No . . . and it leaves the question of "where?" in the "same environment" anyway. I suspect that the dummy head would not hear the same thing flown in the "normal" mic position as it would attached to a mannequin seated out in the audience. That by itself is part of the problem.
But more to "reality", there is a substantial repertoir of orchestral recordings available today, none of which were recorded with a dummy head in any location. There is *some* consistency in the "classical" world that leads to multi-mic'ed recordings being mixed to roughly the same balance as ORTF or Decca Tree or Blumlein, so in that (very limited) realm there is at least some semblance of a "standard" for delivery. We still have to "tune" our reproduction system and environment to make that sound "right", and not all recordings conform. Expand horizons to include studio mixes of other kinds of music (which constitute an overwhelming majority of the current market) and we find little conformance to *any* standard (and most recording engineers and producers have *no* experience with classical, or what sounds "best" for classical, at all. Nor do they care. But what they produce determines what's going to sell in the consumer equipment market.
To the extent that there is *anything* left of the old "high fidelity" market you will find more understanding of its issues in the movie sound world than in the "project studio" with "Pro Tools" and pop music. That's just a reality that we have to learn to live with . . .
This is frighteningly true from my perspective in so many ways.
You should really try and get you mitts on some binaural recordings and a set of quality in ear monitors. Since this is the market nowadays, these should be readily available. Too bad they are so scarce and the ones that are out there are gimmicky.
As far as live classical recordings sounding bright--the mixing/mastering process should take care of it, but it may be hard to translate that into the home environment. Have you thought about ITDG? SL is essentially doing the opposite of its implications. That may well account for the brightness. Problem is, a lot of concert halls are greater the 15 ms to start with. That makes setting up you living room to appreciate the recording pretty difficult. I can't help but think that surround sound and a fairly dead room might be higher fidelity.
I wrote some about it on my blog if you'd like to read:
audio blog: Tightening The Loudspeaker, Recording and Room Connection
A polite critique/comment would be appreciated.
Dan
Last edited:
Wow, a lot going on here today. Need to catch up! Spent the day in a very big room, about 80x90 feet and some 40 feet tall. A luxury aircraft hanger. Weird sound, for sure. Not bad, just weird. But that's another story (I'm in back there tomorrow).
FWIW, I have recorded a bit in the ORTF configuration with mics that actually came from the ORTF (Schoeps tube mics). I remember the recordings sounding darker than the real thing, not brighter. Similar results with other mics. Is it subjective, maybe?
Agree on the 1980s CDs. The often seem a bit "etched" to put it politely.
FWIW, I have recorded a bit in the ORTF configuration with mics that actually came from the ORTF (Schoeps tube mics). I remember the recordings sounding darker than the real thing, not brighter. Similar results with other mics. Is it subjective, maybe?
Agree on the 1980s CDs. The often seem a bit "etched" to put it politely.
Back in the day I had a CD player that included a front panel LED to indicate whether a CD was recorded with pre-emphasis or not. In the 80s I would say the LED was lit with about half the CDs I owned. The pre-emphasis bit flag on the CD engaged a matching de-emphasis filter in the cd player to reduce the high frequency content on discs where pre-emphasis was used. I'm guessing the purpose was to reduce HF noise as with FM broadcast. I owned several CDs afflicted with a decidely "etched" sound all of which did not light the front panel pre-emphasis LED. I began to wonder if perhaps the CD was recorded with pre-emphasis, but the flag was not set correctly on the CD. I added a manual over ride switch so I could turn on or off the de-emphasis filter in the CD at will. When I manually engaged the de-emphasis while playing these "etched" sounding CDs, the spectral balance was tilted to sound neutral to me.
I have often wondered if this was a happy coincidence, or if there actually were many CDs incorrectly flagged for pre-emphasis back in the early days of CDs.
Last edited:
It wouldn't surprise me if there were a few like that, or if some master tapes were digitized without removong the hf boost that had been put in to make up for deficiencies in vinyl and the cartridges of the day. But another issue in the early days of digital was aliasing. Most of the CDs made from analog master tapes sounded pretty good, in part because head gap-length erasure put a hard limit on high frequency response that was typically below the sampling frequency. The direct digital recordings often sounded harsh even if you rolled off the high end, and a good part of that may have been the inadequacy of the (supposedly) brick wall anti-aliasing filters. A lot of condenser mics of the day had nasty peaks in the low 20kHz range, and signal up there folded down into the passband as "noise" and "grunge". As soon as recording sample rates went up (allowing digital filtering before downconversion to redbook) most of the problem went away.
It wasn't all worked out (or even all thought out) when digital was introduced . . .
That's a really interesting question. I've never come across a CD player which had a visible "de-emphasis enabled" indicator, and although I'm aware that pre/de-emphasis is in the CD spec, I had kind of assumed that it was never really used in any great number of discs.
Mastering errors on DVD's where flags are set incorrectly are quite common, particularly the interlace flag is often set wrong.
I found a couple of threads discussing the issue, particularly in relation to software CD rippers and whether they apply the correct de-emphasis when ripping a CD to a file. Apparently iTunes does, but not all ripping software does.
CD/DAT with pre-emphasis: how to de-emphasize correctly? [Archive] - SH Forums
http://www.hydrogenaudio.org/forums/lofiversion/index.php/t60682.html
Exact Audio Copy was also mentioned as one ripper that will indicate whether a given CD has pre-emphasis on it's tracks or not, so I'll go through some of my CD's at some point and check which ones do. So far none of the ones I've tried have pre-emphasis, although they're nearly all discs released (or re-released) since 1995. (For example my thin sounding 1996 re-release of Brothers in Arms doesn't have pre-emphasis flagged, although the original release might well have done)
It seems to me that pre-emphasis on CD's was an ill conceived idea - it's not really needed, and introduces two problems -
1) if you don't encode the flag correctly, and every playback or ripping device doesn't honour it correctly you have a huge error in frequency response (why take the risk)
2) without pre-emphasis you can achieve a dead flat response with PCM, at least up to the point where the low pass filter comes into effect, whereas any de-emphasis curve implemented in the player is likely to have some small errors or deviations either due to it being implemented as an analogue filter with component tolerances (in early CD players I presume) or due to imprecision in the exact specification of the curve, if implemented in the digital domain, which I assume modern players do.
Either way there could be small but important errors in the frequency balance that wouldn't exist without going through de-emphasis/pre-emphasis.
I could be wrong but it seems likely that CD pre-emphasis was used in the 80's but has since been given up as a bad idea. It would be interesting to see if anyone has a recent release using pre-emphasis.
Edit: found another interesting thread. It seems that a few modern players don't get the de-emphasis curve quite right, or don't honour it at all!
http://forum.audiogon.com/cgi-bin/fr.pl?ymisc&1228247162
Also a list of CD's with pre-emphasis and some of those which may not be flagged correctly:
http://www.studio-nibble.com/cd/index.php?title=Pre-Emphasis_(release_list)
Last edited:
Checked and confirmed! - see attached image - especially the Miles Davis CD has HF souds recorded at levels as if to be reproduced for partially deaf audiences or through vintage extended range drivers. Both my Miles Davis and TOTO CD's are on the list.
The TOTO CD has more correction in the upper mids - around 1.5-2 to 6-7 khz
I think that has something to do with the HI-FI standart for allowed maximum distortion where the curve for the level of the distortion looks like a BSC electrical curve - less allowed distortion at higher frequencies, although it is about frequencies up to no more than 8 khz (because 3rd and second harmonics would be well out of the audible range for highed frequencies)
And probably the de-emphasis curve is a relative of the RIAA correction but with different shape.
I am always very glad when the theory and observations confirm each other see post 497 which is mine http://www.diyaudio.com/forums/multi-way/172806-flat-not-correct-stereo-system-50.html#post2487656 I didn't have a clue for any Emphasis or whatsoever, but I'm very happy that my observations were correctly interpreted.
Especially see the last paragraph. - That is a non professional, non informed thinking and conclusions. - That is why it is so important. I have noticed that where the treble was more forward (had more energy) there is some noise together with the musical signal and when the treble is reduced the noise gets less noticeable if noticeable at all... The Miles Davis CD's are a bright example of the above observations and experiments!
The shape of RIAA correction seems to be an answer to many questions.
Best regards!
Attachments
Last edited:
What are you playing the discs on though ? In theory if the player supports de-emphasis the end result should be balanced. Most CD players will do this correctly, but it seems that some PC software will not apply de-emphasis when ripping, with older versions of Exact Audio Copy being fingered as one guilty program - even though it has the ability to report the presence of the pre-emphasis flags on the disc, while iTunes silently does digital de-emphesis if necessary when importing.
So depending on what software you use, if you're ripping CD's to a lossless format like FLAC and then playing them back it's possible that these will be uncorrected and have far too much treble. The difference should be obvious when compared with the same disc played on a compatible CD player though.
As bolserst posted, the pre-emphasis curve is vastly different to RIAA equalization on vinyl, essentially the bass is flat, the response starts to rise near the middle of the midrange, and has risen by about 10dB at the top of the treble range.
Apart form getting the de-emphasis right, I also suspect some stuff has gotten onto modern releases with emphasis. I her track that I just think "that can't be right, why would they master it so hot?"
As mentioned above by dewardh, that hot EQ may be left over from something else. I know this sort of thing happen a lot in the print industry. Some sort of "EQ" gets built in for the system, then it carries along with the file for years, even to many other systems.
As mentioned above by dewardh, that hot EQ may be left over from something else. I know this sort of thing happen a lot in the print industry. Some sort of "EQ" gets built in for the system, then it carries along with the file for years, even to many other systems.
Yes, ever since I got my 42 inch HD TV, I've played DVD's more than I use my CD player anymore. I think the home theater is where people will be gravitating to more and more. My Eric Clapton Crossroads DVD's sound superb to my ears. It may not be perfect, but it's pretty dam close.
Very interesting discussion. Nice to see how many sentences now end with a "?".
Looking for ideal control rooms (or even consistent ones) may be a red-herring.
Look ago, the famous Nagra portable tape recorder (the design of which was a Ph.D. dissertation) came with the famous Beyer DT-48 headphones, with diaphragms said to be machined from blocks of aluminum. They had a lot of low-treble emphasis and served like microscopes to exaggerate stuff that the recordist needed to hear well.
Likewise, engineers often work in all kinds of environments and studios.
I think an engineer can pretty quickly size-up the sound of a control room and as a professional, can anticipate the relationship of that sound to the one he/she wishes to leave on the recording.
The engineer and producer make choices and they are not locked-into the sound of their studio as if they were passive victims of the sound there.
Looking for ideal control rooms (or even consistent ones) may be a red-herring.
Look ago, the famous Nagra portable tape recorder (the design of which was a Ph.D. dissertation) came with the famous Beyer DT-48 headphones, with diaphragms said to be machined from blocks of aluminum. They had a lot of low-treble emphasis and served like microscopes to exaggerate stuff that the recordist needed to hear well.
Likewise, engineers often work in all kinds of environments and studios.
I think an engineer can pretty quickly size-up the sound of a control room and as a professional, can anticipate the relationship of that sound to the one he/she wishes to leave on the recording.
The engineer and producer make choices and they are not locked-into the sound of their studio as if they were passive victims of the sound there.
This may belong as much in the "Godel's" thread.
An excellent test stimulus for evaluating the sound in a music room is music. Too bad "human meters" seem to be essential in the measurement process that uses music because they are more variable than even my mic measurements.
While we have all kinds of time-course and gated measurements. And a variety of test tones and tone-samples, largely identified, like tone controls on pre-amps, for their convenience to manage in electrical engineering at the constructive end or math simplicity at the analysis end.
But the real question is which kind of test stimulus best captures music-in-rooms?
I don't know as sine tones aren't a pretty good simulation of music notes, at least for instruments with little percussion or other onset ticks, etc. Long ago, sine tones were dismissed because they were too "contaminated" by room acoustics. That's true. But are brief snippets drawn from a pink distribution any closer to music-in-rooms?
So maybe we need some kind of test stimulus like "kind of short sine waves but not too short" or maybe a kind of smart-gating?
I bet if somebody could whip-together test tones that worked like music-in-rooms, the gap between mic test results and listener results would narrow.
Footnote: Toole, in a moment of weakness, describes the ideal testing for speakers. It consists of documenting all possible sine wave emanations from a speaker in every possible angle. (I say "weakness" because that is a kind of cowardly retreat to a "why not just measure EVERYTHING point of view". And not very clever.) From that, you can derive or intuit what the results will be like because, obviously, you (think you) know everything the speaker does. I doubt it, even in theory, even believing in Fourier transforms, because there are other qualities not entirely subsumed by sine wave outputs.
An excellent test stimulus for evaluating the sound in a music room is music. Too bad "human meters" seem to be essential in the measurement process that uses music because they are more variable than even my mic measurements.
While we have all kinds of time-course and gated measurements. And a variety of test tones and tone-samples, largely identified, like tone controls on pre-amps, for their convenience to manage in electrical engineering at the constructive end or math simplicity at the analysis end.
But the real question is which kind of test stimulus best captures music-in-rooms?
I don't know as sine tones aren't a pretty good simulation of music notes, at least for instruments with little percussion or other onset ticks, etc. Long ago, sine tones were dismissed because they were too "contaminated" by room acoustics. That's true. But are brief snippets drawn from a pink distribution any closer to music-in-rooms?
So maybe we need some kind of test stimulus like "kind of short sine waves but not too short" or maybe a kind of smart-gating?
I bet if somebody could whip-together test tones that worked like music-in-rooms, the gap between mic test results and listener results would narrow.
Footnote: Toole, in a moment of weakness, describes the ideal testing for speakers. It consists of documenting all possible sine wave emanations from a speaker in every possible angle. (I say "weakness" because that is a kind of cowardly retreat to a "why not just measure EVERYTHING point of view". And not very clever.) From that, you can derive or intuit what the results will be like because, obviously, you (think you) know everything the speaker does. I doubt it, even in theory, even believing in Fourier transforms, because there are other qualities not entirely subsumed by sine wave outputs.
Last edited:
I think the reason is that the perception of the loudness of tones of music or complex sounds is not based wholly on simple tone loudness. Simple as that.
Is there anybody here so ill-informed that they believe the "green color" a piece of green paper will look the same in all settings?
Is the analogy clear... even if none of us can explain what cues lead to the perception of loudness?
Last edited:
I keep coming back to this as your understanding/argument is not what SL wrote. Here's another one among many, quote:
Linkwitz-Links
Experiences with equalizing an initially flat on-axis response to obtain the realistic sounding ORION-3.2 have convinced me that 'flat' is not correct for a stereo system. Here we must create believable phantom sources from two real sources at +/-30 degree angle to the listener, at the 'sweet spot', in the room. Nor is the BBC dip the answer. Instead, the response must be flat above 100 Hz and then gradually drop to a lower level at high frequencies
As JohnK mentioned, the xo changes was to make it flat... then do the shelving.
🙂
Hello Ben
You are talking about a technique JBL uses called "spinarama". It gives you much more information than any simple on axis measurements gives.
How is this a cowardly retreat??
Have you ever seen a measurement set for a speaker measured that way??
Have you ever heard a speaker developed using that measurement technique??
Rob🙂
And here is a rare occurance of the DIN HiFi standard actual parameters - in this case the upper limit of the permited distortion in a loudspeaker...
Look familiar? 😉
A measure to reduce distortion and noise to some accepted level?
1. A 20+ yar old Sony 2x something, a DVD-SACD player Pioneer 6xx something and a Yamaha 993
- Will investigate the exact issues of the CD's - in theory they are original and ordered directly from the publisher, but in practice they might be re issues using unknown master source. - will investigate thoughly as it gets more and more interesting for me.
2. I used "d Relative" for a curve that modifies the energy distribution (balance) across the audible frequencies without discrimination for the actual material ecorded and without relation with the initial energy distribution.
- The resulting energy distribution is just a sum or a product or etc. - a universal and constant correction curve.
Attachments
Nope, incorrect.
It doesn't matter much at this point since you're going to believe what you want to believe. I'm tired of trying to explain it. 🙂
Anyways, regardless of this specific example, there's some good general discussion that has evolved in this thread. (It's one of the few that I have subscribed to.)
Cheers,
Dave.
- Status
- Not open for further replies.