50 kHz Soundstream Files

[Sean B]

Were any of these ever made available to the public? Anyone ever hear one of these recording played back at 50 kHz (or 48 kHz)? Just curious. I was listening to some Michael Murray on the Methuen organ, and thinking about how it might sound without the damaging sample rate conversion in the way.


Sean B

[geoff]

On 2/12/2020 8:52 pm, Sean B wrote:
> I was listening to some Michael Murray on the Methuen organ, and thinking about how it might sound without the damaging sample rate conversion in the way.
>
>

Probably indistinguishable.

geoff

[Scott Dorsey]

geoff > wrote:
>On 2/12/2020 8:52 pm, Sean B wrote:
>> I was listening to some Michael Murray on the Methuen organ, and thinking about how it might sound without the damaging sample rate conversion in the way.
>>
>
>Probably indistinguishable.

There were never any files. When the Telarc records were released, the
conversion was done by playing back the Soundstream machine's analogue output
into the analogue input of a PCM-1630 machine. No computer or workstation
involved.

In 2017, Paul Blakemore demonstrated a restored Soundstream machine playing
back one of the Telarc recordings live at the AES show and it sounded pretty
good. At the time there was some discussion about whether it would be
possible to get a digital output from the Soundstream in order to directly
digitize some of those recordings and the basic answer is that it would be
possible but it might require some software to reproduce some of the
error processing and so forth.

I would suspect that a digital transfer would be a lot better than the CD
releases available, given how horrible the 1630 machine was. Although the
Soundstream used a state of the art ladder converter and sounded far better
than the Sony recorders, I suspect modern conversion would be audibly better.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[geoff]

On 3/12/2020 4:26 am, Scott Dorsey wrote:
> geoff > wrote:
>> On 2/12/2020 8:52 pm, Sean B wrote:
>>> I was listening to some Michael Murray on the Methuen organ, and thinking about how it might sound without the damaging sample rate conversion in the way.
>>>
>>
>> Probably indistinguishable.
>
> There were never any files. When the Telarc records were released, the
> conversion was done by playing back the Soundstream machine's analogue output
> into the analogue input of a PCM-1630 machine. No computer or workstation
> involved.
>
> In 2017, Paul Blakemore demonstrated a restored Soundstream machine playing
> back one of the Telarc recordings live at the AES show and it sounded pretty
> good. At the time there was some discussion about whether it would be
> possible to get a digital output from the Soundstream in order to directly
> digitize some of those recordings and the basic answer is that it would be
> possible but it might require some software to reproduce some of the
> error processing and so forth.
>
> I would suspect that a digital transfer would be a lot better than the CD
> releases available, given how horrible the 1630 machine was. Although the
> Soundstream used a state of the art ladder converter and sounded far better
> than the Sony recorders, I suspect modern conversion would be audibly better.
> --scott
>

So what media is the data retained o internal to the Soundstream unit ?
If still in existance surely is would be a fairly minor DP exercise
these days to somehow access, transcode it to a usable digital format ?

geoff

[Scott Dorsey]

geoff > wrote:
>>
>
>So what media is the data retained o internal to the Soundstream unit ?
>If still in existance surely is would be a fairly minor DP exercise
>these days to somehow access, transcode it to a usable digital format ?

It's a 1" Honeywell instrumentation recorder with 16 bits across the tape
width. As I recall there is no parity, so the decoder stores each word
twice, and on playback if they don't compare identically it interpolates
over the error.

http://www.panix.com/~kludge/soundstream1.jpg
http://www.panix.com/~kludge/soundstream2.jpg

Easiest thing might be to take the raw parallel data off the transport into
a DRV-11 sort of interface and deal with everything in software. Would not
be expensive to do.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Sean B]

The CD booklet claims that the sample rate conversion is performed by a Studer SFC-16.

[Scott Dorsey]

Sean B > wrote:
>The CD booklet claims that the sample rate conversion is performed by a Studer SFC-16.

That would be the smart way to do it and MUCH better than going through
an analogue generation like the older Telarc stuff.

I don't know how tight or accurate the filters on the Studer are... no doubt
modern SRC could be much better than anything done in hardware back then...
but it's going to be a whole lot better than any A/D or D/A at the time and
pretty much blameless.

Hmm... looking it up... and I do actually have a manual for the Studer D19...
it looks like the SFC16 card was really just a couple AD1890 chips and some
glue logic. The AD1890 has been bettered by quite a bit today, but it was
inoffensive and never a problem.

So likely you have something about as good as you're going to get from that
master.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."