And it's Gish gallop time.
Every wondered why no one of the golden ears has ever claimed that they can tell if an ortofon or westrex or xyz cutting head was used on the vinyl. Or that they could hear the clear signature of a particular ADC brand used.
Makes me wonder if anyone could tell soundstream from PCM-1 from the latest and gratest with just a music fragment unless they knew the year of release.
Every wondered why no one of the golden ears has ever claimed that they can tell if an ortofon or westrex or xyz cutting head was used on the vinyl. Or that they could hear the clear signature of a particular ADC brand used.
Makes me wonder if anyone could tell soundstream from PCM-1 from the latest and gratest with just a music fragment unless they knew the year of release.
Maybe. But the point of the book is to distill all of that down to which topologies are the most linear -- inherently linear before FB is applied.
THx-RNMarsh
Last edited:
Hi mmerrill99,
-Chris
That proves you haven't read anything I ever posted. Anything but.
Maybe, but still true. I was in a rush when I dashed that down, so some rewording would probably have turned out better.
Once, when I first got my speakers and before I reworked the same components I have now. It was a riot with a Shure SM-57 and an open reel running at 15 ips that I had calibrated, 456 tape, fresh. Haven't done that since.
Nope. I'm running the same gear and speakers. I did change the TT and cartridge. The change was strongly positive. I think the cartridge made all the difference going from an Ortofon VMS-30E MKII to an Ortofon 540 MKII (new style MM cartridge as opposed to VMS). I moved to a Thorens TD-126 MKII from a TD-125 MKII (which I still have with the old cartridge).
It can get silly loud. And a girl with a guitar does cut it. However, there is no way to capture the full dynamic range of a live event, just in the same way I wouldn't ever want to experience same in my living room. I want to enjoy the music, not be overwhelmed by it.
I've been in enough sessions to have a pretty fair idea. I have some cuts that I was there for. That would seem to me to be an unfair advantage, but there you go. Have you been present for any professional studio work? I spent a lot of time in The MetalWorks here in Canada. Not an unknown studio.
-Chris
Actually I have identified particular gear used in some recordings by their unique sound. It was gear I had worked with so that I learned what some would call it's signature.
Well, I understand. You do have to read between the lines a bit. Some items get awards, others do not. And then you have to look for the more unique statements, or them puchasing it after. They don't like to give real bad press, no mystery there.
Where are the measurements? I haven't seen them posted. But the CH boys don't seem to be the type to not deliver without disclosure.
As "tone" is one of the easiest things to hear it could easily swamp any other change.Destroyer OS said:
Democracy works, to a certain extent, in politics. Market forces work, to a certain extent, in commerce. Neither are good criteria for judging truth in science.
You jump to conclusions. Perhaps you need to read a good book about feedback and servo systems?
You appear to be saying that the only people who don't know how to design good audio electronics are people who know about electronics. That would be daft, so perhaps I have misunderstood you?
Most low distortion 'non feedback' circuits have their feedback removed by the simple act of redefining the word 'feedback' e.g. redefining degeneration as 'not feedback'. I grant you that there are a few precision circuits not relying on feedback, but these are mostly small signal.RNMarsh said:
You seem bent on putting words in my mouth.mmerrill99 said:
In the 1980s I could attend a Prom concert and then the following evening hear the live relay of that evenings' concert. Different event, same hall. The hall sounded the same. To me that implies that the recording technique used was aimed at capturing the event rather than processing it. Limitations of technology may mean that some processing was needed in order to capture it, but the processing was as minimal as possible. This is quite different from a recording engineer/producer in a studio seeking to 'produce' an album for release.
I don't want my system sprinkling distortions over the sound. If I did, I would implement the Wavebourn 'niceness' control.
You are trying to expose perceived contradictions in my position. That is fair enough as a debating technique, but may be a poor way of arriving at understanding or truth.
Dither is used to reduce distortion (not 'correlated noise'), by replacing it with noise. That is the sort of processing which is aimed at improving sound reproduction by specifically targeting a known and quantifiable problem with a precise solution. Nice try, but irrelevant analogy.
No, I don't define realism by measurement. I define it by careful listening tests. What technical criteria result in realism? Having found these, measurements can be used to check that they are being achieved. This has already happened to a large extent so we know in broad terms what is needed in terms of bandwidth, distortion etc. for realism. Future work will simply refine these; that is, we don't yet know everything but we do know something and we probably know most of what is needed. My argument is mainly with those who adopt one of two views:
1. We don't know anything about realism so all the existing criteria can be discarded (e.g. 10% peak distortion and 10kHz rolloff after a 3kHz bump sounds fine to me so that is hi-fi, following the criteria leads to boring clinical sound)
2. The existing criteria are so hopelessly weak that they can be ignored (e.g. we actually need 0.00001% distortion and 3Hz-120kHz bandwidth)
The first one is happy with much less. The second wants much more. They can't both be right, although sometimes people try to keep a foot in both camps ( e.g. by insisting on Brand X resistors (distortion/noise -120dB) in their single-ended feedbackless amp (distortion/noise -30dB) ).
Iirc we have discussed this in other threads already broadly, so i´ll try to keep it short.
In reality we don´t know what approach the producers of the recorded content followed, we only have some recoreded content.
It should be obvious that we had to listen to this recorded content in exactly the same way the producers did if we really want to know what the sound was.
Even then we had to listen in the exact production monitor room and in the room where the final mastering took place (which might be completely different).
As stated in the other threads there are two main approaches and a lot of grades in between; the first one´s goal is that a listener at the reproduction place will have a listening impression most similar (as possible) to the listening impression he would have had if attending the original event.
The other approach tries to enhance the listening experience because the missing visible element of the perfomance has to be taken into account. (so to speak enhancement to make it "bigger than life" )
Even if we could listen under these conditions to the recorded content, it is not ensured that we receive it as intended because the interindividual differences might be so, that what has worked for the recording engineers does not work for us as listeners.
We have to remember that the reproduced thing is a highly distorted version of the "real thing" and the individual reaction to the same cues can be quite ....different.
Although i agree - for obvious reasons 🙂 ) - that all we can do is to replay the recorded content, i´d say we usually don´t know what intention the engineer/producer had, unless they or somebody else tells us.
That´s easy to state, but imo in reality there is no approach widely accepted as the _only_ one. Apparently a lot of records are produced where the intention of delivering high sound quality didn´t get the highest priority. See for example the loudness war.
In the imo quite smaller segment of records where the delivery of the highest sound quality is the intention there still exists a wide range of different recording styles although the goal might be the same.
We know already from experiments that a mix of the same "raw recorded" material in different production environments (by the same person) leads to different results.
We know additionally that listening to the same records (by the same person) in different reproduction environments leads to different impressions.
And of course we know that listening to the same recorded content under the same reproduction conditions by different people can lead to quite different quality ratings.
From Mäkivirta/Anets study we know that measurements of the same loudspeakers in different monitor rooms all over the world show quite large variability wrt to linearity and reverberation time.
Toole coined the term "circle of confusion" for this situation - quality assessment of reproduction by listening to recorded materialand quality assessment of recording by listening through reproduction systems.
One solution might be standardized environments at production and reproduction places, leaving aside for the moment individual differences and
all the thousands of records done before.
Last edited:
Well now. I have read the intro and the table of contents and the bio of Dr. Arto of Tampere Finland who is CEO of: Icraft
I also read the forward by Nelson Pass and noted that Jan Didden is the publisher. And references to Bob Cordell and Doug Self.
It's a small world, isn't it?
I assume you bought this book...in any case, what did you find novel and important about it?
Arto's objectives:
Attachments
Last edited:
It might be better to replace "highly distorted" with 'somewhat modified'. Two reasons for this:Jakob2 said:
1. in the context of audio systems 'distorted' has a particular meaning which is presumably not what you intend here;
2. if the reproduction sounds anything like the original then it cannot be "highly" distorted - yes it has lost much spatial information, and has suffered some bandwidth reduction, but the acoustic signal produced by the speaker is reasonably well-correlated with the acoustic signal received by the microphone.
'Loudness war' and similar recordings have nothing to do with hi-fi sound reproduction.
The discussion seems to be an attempt by various people to justify the point of view which says something like:
unspecified amounts of nonlinear distortion and frequency response modification in the domestic system can be viewed as partly compensating for unspecified and varying amounts of loss of spatial information, loss of dynamic range, loss of bandwidth, differing reverberation times, differing transducer spatial patterns, differing transducer distortion etc.
Let us assume for the moment that they are right. Does that mean that 2% peak distortion (say) can compensate for the recording engineer putting a mike in the wrong place on one recording, while on another recording it can compensate for the studio being too small for the type of music, while on a third piece it can compensate for the conductor insisting on some treble lift against the advice of the recording engineer? Surely at the very least we need adjustable distortion, like tone controls did for frequency response?
Yes, I respect both your methods and attitude.
The problematic attitude I found mostly come from junior amplifier designers I met who lived in their very comfortable and nicely defined little electronic cubbyhole. They show serious lack of understanding to effects coming from the world outside. I hope situation will improve when members of diyAudio show more concern to the performance of the Whole Chain, make public more examples, remarks on the importance and effects of measured quantities and good setups to effect reliable measurements. 🙂
Interesting points. I know how to measure bandwith. Has measurement (even partial or only indicative) of spatial information been established yet?
Last edited:
If I was Scottjoplin, I would ask you is English your first language but I won't go there - I would just simply direct you to the question mark at the end of my sentence - it signifies a question. I restated what I thought was your position, based on the words you posted, & simply questioned if this was the correct restatement - it's a simple Q & A mechanism that is often used to tease out one another's position - not a case of trying to put words in your mouth.
"live relay" meaning radio broadcast or what exactly? But the mic configuration alone is a modification/processing of the acoustic event, irrespective of any subsequent processing.
Again, I'm not putting words in your mouth here, simply trying to understand what you are saying - are you saying you can reliably differentiate sonic differences of two acoustic events separated by a day in time?
No, I'm questioning myself as much as you by following what I see as logic - by playing devil's advocate, I'm looking to find the flaw in my logic & a better understanding of what might be true.
I'm equally divided between believing that reproduction accuracy will deliver realistic perceived soundscapes Vs that some added playback distortion (compensating for what's missing because of limitations of 2 channel stereo recording) will deliver more realistic perceived soundscapes.
I'm hoping that through discussion I can find the way to resolve this dilemma
Dither is the technique to reduce quantization noise which is a noise correlated to signal -which can often result in the perception of large scale patterns such as color banding in images.
OK, I can understand your position, I think?
I don't hold either of those two views - I believe we have progressed a long way & reached a good level of reproduction through listening aided by measurements - pretty much what you state is your approach. But I find that to progress beyond this point requires us to reconsider this approach & try to better understand how our auditory perception evaluates sound & use this knowledge to better focus our playback systems to better match the evaluation criteria that auditory perception uses.
A difficult ask, I know, given that we our whole recording/playback system is inherently limited & what is being created is an audio illusion from 2 channel reproduction. Perhaps we have reached the limits of what is achievable in this configuration but I don't believe we have.
That's why I wonder if some added distortion could actually be perceived as more realistic by auditory perception & I've no problem if that turns out to be the case as the goal, IMO, is to enjoy a more realistic illusion, however that is achieved. But we would need to understand all the parameters & how to apply any such added distortion.
Anyway, it's a thought experiment, not my firm, immutable position, as I said
I don't believe any standardization can realistically happen, just as I don't believe that doing recordings in your playback room is a practically feasible solution.
What I have settled on as a solution is dropping the psychological need for chasing such absolutes & living with the possibility that we can only judge the playback illusion by how realistic it is perceived. This is a relative evaluation without absolutes - A might sound like the best, most realistic sound we have heard until B is heard & so on. This requires auditioning with a large variety of music (as I said, girl with guitar, isn't sufficient) over an extended period of time & one of the reasons I find quick A/B differential listening tests, limited by this very fact.
I agree, "highly distorted" is not how I would categorize it - missing signal data, as you state, is more accurate
As I said, I'm open to the possibility that auditory perception may be better served in its task by giving it an illusion better matching its criteria that it uses to analyze (make sense of) sounds & soundscapes. Discovering these criteria is what I'm open to & how it might modify our playback systems.
I agree, it would probably not be a simple adding of global distortion - auditory perception is highly context sensitive
I see your point, i´ve used the term "very lossy" version of reality before, that might be more suitable wrt to the point of perception; nevertheless i think "distorted" fits by comparing the physical difference between original and recorded soundfields, leaving aside the even more "distorted" recording that will be evantually available at the end of all processing.
Having said in our earlier discussion that my definition of the goal is very much like the first one that i described in my former post (wording according to Wolfgang Hoeg, although translated by me so put the blame on me) i totally agree, but was responding to anatech´s post.
He is following another reproduction approach - if i am not mistaken at this point - where any relation to the original acoustical event is irrelevant as the term "Hi-Fi" is the utmost precise reproduction of the recorded content. So a perfect reproduction of a horrible example of the loudness war is in that sense "Hi-Fi" at its best.
I don´t understand your approach at this specific point. You relate your "Hi-Fi" definition and the case of distorted or just modified reality to the perception (see the first paragraph), but now seems refusing to accept the perception in favour of only concentrating on single technical distortion parameters/figures.
If we are talking about perception we have imo to realize that people are listening to systems (although they often try to draw conclusions about single elements of the systems).
As long as we don´t know what characteristics of the specific system evoke the preference for this specific reproduction all talk about distortion figures is just jumping to conclusions.
If your are talking about distortion figures, maybe it is not the most important variable in _that_ system, it might be something else and it even might be that this specific listener (who is part of the system) would like less distortion more, if just the other more important parameter would remain unchanged.
I still think that the -discussed above- point of distortion/loss compared to reality and the very individually different reaction to this difference is an underestimated variable in the game.
A good example is the, i´m sure i´ve mentioned it before, quite diverging reaction of listeners to big electrostats. If by listening under the same conditions to the same music excerpts the speakers is ranked as the best (not only in the competition but overall in their life) by some while others ranked it as the worst that should tell us something. Even more, if the same group of listener judges quite consistent about other speakers under the same conditions.
If we want to jump to conclusions, we could take the very low distortion figure of that said electrostat and conclude that all other listeners not preferring it obviously just like "high" distortion in their speakers, but i don´t think that would be true......
Last edited:
- Status
- Not open for further replies.