Prepare for Christmas

[PeterSt]

Well, maybe not exactly, but I thought it might be "fun" for (much) later reference to see what was right and wrong about some findings of me.

Already last year's Christmas I was so much surprised that the, say, really old recordings could give so much joy for sound quality. Think about the time recordings started to be in stereo for standard ... which was after the beginnings of The Beatles and the period so many Christmas songs were recorded.

Without now playing such recordings and without reading further, try to imagine why these recordings are perceived as bad and old recordings ...
...
Do you have a clue ?

Ok, because recording techniques weren't on par with todays means, you can say ...
Wrong wrong wrong.

I dare to say : because the recordings techniques from back then implied a "resolution", or a possibly other phenomenon we don't have wordings for yet, were so that they could not be played back 1:1.
And back to the clue I asked about above, this would be my answer from today :

Because of badly represented highs. It sounds raw, roughly smeared, and generally expressed : as bad old recordings.

For a normal human being this leads to nothing else than bad old recordings, but since the earlier versions of Engine#3 showed me at some stage that there's much more in the old cr*p than the old cr*p implies from it's description (old cr*p), it started to intrigue me. Possibly the main reason it started to intrigue is I could not get that back at later Engine#3 versions ...

I think it is one month or so ago, that I started to hear (per the then XXHighEnd version) that something in the highs was changed in such a manner, that highs could be added. Or better : *had* to be added in order to fill up gaps in resolution. Try to imagine : highs started sounding grainy (which might be the opposite of smeared), and the grainy to my perception was caused by missing higher frequency waves. So gaps, where there should be sound. Difficult of course because this is at the micro level, but anyway I perceived it like that.

Well, the logical solution to filling the perceive gaps was easy : crank up the higher frequencies for dB level; if something is there it should come out ...
Note : So this is about mangling with the frequency response curve of the loudspeaker, of which we all say it must be as flat as a pancake.

Not to forget : we are talking about perceived bad old recordings, which at least to my ears could be made concrete by the expression "raw etc. highs".

Ok. No matter how illegal or plain stupid it might be to start such an adventure, I let the frequency curve rise from off 5000Hz so that at 20000Hz there was a 16dB rise ... From 5000 to 20000 quite linearly. Mind you, this implies that the 20000 output is over 5 times as loud as the 5000 output ...

I must honestly say that apart from trying to get away the perceived grainyness, I was attempting to let cymbals and hi-hats output at a level my own drum set creates, keeping the general SPL coming from that in mind, compared with other instruments playing at their normal level with the toms from a drum set as the first reference (those not expressing highs as implied here).

I can tell you ... it worked.
Now an idiot drummer smashing around on the cymbals was perceived as an idiot smashing around on the cymbals. And mind you, listening to the normal frequency curve, this is perceived as -and as you all know- as some cymbals in the back, at far less SPL than reality tells you.
Is that in the recordings ?

Now a few things must be looked at together, and they are not easy to come with a decent explanation, let alone a scientific one :

The increased highs indeed filled up the grainyness.
That by itself *allowed* for increasing the highs in general. I mean, where highs are not good, they should not be over-expressed, right ? In this case they are technically over-expressed, but are perceived as just natural.
There is *no* layer added to voices, as you might expect when highs are well over expressed. This by itself shows that there's just no layer there. But also : the over-expression of highs allowed to happen only just *when* this layer is not there in the base. Appearently XXHighEnd now allows for it ... (and this was not the case in older (one month ago) versions).

To the above list I could add that I cannot imagine that "recording techniques" deliberately squeeze away higher frequencies, although everything is possible.
But now back to the old recordings ...

The first thing old recordings express (normal frequency curve, normal (CD)player) is too much of bad highs.
Well, with the experience now of one month at listening at the uplevelled highs it started to occur to me that old recordings express even more highs. They allow "more fluently" to smash the hi-hat in your face without perceiving it as boosted highs. But the highs are still boosted.
I did not go back to the normal frequency curve to test how the old recordings sound without the high boost, but I would not even like to go back;

Currently I am searching for all the old recordings I have because they sound sooo good and so beautiful. So much more realistic.
The only thing I can think of is that indeed somewhat after the day of poor old recordings, recording engineers started to decrease the level of highs. Mind you, it all fits. With normal playback means, the highs of old recordings sound worse the louder they play. So they really could have done that. But with the "newer" recordings playback of today (well, via XX) you start to hear the gaps. I do. Fill them up again, and all is back to normal.
Play back the old recordings at that same boosted highs level ... to me this doesn't give problems. It is really perceived as a somewhat harder hit at the cymbal.

All is *not* perceived as uneven highs. No hissing, no colouring going away, no layers.

As often, call me crazy ...

Peter


PS: For those who can tweak their frequency curve as indicated, try it ...

PPS: I have no real indication that all should start at 5000Hz. Important or not, no instrument has its fundamental frequency above that (except for maybe a synthesizer and pipe organs).

PPPS: Do note that I apply this at the passive side of the loudspeaker filter, hence it very well can be so that if you pump this kind of crazily boosted highs through your amps things might go wrong *there*.

[soundcheck]

Hi Peter.

Are you findings and conclusions based on listening tests with the AER MD3?

I am just curious.

\Klaus

[BertD]

Peter,

Did you ever thought about the quality of the recordings made in those days?

Microphones with limited high frequency response, old tapes limiting the same, etc. would already indicate a strong roll-off in the high frequencies and THAT is part of the reason why they sound "old"...

I surely believe that such recordings need an EQ (RIAA but then only for high frequencies) but to optimise a speaker for old recordings hardly being a reference seems odd to me...

Bert

[PeterSt]

Klaus, no ... just the very capable Orphean "BMS" ...

Bert, of course I have expected a remark like yours. But although my writings are confusing as always, please read again;
There's contradictions all over, I know. The old recodings contain MORE highs, and by boosting the higher frequencies they turn into GOOD highs.

Btw, nobody told me microphones from those days could not reach frequency levels this is about (which is not about 20K but much much lower). It is about nothing more and nothing less than the grainyness I perceive from ANY recording, but since the old ones have more of it (SPL of high frequencies), they just express worse without filling those gaps.

But as said, my post is about later reference and the (honestly) sense and nonsense of it.
OTOH you could trust me just a little bit on my hearings. I mean, so far I will never let go of this high frequency boost. You (Bert) know what does and what does not hurt my ears. This just does not BECAUSE of the boost.
Isn't that strange ? And I'm using it for a month now ...

[BertD]

Peter,

All seems "weird" to me....perhaps we should check out each others system some day soon so that I know better to understand what you are "doing" and then discuss a bit further on those experiences?

You've lost me at this moment...

Bert

[PeterSt]

Yes Bert, I know. There is *no* logic to this, except then for what I said.
I could be very much wrong, but since I cope with this for such a rather long time ...

Peter

[SeVeReD]

My acouton (sic) super tweeters lie dormant waiting ....

[Gerner]

Christmas will be the right moment Peter for you to place the XX 1.0 under the Christmas tree, and for us to unpack it..... 

Interesting post of yours and also interesting to read about your observations re. the graniness disappears if the highs are cranked up!

I have not had such experiences with any of my numerous speaker constructions and I did not verify your obsrvations om my present speakers either. A thing I would have done time ago if I have had easy acces to the filter. Which I have not.

It's funny, but when you tell us, why shouldn't we beleive your hear it as described? As far as I know your favourit wife agrees. 

I was wondering if you could acheive the same results your place cranking up the treble by means of a graphich equalizer even I know that would not be exactly the same as doing it your way.

But having this in mind that our ear-sensitivity-curve is quite unlinear (remember the good old inegrated amps with a loudness button) actually drag things even more in the way you do it, at the sound levels I know you are playing with, you, relatively understood, expose your brain to an even more exagerated level!!!!!

My conclusion is that we all of us have one non-reference in common: our own ears and brains. They do not perceive things the same way even we agree on so many qualties.

I was just thinking about the ear-frequency curve we are born with and the one we die with. No comparation. But still we can hear and enjoy no matter this fact.


Now being a former drummer myself, I cannot recall I heard any higher frequencies beating those cymbals to Hell. But, my God, sitting 50 cm's from them made me deaf. Bwiiiiing...

For now I have defined graininess rather in the cathegory of *low resolution*. And I am not sure wether old recordings contains lower resolution than newer ones. On the contrary I think. (At least I have some albums of nowadays proving that...Pfew)
If you say we can "create" higher resolution by means of cranking up in the highs, I obey, but don't understand. 

Gerner

[BertD]

Yes Bert, I know. There is *no* logic to this, except then for what I said.
I could be very much wrong, but since I cope with this for such a rather long time ...

Perhaps more on topic but does CD Emphasis/De-emphasis has to do anything with this? My transport has this automatic switch on board, my PC hasn't...

Or something else might be wrong in your system or listening environment compared to most people. Perhaps the total room response is on the dull side as you have a huge and highly dampened listening room.

What works for you in a specific situation might work the opposite in other situations. More highs at my place does not work at all...things get sharp and overly expressed. This sound remembers me of some speakers created in Germany, very much the opposite as the English speakers from those days....

Bert

[soundcheck]

Hi Peter.

IMO - Not any fullrange driver in the world will be able to reproduce the highs the way
you're looking for. Low SPL tweeters and ribbons etc. will suffer the same lack of dynamics.
(That perhaps doesn't have anything to do with the grainy sound you're describing - who knows.)

From my experience it needs a powerful high SPL horn super tweeter to repoduce
the highs dynamically "realistic".

It's not about the frequency response, it's IMO about transient response/energy  (or dynamics)
in the upper range.

Adding some gain through the software to the upper range might lower the deficiencies  on speakers lacking these dynamics. ( Remember -- the Traktor-effect :D  )

Cheers
\Klaus

[PeterSt]

But Klaus ...

I do not have full range drivers ... I have the same as you've heard at Bert's place ... 

This is not about a too much damped room;
This is not about drivers actually being incapable of expressing the highs;
This would be, though, about some "Traktor" effect. So, I must admit that this is actually about "DSP-ing" ... 

Please note, for all who think I am solving some "problem", that without the boost the in-room frequency response is, say, flat. So for Bert : you know that it is so, and actually it is just by your design. No problem with that, and nothing wrong with that. And thus :
after the boost the response is just as I told : an upgoing rise of 16dB starting at 5K and ending at 20K. In-room, anechoic, whatever.

If somewhere else the highs start to sound harsh, it can only be my conclusion that at this somewhere else the "base" for it is wrong. If the base is right, the only thing which happens, is that the SPL of whatever it is, is boosted. Same like cranking up the volume. BUT :

Of course this is theoratically about frequency ranges being out of balance. If that sounds like being out of balance, don't do it. But here it just does not. Keep in mind why I started this : cymbals sound way OUT of balance if the (measured) frequency response is flat. And of course, since we're all used to that, we leave it be. Or, possibly you just don't recognize it. Or ... you recognize it, but kind of know that when highs are cranked up, things indeed become out of balance. Or harsh, because of whatever it is (including theories ) that makes it so.

For Bert : you actually know how this started ... by means of an XX version which suddenly did not show no sissing/hissing from the tweeter, while coincidently I was looking for a perceived out of order tweeter (which was not the case). Before, the sissing/hissing *was* coming from the tweeter, and so easily heard at disconnecting/connecting it. Conlcusion : it just can have it now, and the only thing what happens at boosting those highs, is that NORMAL highs get a higher SPL. This is very very different from plain WRONG highs getting boosted, resulting in all the color going away, and sissing indeed.

Now, since it is looking like that improvements opposed to reality get hard to realize (it becomes hard to recognize what's wrong now), there's a very apparent thing lacking : the SPL of high frequencies opposed to reality. Look in this topic : you actually all agree with that (ok Bert, I did not see it from you I think). When a tom is hit at a realistic SPL level, I expect a similar hit on a cymbal to hurt my ears in the manner they should. And no, this is all not at 50cm distance being the drummer.

What I did, with 5K as the lower boundary, is boosting just the necessary to create realistic SPL of (graduate) frequencies above it. But mind you, at 7K this is already 3 times more SPL (6dB). This is (apparently) so much no sh*it that you think my system or ears is/are wrong and the speakers are under thick blankets. Fine ... 
And Gerner ... you know that I reasoned it out BEFORE I applied it, and it happened to work out as how (and why) I reasoned it. So no explanations after the happenings, but, say, well thought ...  (I was looking for a fishy smiley here, but it isn't there hehe).

A lot is related to the perceived grainyness, which by now has become a too difficult subject, because this is not there anymore (per Gerner's post somewhere about "infinite resolution"). However, this does not say it is not there in the old recordings (which are imcomparable with today's recordings, including those from 30 years ago (!)), and where all actually started around this phenomenon, all is solved by the same means.

All being said ... I *do* emphasize on the out of space stupidity on the idea itself, but since I keep on listening to it ... I must be out of space, or just go there.

Deaf Peter

[Gerner]

Deaf Peter

Hahaha Peter.

I think I grabbed your point. To your ears the up-cranked highs just do the job for your perceiving the energy expressed by splashing symbals with the same energy as a Swartzenegger beat on a tom-tom drum.
For you (and maybe others as well) it does the trick listening to fidele music.
It is just looking freaky on a frequency plot.


As I said I beleive you when you say it's better, but sad I can't verify it. (No filter acces, U know).

Gerner

Edditing added:

Peter taking the risk to offend you  which is not my intension at all, U know. So forgive me if I'm wrong.

I know where you are often listening to your speakers. And that is 15 degrees off axis I beleive? And 8 meters from the speakers I beleive? And mostly with your head in the fridge to look for cold beers  I beleive?

Don't you then think that a close up meassurement should show quite an amplitude rise to cope with the treble loss in air under these conditions?

Gerner

[PeterSt]

Hi Gerner,

Although there is much truth in what you say, including the fridge thing , no, at judging things I'm listening right on-axis which is a necessity especially in this case (the more things could be harsh or unevenly loud, the better it is heard on-axis obviously.
Btw, how can I be offended when it is me stating that the whole idea is rather stupid ...

Talking about offending ...
For you and of course for Bert it may (or does) look like I am trying to solve a problem. A problem with my room ... a problem with the speakers, my ears, listening with my head in the fridge, whatever. But please read more carefully then. There is and was no problem.
I am just trying to "improve" further with say, completely illegal means. I know that, although I have some reasonings (not put here) which could make it legitimate. Could.

I think it is throughout my posts here that it is so as I described. Why do you judge it as solving a problem ?
May you not know it, I'm a guy of experiments. XX tells enough I hope. But also - and this is much more in the line of the subject here - look at my lamphood experiment with the Infinitiy's. If something is "illegal" it would be that. Infinity R90's being well respected speakers known for their VERY fresh highs, showing harshness immediately when something is wrong. So what did I do ? change the 180 degrees radiation of it and changed it into 20 degrees or something in that area. Now do the math on the dB increase ...

Bert, you were witness of that "stupidity". You also were there when we once removed them for fun. The well respected R90's became ... well ... R90's again. With the lamphoods it was a completely different speaker.
Did I try to solve a room problem ? (uhhm, theoretically yes -> less reflections) ... no. I just wanted more fresh highs (and again, do the math 180 degrees versus 20).

Today I act the same. Look back and read again. I said "it can have it".
"It" is everything. The speaker itself, and incurred by the player (since 0.9k or whatever) and the amplifiers.

Not to forget the reason why I started this thread : the generally perceived "too loud bad highs" in old recordings, which actually appear to be good highs with today's XXHighEnd and mainly : my "discovery" that highs must be squeezed in today's (but also 30 year old) recordings. Could be because of compression ... I don't know.

So remember, no problems, just experimenting, and what I hear I like.
And that's subjective ...

Peter

[Gerner]

Hi Peter

Of course I was joking with the fridge thing. 

Crazy is maybe not the word I would use about you. More precise I would call you experimental. Experiments are often leading to theories and later on to science and facts. Facts are objective. As objective as we all know who our mother is. 

My curiosity is a spin off of the fact that I feel no lacks of nothing here in my environment.

And your experiments is of such unorthodocs caracter that it is waking up an old fact-man like me. 

Cheers

Gerner

[BertD]

So remember, no problems, just experimenting, and what I hear I like.
And that's subjective ...

Okay, points taken. Sorry for bothering you with my "logical" way of thinking on the subject. The tweak just does not work out for me...

Enjoy your experiments from which I am sure something logical (to me) and benifical (to all) will arise from it in the future.

Bert