Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Romy The Cat]

maxdimario wrote on Wed, 30 November 2005 17:55

as for ultrasonic frequencies, my ears get very tired after 20 minutes of 78 records. 78 records are very fast and noisy, and I assume that they have noise that goes up over the 100KHz mark. a lowpass filter at 12K does make them more tolerable.

Hmmmm… this is VERY controversial. The lowpass filter the 78s do bring comfort to the recording but only if you do not have a proper way to play 78s. The best 78s transfers that I’ve heard were very much full range.

[Jon Hodgson]

maxdimario wrote on Wed, 30 November 2005 17:55

as for ultrasonic frequencies, my ears get very tired after 20 minutes of 78 records. 78 records are very fast and noisy, and I assume that they have noise that goes up over the 100KHz mark. a lowpass filter at 12K does make them more tolerable.

You put your filter cutoff way down into the audible band, but conclude that it's inaudible frequencies affecting you?

It sounds to me more like you're finding noise in the audible band tiring - nothing new about that, that's why a lot of noise reduction systems are basically some kind of dynamic low pass filter.

Now if you were to sample some 78s at 192kHz, then brickwall filter in the digital domain (no effects in the audible band) at 20kHz, and then double blind A/B tests of those recordngs on the same system showed the unfiltered one consistently made your ears tired faster than the filtered ones... then you might be onto something.

[Ronny]

maxdimario wrote on Wed, 30 November 2005 18:16

78's (78 noise) , analog amps and speakers with tweeters go way over 20k. tests have been performed where exposure to high levels of ultrasound caused ear fatigue and temporary tinnitus if I remember reading correctly. didn't I read on this forum that rupert neve did a test such as this? don't mix my two arguments together please.. the last phrase was just to say that: although we don't hear it, the nervous system is AFFECTED by the ultrasonics... how is a different story...

Rupert also says that Geoff Emerick could hear a +3db boost at 54kHz, due to 3 panels on an AMEK that had some miswired transformers or something like that, but I say it's impossible and he made a mistake as to what Emerick was actually perceiving.

I'm not talking about a different story, I'm talking about what you say your experiences are with 78's. Many old 78's only have a 400Hz to 8k frequency range. Later ones such as Decca used a 6dB slope and were down -10dB at 9k. Most microphones used on 78 recordings, the better ones such as the RCA 44's and 77's only went up to 15k. Other mics were common at 12k max. We really didn't start getting into 20k being common with mics until 78's were well on their way out. ITR, it quite baffles me why you feel frequencies that were never captured nor ever reproduced. Ok, let's say we "feel" freq's above 20k in our nervous system, on 78 rpm's they are never on there to begin with. I suggest that you re-evaluate your testing parameters. You are going to continue hearing what you think you hear, although it's never been there, until you take expectations out of the equation and realize the limitations of the media that you are listening to.

[Romy The Cat]

Ronny wrote on Thu, 01 December 2005 09:24

Rupert also says that Geoff Emerick could hear a +3db boost at 54kHz, due to 3 panels on an AMEK that had some miswired transformers or something like that, but I say it's impossible and he made a mistake as to what Emerick was actually perceiving.

Rony, I do not think that it is +3db boost at 54kHz, but rather +0.1dB at 14kHz.  I made these experiments many times. I have tweeters sitting on transition slope at 60kHz and yes it dose looks like moving them 10kHz up and down produce an auditable result. (The tweeters of course roll of at 15kHz-16kHz). However the have learned that what I was actual hearing was change the 12 kHz -14kHz at many dBs down.


Rgs,
The Cat

[danlavry]

maxdimario wrote on Wed, 30 November 2005 17:55

to me, personally speaking, the timing resolution argument did not totally fall apart. I don't wish to discuss it further because it pivots on arguments that are too difficult to prove. as far as any real system increasing in accuracy with an increase of samples for a given slice of time, that would apply to any sampling system, once the invariable 'unknown element' is considered part of the equasion... but seemingly that does not apply to the world of audio converters.. oh well.. but my listening habits are quite different than 99.9999% of the population, so I may be focusing in on minutiae. I do know that absolute precision in the attack transients in the time domain, relative to a fixed timeline (phase shift not included, phase distortion and ringing included) is something I definetly hear and I am so glad that i do because it adds to my listening enjoyment very much, especially with recordings done on low feedback tube equipment. as for ultrasonic frequencies, my ears get very tired after 20 minutes of 78 records. 78 records are very fast and noisy, and I assume that they have noise that goes up over the 100KHz mark. a lowpass filter at 12K does make them more tolerable. even though our brains do not register over 20KHz, the immediate nervous system near the ear seemingly does.. otherwise ears would not get tired with ultrasonics.

You are at "square 1". You did not read my paper, nor did you follow any of the explanations.

You said: "to me, personally speaking, the timing resolution argument did not totally fall apart. I don't wish to discuss it further because it pivots on arguments that are too difficult to prove."

The subject has been covered over and over! The time resolution (minimum impulse width) IS THE SAME AS BANDWIDTH. It is a FUNDAMENTAL BASIC FACT OF PHYSICS AND MATH. The minimum impulse width of a system is LIMITED BY THE LOWEST PART IN THE CHAIN (be it mic, speaker or ear...)

You said: "as far as any real system increasing in accuracy with an increase of samples for a given slice of time, that would apply to any sampling system, once the invariable unknown element' is considered part of the equation... but seemingly that does not apply to the world of audio converters.. oh well.."

Well you do not understand Nyqusit theory. More dots is NOT MORE DETAIL. In theory, there is NOTHING TO BE GAINED by sampling faster the twice the highest frequency of the content, be it audio, video or what not. Audio IS NOT DIFFERENT!
In practice, one may need some reasonable margin above the theoretical X2 factor, all that is well understood.

I will not comment on the rest of your statements. Please look at the rules of this site. We try to avoid subjective remarks.

Regards
Dan Lavry
www.lavryengineering.com

[AndreasN]

Hi!

Made a picture to show where the inter-sample phase information is lurking.

The rising wavefront is a drum attack, chopped off before the real attack begins at all. Taking only the very first tiny little beginning of the wave in consideration. The end is suddenly gated, leaving a half cycle pulse. Upper picture is 44kHz, lower is a resampling of this to 192. The time grid lines are set at one millisec.



The three lumps of sound was first edited in 192khz. The last two pulses being offset 2 and 6 samples forward. Resampling this to 44kHz gives sub-sample offsets on the transients. Is this what you're looking for? Notice how all the peaks look different. Still, they represent the exact same waveform, at different phase relationship in regards to the sample dots.

Quadrupling the number of dots in the second picture gives a more comforting picture. The wave is still the same, with new offsets in regard to the 192kHz dots. Going to 960000kHz would give very accurate visual representation, as a digital oscilloscope does. Better editing too, btw, in lack of sub-sample edits, but the 22kHz limited waveform would be the same all the time.


Hope this may help someone out there. The issue sure used to confused me!

[Ronny]

Romy The Cat wrote on Thu, 01 December 2005 13:31

Ronny wrote on Thu, 01 December 2005 09:24 Rupert also says that Geoff Emerick could hear a +3db boost at 54kHz, due to 3 panels on an AMEK that had some miswired transformers or something like that, but I say it's impossible and he made a mistake as to what Emerick was actually perceiving. Rony, I do not think that it is +3db boost at 54kHz, but rather +0.1dB at 14kHz.  I made these experiments many times. I have tweeters sitting on transition slope at 60kHz and yes it dose looks like moving them 10kHz up and down produce an auditable result. (The tweeters of course roll of at 15kHz-16kHz). However the have learned that what I was actual hearing was change the 12 kHz -14kHz at many dBs down. Rgs, The Cat

Romy, I was specifically talking about what Rupert said in a PSW interview, different than what you are relating to. I was a panelist and linear phase digi eq's were being developed by a few companies at the time, that was my particular interest in getting Ruperts input. Do a search in PSW archives, Rubert Neve Interview hosted by Fletcher, for the Geoff Emerick story.

[maxdimario]

Quote:

I'm not talking about a different story, I'm talking about what you say your experiences are with 78's. Many old 78's only have a 400Hz to 8k frequency range.

I wrote  noise...noise...schellack record scratching at high speed you know?? shhhhhhhhhhhhkhkhkhkhkhhhh! scratchy 78 records.

lots of nasty unwanted noise well into the 100 KHz region.

anyway ronnie.... since you do mastering... did you hear the difference between 48 and 96KHz on the same machines for the 'PT loss of bass' test?

I sure as hell did, on my laptop computer monitor... which makes the difference pretty obvious.. huh?.. regardless of information over 20 KHz.. which is over the hearing range.

[danlavry]

AndreasN wrote on Thu, 01 December 2005 23:30

Hi! Made a picture to show where the inter-sample phase information is lurking. The rising wavefront is a drum attack, chopped off before the real attack begins at all. Taking only the very first tiny little beginning of the wave in consideration. The end is suddenly gated, leaving a half cycle pulse. Upper picture is 44kHz, lower is a resampling of this to 192. The time grid lines are set at one millisec. The three lumps of sound was first edited in 192khz. The last two pulses being offset 2 and 6 samples forward. Resampling this to 44kHz gives sub-sample offsets on the transients. Is this what you're looking for? Notice how all the peaks look different. Still, they represent the exact same waveform, at different phase relationship in regards to the sample dots. Quadrupling the number of dots in the second picture gives a more comforting picture. The wave is still the same, with new offsets in regard to the 192kHz dots. Going to 960000kHz would give very accurate visual representation, as a digital oscilloscope does. Better editing too, btw, in lack of sub-sample edits, but the 22kHz limited waveform would be the same all the time. Hope this may help someone out there. The issue sure used to confused me!

You issue is still confusing you.

The reason the pictures look different is simple: You are violating Nyquist. A sudden change in the input voltage wave means having high frequency content. The ear does not react to the high frequency energy due to sudden changes, above a certain frequency (be it 20KHz or what not). We can and need to agree on some reasonable bandwidth (per application, such as audio, video, instrumentation...). Once we know the bandwidth, it is our responsibility to sample faster then twice the bandwidth. Differently stated - we need to make sure that our AD does not process signals above half the sample rate. We often use a filter to remove energy above Nyquist. Note that modern AD's have Nyquist at the MHZ range, making the filtering easier.

At the DA side, one needs a similar anti imaging filter to ensure proper conversion from digital to analog.

Try to redo what you did with proper filtering prior to the AD process, and after the DA, and the pictures will all look the same. Alternativly, go to www.lavryengineering.com then to Forums, then to Tutorials, and look at the paper "Sampling Theory" for some more detailed explanation.

A wave form that has been properly sampled (allow low pass energy below Nyquist, reject energy above Nyquist), and properly reconstructed (again, allow low pass energy below Nyquist, reject energy above Nyquist) will look THE SAME. You can increase the sampling frequency all you want and it will make ZERO DIFFERANCE.

That is the whole concept behind digital representation of analog. We can not represent ANY ANALOG by digital means, that would take infinite points... But we can represent a RESTRICTED analog signal with digital signals,. We obey Nyquist, and the representation is ERROR FREE. It is not an aproximation with an error, it is PERFECTION.

The real world deviations from perfections are due to imperfect filters, jitter and other practical sources of error, but they are very small. Certainly an 8-10 bit simulation is far better then a time domain picture on a scope (which is roughly a 1% instrument...)

Also, note that it is not a simple matter to simulate such things on a computer. The computer does not Handel continues analog waves. The computer itself takes the analog signal and makes it into data point. In my tutorials, I start by approximating the analog as a signal by oversampling it by a large factor, then I get an approximate result, good enough for visualizing scope like pictures...

How many "dots" does one need to show a digitized 44.1KHz signal? You need 44100 dots per second. How many "dots" does one need to show an analog signal? It depending on screen resolution, but figure on a huge factor. Viewing an "analog signal" on a computer does require filling the "in between" to a resolution "good enough" for the eye. A crude computer modeling of analog amounts to violation of Nyquist, before you even began the analysis. So take care in with you do, when presenting analog on a computer...  

Regards
Dan Lavry
www.lavruengineering.com    

[Ronny]

maxdimario wrote on Fri, 02 December 2005 06:17

Quote: I'm not talking about a different story, I'm talking about what you say your experiences are with 78's. Many old 78's only have a 400Hz to 8k frequency range. I wrote  noise...noise...schellack record scratching at high speed you know?? shhhhhhhhhhhhkhkhkhkhkhhhh! scratchy 78 records. lots of nasty unwanted noise well into the 100 KHz region. anyway ronnie.... since you do mastering... did you hear the difference between 48 and 96KHz on the same machines for the 'PT loss of bass' test? I sure as hell did, on my laptop computer monitor... which makes the difference pretty obvious.. huh?.. regardless of information over 20 KHz.. which is over the hearing range.

The noise you are talking about has to do with the playback devices and the abuse to the record over the years. None of these freq's are recorded and aren't relative to the music, so we are talking about two different things. Still the speaker cabinets that you have won't reproduce any mechanical noise at 100k.

I'm not familiar with the PT loss of bass test, but I was a participant on Lynn Fustons PT Bloodhound CD. It was supposed to put an end once and for all, to the PT having a sonic footprint debate, that went on in the 90's. I picked out the PT mixes 70% of the time in the blind, that was the highest percentage, IIRC, Dave Davis from QCA and I tied with the most, or he might have gotten 71% right, can't remember, it was so long ago, but most of the older engineers scored above 50%, which may leave one to less than speculate that PT had a detectable sonic character at one time. However, I feel that much of the talk about BTD sounding different or losing chorusing effects and such, is due to pilot error and not so much the rewrite being corrupted, when things are operating correctly.
 

[Schallfeldnebel]

And what about analog tape? I have never heard anybody got tired from analog tape. There is a residue from a 120kHz bias tone left when playing back tapes,I rather heard the opposite, people getting tired from digital. And what about watching television, or listening to FM radio or 38kHz pilottone quadro CD-4 vinyl discs?

Erik Sikkema

[maxdimario]

this is true, but 78 noise is nasty...and loud. my speakers do something up to 40 Khz at least.

the noise on some of those old scratchy records is VERY loud and obnoxious, changing constantly.

tape bias noise was usually filtered out somewhat, and I don't think it was as severely nasty as a scratchy shellack record spinning at 78.

...I try to listen to the 78's in better condition obviously.

as far as the difference between 48 and 96 it seems obvious sonically to me, but I do not want to claim that I know exactly why, as we've been through this before.

can you hear it on the PT bass test?

[Jon Hodgson]

maxdimario wrote on Sat, 03 December 2005 11:25

can you hear it on the PT bass test?

If I do a spectral analysis on those files I can SEE a difference below 20kHz.

What does that tell me?

It tells me that for whatever reason (might be the converters, might be the filters, might be some other variable, say in the 2", or someone just knocked a knob), the capture of information below 20kHz is not identical.

Sample rate was not the only thing that changed.

[Schallfeldnebel]

Max wrote:"as far as the difference between 48 and 96 it seems obvious sonically to me, but I do not want to claim that I know exactly why, as we've been through this before."

When I compare 48 and 96 with my converters, it is obvious the 3-4K Hz region is sounding less stressed when using 96. Have you read this?

http://www.digitalaudio.dk/technical_papers/aid.pdf

Erik Sikkema

[AndreasN]

danlavry wrote on Fri, 02 December 2005 19:27

You issue is still confusing you. .... Try to redo what you did with proper filtering prior to the AD process, and after the DA, and the pictures will all look the same. .... Also, note that it is not a simple matter to simulate such things on a computer. The computer does not Handel continues analog waves. The computer itself takes the analog signal and makes it into data point. In my tutorials, I start by approximating the analog as a signal by oversampling it by a large factor, then I get an approximate result, good enough for visualizing scope like pictures...

Was talking strickly about digital visual representation, not what's in the other end(analog wave), which we agree is the same wave comming in and out no matter what it looks like on the computer screen. The 22kHz bandwith limited signal which was resampled to 192kHz, in the second row of waveforms, was included to illustrate how things may look better, yet represent the exact same waveform.

It sure is confusing, yes! If I managed to confuse you, everyone else must have been as well. Thanks for clearing things up.



Good weekend to all!


Cheers,

Andreas Nordenstam