Sound issue
in [UK Mac]
|
From: J. J. Lodder on 25 Dec 2009 07:23 Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > > > J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > > > > > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > > > > > > > T i m <news(a)spaced.me.uk> wrote: > > > > > > > > > > > real-address-in-sig(a)flur.bltigibbet.invalid (Rowland McDonnell) wrot e: > > > > > > > > > > > > >300Hz is a `typical highish voice tone' (I hope) > > > > > > > > > > > > > When I was with BT we talked of 'voice' paths being from 300Hz to > > > > > > 3300Hz suggesting 300Hz was a pretty low voice tone (if we are > > > > > > talking about the same things)? > > > > > > > > > > I've no idea what a `voice path' might be - but 3kHz is `overtone' > > > > > territory for the human voice, not a fundamental. At least, not for > > > > > anyone pitching their voice in the usual range. > > > > > > > > > > Hmm. > > > > > > > > > > "In telephony, narrowband is usually considered to cover frequencies > > > > > 300-3400 Hz." > > > > > > > > > > Okay, so I dunno. 400Hz sounds pretty high pitched to me and > > > > > telephone voices do seem to be missing the lower frequency part of > > > > > the voice sounds, so I've thought for as far back as I can recall. > > > > > Don't have a spectrum analyzer handy - if I did, I'd have a look. > > > > > > > > For the stereo signal 400 Hz is the roll-over point. > > > > > > What roll-over point? > > > > <http://en.wikipedia.org/wiki/Low-pass_filter> > > But that does nothing to answer my question. > > > > > below that the S-component drops with the usual 6 dB/octave. > > > > > > S-component? > > > > Stereo component? > > Uhuh. > > > M = L + R > > S = L - R > > > > L = M + S > > R = M - S > > (insert /2 where wanted) > > > > So, supposing you start with L and R signals, > > from a stereo microphone for example. > > Before publishing on disk one would > > 1) transform L and R to M and S > > Feed S through a high-pass filter, > > roll-off 6 dB/oct below 400 Hz, to obtain S'. > > Ah! *THAT* roll-over point - right, now you've explained what the hell > you're talking about. Sorry, I was under the impression that I was talking to a person with a basic technical understanding of the subect. > > 2) recombine M and S' to L' and R' > > Why? > > > This has the effect of making low frequencies mono only. > > (to be cut horizontally into the vinyl) > > But why? > > > No real loss, metre waves > > don't carry directional information, > > given the size of an average room. > > Why do that at all? Told you several times already. For good technical reasons a low frequency S-component is undesirable. 1) (old) Record players can't handle it well. 2) Many amplifiers/speakers are underdimensioned, yet their owners want max boom without clipping. 3) the low frequency components don't contribute much (if at all) to the perceived stereo effect. And apart from the good reasons, that's the way it is done. > > That's also why there is little loss > > in systems that use a single sub-wooofer. > > What do you mean by that? You lose little in perceived quality in systems with a sub-woofer in the centre and smaller satellite speakers for the middle and high. (as first commmercialised by Bose, and used by many computer speaker systems) Even with all the low coming from the centre humans still pick up the directional cues from the middle and high. > > > > Bass singers are therefore hardly affected, > > > > > > Er? > > > > They produce no significant power below 400 Hz. > > Even if they can go lower > > almost all the power is in the overtones > > You made this statement: > > "For the stereo signal 400 Hz is the roll-over point. > below that the S-component drops with the usual 6 dB/octave. > Bass singers are therefore hardly affected," > > Your explanation is no explanation at all. Sorry, can't make it clearer than this, (you do know the range of a bass singer?) Jan
From: Adrian Tuddenham on 25 Dec 2009 17:16 J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > A listen to a classic Decca recording will show the fallacy of all those > > assumptions. > > Try feeding the signals through a low-pass filter, > and see how much S-component you've got left. > But you cleary never tried it. > The pick up elements of those days > just can't handle the large vertical amplitudes. > > BTW, you can even see it with a magifying glass. > The long wavelength wiggles are purely horizontal, During the era of changeover from mono to stereo when both types of record were on sale, Decca produced two completely independent mixes for the two issues. The limitation on the vertical channel with a grooved medium was more due to the slew rate than to pure amplitude, so it was the high frequencies that were badly affected. The other reasons you give for central bass were more-or-less correct for commercial pop records, but they didn't apply to good quality orchestral recordings. With really good studio monitor speakers and a proper stereo recording, the 'bass' can appear to be be directional, mostly because the harmonics give the position away. (Example: The recent Radio 4 broadcast of the BBC Philharmonic Orchestra playing the "Upstairs-Downstairs" theme showed a very clear stereo image with quite noticeably directional bass.) If the bass is mono and pan-potted, it is usually central, but that is not a proper stereo recording. If a variation on the Blumlein shuffler is used to try to bring the bass into the middle of a genuine one-mic-pair stereo recording, the result can be confusing if the fundamental moves but the harmonics stay put. -- ~ Adrian Tuddenham ~ (Remove the ".invalid"s and add ".co.uk" to reply) www.poppyrecords.co.uk
From: J. J. Lodder on 26 Dec 2009 10:26 Adrian Tuddenham <adrian(a)poppyrecords.invalid.invalid> wrote: > J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > > A listen to a classic Decca recording will show the fallacy of all those > > > assumptions. > > > > Try feeding the signals through a low-pass filter, > > and see how much S-component you've got left. > > But you cleary never tried it. > > The pick up elements of those days > > just can't handle the large vertical amplitudes. > > > > BTW, you can even see it with a magifying glass. > > The long wavelength wiggles are purely horizontal, > > During the era of changeover from mono to stereo when both types of > record were on sale, Decca produced two completely independent mixes for > the two issues. Yes. Why wouldn't they? > The limitation on the vertical channel with a grooved medium was more > due to the slew rate than to pure amplitude, so it was the high > frequencies that were badly affected. The other reasons you give for > central bass were more-or-less correct for commercial pop records, but > they didn't apply to good quality orchestral recordings. Disagree. Large low frequency vertical amplitude is a problem. See for example <http://www.gzcd.cz/en/vinyl/customer-area/production-materials/master-f or-vinyl/> points 4 and 5 for the same recommendation. In extreme cases the needle may even jump. Other problems at the high frequency end are irrelevant for this point. > With really good studio monitor speakers and a proper stereo recording, > the 'bass' can appear to be be directional, mostly because the harmonics > give the position away. That's what I said already. And (also as I said already) it's the reason why a single central subwoofer with good satellite speakers works nearly as well. (with a crossover well below 400 Hz) > (Example: The recent Radio 4 broadcast of the BBC Philharmonic > Orchestra playing the "Upstairs-Downstairs" theme showed a very clear > stereo image with quite noticeably directional bass.) > > If the bass is mono and pan-potted, it is usually central, but that is > not a proper stereo recording. If a variation on the Blumlein shuffler > is used to try to bring the bass into the middle of a genuine > one-mic-pair stereo recording, the result can be confusing if the > fundamental moves but the harmonics stay put. Then there are -a lot- of non-proper stereo records on the market, Jan
From: Adrian Tuddenham on 26 Dec 2009 11:52 J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > Adrian Tuddenham <adrian(a)poppyrecords.invalid.invalid> wrote: > > > J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > > > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > > > > > > A listen to a classic Decca recording will show the fallacy of all those > > > > assumptions. > > > > > > Try feeding the signals through a low-pass filter, > > > and see how much S-component you've got left. > > > But you cleary never tried it. > > > The pick up elements of those days > > > just can't handle the large vertical amplitudes. > > > > > > BTW, you can even see it with a magifying glass. > > > The long wavelength wiggles are purely horizontal, > > > > During the era of changeover from mono to stereo when both types of > > record were on sale, Decca produced two completely independent mixes for > > the two issues. > > Yes. Why wouldn't they? Some record companies just combined the two channels, with or without filtering, to produce their mono versions, it was a lot cheaper and less time-consuming. The proportion of reverberation has to be different wth mono and that can't be properly controlled by just knocking the 'S' signal out of a stereo recording, which was why Decca took the extra trouble to get it right. > > > The limitation on the vertical channel with a grooved medium was more > > due to the slew rate than to pure amplitude, so it was the high > > frequencies that were badly affected. The other reasons you give for > > central bass were more-or-less correct for commercial pop records, but > > they didn't apply to good quality orchestral recordings. > > Disagree. Large low frequency vertical amplitude is a problem. > See for example > <http://www.gzcd.cz/en/vinyl/customer-area/production-materials/master-f > or-vinyl/> points 4 and 5 for the same recommendation. That refers to extremely high antiphase amplitudes at 20 c/s, which are rarely found in orchestral or natural instrumental recordings. (Except organs in reverberant surroundings) > In extreme cases the needle may even jump. The extreme cases are those resulting from high levels of modulation on pop records combined with poor playback equipment, they almost never occur in classical orchestral recordings intended for playback on good equipment. > Other problems at the high frequency end are irrelevant for this point. High frequencies are the end of the spectrum where the limitations are imposed by the inherent geometry of the medium, rather than by deficiencies in the playback equipment. > > > With really good studio monitor speakers and a proper stereo recording, > > the 'bass' can appear to be be directional, mostly because the harmonics > > give the position away. > > That's what I said already. > And (also as I said already) > it's the reason why a single central subwoofer > with good satellite speakers works nearly as well. > (with a crossover well below 400 Hz) The problem with that setup on real instruments is that the fundamental and some harmonics may be below the crossover but other harmonics may be above it. This 'splurges' the image of the instrument across the stereo stage. Whilst it can work with some instruments, it might not work with all of them, so it is an unnecessary limitation of the playback system (unless space or price constraints oblige you to use a poor quality set-up, where you must accept those restrictions). On one occasion only, I have selected-out the bass and boosted it to the opposite channel to move the image of a double-bass nearer to the centre of the stereo stage. It was done because the instrument had become damaged before the recording session and was not producing its fundamental properly - also it was partly hidden behind the piano due to limitations in the recording studio layout. The process had to be done very cautiously and only at certain key points in the recording; the amount of boost was never more than 3dB. This was a jazz band session with only one bass instrument, I would never contemplated doing that to an orchestral recording. (Before you ask, the CD is not yet on sale because I am waiting for the band to arrange a second recording session to complete it - they might have repaired the double-bass by then.) The record producer who is aiming for a good realistic stereo result will not want all the bass instruments confined to the centre of the stage. It is much better to spread them out into their natural positions - or, at least, keep them separate so that they don't just merge into one droning hubbub. With pop records, the wanted effects are artificial and the producer may choose different techniques for the reasons you gave earlier. > > > (Example: The recent Radio 4 broadcast of the BBC Philharmonic > > Orchestra playing the "Upstairs-Downstairs" theme showed a very clear > > stereo image with quite noticeably directional bass.) > > > > If the bass is mono and pan-potted, it is usually central, but that is > > not a proper stereo recording. If a variation on the Blumlein shuffler > > is used to try to bring the bass into the middle of a genuine > > one-mic-pair stereo recording, the result can be confusing if the > > fundamental moves but the harmonics stay put. > > Then there are -a lot- of non-proper stereo records on the market, Yes. It has never been easier to make good recordings - and there have never been so many bad recordings made and sold to the public. -- ~ Adrian Tuddenham ~ (Remove the ".invalid"s and add ".co.uk" to reply) www.poppyrecords.co.uk
From: Adrian Tuddenham on 26 Dec 2009 11:52
Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: > Adrian Tuddenham <adrian(a)poppyrecords.invalid.invalid> wrote: > > > J. J. Lodder <nospam(a)de-ster.demon.nl> wrote: > > > > > Rowland McDonnell <real-address-in-sig(a)flur.bltigibbet.invalid> wrote: [...] > > During the era of changeover from mono to stereo when both types of > > record were on sale, Decca produced two completely independent mixes for > > the two issues. > > ISTR reading that more than Decca produced two different mixes, and not > merely for the change-over period (which never really ended, did it?). John Culshaw, Decca's A&R man, gives a detailed account of the period in his book "Putting the Record Straight". > > The limitation on the vertical channel with a grooved medium was more > > due to the slew rate than to pure amplitude, so it was the high > > frequencies that were badly affected. > > Werl, yeah. Anyway, doesn't the RIAA curve deal with the bass amplitude > problem completely? It certainly prevents the groove inter-cutting which would result from constant-amplitude recording characteristics, but the majority of the bass problem arises from pickups having low compliance stylus mountings. The earlier mono cartridges (e.g. TC8) had almost no vertical compliance at all. Heavy vertical LF modulation would make them jump, so a record producer who made stereo recordings with the idea that they might be played on cheap mono 'compatible' equipment, would have to make sure they were really compatible by constraining the LF to horizontal modulation. Even then, really heavy modulation or a frequency which coincided with a badly-damped arm resonance could well fling the pickup out of the grooves sideways; so the problem was mainly one of cheap and nasty playback equipment, not one of which direction the bass was recorded. A different problem was caused by playing records with high levels of HF on cheap equipment. The effective stylus tip mass was high and placed acceleration loads on the groove walls which broke them down (see John/Jean Walton's articles in Wireless World). The owner of the cheap equipment didn't find this a problem because it wouldn't reproduce the resulting distortion anyway - but when the damaged records were subsequently played on good equipment, it was revealed that they had been ruined. > > With really good studio monitor speakers and a proper stereo recording, > > the 'bass' can appear to be be directional, mostly because the harmonics > > give the position away. > > Oh! > > > (Example: The recent Radio 4 broadcast of the BBC Philharmonic > > Orchestra playing the "Upstairs-Downstairs" theme showed a very clear > > stereo image with quite noticeably directional bass.) > > I'd love to have heard stereo imaging from my bedside radio-alarm clock > ;-) On my studio monitors it was fairly breathtaking. As it was a voluntary tribute to the composer and was done 'on the cheap' at the end of another recording session, I suspect they used a single pair of close-spaced mics and didn't do anything to 'improve' it. ... Other recording engineers please note. -- ~ Adrian Tuddenham ~ (Remove the ".invalid"s and add ".co.uk" to reply) www.poppyrecords.co.uk |