Is the BIS 'warning' on their early discs for real?

[pastoralguy]

I've often expressed a great love for the Dvorak 'cello concerto and, having listened to MANY recordings, my favourite is Frans Helmerson with Neemi a Jarvi and the Gothenburg symphony orchestra. The cd I have is a recent one so no problems there. However, my local Oxfam had the original release for £2.99 so, intrigued by the 'Warning!' I bought it.

I live four in a block so cranking the volume up to full isn't really an option. However, all three sets of neighbours are away just now SO...

I'm intrigued by the statement 'Contrary to established practice(,) this recording retains the staggering dynamics of the Original performance. This May damage your loudspeakers, but given first-rate playback equipment you are guaranteed a truly remarkable musical and audio experience. Good luck!

Opinions much appreciated.

[Eine Alpensinfonie]

I would suggest it would damage your ears before damaging your speakers.

[pastoralguy]

I would suggest it would damage your ears before damaging your speakers.

Pardon.

[Stunsworth]

The problem would be caused by the amplifier clipping because the listener had turned up the volume to hear the quiet bits. That can destroy tweeters very quickly - I found this out the hard way many years ago when I played an electric guitar through a pair of Kef Corellis.

[johnb]

I'm intrigued by the statement 'Contrary to established practice(,) this recording retains the staggering dynamics of the Original performance. This May damage your loudspeakers, but given first-rate playback equipment you are guaranteed a truly remarkable musical and audio experience. Good luck!

This is very similar to the statement on some of the BIS Nielsen/Chung recordings. When I looked at what was going on it became very apparent that the recording engineers had processed the sound so that the levels were higher and this resulted in an enormous amount of clipping on ALL the loud passages - giving the overall sound a coarseness. I thought it was appallingly engineered - reminiscent of the worse practices of the pop industry.

[johnb]

PS This gives my thoughts on one of the BIS Nielsen/Chung, complete with graphical representations which show the clipping and a comparative recording by Blomstedt.

http://www.for3.org/forums/showthread.php?7051-Carl-Nielsen-Symphonies&p=244675#post244675

[Gordon]

Whilst the warning has the threats highlighted above, what they are really saying is that they haven't been "interventionist" about tweaking the dynamics the way engineers had to in the early-mid analogue era, when tape machines were a bit short of DR, especially when considering the later task of the disc cutting engineer, riding the gain for quiet bits and then pulling back in the climaxes. A practice that was not desirable but perhaps necessary then. Perhaps this should be seen as a bit of marketing rather than anything of consequence? No one should drive their HiFi at such a level as either to deafen them or lead to damage. Your ears should tell you what's going on!! If you are a heavy metal fan, techno-beat club music or the like then make sure your system components have the caacity to deal with it.

All engineers would be doing what BIS claim in the digital era [and even in the later analogue] because the headroom is there in the medium AND they have to be very careful of clipping. johnb's comment in #5/6 above suggests bad engineering and I agree. Some while ago we had a spate of posts in a thread about some strange goings on in some download files and even some CDs where an Audacity plot showed v bad clipping.

Nasty stuff clipping and to be avoided at all costs - it can occur in analogue too but not nearly so abruptly. Excessive level in analogue systems leads progressively into soft limiting which was used by some analogue engineers to "brighten" the sound with a whiff of distortion!! Whilst loudspeakers have a power rating that would seem to be very large, most of the power goes to the mid range/bass drivers and relatively little to the tweeter so its power rating is often quite low thus making it very vulnerable to heavy driving, whether because of sheer linear level or because of amp clipping.

It is arguable whether, even in this digital era, the full dynamics of a large orchestra and chorus should be retained for playback in a small living room. But then why compromise when some few people do have a large space? Personally I do find the DR of some recordings - some of Rattles eg - too wide for home listening, regardless of damage risk to hardware.

[mathias broucek]

Personally I do find the DR of some recordings - some of Rattles eg - too wide for home listening, regardless of damage risk to hardware.

Agreed. One of my fondest memories from the concert hall is the pppppp that Rattle used to conjour from the CBSO - for example at the start of Mahler 1. However it's not really practical for home listening.

[Dave2002]

This is very similar to the statement on some of the BIS Nielsen/Chung recordings. When I looked at what was going on it became very apparent that the recording engineers had processed the sound so that the levels were higher and this resulted in an enormous amount of clipping on ALL the loud passages - giving the overall sound a coarseness. I thought it was appallingly engineered - reminiscent of the worse practices of the pop industry.

There are several issues here. It does not follow that BIS were making erroneous statements about the dynamic range. What is sad is that they appear, in some cases, to have done the transfer to CD at too high a level, with the result that the loud passages clip, which as stated, is pretty appalling. I wonder how extensive the problem is on those early BIS CDs. I had some, and thought they were OK, and perhaps if only a few of the climaxes had the clipping the overall effect wouldn't have been too bad. Otherwise, and if the clipping were hard, the results could have been appalling. It may be that the dynamic range of the performances would barely match to that possible with a 16 bit communications medium.

[Gordon]

It may be that the dynamic range of the performances would barely match to that possible with a 16 bit communications medium.

I would have expected the masters to have been captured at higher than 16 bit giving plenty of headroom to control clipping at that stage and retain a natural DR. I think BIS used the Sony 20 bit mastering system but I'm not absolutely sure of that. Also they may have mastered in DSD [and then reduced to CD for the hybrid layer] which makes some of the numbers regarding samling and DR to be a bit more difficult to get.

If so the clipping was introduced in the digital world which is hard to do!! An EBU/AES 24 capture gives an enormous DR. Theoretically the Pk-Pk Signal/RMS Noise [near enough the same as DR] for a linear quantiser is 20 log (2^N times root 12) which for 24 bits is 155 dB. Knock off a couple of bits for headroom [-12dB] , and measure the signal as RMS too [2 root 2 less, another 9dB], and the result is about 134dB. Theshold of pain to threshold of hearing is a lot less and a good value might be about 110dB. Provided your equipment is calibrated correctly [monitor the digital numbers and avoid FFFFFF] there is absolutely no excuse for getting your analogue/digital capture wrong. Once you do that reducing for CD masters, including noise shaping etc in the decimation, should not corrupt the DR.

Do the sums for 16 bits and you get a "DR" of 107 dB or about 96dB in practice but with no room to provide a large headroom so be very very careful with your metering. Some of this is all to do with that metering. If using digital peak detectors every sample can be examined and a "sample and hold" ensures that the engineer sees anything that momentarily gets near to the clippers, unlike the old days of Volume Unit [VU, a sort of weighted average over a time constant in an attempt to measure the ear's loudness profile] or even PPMs using mechanicla deflection needle displays. So if the metering is not precise then peaks can get into the clippers without them being detected except by the ears of the engineers and producers.

Message is: if mastering at 16 you can't get a full "natural" DR that matches Pain/Silence thresholds. Nevertheless 96 dB is enormous and more than enough for a living room and probably enough for a quiet concert hall too. In concert/studio one never gets to the pain level and the background noise of audience and air-con etc sets the "silence" so 96 is not far off. So broadly speaking, it must be finger trouble at BIS, or wherever, that caused that gross clipping. I can't believe that it's in the masters, engineers should be glued to decent meters.

[johnb]

I've often expressed a great love for the Dvorak 'cello concerto and, having listened to MANY recordings, my favourite is Frans Helmerson with Neemi a Jarvi and the Gothenburg symphony orchestra. The cd I have is a recent one so no problems there. However, my local Oxfam had the original release for £2.99 so, intrigued by the 'Warning!' I bought it.

It would be interesting to hear your comparison of the two discs. They might be identical of course, or the recording on the newer CD might have been remastered. If the original release is anything like the BIS Nielsen/Chung you will soon know - the average level is *much* higher than normal.

[Stunsworth]

There are several issues here. It does not follow at BIS were making erroneous statements about the dynamic range. What is sad is that they appear, in some cases, to have done the transfer to CD at too high a level, with the result that the loud passages clip, which as ststed, is pretty appalling

My memory of the few early BIS recordings that I have is that there could be a large dynamic range. I have a percussion CD that features the above mentioned warning. As I say, I think it was more to do with issuing a warning about amplifier clipping rather than clipping in the recording itself.

[Stunsworth]

Could the clipping show in the link above be due to the conversion from 24 to 16 bit for the CD master?

[Bryn]

It would be interesting to hear your comparison of the two discs. They might be identical of course, or the recording on the newer CD might have been remastered. If the original release is anything like the BIS Nielsen/Chung you will soon know - the average level is *much* higher than normal.

John, I can't recall whether you wrote to Robert von Bahr Re. the clipping problem on those BIS CDs. I have found him to be ready to respond personally to specific enquiries regarding BIS disc issues.

[Dave2002]

Re Msg 10
Gordon

Agreed. The masters should have been made using at least 20 bits or equivalent resolution. Once the data has been captured it should have been possible to post process it to fit into 16 bits for CD without introducing clipping, or going down into the noise floor - dithering should help. The dynamic range should be good enough for most material. The graphs johnb provides indicate that something has gone wrong. Presumably the software used to produce those is OK, and corresponds to the listening experiences, otherwised we are not discussing the correct version. Was the ripping to capture the data done correctly?