Do they think we do not hear the difference?

[Gordon]

I like it, but we must not tempt fait!

Thank you for clarifying the Nicam feeds to transmitters. If you look at this site you will see some photographs that I took of the Nicam feeds, amongst other things, at the North Hessary Tor transmitter when Radio 1 started to broadcast in FM from that transmitter. I was friendly with the engineers of the time and they invited me to the opening. My interest was always with Radio 3 I hasten to add!

http://tx.mb21.co.uk/gallery/gallerypage.php?txid=154&pageid=1131

NHTL: If you look here:

http://www.bbceng.info/Install/comms...changeover.htm and here:

http://www.audiomisc.co.uk/BBC/PCMandNICAM/History.html

you will find some information about NICAM system design and installation, finishing in 1993, as well as some further links to details of the various NICAM equipments and a picture gallery of the team. All FM listeners have been hearing their FM via NICAM 3 since then but those around London since 1972, starting with NICAM 2.

Bearing in mind the discussion in this thread, note that the PCM parameters include the quantising accuracy to 14 bits [earlier it was 13] companded to 10 and a sampling frequency of 32 kHz, a bit rate of 320 kBit/s per channel. A stereo pair plus its housekeeping overhead is 676 kBit/s. 12 pairs like this, with some more overhead, makes 8.448 MBit/s which is/was an international digital telephony standard bit rate.

That sampling frequency means a sharp anti-alias filter to avoid alias at 16 kHz, never mind the pilot tone at 19, if the required 15 kHz bandwidth is to be maintained. These sharp filters will impart the same transient issues as we have been laboriously discussing here.

[NHTL]

NHTL: If you look here:

http://www.bbceng.info/Install/comms...changeover.htm and here:

http://www.audiomisc.co.uk/BBC/PCMandNICAM/History.html

Thanks Gordon for the wonderful links.

[Resurrection Man]

Likewise, Gordon, for one of the names was at my wedding! That took me back a few years or more.......

[An_Inspector_Calls]

Gordon,

Thanks for the lengthy reply. I don't have some fond idea that LP is phase neutral!! But listening to some LPs from the 80s, some of these have stunning clarity of transients that I've never heard matched by CD. Maybe they used digital recording for those sessions (which blows my argument apart) but at what sampling and bit density?

I agree about sneak phase changes scattered down the audio hi-fi path, and you mention phase changes in cross-overs; an interesting take on that is ATC's set-up of their active cross-overs in the SCM loudespeaker series where they actually adjust for common phase response when 'tuning' the cross-over. So clearly, they're bothered about phasing (and even more bothered about second harmonic distortion of course - see the super-linear mid-range units they make).

[Gordon]

Gordon,

Thanks for the lengthy reply. I don't have some fond idea that LP is phase neutral!! But listening to some LPs from the 80s, some of these have stunning clarity of transients that I've never heard matched by CD. Maybe they used digital recording for those sessions (which blows my argument apart) but at what sampling and bit density?

I agree about sneak phase changes scattered down the audio hi-fi path, and you mention phase changes in cross-overs; an interesting take on that is ATC's set-up of their active cross-overs in the SCM loudespeaker series where they actually adjust for common phase response when 'tuning' the cross-over. So clearly, they're bothered about phasing (and even more bothered about second harmonic distortion of course - see the super-linear mid-range units they make).

Multi amp loudspeaker with specific correction is of course a good thing but a bit expensive, but I suppse one loses the expense of a main power amp!! As regards LP transients in general, there is a clue in the response of cartridges - their tip mass/compliance resonance is usually too close to the top the audio band for comfort [your remarks about in band phase again] and so perhaps [I've no proof] that, and a whiff of distortion, gives a false dynamic feel to the sound??!? We have got very used to electronically reproduced sound and so, when we hear it sans distortion, etc we think it a bit flat? I always feel this after a concert!!

Anyway, to your query: see here for details of the Decca recording system designed by ex BBC R&D Tony Griffiths and his team.

http://theartofsound.net/forum/showthread.php?t=8386

I remember meeting him in about 1978 when I was working on digital video recording [and hence also audio] and we compared notes because we of course were working with modified video machines and he eventually used a compact IVC one for the Decca audio system [picture in link]. They also played with an early optical machine but I don’t know what happened to that because they used the IVCs throughout the 80s and there are still [?] some at Phonogram for remastering purposes but I suppose all those masters have been format converted by now.

R&D was going on in EMI on digital recording in the mid 70s too. I also knew [still do!] one of the team working at the same time on that system so we had to keep mum when talking with each of them!! Back then high precision ADCs were hard to make with full resolution ie a nominal 16 bit device would probably work out nearer 14 in practice and a bit non-linear with it. I built an 8 bit device for video [6 MHz bandwidth] in 1968 and although it tested out statically at 8, dynamically it only came in at 6 and a bit!!!

The Decca sampling was at 48 kHz and quantising precision was 16 bits at first and later on 20 bits. From Spring 1979 [Vienna New Year concert] Decca took a lead followed by EMI’s Previn disc of Debussy etc [there’s a picture somewhere of Previn leading on the recorder].



Each of these was a one off ie no or very simple but editing until the editing systems came along later. The Stockham Soundstream system [USA] used 50kHz, not sure about the bit precision [it was 16 eventually], and 3M had licensed some BBC technology for its machine. I remember a demo of the BBC R&D machine in the early 1970s. All the early machines were two channel only, but developed for multichannel later, the mixing having reduced multiple channels to 2. Because of the limited bit precision some of them used pre-emphasis but that was dropped later. Denon [Nippon Columbia] also was a pioneer of digital recording, using 42.75/13 in 1972, as was NHK [Japans's BBC] in 1969 using 32/13. For a history see here:

http://www.aes.org/aeshc/pdf/fine_dawn-of-digital.pdf but there are some errors in it.

it claims that Bell invented PCM - well it was an Englishman, Alec Reeves who wrote the patent, working in Paris for ITT, once the international arm of Bell, but not Bell any more and not in the US. There is also a mention of the famous Bert Whyte. No mention of EMI despite the fact that the AES publshed papers about it.

I have an article about the EMI system that states that the sampling frequency was 50kHz but the quantising precision [on the recorder itself] was only 14 bits and used a scheme similar to the BBC's NICAM [12+2] although the data fields allowed for 16. The later editing system probably used more to deal with mixing issues from the 16 channel mixer [an AES paper from 1981 describes it and confirms that the arithmetic supported final 16 resolution].



As an aside arising from your comment about transients in LP, see this extract:



this comes from a paper on Decca’s recording standards in the mid 1960s when they had adopted Dolby A processing for mastering. This gave their analogue machines a S/N getting on for 80+dB. The digital ones are much better than this in principle and so apart from band limitations [ca23kHz] and the issues we have been talking about here they should have been good enough for Wilkie!! Note how late these new cutters were, that new cutter system in 1981 would have had extra challenges, all because of digital. By this time DMM was also in play and so LP was getting a lift in quality, again as a result of the challenge from digital - and THAT is another story!!

And here's a quote for Michael Mailes, a Decca balancer, about digital recording:

Record buyers heard only the sound through their own audio equipment; they did not have the opportunity
to hear the sound direct from the recording console. We as engineers could compare the console output to
the off-tape sound. Comparing off-tape digital sound/off-tape analogue sound and direct from the console
sound it was obvious that Digital won hands down. Of course the first Digital records were issued on
vinyl. Vinyl versus Compact Disc, another controversial point for discussion! There were many critical
comments made about 'Digital' mostly totally invalid. If people preferred analogue sound it was not
generally because of technical reasons it was just 'The sound' that they were used to.

I remember after a playback on one of the early digital recording sessions in Chicago, (Solti/Chicago
Symphony Orchestra) one of the woodwind players asking, "was that a digital playback?" on being told it
was, he said "I thought so, the woodwinds and upper strings sound much cleaner".

[An_Inspector_Calls]

Gordon,

That's quite a formidable reply!

The resonance caused by stylus mass/compliance can give a peak at the top end of the audio band. I think the Shure V15 range suffered from this (albeit, slightly), but that's where a good designer will attempt to shift the resonance higher. Many modern Ortofon cartridges show this as a 40 kHz resonance, comfortably out of the audio band.

Re PCM: oh dear, I have to confess I was working on PCM at Dollis Hill (GPO) - but we'll move on!

Also in my early days was work on a bank of 64 ADCs (10 bit, 1 volt standard input voltage) converting an array of 4,096 power station analogue inputs every 50 mS (the time stamping had to be that fast and accurate time stamped because it was necessary to resolve cause of trips accurately). The digitisers were comparator/dividers (rather than comparator/ramp) which meant every digitisation took exactly the same time; all done in germanium transistors. Again, all that was long ago, we'd better move on!

[Gordon]

....Re PCM: oh dear, I have to confess I was working on PCM at Dollis Hill (GPO) - but we'll move on!

Ha!! Confessions eh! In the late 60s I knew some people at DH but in the TV area. Ian McDiarmid was Head of the TV group and John Thompson was the deputy and he went on to great things in BT. They contributed at lot to TV measurments for the network and also to early digital TV coding. I don't suppose you came across either of them, DH was a big place!!

I was not far away at GEC Telecomms R&D in N Wembley in the late 60s working on ADCs for digitising supergroups etc as well as video. Analogue PAL TV signals gave some interesting problems. Already at that time PCM telephony was operational at 1.536 MBit/s in the GPO network carrying 24 channels between exchanges. I had also worked at BBC R&D Kingswood when Sound in Syncs was being trialled and then made operational. Hard to think it was almost 50 years ago. I moved to the GPO headquarters in the city area in 1972 to work on digital transmission, telephony and video, on satellite systems which were exclusively analogue FM at that time. Heady days!!

Also in my early days was work on a bank of 64 ADCs (10 bit, 1 volt standard input voltage) converting an array of 4,096 power station analogue inputs every 50 mS (the time stamping had to be that fast and accurate time stamped because it was necessary to resolve cause of trips accurately). The digitisers were comparator/dividers (rather than comparator/ramp) which meant every digitisation took exactly the same time; all done in germanium transistors. Again, all that was long ago, we'd better move on!

Interesting work by the sound of it!! Different challenges to ours. Our ADC had to work at sample rates around 16 MHz and quantise to 8 bits - we used a strange but elegant design known as a "folding" coder due to a chap called Waldhauer at Bell in the US. Trouble was if we had used it commercially we'd have had to pay Bell royalties - so we started looking for other ways to do it too. The BBC used a more conventional design involving ramps and comparators. And those Geranium transistors bring back memories too, all that leakage to worry about, thermal runaway in power stages if they got too hot. That early transistor amplifier from Leak - the Stereo 30 - was full of them!!

[David-G]

I tried looking at the Art of Sound forum. Virtually every page was flagged up by my antivirus software as containing a malicious URL.

[Gordon]

I tried looking at the Art of Sound forum. Virtually every page was flagged up by my antivirus software as containing a malicious URL.

That's odd because my system doesn't and I use McAfee neither does the laptop using Norton. Perhaps it's the settings? I have never changed my alert settings from the defaults. I suppose that the anti virus system has a list of suspect sites given out during regular updates.

[Dave2002]

I like it, but we must not tempt fait!

Thank you for clarifying the Nicam feeds to transmitters. If you look at this site you will see some photographs that I took of the Nicam feeds, amongst other things, at the North Hessary Tor transmitter when Radio 1 started to broadcast in FM from that transmitter. I was friendly with the engineers of the time and they invited me to the opening. My interest was always with Radio 3 I hasten to add!

http://tx.mb21.co.uk/gallery/gallerypage.php?txid=154&pageid=1131

Wow. Is/was the kit really that big? I had thought it is possible to do everything with something the size of a small suitcase these days!

[Bryn]

... That early transistor amplifier from Leak - the Stereo 30 - was full of them!!

I'd rather forget how many times I had to replace the OC28 output germaniums of my teenage home built amp.

[Gordon]

Wow. Is/was the kit really that big? I had thought it is possible to do everything with something the size of a small suitcase these days!

It probably is now but the elements of the present design of NICAM go back a long way - the NHT was installation was done 20 years ago. If the NICAMs got too unreliable and/or spares were hard to find they might have to be replaced and one doubts whether a new design of NICAM would be used but it is not impossible and it would be a lot smaller!!

[Gordon]

I'd rather forget how many times I had to replace the OC28 output germaniums of my teenage home built amp.

Yes indeed!! And it was a good thing back then to match the output pairs if you could. The Leak S30 had AD140s which were pnp types and rather expensive.

Just been reading about Higgs of boson fame, I learn that his HiFi was vintage Leak - why not Bose?!!

[gradus]

I too owned a Leak Stereo 30 and it was trouble free for at least 20 years of constant use, including at one point deputising at full whack for a hefty valve amp that had mysteriously given up the ghost when washed in a pint of beer at a party.
Wish I still had my V15 111 working, sadly hors de combat owing to the indisposition of my Systemdek.

[Frances_iom]

...Just been reading about Higgs of boson fame, I learn that his HiFi was vintage Leak - why not Bose?!!

maybe he preferred a model without spin ?