 audio in my studio.
>
>Due to logistics, I plan to do a casual listening of all 3 on my home
>system at various levels.
-- Best if you do it the next morning after the evening work on
transfer, your ears will be fresh.
>What I hope to discover is if I can:
>
>1. Hear any difference between 16/44 and 24/96.
-- Hmm I think you should.

>2. If the translation of vinyl to digital loses anything or leaves
>audible artifacts.

Digital will sound different to a degree but a properly digitized
analog recordings, with modern A/D converters, doesn't always have to
be sounding sharp and grainy.

>This is NOT a "which format is better sounding",nor is it a blind test
>to discover which is which. It is an experiment to listen and as
>honestly as possible, understanding the obvious biases that are
>interjected by a non blind test, to see if I can hear any differences.
>
>I am a degreed EE but have never worked as one because my love of music
>has lead me into other fields as a musician, jazz pianist to be exact.
>
>Obviously not quantifying measurements and relying on just my ears is
>going against my grain as an engineer, but as a musician, my ears and
>sense of musicality is a better tool for this type of highly subjective
>experiment. I am 45 years young and have a decent set of ears, although
>not as good as when I was 18 and wild!
>I have a bias toward CD and while vinyl has a special place in my
>heart, I often wonder what some of my favorite vinyl recordings would
>have sounded like had they been digitally recorded.

Why not, if one is careful enough? But yes, a cleaned record is
sounding much better, gives you less trouble after it is digitized and
with a properly set up Shure V15VxMR you sholud not have much problems
as to tracking distorsion. If some of louder clicks and pops are
objectionable, in many cases you can remove them uneventfully by
interpolation, replacing ot channel swap in the Sound Forge. Undo and
try several selected lenghts of the waveform if you find a large click
leaves pops and thumps; if you can't mend it enough, leave it. Maybe
using a dynamic stabilizer at the Shure V15 could shorten the low
frequency portion of the click, making it somewhat easier to filter it
out by replace function. It depends of many things.

>Again, I am not trying to determine whether vinyl sounds better or
>worse but if the digital transfer and or down sampling has an audible
>effect

>So, can anyone offer any advice for me to try and make this experiment
>at least technically valid as far as the transfer, down sampling and
>such is concerned?
>
>ChowForNow
>
>Allen Catalano
Good luck, Edi Zubovic, Crikvenica, Croatia

Archived from groups: rec.audio.pro (More info?)

"Allen Catalano" <allen_catalano@yahoo.com> wrote in message
news:1120166806.644030.304320@g43g2000cwa.googlegroups.com

> I have a copy of the KJV along with several other
direkt2disk
> recordings from the 70's and 80's. They are in, semi
decent
> shape but good enough for some basic testing as the ticks
and
> pops will be on the digital copy as well. Unfortunately
the
> systems I was using when I first purchased those records
were
> less than great, Dual/Empire/Pickering/Stanton681EEE and
so
> forth in addition to the fact that I played them over and
over
> again

> I plan to transfer into Sound forge 8 via the Creek and do
> both a 16/44 and 24/96 transfer. I will then burn 2 audio
> CD's, one direct 16/44 and the other down sampled from the
> 24/96 using Sound forge.

IMO it would be entirely sufficient to transcribe the
recording once at 24/96, and base your other files on it.

> QUESTION: Should I do the bit depth conversion before or
after
> the sample rate conversion?

Probably best if you do the sample rate conversion first, so
that as much processing as possible is done with 24 bit
calculations.

> QUESTION: Should I use anti-aliasing?

Yes.

> I plan to do a level matched listening comparison between
vinyl
> playing(pre converter), the 16/44 and 24/96 audio in my
studio.

One problem with this plan is that each playing of a LP
degrades it a bit, often audibly. Therefore the LP you
compare your transcription to is not quite the LP that that
you transcribed from.

> Due to logistics, I plan to do a casual listening of all 3
on
> my home system at various levels.

> What I hope to discover is if I can:

> 1. Hear any difference between 16/44 and 24/96.

That requires no additional playback of the LP at all. First
make your 24/96 transcription. Then downample it to 16/44.
Finally upsample the 16/44 back to 24/96 so that you hold as
many variables (The LP being digitized and DAC performance
as examples) constant.

The best way to compare digital files is to use a comparator
program such as one of the ones you can freely download from
www.pcabx.com . Be sure that the files you have have the
same level and the same begin and end points.

> 2. If the translation of vinyl to digital loses anything
or
> leaves audible artifacts.

Since sucessive playback of vinyl loses something, this is a
very hard experiment to do without introducing other
variables.

> Obviously not quantifying measurements and relying on just
my
> ears is going against my grain as an engineer,

I don't know why. Well-done listening tests are highly
definitive. It's just that there aren't as many well-done
listening tests as would be optimal. It's all the casually
done tests that are touted as being definitive or even just
indicative that has given listening tests a bad name among
many engineers.

Archived from groups: rec.audio.pro (More info?)

"Mark" <makolber@yahoo.com> wrote in message
news:1120224973.154682.164990@o13g2000cwo.googlegroups.com

> ...and when you have completed the listening tests, let
the
> engineer in you also do the follow up and use objective
> measurements to try to identify the cause of any
difference
> that you may hear.

Measurements on vinyl are *always* disappointing.

> Also consider this, you may be able to set up a system
that
> takes your analog input signal and passes it though an A/D
and
> directly back out through a D/A. Then you can do "real
time"
> comparisons without actually burning to a CD.

Been there, done that several different ways.

> This is also a handy way to make the objective
measurements where you can
> directly compare the in and out.

Ditto.

> I agree..... many CD recordings sound harsh and ugly, but
as
> someone else has said, and this is a wonderful point, if
> there is even ONE CD that does not sound harsh and ugly
that
> indicates the harsh and ugly is not the fault of the CD
> medium. I have a least one CD that sounds great. So lets
> turn our attention to why the others do sound harsh and
ugly.

It's all about care at every stage of the production
process.

Archived from groups: rec.audio.pro (More info?)

"Allen Catalano" <allen_catalano@yahoo.com> wrote in message .
>
> I will use no noise reduction, normalizing etc.
>
> QUESTION: Should I do the bit depth conversion before or after the
> sample rate conversion?

Try one of each !

> QUESTION: Should I use anti-aliasing?

Absolutely.

geoff