As I tried to be clear about, I didn't say it wasn't audible, I just said I hadn't heard it. Lots of people can hear things that I don't hear and I seem to be able to hear things that others can't hear. It's the nature of the beast. To me, absolute phase being correct or reversed is quite audible but a lot of engineers have told me that I'm nuts and that it makes "absolutely" no difference. Oh, and their logical reasoning? That all complex sounds are simply made up of various sine waves and sine waves have no absolute polarity.
db
On Saturday, December 13, 2014 12:58 PM, Dave Glasser <[log in to unmask]> wrote:
It is absolutely audible! Comparing acoustic delay and electrical delay is apples and oranges. The stereo image snaps into better focus when the 1/2 sample delay is corrected. And since it's so easy to do, why not do it?
Untitled DocumentDavid Glasser
Chief Engineer
Airshow
3063 E Sterling Circle #3
Boulder, CO 80301
303-247-9035
[log in to unmask]
On Dec 13, 2014, at 10:42 AM, DAVID BURNHAM <[log in to unmask]> wrote:
I've never noticed the loss of hi-end when listening to an F1 recording through a non-corrected play-back system in mono, (but that's not saying that it isn't there). If, as Ted points out, the error at 11khz is 90 degrees, doesn't that mean that the lowest cancelled frequency is 22khz? Well beyond the pass-band of 44.1 khz digital systems. That, (as I think Rob pointed out), also represents the length of time it takes for sound to travel a small fraction of an inch. No coincident stereo mike that I know of has the capsules that close together and the distance between your ears and each speaker will always vary by a much larger value than that.
db
On Saturday, December 13, 2014 10:34 AM, Ted Kendall <[log in to unmask]> wrote:
The inter-channel time difference is about 11uS, which translates into
90 degrees at 11kHz, if memory serves. That is enough to cause audible
top loss when combined to mono, so should be corrected. The system is,
however, coincident when used end-to-end in analogue.
What originally threw this into sharp relief was the common practice of
using F1s for mobile use and transcoding to 1610 for editing. Audio and
Design came up with an interface box which put a delay in the
appropriate channel, and neatly solved the problem. Unfortunately, when
they introduced a "professional" version of the 701, they put the delay
in the wrong channel. The resulting mess can be imagined.
To be fair to the F1 system, if you actually treat it as Sony
recommended and use sensible video carriers (ie, not slow speed and
preferably Beta), it's pretty robust. Lower bandwidth and increased
dropout on the recording mediium translate into higher error rates, as
they would on any system. We used it in the field for many years with
few problems.
On 13/12/2014 02:36, Ellis Burman wrote:
I agree with Rob about the 1/2 sample error. 1 sample at 44.1K is clearly
audible. 1/2 sample, not as much, but still audible.
Does F1 work such that there is a 1/2 sample delay upon A/D, and another
one, but in the other channel, upon D/A, such that the delay cancels in the
analog domain (A-A)? If that is true, then only the digital output has the
delay, and analog transfers are correct as-is.
Ellis Burman
www.audiomechanics.com
On Fri, Dec 12, 2014 at 5:08 AM, Rob Poretti - Cube-Tec <
[log in to unmask]> wrote:
Hi Mathew,
Below is a link to an F1 manual that illustrates the differences between
the
14-bit and 16-bit formats:
http://www.manualslib.com/manual/466194/Sony-Pcm-F1.html?page=26#manual
As you can see, the 16-bit format has one error correction word: "P" -
while
the 14-bit has tow "P" & "Q"
Earlier in the manual, they talk about the fact that a 16 horizontal line
burst-error can be accommodated in 16-bit - and 32 H lines in 14-bit.
In terms of correcting channel delay - IMHO: yes!
(Ask any mastering engineer.)
Cheers!
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
_/
Rob Poretti - Sales Engineer - Archiving
Cube-Tec North America LLC
Vox.905.827.0741 Fax.905.901.9996 Cel.905.510.6785
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
_/
-----Original Message-----
From: Association for Recorded Sound Discussion List
[mailto:[log in to unmask]] On Behalf Of Matthew Sohn
Sent: December 12, 2014 12:19 AM
To: [log in to unmask]
Subject: Re: [ARSCLIST] Sony 601-esd pcm audiio processor decoder
John, thank you for your detailed answer. My few experiences with F-1 tapes
have been nightmarish. Everything is fine, until you hit that one spot
where
it just won't track..I applied the skew method. twist the knob left or
right
until the lights look good, remember where you started, and do it again for
real. When it works, it sounds great, when it doesn't... Ugh..
-Matt Sohn
On Thursday, December 11, 2014 10:10 PM, John Gledhill <[log in to unmask]>
wrote:
Tom -
o The tape are recorded in the 14 bit mode with TOSHIBA DX-900. They
actually have FM tracks of the same material
o I am capturing both the FM and PCM
To David Glasser
o I understand re the sample rate.
o live performance recordings - 6 hours per tape
I have the 14 bit recovery working well with the EP tapes and have a solid
right green on (or flashing) MOST of the time.
I just wanted to know the significance of the LEDs
I actually record a new tape through the coax in from a CD player to the
601-esd (16 bit) , and then played the tape back and captured through spdif
to compare the waveforms. I think the the 601 saturates slightly before
0x7ffff (around 0x7fef) but the lower values were a bit for bit match.
To Rob Poretti
o I agree re the 14 bit versus the 16 bit - if the 601-esd does not like
the VCR then more error correction will win.
o However, from watching the screen it is no 1 party bit versus two parity
bits.
I think all of the stuff on the right hand side of the screen is error
correction data. Each line has 3 word pairs and a huge pile of ECC .
Roughly 3 * 15.75 minus the time lost during the vertical will give you the
44.056 This is also confirmed by the fact I can pop up the on-screen
diiplay
from the VCR and mot hear any errors,. Like a CD the ECC is spread over
tape
space = time to work around dropouts
Is it important to correct for a 1/2 sample offset. Is this not the same
as
moving one speaker in a stereo pair 3mm further back.
Slightly interesting aside.
While doing my experimenting for this I tried feeding the video from a vcr
through a time base corrector to clean it up before going to the 601-esd.
Video looked much more stable on the screen but the 601-esd did not like
this arrangement and the tracking bar stayed on the left.
Perhaps it was the digital re-sampling in the TBC that cause this result.
Anyways - I am getting very clean audio back from the ep tapes and wanted
to
know about the lights.
I might guess that the guy who designed the LSI to do the decoding put the
logic outputs there for his own diagnostic purposes and a bright marketing
guy said "we have left oer LEDs - lets use em" with no one really writing
down what they meant.
On 12/11/2014 7:44 PM, Tom Fine wrote:
Hi John:
My experience is that NO F1 decoder works well with 16-bit recordings
made at EP speed. I also think that later VHS machines with
auto-tracking may not align best for F1 recovery because they were
designed to also take into account VHS-HIFI signal. A
professional-quality VHS deck, even a late-era VHS-HIFI deck, will
have a tracking control, but the consumer models lost the tracking
control shortly after VHS-HIFI was developed.
I'm curious, what sort of material are you working with and how many
tapes?
-- Tom Fine
----- Original Message ----- From: "John Gledhill" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, December 11, 2014 4:45 PM
Subject: [ARSCLIST] Sony 601-esd pcm audiio processor decoder
I am recovering sony F1 type audio (PCM and ECC in the video image)
from some VHS tapes (ep speed 6 hr per tape) - it was actually
recorded with a Toshiba version.
I have the manual for the Sony 601-esd and I think is says the boobie
lights (Red, FlashRed, FlashGreen, FlashGreen, SteadyGreen) is a
logical procession from poor signal to good signal.
Additionally there is a tracking indicator (bar graph moving to the
right)
I can see from the schematic the LED's are driven from the decoding
logic (didn't really need a schematic for that one).
However, I an not find out at which point errors are still being
corrected (apparently a few red flashes are fine) and at what point I
can not count on the data.
I am recovering through the spdif -> PC.
I am hoping there there is a audio archivists list with somebody who
used these 30 years ago and thought to ask Sony exactly what was
being measured with the lights.
P.S. I found the Sony 601-esd encoder/decoder is not a good match
with most later model VCR's and ep tapes. (the data on the back of a
VCR switch disappeared years ago).
--
John Gledhill
BIT WORKS Inc.
905 881 2733
[log in to unmask]
--
John Gledhill
BIT WORKS Inc.
905 881 2733
[log in to unmask]
---
This email has been checked for viruses by Avast antivirus software.
http://www.avast.com
