• "Front Row" (2025-7-28 19:15) technical oddities

    From J. P. Gilliver@21:1/5 to All on Mon Jul 28 23:15:02 2025
    (https://www.bbc.co.uk/sounds/play/m002gfzv)

    As I wanted the Richard Stilgoe tribute to Tom Lehrer, I've extracted
    the audio from this programme. It's nominally MPEG4 AAC, 48 kHz sample
    rate (so in theory up to 24 kHz), 320 kbps, stereo.

    1. It's definitely mono; my X-Y display is a straight diagonal line.
    Apart from the "BBC Sounds" sting in the first four seconds, the
    musician at 0:31 to 0:50, the Lehrer recordings at 4:19-4:48 and
    6:15-6:39, Cleo Laine 8:10-8:38, and Nick Drake 10:46-11:~11:13,
    12:31-12:5x*, and 15:13-15:39*. 17:05-17:27. 22:45-23:02 drama extract.

    Basically, only the musical inserts are stereo (and the *d ones only
    slightly so, more like a mono LP played on a stereo pickup). I did
    notice that the _angle_ of my X-Y plot varied between when different
    people were speaking, so sent in different proportions to left and right
    - so speakers _would_ appear to be in different places; however, clearly
    _not_ pickup from two or more microphones. Samira centre.

    Maybe this is normal (maybe has been for decades?) - record individual
    speakers with a mono mic., but mix (really, weight) them so they appear
    (only very slightly) at different positions? With the presenter centre.
    I wouldn't really have noticed the varying position, but I _am_
    listening on a laptop's internal speakers, so that's probably
    understandable; I only really noticed when the mono line varied in angle slightly between speakers.


    2. A lot of it is sort of soft-clipped on the negative side: I can't
    _hear_ any distortion, but a lot of the waveform is noticeably limited
    to about 0.25 in the negative direction, where positive peaks regularly
    exceed 0.5. Noticeable in the waveform, and in the X-Y display. (Not a
    DC bias; when quiet, it's still around the middle.) Most noticeably -
    maybe only - when Cally Callender (?) is speaking.


    3. The spectra for the various speakers have very different
    characteristics! This may be quite normal for such a programme; I've not
    looked at one before.

    Samira Ahmed (presenter): brickwalled at 17 kHz, as were the Lehrer
    extracts.
    Richard Stilgoe: brickwalled at 12 kHz, though he joined "on the line",
    so that may have been that. Noticeable (visually, not audibly) notches
    about 3.8-4kHz, and 6.8-7.
    Nick Drake archivists (Cally Callender and Laura Barton?) - same as Samira. Petra Volpe: as Richard Stilgoe - "from New York", so probably same.

    The 3.9 and 6.9 kHz notches on the remote line(s) intrigue me!

    --
    J. P. Gilliver

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From John Williamson@21:1/5 to J. P. Gilliver on Tue Jul 29 19:38:00 2025
    On 28/07/2025 23:15, J. P. Gilliver wrote:
    (https://www.bbc.co.uk/sounds/play/m002gfzv)

    As I wanted the Richard Stilgoe tribute to Tom Lehrer, I've extracted
    the audio from this programme. It's nominally MPEG4 AAC, 48 kHz sample
    rate (so in theory up to 24 kHz), 320 kbps, stereo.

    1. It's definitely mono; my X-Y display is a straight diagonal line.
    Apart from the "BBC Sounds" sting in the first four seconds, the
    musician at 0:31 to 0:50, the Lehrer recordings at 4:19-4:48 and
    6:15-6:39, Cleo Laine 8:10-8:38, and Nick Drake 10:46-11:~11:13, 12:31-12:5x*, and 15:13-15:39*. 17:05-17:27. 22:45-23:02 drama extract.

    Maybe this is normal (maybe has been for decades?) - record individual speakers with a mono mic., but mix (really, weight) them so they appear
    (only very slightly) at different positions? With the presenter centre.
    I wouldn't really have noticed the varying position, but I _am_
    listening on a laptop's internal speakers, so that's probably
    understandable; I only really noticed when the mono line varied in angle slightly between speakers.

    The normal layout for recording these shows is with the participants sat
    round a table and each has their own microphone. The inserts are added
    by the engineer. Quiz shows such as "Have I Got News For You" use a
    table for each team and one for the judge and scorer, with a microphone
    for each participant. There are photographs and videos available on line
    which show this. As you have noticed, stereo is simulated by panning the signals left and right in varying proportions. There are very good
    reasons for this including room noise and keeping relative levels steady.


    --
    Tciao for Now!

    John.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. P. Gilliver@21:1/5 to David Paste on Tue Jul 29 21:31:50 2025
    On 2025/7/29 19:20:49, David Paste wrote:
    On 28/07/2025 23:15, J. P. Gilliver wrote:

    Samira Ahmed (presenter): brickwalled at 17 kHz, as were the Lehrer
    extracts.
    Is this not an MP4 'thing'? If you download a piece of music from
    YouTube with yt-dlp so you can choose the format, the MP4 ones all have
    this 17 kHz limit whilst the Opus files don't.

    Ah, may well be! I will admit I just used y with no parameter (other
    than the URL, obviously), then used something to "extract original audio stream" from the mp4. (I use Pazera, as I can be sure with that that
    it's not doing any transcoding when I do the extraction. There are many
    other extractors available.) Though this wasn't from YouTube. But it may
    still be an mp4 "thing".

    I can't hear above 8 kHz anyway, but if I see in GoldWave that something
    has content higher, I try to preserve it. Samira (and the other person I
    said had the same spectrum) definitely had content up to 17 - the
    brickwall was very noticeable. But as you say may be an mp4 artefact
    rather than a BBC one.
    --
    J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

    This was before we knew that a laboratory rat, if experimented upon,
    will develop cancer. [Quoted by] Anne ([email protected]), 1997-1-29

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. P. Gilliver@21:1/5 to John Williamson on Tue Jul 29 21:37:24 2025
    On 2025/7/29 19:38:0, John Williamson wrote:

    []

    The normal layout for recording these shows is with the participants sat round a table and each has their own microphone. The inserts are added
    by the engineer. Quiz shows such as "Have I Got News For You" use a
    table for each team and one for the judge and scorer, with a microphone
    for each participant. There are photographs and videos available on line which show this. As you have noticed, stereo is simulated by panning the signals left and right in varying proportions. There are very good
    reasons for this including room noise and keeping relative levels steady.

    But each participant just collected with a single (mono) mic., which is
    then mixed in fixed proportion to the two channels, with the proportion differing from one speaker to another (presenter usually being 1:1,
    i. e. centre). Interesting.>
    --
    J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

    This was before we knew that a laboratory rat, if experimented upon,
    will develop cancer. [Quoted by] Anne ([email protected]), 1997-1-29

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From John Williamson@21:1/5 to J. P. Gilliver on Tue Jul 29 21:48:34 2025
    On 29/07/2025 21:37, J. P. Gilliver wrote:
    On 2025/7/29 19:38:0, John Williamson wrote:

    []

    The normal layout for recording these shows is with the participants sat
    round a table and each has their own microphone. The inserts are added
    by the engineer. Quiz shows such as "Have I Got News For You" use a
    table for each team and one for the judge and scorer, with a microphone
    for each participant. There are photographs and videos available on line
    which show this. As you have noticed, stereo is simulated by panning the
    signals left and right in varying proportions. There are very good
    reasons for this including room noise and keeping relative levels steady.

    But each participant just collected with a single (mono) mic., which is
    then mixed in fixed proportion to the two channels, with the proportion differing from one speaker to another (presenter usually being 1:1,
    i. e. centre). Interesting.>

    It also means that the vocal bits are 100% mono compatible with none of
    the phasing effects you would get using, say, a spaced pair.
    --
    Tciao for Now!

    John.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. P. Gilliver@21:1/5 to John Williamson on Tue Jul 29 23:37:28 2025
    On 2025/7/29 21:48:34, John Williamson wrote:
    On 29/07/2025 21:37, J. P. Gilliver wrote:

    []

    But each participant just collected with a single (mono) mic., which is
    then mixed in fixed proportion to the two channels, with the proportion
    differing from one speaker to another (presenter usually being 1:1,
    i. e. centre). Interesting.>

    It also means that the vocal bits are 100% mono compatible with none of
    the phasing effects you would get using, say, a spaced pair.

    Good point I hadn't thought of! Similar to the slight variable-comb
    effects used when combining the channels from a stereo pickup playing a
    mono record on a non-linear-tracking turntable.
    --
    J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

    Of course some of it [television] is bad. But some of everything is bad
    - books, music, family ... - Melvyn Bragg, RT 2017/7/1-7

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to J. P. Gilliver on Wed Jul 30 08:31:10 2025
    J. P. Gilliver <[email protected]> wrote:

    On 2025/7/29 21:48:34, John Williamson wrote:
    On 29/07/2025 21:37, J. P. Gilliver wrote:

    []

    But each participant just collected with a single (mono) mic., which is
    then mixed in fixed proportion to the two channels, with the proportion
    differing from one speaker to another (presenter usually being 1:1,
    i. e. centre). Interesting.>

    It also means that the vocal bits are 100% mono compatible with none of
    the phasing effects you would get using, say, a spaced pair.

    Good point I hadn't thought of! Similar to the slight variable-comb
    effects used when combining the channels from a stereo pickup playing a
    mono record on a non-linear-tracking turntable.

    ...made even worse on early recordings by the recording engineer
    twisting the recording stylus to help throw the swarf to one side. The
    two 'channels' can be out of step by an amount which depends on how far
    up the groove walls the elliptical playback stylus makes contact, so
    even a parallel-tracker won't play them correctly unless the cartridge
    is mounted on a swivel and can be offset to the correct angle.

    At about 8 Kc/s, the HMV frequency test record can give almost purely
    circular movement to the stylus tip; the two groove walls are 90-degrees
    out of phase with a 0025" radius elliptical stylus mounted orthogonally.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. P. Gilliver@21:1/5 to Liz Tuddenham on Wed Jul 30 13:48:34 2025
    On 2025/7/30 8:31:10, Liz Tuddenham wrote:
    J. P. Gilliver <[email protected]> wrote:

    On 2025/7/29 21:48:34, John Williamson wrote:

    []

    It also means that the vocal bits are 100% mono compatible with none of
    the phasing effects you would get using, say, a spaced pair.

    Good point I hadn't thought of! Similar to the slight variable-comb
    effects used when combining the channels from a stereo pickup playing a
    mono record on a non-linear-tracking turntable.

    ...made even worse on early recordings by the recording engineer
    twisting the recording stylus to help throw the swarf to one side. The

    I wasn't aware of that!

    two 'channels' can be out of step by an amount which depends on how far
    up the groove walls the elliptical playback stylus makes contact, so
    even a parallel-tracker won't play them correctly unless the cartridge
    is mounted on a swivel and can be offset to the correct angle.

    (Which mine - Marantz TT520 - certainly isn't.)>
    At about 8 Kc/s, the HMV frequency test record can give almost purely circular movement to the stylus tip; the two groove walls are 90-degrees
    out of phase with a 0025" radius elliptical stylus mounted orthogonally.


    I presume that's a test record made before stereo? (What date _is_ it,
    and what frequencies does it contain [assuming it has tones or a sweep],
    out of interest?)
    --
    J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

    just because you are offended - doesn't mean you are right

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to J. P. Gilliver on Wed Jul 30 16:49:55 2025
    J. P. Gilliver <[email protected]> wrote:

    On 2025/7/30 8:31:10, Liz Tuddenham wrote:
    J. P. Gilliver <[email protected]> wrote:

    On 2025/7/29 21:48:34, John Williamson wrote:

    []

    It also means that the vocal bits are 100% mono compatible with none of >>> the phasing effects you would get using, say, a spaced pair.

    Good point I hadn't thought of! Similar to the slight variable-comb
    effects used when combining the channels from a stereo pickup playing a
    mono record on a non-linear-tracking turntable.

    ...made even worse on early recordings by the recording engineer
    twisting the recording stylus to help throw the swarf to one side. The

    I wasn't aware of that!

    two 'channels' can be out of step by an amount which depends on how far
    up the groove walls the elliptical playback stylus makes contact, so
    even a parallel-tracker won't play them correctly unless the cartridge
    is mounted on a swivel and can be offset to the correct angle.

    (Which mine - Marantz TT520 - certainly isn't.)>
    At about 8 Kc/s, the HMV frequency test record can give almost purely circular movement to the stylus tip; the two groove walls are 90-degrees out of phase with a 0025" radius elliptical stylus mounted orthogonally.


    I presume that's a test record made before stereo? (What date _is_ it,
    and what frequencies does it contain [assuming it has tones or a sweep],
    out of interest?)

    One side of HMV DB4037 is a set of bands at different frequencies with
    the highest, 8.6 Kc/s, on the outside to take advantage of the higher
    surface speed. There are other sides in the set with fixed and sweep
    tones, but none above 8.6 Kc/s. It was monophonic and was cut in 1936
    with the Blumlein mono cutterhead, which had a particular problem with 'azimuth' offset. They were all recorded with a 'U'-bottomed groove,
    which needs a truncated elliptical playback stylus for the best results
    - and that is the type most affected by theoffset.

    The Blumlein cutterhead had the cutting tip hung on the end of a
    trailing cantilever which was swung from side to side by rotary movement
    of a near-vertical shaft. If there was any misalignment of the cutting
    face, the force it generated by throwing the swarf off to one side would
    push it sideways and cause an even bigger misalignment. Because the
    restoring springs on the vertical shaft were not very stiff (to achieve
    the correct resonant frequency), this misalignment could be quite
    considerable.

    Another problem caused by the compliance of the springs was that the
    rush of air into the suction pipe (which was fitted to remove the swarf)
    had to be limited in order to prevent it disturbing the cutter; on
    Columbia DX73 this is very obvious as a constant roaring background
    noise. I imagine the operator would have been very tempted to
    deliberately skew the cutter to aid swarf removal - although I have no
    proof of that, other than a large number of discs recorded with the
    Blumlein cutter which have considerable 'azimuth' errors.


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. P. Gilliver@21:1/5 to Liz Tuddenham on Thu Jul 31 00:43:37 2025
    On 2025/7/30 16:49:55, Liz Tuddenham wrote:
    J. P. Gilliver <[email protected]> wrote:

    []

    I presume that's a test record made before stereo? (What date _is_ it,
    and what frequencies does it contain [assuming it has tones or a sweep],
    out of interest?)

    One side of HMV DB4037 is a set of bands at different frequencies with
    the highest, 8.6 Kc/s, on the outside to take advantage of the higher
    surface speed. There are other sides in the set with fixed and sweep
    tones, but none above 8.6 Kc/s. It was monophonic and was cut in 1936
    with the Blumlein mono cutterhead, which had a particular problem with 'azimuth' offset. They were all recorded with a 'U'-bottomed groove,
    which needs a truncated elliptical playback stylus for the best results
    - and that is the type most affected by theoffset.

    So 8.6 k (strange number!) was the highest it was thought worth worrying
    about in 1936. Interesting. (Do you have a chart of what was considered
    a reasonablr bandwidth at various dates?) I'd have thought that rather
    low for 1936, though couldn't justify that feeling if challenged.>
    The Blumlein cutterhead had the cutting tip hung on the end of a
    trailing cantilever which was swung from side to side by rotary movement
    of a near-vertical shaft. If there was any misalignment of the cutting
    face, the force it generated by throwing the swarf off to one side would
    push it sideways and cause an even bigger misalignment. Because the restoring springs on the vertical shaft were not very stiff (to achieve
    the correct resonant frequency), this misalignment could be quite considerable.

    Another problem caused by the compliance of the springs was that the
    rush of air into the suction pipe (which was fitted to remove the swarf)
    had to be limited in order to prevent it disturbing the cutter; on
    Columbia DX73 this is very obvious as a constant roaring background
    noise. I imagine the operator would have been very tempted to
    deliberately skew the cutter to aid swarf removal - although I have no
    proof of that, other than a large number of discs recorded with the
    Blumlein cutter which have considerable 'azimuth' errors.


    Thanks - you rarely fail to be fascinating!
    --
    J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

    As for cooking, what a bore that is. It's such a faff, thinking of what
    to have, buying it and cooking it and clearing up, then all you do is
    eat it - and have to start all over again next day.
    - Hunter Davies, RT 2017/2/4-10

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Liz Tuddenham@21:1/5 to J. P. Gilliver on Thu Jul 31 09:15:20 2025
    J. P. Gilliver <[email protected]> wrote:

    On 2025/7/30 16:49:55, Liz Tuddenham wrote:
    J. P. Gilliver <[email protected]> wrote:

    []

    I presume that's a test record made before stereo? (What date _is_ it,
    and what frequencies does it contain [assuming it has tones or a sweep], >> out of interest?)

    One side of HMV DB4037 is a set of bands at different frequencies with
    the highest, 8.6 Kc/s, on the outside to take advantage of the higher surface speed. There are other sides in the set with fixed and sweep tones, but none above 8.6 Kc/s. It was monophonic and was cut in 1936 with the Blumlein mono cutterhead, which had a particular problem with 'azimuth' offset. They were all recorded with a 'U'-bottomed groove,
    which needs a truncated elliptical playback stylus for the best results
    - and that is the type most affected by theoffset.

    So 8.6 k (strange number!) was the highest it was thought worth worrying about in 1936. Interesting. (Do you have a chart of what was considered
    a reasonablr bandwidth at various dates?) I'd have thought that rather
    low for 1936, though couldn't justify that feeling if challenged.>

    I think it was sufficient to cover the top resonance of the pickups of
    the day, which was a major concern - and, yes, it was probably the
    highest frequency they could sensibly record with the equipment they had
    at EMI at that date. Higher frequencies than that were recorded by
    slowing down the recording lathe, but this wasn't normal practice.

    It wasn't until Arthur Haddy of Decca produced the FFRR recording system
    that higher frequencies were recorded on commercial records (which made Mantovani's strings famous). Haddy's cutterhead was copied from Voigts
    moving coil design without acknowledgement; when Sugden later copied it
    from Haddy, Haddy was furious but couldn't do anything about it.

    For a much fuller explanation look at Peter Copeland's "Manual of
    Analogue Audio Resoration Techniques". It is supposed to be on the
    British Library's website but the page has vanished so you may have to
    download an archive copy. [That should keep you busy until I come back
    from holiday.]


    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. P. Gilliver@21:1/5 to Liz Tuddenham on Thu Jul 31 14:17:35 2025
    On 2025/7/31 9:15:20, Liz Tuddenham wrote:
    J. P. Gilliver <[email protected]> wrote:

    []

    So 8.6 k (strange number!) was the highest it was thought worth worrying
    about in 1936. Interesting. (Do you have a chart of what was considered
    a reasonablr bandwidth at various dates?) I'd have thought that rather
    low for 1936, though couldn't justify that feeling if challenged.>

    I think it was sufficient to cover the top resonance of the pickups of
    the day, which was a major concern - and, yes, it was probably the

    Ah yes, that's something (probably one of many!) I hadn't thought of.

    If that _was_ a major concern, did (the more expensive - which probably includes _all_ electronic ones in those days) systems (record player
    would be part of same piece of furniture as amplifier) contain a notch,
    or at least a low-pass filter, to cover the mentioned resonance? (I
    doubt it - certainly never heard any mention.)

    highest frequency they could sensibly record with the equipment they had
    at EMI at that date. Higher frequencies than that were recorded by

    Interesting - and I remain surprised.

    slowing down the recording lathe, but this wasn't normal practice.

    Wasn't that technique re-visited by some of the labels associated with
    high quality, in the '70s-'90s? Like maybe Deutsche Grammophon, Telarc..>
    It wasn't until Arthur Haddy of Decca produced the FFRR recording system
    that higher frequencies were recorded on commercial records (which made

    I hadn't realised that was an actual thing, rather than just marketing.
    (FFRR for me - Flanders & Swann, "Song of Reproduction", 1959-5-2: https://www.youtube.com/watch?v=EL5SzTSMxLU.)

    Mantovani's strings famous). Haddy's cutterhead was copied from Voigts

    Didn't Mantovani also sneak in a bit of double (or more) tracking?

    moving coil design without acknowledgement; when Sugden later copied it
    from Haddy, Haddy was furious but couldn't do anything about it.

    For a much fuller explanation look at Peter Copeland's "Manual of
    Analogue Audio Resoration Techniques". It is supposed to be on the
    British Library's website but the page has vanished so you may have to download an archive copy. [That should keep you busy until I come back
    from holiday.]


    Ah, once I'd put the missing t in, Google found it for me: <https://cool.culturalheritage.org/byform/mailing-lists/arsclist/2008/09/msg00144.html#:~:text=Peter%20Copeland%2C%20Conservation%20Manager%20at,technicians%20working%20in%20digitisation%20programmes.>Oh.
    404 from the link there.
    But the link on
    <https://www.vintage-radio.net/forum/showthread.php?t=165910> works,
    today anyway! Got it.Have a good holiday!
    --
    J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

    I don't see the requirement to upset people. ... There's enough to make
    fun of without offending.
    - Ronnie Corbett, in Radio Times 6-12 August 2011.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)