Wimbledon and the PPM

[SimonM]

It takes a lot to get my better half at all
interested in sound.

Last night's Norrie match was one such, however.

If I had to sum up the mix in one word, it would
probably be unprintable.

It's fair to point out that the umpire didn't do
much to control the spectators' rather undignified
yelling, however, the sound mixer has mics in
helpful positions, faders, and, one hopes, also
some discretion.

Neither of the latter two seem to have been used.

There were several points that the umpire ordered
to be replayed for reasons unknown (he was
unintelligible). On other occasions, the
commentators were also unintelligible, because of
the yelling (from separate mics -- they are, after
all probably using 4104s and behind glass).

If I wanted to be deafened by uncouth people, of
course I'd go there in person.

I really don't need an apparently amateur sound
super to imitate the 'being there' discomfort to
the fullest extent possible. I expect if I had 5.1
surround I could make out where each individual
hooligan was sitting, etc. What a great technical
improvement.

I obviously haven't 'got' with the modern age,
where evidently-drunk fools' utterances are more
important than the actual tennis. I'm afraid I
come from a past where one routinely made sure the
important stuff could actually be heard.

Does anybody else remember checking for mono
compatibility, and using 'squawk-boxes' to confirm
how things sounded on tinny speakers? We used to
think it mattered if the bulk of our audience
could hear things clearly.

I know: experienced hands retire, and in the brave
new freelance world, I fear the directors get what
they ask for, without any counter arguments being
put to them. Nobody dares any more.

I am reminded of the mythical 'directors' boxes'
in dubbing - with unused knobs they could twiddle
pointlessly, whilst the mixer got on with the real
job. There must be room for something similar in
those modern trucks with the clever expanding
sides, surely?

The BBC managed to do this well for decades. Now
quite frankly it should be ashamed of itself.

[Scott]

I don't know if this is practical but I have always thought there
should be two soundtracks (using the language option) for a choice
between studio sound and stadium sound. For football they could offer
a choice of Rangers commentary or Celtic commentary.

[Brian Gaff]

Yes you get similar issues in horse racing as well. Suddenly you hear the
level reduced as if the mixer only had switches and no pots.
Still a couple of days ago they let Cliff sing again at Wimbledon. A bit
flat though.
Brian

--

--:
This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"SimonM" <some...@large.in.the.world> wrote in message
news:ta3gf6$2ia$1...@dont-email.me...

[Brian Gaff]

I know on football you often find people listening to the radio for the
commentary and watching the screen for pictures.

Not that I care, but I did find it annoying on F1 this weekend when the
chaps talking in the studio kept making comments and nobody thought to tell
the blind what actually had occurred until much later on.

Brian

--

--:
This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!

"Scott" <newsg...@gefion.myzen.co.uk> wrote in message
news:jpiach1vsqsusomlt...@4ax.com...

[charles]

In article <ta3gf6$2ia$1...@dont-email.me>,

As I understod it, OBs have been "contracted out" for many years.

--
from KT24 in Surrey, England
"I'd rather die of exhaustion than die of boredom" Thomas Carlyle

[SimonM]

On 06/07/2022 09:46, Scott wrote:
> I don't know if this is practical but I have always thought there
> should be two soundtracks (using the language option) for a choice
> between studio sound and stadium sound. For football they could offer
> a choice of Rangers commentary or Celtic commentary.

That would be wonderful - they might even have a
"partisan/balanced" switch in a telly menu
somewhere (with a submenu to select your team,
obviously), and for Celtic/Rangers a choice of
translation... :-)

[Mark Carver]

Yes, but the production staff on board the OB trucks are still manned by
the client broadcaster.

In the case of Wimbledon the AELTA decided to take away the BBC's status
as host broadcaster about five years ago ,
and created an in house operation, WBS. NEP provide the technical
facilities, WBS produce the coverage.
However, the production staff for WBS were recruited from BBC Sport.

[Andy Burns]

Brian Gaff wrote:

> I did find it annoying on F1 this weekend when the
> chaps talking in the studio kept making comments and nobody thought to tell
> the blind what actually had occurred until much later on.

They didn't tell the sighted what had occurred either, and certainly didn't show
replays or alternate camera angles for a long time ... twitter did though.

[Scott]

And subtitles like the karaoke machine to sing along :-)

[Mary Wolstenholme]

What's the PPM?

[J. P. Gilliver (John)]

On Wed, 6 Jul 2022 at 09:46:54, Scott <newsg...@gefion.myzen.co.uk>
wrote (my responses usually FOLLOW):
[]

>I don't know if this is practical but I have always thought there
>should be two soundtracks (using the language option) for a choice
>between studio sound and stadium sound. For football they could offer
>a choice of Rangers commentary or Celtic commentary.

I haven't looked for the current Wimbledon - I'm not that interested in
the sport, though sometimes have left it on after the lunchtime news as
I find the pok, pok, pok, cheer noises restful and summery (I'm not
usually actually _watching_!) - but don't they use the multiple-sound
facilities to offer a commentary-free version? I thought they did, and
for some football matches and other sports too. (Not that that would
overcome the criticisms of the OP here, that louts in the crowd were
intrusive and the umpire was inaudible.)
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Actors are fairly modest...A lot of us have quite a lot to be modest about. -
Simon Greenall (voice of Aleksandr the "Simples!" Meerkat), RT 11-17 Dec 2010

[J. P. Gilliver (John)]

On Wed, 6 Jul 2022 at 10:47:08, Andy Burns <use...@andyburns.uk> wrote
(my responses usually FOLLOW):

>Brian Gaff wrote:
>
>> I did find it annoying on F1 this weekend when the
>> chaps talking in the studio kept making comments and nobody thought to tell
>> the blind what actually had occurred until much later on.

To be fair, AD needs a gap in the speech; I know for prerecorded prog.s,
if there's a long bit of dialogue with no gaps, they sometimes insert
the AD _before_ it, but of course they can't do that in a live
situation. (I didn't see this coverage though so don't know if that was
the reason.)

>
>They didn't tell the sighted what had occurred either, and certainly
>didn't show replays or alternate camera angles for a long time ...

Suggests some incompetence, certainly. Assuming the prog. producers
actually have all the feeds. (And if they don't, then the buck just
passes to those who do.)

>twitter did though.
>
(-:

[Andy Burns]

J. P. Gilliver (John) wrote:

> Andy Burns wrote:
>
>> They didn't tell the sighted what had occurred either, and certainly didn't
>> show replays or alternate camera angles for a long time ...
>
> Suggests some incompetence, certainly.

They were just making sure nobody had died ...

[Roderick Stewart]

On Wed, 06 Jul 2022 12:51:24 +0100, Mary Wolstenholme
<ma...@easynn.com> wrote:

>What's the PPM?

Peak Programme Meter.

A meter standardised by the BBC to enable consistent control of audio
levels. It was developed in the 1930s to do something that some
apparently still have difficulty with today.

Rod.

[J. P. Gilliver (John)]

On Wed, 6 Jul 2022 at 14:01:42, Roderick Stewart
<rj...@escapetime.myzen.co.uk> wrote (my responses usually FOLLOW):

Sometimes with two pointers, one with a slower decay than the other; LED
equivalents sometimes (I think virtually always in the case of sound
software that _simulates_ an LED "meter") have a bar for the fast
movement and a dot for the slow.

I think when they have only one pointer, if it's a BBC-type PPM, it's
the slow one, so it shows (prolongs) the peaks, hence the name.

[Roderick Stewart]

On Wed, 6 Jul 2022 14:20:32 +0100, "J. P. Gilliver (John)"
<G6...@255soft.uk> wrote:

>On Wed, 6 Jul 2022 at 14:01:42, Roderick Stewart
><rj...@escapetime.myzen.co.uk> wrote (my responses usually FOLLOW):
>>On Wed, 06 Jul 2022 12:51:24 +0100, Mary Wolstenholme
>><ma...@easynn.com> wrote:
>>
>>>What's the PPM?
>>
>>Peak Programme Meter.
>>
>>A meter standardised by the BBC to enable consistent control of audio
>>levels. It was developed in the 1930s to do something that some
>>apparently still have difficulty with today.
>>
>>Rod.
>
>Sometimes with two pointers, one with a slower decay than the other; LED
>equivalents sometimes (I think virtually always in the case of sound
>software that _simulates_ an LED "meter") have a bar for the fast
>movement and a dot for the slow.
>
>I think when they have only one pointer, if it's a BBC-type PPM, it's
>the slow one, so it shows (prolongs) the peaks, hence the name.

I thought the ones with two pointers were for stereo? I've also seen
side by side pairs of double pointer PPMs, one with red and green
pointers for left and right, and the other with different colours
(white and yellow?) for sum and difference.

Rod.

[Scott]

On Wed, 06 Jul 2022 14:01:42 +0100, Roderick Stewart
<rj...@escapetime.myzen.co.uk> wrote:

>On Wed, 06 Jul 2022 12:51:24 +0100, Mary Wolstenholme
><ma...@easynn.com> wrote:
>
>>What's the PPM?
>
>Peak Programme Meter.
>

Maybe not quite the same but I remember when tape recorders had
similar meters. It was never clear whether the aim was to keep the
needle out of the red or out of the red most of the time with
occasional peaks in the red area. I was told the latter, on the basis
that the main aim was to keep the sound level subjectively constant.

[SimonM]

On 06/07/2022 14:20, J. P. Gilliver (John) wrote:

> Sometimes with two pointers, one with a slower
> decay than the other; LED equivalents sometimes (I
> think virtually always in the case of sound
> software that _simulates_ an LED "meter") have a
> bar for the fast movement and a dot for the slow.
>
> I think when they have only one pointer, if it's a
> BBC-type PPM, it's the slow one, so it shows
> (prolongs) the peaks, hence the name.

Wrong at many levels. I'm sure Wikipedia has
something about this but anyway:

PPMs were/are used by many broadcasters, and there
are international standards describing their
behaviour.

Various printed scale types exist, including the
TPM ("Test Programme Meter"), and the EBU variant,
both of which have more scale divisions between
certain numbers (they look very similar), but the
idea is exactly the same.

PPMs are voltage-measuring devices and have high
impedance - any matching termination is external
to the meter. In contrast, the American-favourite
VU meters are merely bridge rectifiers and
originally had to be used in a circuit, which they
loaded. PPMs go across circuits, logically if not
actually.

The scale is 0 to 7 and all bar 0 to 1 are 4dB
between divisions. I believe valve PPMs had 1 to 2
as 6dB. The needle also rests on the right (beyond
"7") when not powered as the spring of the moving
coil movement was used to provide the rise time.
Solid-state PPM drivers use conventional moving
coil meters, but without the copper cylinder found
in the middle that damps AVOs and the like. Both
rise time and decay are defined in the standard(s).

Line-up level is PPM 4, maximum permitted signal
(normally) is PPM 6, which I think nowadays is
-12dB WRT 100% of a digital system. In the BBC
television chain (possibly radio also) there was
an extra 2dB above 6 before the brutal
transmission limiters kicked-in. This was fun to
exploit sometimes!

The only "slugged" ones I'm aware of were the huge
meters across the (then) four networks in London
Control Room (BH). I've never seen one anywhere
else (possibly transmitter stations) and can't
imagine why you'd want one.

One might occasionally come across "yellow spot"
ones (a spot on the meter's scale), meaning a
simplified driver circuit was being used, one with
non-standard ballistics. These were a cost saving
exercise, nominally for line-up purposes only, but
usually good enough for general use.

Twin PPMs:

Either A and B legs of a stereo signal, or sum and
difference. The difference needle usually had a
+20dB pushbutton, so that it could be used for
precise stereo lineup.

Normal use was red and green needles for A and B,
white for sum (i.e. mono) and yellow for
difference, but most sets of dual needle PPMs
allowed switching between modes.

There were other non-standard uses, for example
carrying record send and replay on a pair of
needles, for film dubbing (to help match signal
for opt-ins), and a similar purpose for matching
local output to network for regional telly opt-outs.

After my time, there was a change to the standard
slightly: in my day radio led with stereo. There
was a 3dB pad in the sum needle so that the meter
reflected the levels perceived by the ear: with
tone, 4 on the white needle equated to 3 1/4 on
the red and green ones. This meant as you panned
across a stereo soundstage, the needles would more
reasonably represent what was going on.

At some point in the 1990s, the 3dB pad was
altered to 6dB, so that A, B, and M all read the
same when tone is applied to both channels, and
thus the A and B channels peak an apparent 3dB
higher than before. It wasn't broke, dunno why it
needed 'fixing'. I have a six-needle set (4
needles plus an auxiliary twin PPM) that has a
push button to alter this pad between the two
standards.

Most PPM circuits work from a 15-0-15 supply and
accept balanced inputs. The last manufacturer of
the twin needle movements was Sifam (Earnest
Turner made the first ones), and they stopped
manufacture probably a decade ago.

There have been other versions such as edge-meters
and optical projection systems, but they have had
the same impedance, calibrated to the same values
and the same 0-7 scales.

I think the 'modulometers' on Nagras had a similar
range and ballistics, but the scale markings were
more akin to VU meters.

Finally, true PPMs do not catch really fast loud
transients - they are designed not to, as the
result of BBC experiments done in the 1930s. This
is supposed to be a bad thing in the digital age,
however in practice it all depends on the headroom
allowed for in a digital system. Personally I hat
'bouncy' peak meters and would use a mechanical
meter any day...

[SimonM]

There is no 'red' on a PPM's scale anywhere -
almost always white numbering on black (see my
other post).

It's a broadcast meter, not usually found in
domestic kit (High-end Revox a possible exception)
although excellent for recordings too.

The design brief in the 1930s was to make best use
of the roughly 26dB dynamic range available in the
transmission chain. That really hasn't changed.
the chain may have a better s/n ratio, but the
places we listen in don't.

[Mark Carver]

You're thinking of a VU meter, and its cousins. As Simon says, PPM
meters are used in broadcasting, and not just here but in other
countries too, but notably NOT in North America.

There is one story I heard, that when CBS hired Studio 1 at The London
Studios (nee LWT) for a week's worth of David Letterman's chat show,
they requested to have all the PPM meters on the audio desk replaced
with VUs !

[williamwright]

On 06/07/2022 15:40, Mark Carver wrote:
> There is one story I heard, that when CBS hired Studio 1 at The London
> Studios (nee LWT) for a week's worth of David Letterman's chat show,
> they requested to have all the PPM meters on the audio desk replaced
> with VUs !

Did they want them calibrating in feet and inches?

Bill

[MB]

On 06/07/2022 09:46, Scott wrote:

> I don't know if this is practical but I have always thought there
> should be two soundtracks (using the language option) for a choice
> between studio sound and stadium sound. For football they could offer
> a choice of Rangers commentary or Celtic commentary.

I thought they alread did that,

But if you had to very partisan commentaries the you would need a third
for normal people,

Perhaps should have male and female commentator channels for the women's
football tournament. So the men can moan about how much better male
footballers are and the women could point out the male footballers would
spend their time being stretchered off for broken finger nails.

[Liz Tuddenham]

SimonM <some...@large.in.the.world> wrote:

> ...The needle also rests on the right (beyond

> "7") when not powered as the spring of the moving
> coil movement was used to provide the rise time.

The rise time was determined partly by the time constant of the charging
circuit for the reservoir capacitor and partly by the inertia of the
moving components (whaich was made as low as possible). The pointer was
mounted flat-ways-on to the operator, so as to minimise its air-braking
effect and prevent it bending if the coil 'kicked'.

The reason for the RH zero was because the original circuit relied on
the square-law characteristics of a particular pentode, which were only
correct if the meter scale zero corresponded to maximum anode current -
a larger signal gave a lower current. This had the added advantage that
in the event of a really huge signal cutting off the valve completely,
the meter never got driven past the mechanical zero, so it was
overload-proof.

To get an accurate square-law characteristic, the valve had to have the
right voltage on the screen grid. This was usually derived from a
supply which was stabilised by a neon tube and had to be adjusted
whenever the valve was changed. Manufacturing tolerances meant that
some valves had to be rejected if they would not line up properly.

Setting up a valve PPM is quite an art, even with the correct signal
sources and the pots working smoothly (which they often don't, now they
are 70 years old); all the adjustments interact.


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

[tony sayer]

In article <ta46g0$278b$1...@dont-email.me>, SimonM
<some...@large.in.the.world> scribeth thus

If your really bored or have too mucho time on your hands Orban have
IIRC a PPM meter of sorts thats for your computer screen its a Loudness
meter in the downloads but someone did a PPM PC version can't remember
who now!..


http://www.orban-europe.com/downloads/

--
Tony Sayer


Man is least himself when he talks in his own person.

Give him a keyboard, and he will reveal himself.

[tony sayer]

In article <1puolr4.1m9bjly99h208N%l...@poppyrecords.invalid.invalid>,
Liz Tuddenham <l...@poppyrecords.invalid.invalid> scribeth thus

Christ! remember doing that many years ago, adjust this twiddle that
mind the interactions and rinse and repeat!.

Was at PYE TvT IIRC...

[tony sayer]

In article <5g3bch5bb0hlgvlh1...@4ax.com>, Roderick
Stewart <rj...@escapetime.myzen.co.uk> scribeth thus

Yes, got some here:)...

a bit more info here;!...


https://en.wikipedia.org/wiki/Peak_programme_meter

[Roderick Stewart]

On Wed, 6 Jul 2022 23:03:48 +0100, tony sayer <to...@bancom.co.uk>
wrote:

I never had to do this but was told once by somebody who did that the
best way was to persuade BBC Tech Stores to trust you with the entire
box of these pentodes so you could select a good one by trying them in
the meter and then give the rest of the box back. This involved less
paperwork and less walking than signing out one at a time and
returning them if they didn't give a good match to the scale markings,
but it did mean that it became gradually less likely that you'd find a
good match because eventually all the good ones would be taken and the
box would be full of rejects. I think (though couldn't possibly prove)
that a reasonable throughput of new valves was ensured by contriving
occasional strategic accidents so that the dropped ones would have to
be replaced.

Rod.

[jon]

On Wed, 06 Jul 2022 09:52:34 +0100, Brian Gaff wrote:

> Yes you get similar issues in horse racing as well. Suddenly you hear
> the level reduced as if the mixer only had switches and no pots.
> Still a couple of days ago they let Cliff sing again at Wimbledon. A
> bit
> flat though.
> Brian

I have always thought that narration and background music should be on
separate stereo channels, so that the background level music can be turned
down to enhance the narration in documentaries.

[Brian Gaff]

I remember having an Amstrad deck, and a Tandberg deck. On the Tandberg
which had peak meters you could get a clean nice recording with just a tiny
bit into the read. On the Amstrad Us you had to keep the meters about a
third of the way up to get the same recording levels, and Sony were not a
lot better, but Technics on the other hand were like the Tandberg and worked
best of all, in my view. Surely the best of all metering should be peak hold
ones as we started to see when they using real moving coil meters, just
before my sight went.

Brian

--

--:
This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!

"williamwright" <wrights...@f2s.com> wrote in message
news:jilmc8...@mid.individual.net...

[Roderick Stewart]

The BBC PPM is effectively peak hold, but with a numerical definition
of what counts as a peak. It's been embodied in British Standards
since the 1930s. I'd have to look it up, but I think it says that a
2ms burst has to read a certain percentage of the actual peak value.
The inertia of a galvanometer pointer would normally reduce the
reading for shorter bursts of sound even if they were the same
amplitude, so the electronics that drives the meter is designed to
store the value long enough for the meter to display it properly.
Bursts of shorter duration would be missed by the meter, but it was
established experimentally that they could be curtailed by a fast
acting limiter without audible detriment.

The intention is of course to produce a workable compromise between
adequate level to cover background noise while avoiding
overmodulation, and also presenting credible differences in sound
levels consistent with the type of programme material. All of this is
open to interpretation, but the important thing is that all the
characteristics of the meter are standardised, so that suitably
trained operators can make consistent judgements in any BBC control
room equipped with these meters. Regardless of the mechanics of the
actual meters or the details of the electronics driving them, they are
all specified to behave visually in the same way.

Rod.

On Thu, 7 Jul 2022 15:20:52 +0100, "Brian Gaff" <brian...@gmail.com>
wrote:

[Liz Tuddenham]

Roderick Stewart <rj...@escapetime.myzen.co.uk> wrote:

[...]

> The intention is of course to produce a workable compromise between
> adequate level to cover background noise while avoiding
> overmodulation, and also presenting credible differences in sound
> levels consistent with the type of programme material. All of this is
> open to interpretation, but the important thing is that all the
> characteristics of the meter are standardised, so that suitably
> trained operators can make consistent judgements in any BBC control
> room equipped with these meters.

In addition to objective measurements, there was a large element of 'it
has to show what you hear subjectively' involved in the design. This
was extremely successful and is very helpful at showing up hearing
fatigue when doing difficult transfer work.

If the PPM doesn't appear to be in agreement with your hearing, go for a
long walk, then come back and check again. To your amazement, it will
now agree.

[Roderick Stewart]

On Thu, 7 Jul 2022 18:19:59 +0100, l...@poppyrecords.invalid.invalid
(Liz Tuddenham) wrote:

>If the PPM doesn't appear to be in agreement with your hearing, go for a
>long walk, then come back and check again. To your amazement, it will
>now agree.

:-)

Rod.

[NY]

"Roderick Stewart" <rj...@escapetime.myzen.co.uk> wrote in message
news:k5tdch5ihn6lma43q...@4ax.com...

> The BBC PPM is effectively peak hold, but with a numerical definition
> of what counts as a peak. It's been embodied in British Standards
> since the 1930s. I'd have to look it up, but I think it says that a
> 2ms burst has to read a certain percentage of the actual peak value.
> The inertia of a galvanometer pointer would normally reduce the
> reading for shorter bursts of sound even if they were the same
> amplitude, so the electronics that drives the meter is designed to
> store the value long enough for the meter to display it properly.
> Bursts of shorter duration would be missed by the meter, but it was
> established experimentally that they could be curtailed by a fast
> acting limiter without audible detriment.
>
> The intention is of course to produce a workable compromise between
> adequate level to cover background noise while avoiding
> overmodulation, and also presenting credible differences in sound
> levels consistent with the type of programme material. All of this is
> open to interpretation, but the important thing is that all the
> characteristics of the meter are standardised, so that suitably
> trained operators can make consistent judgements in any BBC control
> room equipped with these meters. Regardless of the mechanics of the
> actual meters or the details of the electronics driving them, they are
> all specified to behave visually in the same way.

I know this tread is about *sound* levels, but I'd like, if I may, to widen
it to exposure of pictures, which is a similar problem of peak levels and
avoiding maxing-out and hard-limiting.

I was watching the Djokevic/Norrie tennis match this afternoon (on BBC1 HD,
terrestrial) and I was rather horrified at the degree of peak-white clipping
on pictures. Now it was a very sunny day, and by later afternoon the light
was very one-sided. And of course the players are wearing white. But I
noticed that one half of both competitors was featureless white - no detail
in creases in the clothing etc. I wonder how the racks people are setting
the exposure. It looks as it if could have done with a good 1/2 to 1 stop of
reduced exposure to try to restore a bit more highlight detail to prevent
the featureless blob effect. How is "correct" exposure judged in extreme
conditions like this? Do they measure average exposure (either with incident
or reflected light meters) or do have equipment that looks at the histogram
(proportion of pixels in a frame with each brightness value) and adjusts the
exposure to reduce the level at the highlight end to below some acceptable
threshold to expose for the highlights?

It seemed to affect the high-angle camera that looks from behind one
competitor more than the lower-angle cameras that tended to be used for
close-ups during action replays, though some of the wide-angle shots of the
spectators on the sunny side of the stadium were truly ghastly.

I did wonder whether my TV had (unknowingly) some sort of "vibrant picture"
setting enabled which deliberately crushes the top n% and bottom m% into
peak white and "peak" black, in the same way that shops which developed and
printed negative film into prints tended to sacrifice a bit of highlight and
shadow detail to make the rest of the picture look vibrant: I remember when
I first tried scanning a colour neg with a computer scanner how much extra
highlight detail there was which had been lost in the print as I assumed had
been lost on the neg as well.

So I recorded a bit and watched it on my PC and its monitor, where I'm more
confident that the picture is not being "mucked about with" (to use a Terry
Pratchett phrase). From memory, I think there may have been a tad more
highlight detail but it was still not good.

Does the compression of studio-quality pictures into those which are
broadcast on terrestrial/satellite affect the dynamic range and degree of
highlight/shadow clipping, or does it just affect the degree of compression
artefacts?

[J. P. Gilliver (John)]

On Fri, 8 Jul 2022 at 22:10:24, NY <m...@privacy.invalid> wrote (my
responses usually FOLLOW):
[]

>I know this tread is about *sound* levels, but I'd like, if I may, to

(I keep seeing the subject line and automatically adding "their second
album was rubbish".)

>widen it to exposure of pictures, which is a similar problem of peak
>levels and avoiding maxing-out and hard-limiting.
>
>I was watching the Djokevic/Norrie tennis match this afternoon (on BBC1
>HD, terrestrial) and I was rather horrified at the degree of peak-white
>clipping on pictures. Now it was a very sunny day, and by later

[]

>how the racks people are setting the exposure. It looks as it if could
>have done with a good 1/2 to 1 stop of reduced exposure to try to

[]

>Does the compression of studio-quality pictures into those which are
>broadcast on terrestrial/satellite affect the dynamic range and degree
>of highlight/shadow clipping, or does it just affect the degree of
>compression artefacts?

As an outsider: are we talking cameras with remote-control irises (or
whatever the modern equivalent is)? Unless that is the case, presumably
there's nothing the racks people can _do_ if the camera itself is into
clipping.

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

Security is the perfect excuse to lock you out of your own computer.
- Mayayana in alt.windows7.general, 2015-12-4

[MB]

On 08/07/2022 23:20, J. P. Gilliver (John) wrote:
> As an outsider: are we talking cameras with remote-control irises (or
> whatever the modern equivalent is)? Unless that is the case, presumably

> there's nothing the racks people can_do_ if the camera itself is into
> clipping.

I would not have thought there is a great deal that can be done on a
fast moving live action programme when half the tennis court is in
bright sunshine and the other half in the shade?

I take all my still photographs in RAW mode so I can at least choose to
make the detail in shadows visible but not a lot that be done about the
highlights.

[Roderick Stewart]

On Fri, 8 Jul 2022 22:10:24 +0100, "NY" <m...@privacy.invalid> wrote:

[about vision control with tennis players wearing white]

>How is "correct" exposure judged in extreme
>conditions like this? Do they measure average exposure (either with incident
>or reflected light meters) or do have equipment that looks at the histogram
>(proportion of pixels in a frame with each brightness value) and adjusts the
>exposure to reduce the level at the highlight end to below some acceptable
>threshold to expose for the highlights?

There's no need for light meters with television cameras, because a
television camera effectively *is* a light meter, i.e. a device that
converts light levels into a visual indication of some sort. In this
case, the visual indication is the video signal itself, so that's what
they use to make their judgements.

Unless the practice has changed drastically since I left the biz, the
vision operator (or vision engineer, or whatever the production
company calls them) will have a picture monitor and a waveform
monitor, so they can judge subjectively what the picture looks like,
and also have an objective measure of modulation level for major
picture elements. The vision operator will have remote servo control
of the iris in the camera lens, and black level, and maybe a few other
electronic adjustments depending on the particular cameras in use. As
with PPM control of sound levels (which is what began this discussion)
there are no absolutely "correct" values, just guidelines, and
judgement, hopefully based on experience and common sense.

I haven't been involved with this for a while, so if they use
something other than a waveform monitor with digital cameras, maybe
someone can enlighten me? A histogram would show the proportions of
the signal at various amplitudes, but no indication of which parts of
the picture they belong to, so in my opinion not the best thing to use
for live television.

People wearing white on a sunny day against a darker background that
may be partially in shade is a severe test of the system, even with
modern cameras. Some cameras have a "soft clipper" circuit that can be
adjusted to start reducing gain at some level below 100% signal
voltage, which is a distortion of course because it affects other
piture components than the white objects that are causing the problem,
but it can sometimes give subjectively better results. Full adjustment
of this may or may not be available to the operator, so they may just
have to make the best compromise with what they've got.

Rod.

[SimonM]

On 09/07/2022 10:23, Roderick Stewart wrote:
> I haven't been involved with this for a while, so if they use
> something other than a waveform monitor with digital cameras, maybe
> someone can enlighten me? A histogram would show the proportions of
> the signal at various amplitudes, but no indication of which parts of
> the picture they belong to, so in my opinion not the best thing to use
> for live television.

It would make some sense to stick with waveform
monitors, since the sensor arrays are analogue,
digitized when they are read out.

> People wearing white on a sunny day against a darker background that
> may be partially in shade is a severe test of the system, even with
> modern cameras. Some cameras have a "soft clipper" circuit that can be
> adjusted to start reducing gain at some level below 100% signal
> voltage, which is a distortion of course because it affects other
> piture components than the white objects that are causing the problem,
> but it can sometimes give subjectively better results. Full adjustment
> of this may or may not be available to the operator, so they may just
> have to make the best compromise with what they've got.

It's a form of peak compression, functionally
similar to the audio equivalent. Of course you can
do the same clever stuff with RGB you hinted at
WRT raw images from stills cameras - not all
channels will necessarily overload together, and
the eye/brain tends to regard all really bright
things as white.

Also, of course, many cameras have a neutral
density wheel (or a single filter), between the
lens mount and the sensor block. In my day this
was done by the cameraman, but it might be
automated now - of course it won't help extreme
contrast in the way a signal 'compressor' might,
but it can get overall high light levels down to
manageable.

I doubt you necessarily want to squash the entire
contrast range. If, say, Caucasian skintones are
pegged at 17-18%, you'd expect too-dark shadows on
a sunny Wimbledon afternoon. The picture needs to
look natural (after a fashion), and you still need
to match cameras, too.

The other issue that still affects OBs is zoom
ramping - the change of effective aperture,
usually as the focal length increases* - but I'd
expect there must be a stored lens profile
nowadays that will tweak the iris as the focal
length is changed by the operator. I can't believe
this isn't accommodated in lens or camera firmware
(or both). It probably doesn't affect Wimbledon
much, because of the subject matter, but the
mega-expensive mega-zooms are very unlikely to be
constant aperture.

S.


*generally speaking, but on some complex zooms
this isn't even a smooth change.

[NY]

This wasn't the problem where the court was half in sunlight and half in
shadow. As the match progressed a shadow was shading one corner of the
overall grassed area but barely touched the region of play within the
external lines of the court. Had the match (and coverage of it)
continued much longer, it could have been a problem with competitors
moving from sunny to shaded part of the court.

This was extreme contrast between the lit and unlit side of a person who
was fully in the sun. They chose the exposure so the shaded side was
correct and the sunlit side was several stops overexposed, rather than
the sunlight side was correct and the shaded side was under. Neither
solution is perfect.


This was a problem even at the start of the match at about 15:00, when
the sun was still high in the sky and the whole of the grassy area and
both sides of the spectators' stands were full sun.

I wonder if part of the exposure decision is to avoid partial clipping,
where one or two channels, but not all three, clip - that can cause much
more intrusive problems with strongly coloured featureless areas rather
neutral white coloured areas. You see coloured clipping a lot on
documentaries shot on cheaper cameras (eg given to participants to
self-shoot): blank orange areas of skin on faces, or garish featureless
hi-vis jackets (eg on programmes about the running of the railways).
Hi-vis jackets are a particular problem because some have fluorescent
inks which convert UV into visible light and so emit more light than is
actually shining on them, or else they have a beaded surface like a
projector screen which concentrate the reflected light in one direction.

[Roderick Stewart]

On Sat, 9 Jul 2022 12:47:09 +0100, SimonM
<some...@large.in.the.world> wrote:

>The other issue that still affects OBs is zoom
>ramping - the change of effective aperture,
>usually as the focal length increases* - but I'd
>expect there must be a stored lens profile
>nowadays that will tweak the iris as the focal
>length is changed by the operator. I can't believe
>this isn't accommodated in lens or camera firmware
>(or both).

It usually is included as part of a zoom lens mechanism. The actual
aperture size increases as you zoom "in" (i.e. longer focal length) so
that the aperture ratio (f number) remains constant, but it usually
runs out of range before maximum zoom because the lens isn't
physically big enough to track for all f number settings. This is
generally accepted as a compromise, otherwise you'd never be able to
make a lens with a useful zoom range.

Think about the numbers. For a zoom lens with a maximum focal length
of 1000mm, if you wanted the aperture to track all the way at f2, the
diameter of the objective would have to be 500mm, which would be a
pretty hefty lens, but if f8 is acceptable, it only has to be 125mm.

Rod.

[Mark Carver]

On 08/07/2022 22:10, NY wrote:
>
>
> Does the compression of studio-quality pictures into those which are
> broadcast on terrestrial/satellite affect the dynamic range and degree
> of highlight/shadow clipping, or does it just affect the degree of
> compression artefacts?

No, emission compression has little to no effect on the highlights handling.

Don't forget Wimbledon is now being racked for HDR, have you taken a
look at the UHD/HDR stream on iplayer (assuming you have a compatible
telly)?

This is exactly the problem  HDR helps solve.

The sports pitch being half in bright sunshine, and half in shade, is
nothing new, happens with winter afternoon football every week, it's
something racks engineers have had to deal with for decades.

[NY]

"Mark Carver" <mark....@invalid.invalid> wrote in message
news:jivlqk...@mid.individual.net...

Ah, of course, HDR is coming. I don't have Sky, only Freesat and Frreview,
so the only UHD I've seen is the test transmissions on 12441V which are
impressive resolution: sharper that HD, even when down-scaled to a 1920x1080
computer monitor. I wonder if those use HDR.

I remember the winter football/rugby matches where half the pitch is
overexposed and half is in gloomy shadow. That made the racks people really
earn their money, constantly adjusting the iris as play moves between one
and the other. In a situation like that, I presume they try to adjust all
the cameras simultaneously so the picture looks similar when the vision
mixer switches from one camera to another, maybe with one camera kept at the
"previous" setting and trained on the "previous" side of the pitch in case
there is a need to switch back to action in the shadow when most of the play
is now in the sun, and vice versa.

[Mark Carver]

On 10/07/2022 13:38, NY wrote:
> "Mark Carver" <mark....@invalid.invalid> wrote in message
> news:jivlqk...@mid.individual.net...
>> On 08/07/2022 22:10, NY wrote:
>>>
>>>
>>> Does the compression of studio-quality pictures into those which are
>>> broadcast on terrestrial/satellite affect the dynamic range and
>>> degree of highlight/shadow clipping, or does it just affect the
>>> degree of compression artefacts?
>>
>> No, emission compression has little to no effect on the highlights
>> handling.
>>
>> Don't forget Wimbledon is now being racked for HDR, have you taken a
>> look at the UHD/HDR stream on iplayer (assuming you have a compatible
>> telly)?
>>
>> This is exactly the problem  HDR helps solve.
>>
>> The sports pitch being half in bright sunshine, and half in shade, is
>> nothing new, happens with winter afternoon football every week, it's
>> something racks engineers have had to deal with for decades.
>
> Ah, of course, HDR is coming. I don't have Sky, only Freesat and
> Frreview, so the only UHD I've seen is the test transmissions on
> 12441V which are impressive resolution: sharper that HD, even when
> down-scaled to a 1920x1080 computer monitor. I wonder if those use HDR.

If you have a suitable telly, Wimbledon is streaming right now on BBC
iplayer in UHD and HDR.

[Tony Gamble]

On 10/07/2022 14:10, Mark Carver wrote:

>
> If you have a suitable telly, Wimbledon is streaming right now on BBC
> iplayer in UHD and HDR.

I do have a suitable one and am annoyed I did not see this post when
Mark made it.

How can I find out when the iPlayer is streaming, or replaying,
something else in HD?

Tony

[Tony Gamble]

UPDATE. I found a list of repeats such as Blue Planet that are supposed
to be available on HD. I have set my iPlayer Beta to use the highest
quality but I doubt that is what is being sent to my LG tv via my Nvidia
Shield.

Any advice folks?

Tony

[Tony Gamble]

On 11/07/2022 09:33, Tony Gamble wrote:

>>
>
> UPDATE. I found a list of repeats such as Blue Planet that are supposed
> to be available on HD. I have set my iPlayer Beta to use the highest
> quality but I doubt that is what is being sent to my LG tv via my Nvidia
> Shield.
>
> Any advice folks?
>
> Tony
>

I am still waiting for someone to tell me how you folk learn about an
event being streamed live in UHD?

Tony

[Robin]

I'd imagine different folk use different methods: visit the BBC site,
read the emails you can sign up for, closely examine the entrails of the
animal of your choice, pay your local seer, etc etc. (I can only vouch
for some of these.)

https://www.bbc.co.uk/iplayer/help/questions/features/uhd-connected-tv/#/Notification

--
Robin
reply-to address is (intended to be) valid

[MB]

On 12/07/2022 14:28, Tony Gamble wrote:

> I am still waiting for someone to tell me how you folk learn about an
> event being streamed live in UHD?

"Want to know when new Ultra HD content is available?

We'll keep this page up to date with the latest Ultra HD content, but if
you scroll to the bottom of this page - or click here - you can add your
email address to our notification list. We'll let you know when new UHD
programmes are ready to watch! "