How do you know what an accurate reproduction of a recording is? Using monitors someone has given the label of "accurate" to can be part of the prcoess - but no guarantee! My personal experience of monitors is that, as a group, they are pretty dreadful as a means of hearing what's in a recording ...
There is at least one other option - neutral, revealing and extremely enjoyable, all in one package. This is difficult to achieve, and that's why it's rarely heard - but that doesn't mean it doesn't exist. If one wants to pursue it rigorously, there is only one presentation, that which has been captured in the recording - irrespective of what any of the components I've used are, including the speakers, what has always happened, over the years, is that as the setup is optimised the sound of a particular album steadily gravitates towards a single point - the intrinsic sound of the recording itself. It's when I put a familiar recording on some curious, usually expensive setup - and a weird contortion of the music emerges, I can only scratch my head ...
Last edited:
Accur-it?
I live a block and a half from a retro style coffee bar; the owner provides a venue for musicians to perform there. Basically for whatever's in a hat passed around. What I hear there is what I attempt to have every pair of speakers I make emulate .
So far I haven't gotten it close with regular styled box speakers.
The "monitor" speakers I mention are my twist on an old M&K idea , and will be the first I've attempted with anything resembling measurements, albeit only with a sound meter .
I am by no stretch an engineer, and of course couldn't know what was on original recorded performances.
The only analysis I can do is with a few recordings I know to have been simply recorded with no or very few overdubs, like Rita Chiarelli, some Audioquest blues selections, Jazz at the Pawnshop, etc.
Mostly I reference to live sound.
I live a block and a half from a retro style coffee bar; the owner provides a venue for musicians to perform there. Basically for whatever's in a hat passed around. What I hear there is what I attempt to have every pair of speakers I make emulate .
So far I haven't gotten it close with regular styled box speakers.
The "monitor" speakers I mention are my twist on an old M&K idea , and will be the first I've attempted with anything resembling measurements, albeit only with a sound meter .
I am by no stretch an engineer, and of course couldn't know what was on original recorded performances.
The only analysis I can do is with a few recordings I know to have been simply recorded with no or very few overdubs, like Rita Chiarelli, some Audioquest blues selections, Jazz at the Pawnshop, etc.
Mostly I reference to live sound.
So when you tune a system to sound the way you like, that is always the intrinsic sound of the recording? How do you know this? Were you present when the recording was made? Were you in the studio with the musicians, or at the console during the final mix?
System 'tuning' usually does not directly involve frequency response. At least it is not usually possible to measure any significant difference over a large bandwidth up to the speaker terminals.
People should listen to LIVE recitals or performances of the music they like, and get into your memory what good musical instruments sound like. Then you can compare your system to others or to a different configuration of your system. I never deliberately modify the frequency response. I don't know how I could with the designs I use. YET, my designs sound somewhat different after listening over a period of time with each. Those of you who can't hear the difference should stay with what you have.
People should listen to LIVE recitals or performances of the music they like, and get into your memory what good musical instruments sound like. Then you can compare your system to others or to a different configuration of your system. I never deliberately modify the frequency response. I don't know how I could with the designs I use. YET, my designs sound somewhat different after listening over a period of time with each. Those of you who can't hear the difference should stay with what you have.
Here we go again 🙄. So your subjective listening is way better than measurements? Hooray for Frank's golden ears 😱. I say you are fooling yourself but hey if you want to continue down the path of make believe then do so.
Meaningful results for yourself but no one else. Where's the science in that? It's years and years of subjective sighted bias with zero controls to see if anything actually changed or not 🙄
After what Frank, 20 or 30 something years of doing this you still can't show anyone any specifics to a particular circuit? 😕
Okay Frank here ya gofiddling.
FAR better to be meaningful for one than for a [probably inaccurate] test bench. What has to get the accolade is that which sounds and emotes closest to the real sound………and don't forget - much music is recorded in studios [or even garages] which sound nothing whatsoever like a normal live performance OR an average record listener's room. So, on that basis, where does measurement come into it? I suspect that you have little experience of live music performance or, come to think of it, in music as other than a complex test signal.😉
Regarding specific circuitry I haven't seen much from you either. Why don't you post one or two showing a bit of kit which measures well AND sounds convincing. [Posting any of JC's designs would be shooting yourself in the foot! 😉]
[I cannot help but think that I am feeding a Troll 😱]
If on axis measurements are about the same , but sound different, it must be your speakers differing dispersion reacting with your room, no?
And even though most of my golden eared friends say live music is bright, I too have to agree it's the best reference.
Part of the difficulty is my ears seem to have a short memory for absolute sounds, and quickly seem to hear "around" a systems flaws, if I'm not careful.
AGREE 1000%. I would add at the end of the coloured part a condition being that the instruments are being well played.
Because I am not "tuning" it! What I am doing is locating weaknesses in the particular implementation, and resolving them, "hardening" the system against interference effects. Most systems are riddled with little weaknesses, all over the place, which add up to give that system its particular 'colour'; so the attitude I bring to the party is to find and eliminate each one of those gremlins, one by one. And as you do so, automatically the system sound gravitates to the "true" nature of the recording, the raw sound of what went into the final mix, irrespective of what the sound engineer thought he was capturing.
Another way of thinking of it is that I'm troubleshooting: I listen, the sound is not right, reproduction distortion artifacts are clearly obvious; so something is faulty, somewhere - therefore, one's job is to sort out the issue. And if your reaction is, "How do you know that the distortion is not part of the recording?" it's because I've done this over and over again, dozens, almost hundreds of times - and I have yet to find a recording, no matter how 'impossible' it sounds, that can't be 'recovered' by fixing up the replay of it.
A simple example: a recording, where one section sounds messy, with peculiar vocals on it - as the system improves that segment resolves into multiple parts, several sound elements which are completely cleanly recorded, and the vocals, which have been put through a studio distortion effects unit to give them a flavour. The mulched vocals now stand out as a completely separate sound element from the undistorted other instruments - you are literally able to hear, separate each microphone and input plug.
Because the sound will be different, in some way - it has to be, because one source is a live performance - if you ask the musicians to repeat a section, it will be different from the version they played some minutes earlier. So what you're trying to compare is the overall 'flavour', the quality of the experience of what you're hearing.
Yes, it is easy to start deceiving yourself, if you choose the wrong sort of material to use for evaluating the current status of the sound. The solution, for me, is to use 'dodgy' recordings - ones that are on the borderline of being tolerable to listen to, these will instantly give strong feedback as to where things are at - your hearing will give a thumbs up or down before your mind has a chance to adjust, and you've got your answer.
Yes, live music is bright, which means reproduction needs to be bright, to be convincing! The flaw I hear in many systems is that the sound is deadened, because the "brightness" they would otherwise have highlights distortion artifacts - the sound is impossible to listen to, long term. But the easy fix, of killing the treble intensity, is a wrong move - it's equivalent to dropping a blanket over the tweeter; the productive solution is to find out why the brightness is unpleasant, and resolve the underlying issue ...
Dead right! fas42. TUNING in our case is like tuning an engine or a race car. The design starts a bit rough and it has to be optimized to do its best. It has NOTHING to do with changes in frequency response or adding distortion.
I just found my reference "indestructable" recording, the original Polydor CD of "Machine Gun" from the "Band of Gypsies". Before the manipulations at the hands of the Hendrix Family LLC Trust and their "engineers".
You could probably record this onto a vintage WWII magnetophon and broadcast it on AM radio and it would still sound fine.
A detailed case study would probably help here.
Again, reference to the details of the specific example would be helpful. This reads as hand-waving atm,
What was the recording the changes were based on?
What was the identified sound signature that presented as a problem?
What changes did you make to the system and what occurred after each change?
When did you stop making changes, and why?
Once the changes were made to optimise that recording, were they transferable to other recordings? Other sources via the same amp->speaker chain?
I'm not so impressed by the ability to hear down to the level of "each microphone and input plug" given the number of mixes that most recordings go through these days that is kinda irrelevant even if it were demonstrable.
Whatever, glad you and Frank are good pals you deserve each other . Yeah, right you know all about my live music experience 🙄
Frank is the one talking all about his years of circuit fiddling yet can't show us any examples at all.
🙄 Fiddle away. By the way I blocked Frank, his opinions have little merit in the real world.
If you wanted to find a troll then Frank is your man.
Subjective blather without any objective controls are useless. Might as well toss a dart at a wall with "tweaks" posted on it and live that way.
"Indestructable" meaning that the sound is so bad that it can't be damaged any further? Don't have the album, just looked it up on YouTube, there are some versions which are excellent, there would be no trouble in generating a very impressive playback - would there be one there that matched what you hear on the Polydor?
The cymbals on that are an excellent measuring stick, recovering the correct sound from them would be a first step, and then maintaining that quality up to full strength volume as the next part ...
I've already described in reasonable detail what was done to the PC speakers in a posting, I can elaborate a bit more if you wish.
I was thinking specifically here of a classic 80's New Wave, heavily processed female vocals track, can't recall the name and performers at the moment - I was using this item as a 'workpiece' many, many years ago.
The sound signature of the problem is that the strands of sound that make up a difficult part of a track are not clearly perceivable and separable - the goal is that, irrespective of how 'busy' the sound becomes, that the individual components that make up the sound totality remain differentiated at all times.
I've mentioned what I address numerous times, no point in repeating - each case is different, experience guides me on the best areas to tackle first. What results, if there was a significant problem there, is that the sound 'cleans up' somewhat, the level of 'messiness' decreases, a passage that was 'impossible' to listen to starts to show signs of life, the content within becomes clearer.
In essence, there are almost no limits to how far one can push the optimising; a general goal is to have the speakers become subjectively completely invisible for all recordings, but it can be taken further. Time, and effort required are the constraints ...
If the optimising was done correctly then the benefits will also occur for all other recordings. The process is usually to take a halfway house difficult recording, bring that up a better level of replay; then go to a much more difficult recording and see where that stands. Sometimes another recording can sound worse in various ways than earlier, because more detail is being revealed; here experience guides one as to whether one made a mistake earlier, or just that a further round of optimising is required.
I always treat a system as a totality - just changing the source is equivalent to dealing with a completely different system.
That's just a way of expressing that the contribution of each sound element can be clearly discerned, if the listener chooses to focus on that.
A far better item, for me, would be something like this, http://http://www.diyaudio.com/forums/music/6958-now-playing-what-you-listening-472.html#post3917196. This defeats my PC setup in its current state, makes the Hendrix track sound like an audiophile recording. Engineered from the core out to be 'dirty', I had to drop the volume to a quarter of the setting used for Machine Gun, otherwise the system overloaded. Every element of the recording is sludgy, trying to be old-fashioned, and it's compressed to the max as well - when high level distortion is deliberately mastered in, it's much, much harder to extricate the good stuff ...
- Status
- Not open for further replies.