[analogsurviver]

   
Finally, something we can agree on. Digital crossovers are seriously good, eliminating all sorts of problems in traditional passive crossover, simple amp configurations. Only thing holding me back is the considerable extra cost.
 
 
This is also a false conclusion. If it were true, then every IEM would demonstrate exceptional impulse response. Size does not necessarily make a driver faster or slower - there are other factors involved.
 
 
No, I literally posted the first high end speaker I found.

Finally, something we can agree on. Backed by the cost con$ideration$

. Were it not for the money, I would already have that digital crossover - the measurements of finished loudspeakers looked to die for.
 
You are wrong regarding the structural resonances in ESL drivers - given the same size thickness of the perforated plates, the smaller one will be resonating at higher frequency(es) - thus less likely to be  exited by music "overtones" than "fundamentals" of the larger drivers. A somewhat cheesy but effective (if implemented better) solution would be to use conductive textile mesh for stators - as used by Sennheiser in their Unipolar 2000 and 2002 electret headphones - FAR less "stator talk" than thin metal used by Stax. To those in possesion of Martin Logan ESLs : tap with a fingernail (or plastic pencil, to make 1000 % sure no electrical shock can occur ) on the stators (or better yet, have someone tapping on the stators while you are in the listening position ) with no music playing - whatever you will hear, SHOULD NOT BE THERE IN THE FIRST PLACE. It is true that ESL diaphragm does not have resonances - but it can not but follow the stators. Which DO have structural resonances ...- which are added to the original signal. This is the path a very well executed dynamic driver might be able to use to surpass the performance of ESLs.
 
I did mention the International audio Review's review of Acoustat vs Bowers & Wilkins speakers; it presented the same type of measurements, both for pulse and square wave. Acoustat (full range ESL ) was correct (with understandable limitations ), B&W was even worse than Utopia - going from positive and negative response in 20-20 kHz range 5 times ! As IAR is adamant about copyrights etc, I can not publish these decades old results; it remains the only audio publication that does not accept any advertising - and thus CAN publish whatever the real findings. But subscriptions are steep - 200$+/year. I only check a few snippets available for free online.

 

[lamode]

  You are wrong regarding the structural resonances in ESL drivers ..

 
Hold on... I was talking about impulse response and now you are suddenly talking resonances.
 
 
 

 

[analogsurviver]

   
Hold on... I was talking about impulse response and now you are suddenly talking resonances.
 
 
 

And what impedes better impulse response if not structural resonances - at least in ESLs ? 
 
In Stax 009 there is a good initial response, marred by "later" - and that is stators flexing. Diaphragm is so light it is more than critically damped by the air enclosed in gaps and perforations of the stators, amplifier does not ring like that, what remains are stators - which diaphragm has no other way but to follow. 
 
In good and bad...

 

[castleofargh]

well if the damping isn't electrical, it's mechanical. but guessing who's who from an impulse is beyond me I must say  
maybe one day when I have more experience with those stuff.

 

[lamode]

  And what impedes better impulse response if not structural resonances - at least in ESLs ? 
 

 
Resonances, frequency response, damping, etc... many factors play a role.

 

[Don Hills]

  Ever listened to a real drum kit ? Live, acoustic, not over speakers - not a recording over youtube ?
 
Then try to squeeze that into CD redbook ... - and see if it is still "realistic". 
 
DSD128 comes close - but even more would be required for the sound that no longer can be discerned from live. CD has absolutely no chance with percussion, cymbals being only the most problematic/critical part.
Rise times on say a rim shot are extremely fast - CD can NEVER reach the proper amplitude of this pulse, it is simply too slow.

 
I listened to live drums often in my youth. I even became quite good at tuning kits. I quit in disgust in the 80s when the "whacking a cardboard box with a wooden spoon" sound came into fashion and people wanted their drums to make that studio sound when playing live.
 
As for fitting the sound on a CD, it's not the CD that limits it - it's almost always the amps and/or the speakers. A realistic rimshot requires an extraordinary peak output. I have some locally produced CDs, done without compression, where some snare hits and rimshots are 20 to 30 dB higher than the average program level yet don't sound unusually loud. (And yes, they sound as they did when being recorded.)
Regarding the frequency response of a CD, I simply don't see the necessity of attempting to reproduce something you can't hear. I do agree that a higher sampling rate eases the design of the necessary filters.

 

[analogsurviver]

   
I listened to live drums often in my youth. I even became quite good at tuning kits. I quit in disgust in the 80s when the "whacking a cardboard box with a wooden spoon" sound came into fashion and people wanted their drums to make that studio sound when playing live.
 
As for fitting the sound on a CD, it's not the CD that limits it - it's almost always the amps and/or the speakers. A realistic rimshot requires an extraordinary peak output. I have some locally produced CDs, done without compression, where some snare hits and rimshots are 20 to 30 dB higher than the average program level yet don't sound unusually loud. (And yes, they sound as they did when being recorded.)
Regarding the frequency response of a CD, I simply don't see the necessity of attempting to reproduce something you can't hear. I do agree that a higher sampling rate eases the design of the necessary filters.

Well, good to know. I agree that almost any commercially available recording of drums is "doctored" in order to accommodate playback on usually available equipment. 
 
I am VERY aware of the fact that it does not sound unusually loud - yet it requires 120 dB + SPL capability on playback. It also means that any "average" music that would otherwise get recorded close to 0 dB is by now at -20 to - 30 dB down in level - decreasing the bit depth available for "music without rimshots" by 4 or 5 bit - which IS objectionable - and is a case for 24 bit recording and not only extended frequency response.
 
Try to record once drum kit with at least 88.2 kHz sampling - and the recording should appreciably gain in "immediacy". The catch lies in rise times - put it simply, 88.2 allows for twice faster rise times - much closer to the real thing. Going to 192 or 384 (or DSD128 and up ) is a much more subtle difference. It works even if microphones do not extend much beyond 20 kHz - and really comes to life with something that goes at least to 40 kHz.
 
Regardless of  our hearing limited to 20 kHz or less...
 
The filters that can be made much more simple IS the reason why converting 44.1/whatever (can be MP3 as well..) to DSD64 or higher WILL open up in stage etc by
a definitely audible margin. Although I try to stay within the original format as much as possible ( DSD to PCM and vice versa is a lossy process and should be avoided ) , there are recordings that benefit from this - sometimes enormously so. Conversion can also be on the fly ( using jRiver 19 for example ).
 
Simple - use your ears ...

 

[StanD]

Unless one is listening to recorded drum solos, the available DR must be shared with the entirety of the music. If everyone always listened to volume the DR found on stage, there would be a lot of people stumbling around with hearing loss. Reality strikes.
Gee, I thought Soundstage was due to recorded spacial queues and the ability of speakers/headphones to reproduce that. Amps and DACs are transparent and shouldn't be modifying it. Since we can't tell the difference between formats, bit depths and sample rates, that's not going to affect it.

 

[analogsurviver]

  Unless one is listening to recorded drum solos, the available DR must be shared with the entirety of the music. If everyone always listened to volume the DR found on stage, there would be a lot of people stumbling around with hearing loss. Reality strikes.
Gee, I thought Soundstage was due to recorded spacial queues and the ability of speakers/headphones to reproduce that. Amps and DACs are transparent and shouldn't be modifying it. Since we can't tell the difference between formats, bit depths and sample rates, that's not going to affect it.

Even if that real life DR would be listened to on regular basis, it is still LESS than compressed masters of today - some with the whole DR of - 3 dB or less. That means NEVER below 90 dB or so  for the duration of the recording. This IS much more dangerous for hearing loss compared to something that is on average much more quiet and peaks of 120 dB represent maybe 1% of the time of the entire recording...
 
Amps and DACs DO modify the soundstage - IF they are, for whatever reason, limited in frequency response. It most directly affects DEPTH. One difference that would be immediately perceived between DSD64 and DSD128 ( using amplification with extended frequency response > 100 kHz ) is precisely that - DSD128 allows for MUCH better depth information - it will allow one to follow the initial strike of say percussion instrument from the initial strike all the way to the boundaries of the recording venue - and will not be pancake flat as with redbook, with depth info limited to very shallow "window" - at best. If tennis reference works for you, it is like depth is limited to say a foot or two in front and behind the net on redbook - and at least to the limit lines of the field with DSD128. That is BIG difference.
 
Clear enough? 
 
Compared to live feed from mike, on location, to recorder(s) set to various DSD and PCM resolution(s) - not a recording done by others and without reference to actual live sound.
 
And this is why I insist on frequency response beyond 20 kHz.
 
It is most audible on binaural recordings. Particularly those done by what I have termed Binaural Natural - mics worn on one's own head/ears. It just does not get more realistic than that.

 

[lamode]

Originally Posted by analogsurviver /img/forum/go_quote.gif
 
The catch lies in rise times - put it simply, 88.2 allows for twice faster rise times - much closer to the real thing.

 
Digital audio doesn't have a "rise time". 16/44 can go from 0 to 100% signal in one sample, as can all PCM.

 

[lamode]

  Even if that real life DR would be listened to on regular basis, it is still LESS than compressed masters of today - some with the whole DR of - 3 dB or less. That means NEVER below 90 dB or so  for the duration of the recording. This IS much more dangerous for hearing loss compared to something that is on average much more quiet and peaks of 120 dB represent maybe 1% of the time of the entire recording...

 
Yes, it's true that compressed recordings are more likely to inflict hearing damage due to the higher average SPL.
 

Originally Posted by analogsurviver /img/forum/go_quote.gif
 
it will allow one to follow the initial strike of say percussion instrument from the initial strike all the way to the boundaries of the recording venue - and will not be pancake flat as with redbook,

 
False. There is nothing audible which DSD128 can capture which Redbook can't, whether it's a direct sound of an instrument or reverberation.

 

[analogsurviver]

   
Digital audio doesn't have a "rise time". 16/44 can go from 0 to 100% signal in one sample, as can all PCM.

False. 
 
Digital yes - but 44.1/16 has to go trough output filter - and 22.1 kHz brick filter limits rise time to approx 14 microseconds. It is problematic enough that Legato Link filtering ( basically less sharp filter with effect below 20 kHz , but allowing less sharp filtering above, yielding a small but audible improvement in speed/rise time ) was developed. Even with Legato Link, CD is at least twice too slow compared to good analog or hirez.
 
More about format comparisons here : http://www.lindberg.no/english/collection/004.pdf

 

[RRod]

  False. 
 
Digital yes - but 44.1/16 has to go trough output filter - and 22.1 kHz brick filter limits rise time to approx 14 microseconds. It is problematic enough that Legato Link filtering ( basically less sharp filter with effect below 20 kHz , but allowing less sharp filtering above, yielding a small but audible improvement in speed/rise time ) was developed. Even with Legato Link, CD is at least twice too slow compared to good analog or hirez.
 
More about format comparisons here : http://www.lindberg.no/english/collection/004.pdf

 
It's not like higher res formats get better rise times from magic; they add in frequency content, at frequencies that we can't hear.

 

[StanD]

 
Amps and DACs DO modify the soundstage - IF they are, for whatever reason, limited in frequency response. It most directly affects DEPTH. One difference that would be immediately perceived between DSD64 and DSD128 ( using amplification with extended frequency response > 100 kHz ) is precisely that - DSD128 allows for MUCH better depth information - it will allow one to follow the initial strike of say percussion instrument from the initial strike all the way to the boundaries of the recording venue - and will not be pancake flat as with redbook, with depth info limited to very shallow "window" - at best. If tennis reference works for you, it is like depth is limited to say a foot or two in front and behind the net on redbook - and at least to the limit lines of the field with DSD128. That is BIG difference.
 
Clear enough? 
 
Compared to live feed from mike, on location, to recorder(s) set to various DSD and PCM resolution(s) - not a recording done by others and without reference to actual live sound.
 
And this is why I insist on frequency response beyond 20 kHz.
 
It is most audible on binaural recordings. Particularly those done by what I have termed Binaural Natural - mics worn on one's own head/ears. It just does not get more realistic than that.

Yes batlike hearing is a requirement, sheesh. DACs do not remove spacial queues and if you bought a DAC with an inferior FR that chops off treble, then you must have found a real dud on fleaBay. Only people with alien DNA can hear above 20 kHz.

 

[manbear]

Analog, what headphones/ speakers do you use to play back that content over 100khz on your DSD128 recordings?