I have an old wind up record player that plays 70rpm records, keep it around for perspective.
Have to admit its a seductive process, removing the rcord from the outer sleeve, then the inner, lightly blowing off any dust...
The modern alternative is to just insert a memory stick into my pc.
Have to admit its a seductive process, removing the rcord from the outer sleeve, then the inner, lightly blowing off any dust...
The modern alternative is to just insert a memory stick into my pc.
It appears on the low voltage (45V) MPSA18 as well.
Attachments
Douglas Self will be presenting his Devinyliser concept to the Audio Engineering Society in London at 6:30pm on Tuesday 11th October at Kings College, London. This will be a longer version of the short presentation he gave to the Paris AES convention in June this year.
The event is free to all, and we hope to see some of you there.
The event is free to all, and we hope to see some of you there.
Adding to this discussion, I just read a number of interesting articles on MQA.
This newly introduced innovative algorithm is supposed to remove the temporal smearing caused by the steep filtering steps in the whole digital process.
A high frequency high bit master recording processed by MQA and played at 44.1 Khz is reported to sound better as the master recording.
An inconvenient truth: MQA sounds better! | DAR__KO
Since steep filtering and the consequential temporal smeering is no part of Analogue recorded LP's, this might lead to understanding why some prefer vinyl.
It could also explain why I prefer the direct sound of an analogue recorded LP versus the sound of the very same LP signal after being converted to digital in 192Khz/24bit and back to analogue in real time.
Hans
This newly introduced innovative algorithm is supposed to remove the temporal smearing caused by the steep filtering steps in the whole digital process.
A high frequency high bit master recording processed by MQA and played at 44.1 Khz is reported to sound better as the master recording.
An inconvenient truth: MQA sounds better! | DAR__KO
Since steep filtering and the consequential temporal smeering is no part of Analogue recorded LP's, this might lead to understanding why some prefer vinyl.
It could also explain why I prefer the direct sound of an analogue recorded LP versus the sound of the very same LP signal after being converted to digital in 192Khz/24bit and back to analogue in real time.
Hans
I certainly respect the credentials of the folks involved but if there is not some poetic licence in coming up with these msec numbers I would like to see a better explanation than what I have seen.
An ordinary USB sound card using conventional sigma delta A/D's is perfectly capable of capturing the impulse response of a microphone at the usec level. Picture below is from an old experiment where I used a spark discharge (which actually produces a doublet) and a fairly good 1/4 mic for which I had a calibration chart produced by a conventional B&K test set up (all analog). The initial part is actually a shockwave produced by the spark heating the air and the second is the thermal relaxation. The minute details of the math are complex but the literature covers deriving the impulse response from this data. The result was an almost perfect fit between the two. This is 96k and not 44.1 but the scale is 50usec/div. Even the small ripples that trail off are mostly due to unavoidable early reflections from my crude spark generator, the starter for an outdoor grill.
Attachments
IT IS A FACT that the steeper the filters, the more ringing is added to the original signal.
In case of analogue filtering it is postringing and in case of digital filtering it results in either pre plus postringing with linear phase filters , or just postringing with minimum phase filters such as apodizing filters.
The effect of added ringing to the signal is also called "time smearing",
The frequency and duration of this ringing is dependant on the specific filter and on the sampling frequency.
In general one can say that the higher the sampling frequency, the less steeper the filters have to be to prevent aliasing, leading to less ringing and at the same time having a higher (ultrasonic) frequency.
That results technically in less time smearing.
WHAT IS NOT AT ALL PROVEN, is whether one can hear this time smearing.
On one hand there are interesting blind tests, telling that no difference can be heard between 176.4/24 PCM and DSD, others mention that no difference can be heard between 96/24 and 192/24 PCM and last but not least there are papers telling that with a top quality DAC, no difference can be heard between 44.1/16 and 192/24 PCM.
On the other hand, modern high end DAC's equiped with different filters to choose from, are told to be sounding better with an apodizing filter (i.e. no pre ringing) than with linear phase filters, but it is just the opinion of the tester of that audio magazine
Fact however is that 44.1/16 PCM's are already processed at the recording stage and will be suffering from time smearing, which can never be removed at playback time.
MQA tries to adress the streaming world, where it is highly unpractical to stream DXD with 352.8/24 PCM and that's why they claim to have come with a compression solution that sounds just as good as the original DXD recording.
To avoid the time smearing when going to 44.1/16 or 48/16, their algorithm is supposed to minimise / reduce or even eliminate the so called time smearing.
Very little is known about this algorithm, and yes it is very good to remain sceptical, but here is some information found on Wiki.
While the technology has received little comment in the general and mainstream press, it has been exalted by the audiophile and hi-fi press. Robert Harley, editor of The Absolute Sound has referred to it as "The most significant audio technology of my lifetime".[11] Editor John Atkinson writing in Stereophile magazine following the UK launch in December 2014 wrote "In almost 40 years of attending audio press events, only rarely have I come away feeling that I was present at the birth of a new world."[12]
Some critical but primarily speculative comments have been made in online forums such as the Computer Audiophile forum[13] and in audio magazine website comments, and a few writers have expressed concern in some areas. Over 80 detailed questions, some of which voiced these concerns, were submitted to the editors of the Computer Audiophile forum and subsequently addressed in detail by the creator of MQA, Bob Stuart, in an extended question-and-answer art
A Comprehensive Q&A With MQA's Bob Stuart
[URL]http://www.computeraudiophile.com/content/694-comprehensive-q-mqa-s-bob-stuart/[/URL]
I can only say that in an experiment that I performed, I could very well hear the difference between the sound of analoque recorded LP's played over an analoque chain of TT/Phono pre-amp / pre-amp / main amp / ESL63 versus
TT/phono pre-amp/24-192 AD/24-192 DA / pre amp / amp/ ESL63, being in favour of the first.
The AD/DA processed LP's sounded like CD's, very clean but less life.
Playing digital recorded LP's however did not show a difference in sound.
I know that this is absolutely no prove at all, but for me it is nothing but an indication that it COULD be that the process of digitizing plus filtering is acoustically not as transparant as it is supposed to be.
So maybe in the end, MQA might prove to be just a commercial gadget, but one should always have an open mind and accept that almost everything is possible.
So let's wait and see.
Hans
In case of analogue filtering it is postringing and in case of digital filtering it results in either pre plus postringing with linear phase filters , or just postringing with minimum phase filters such as apodizing filters.
The effect of added ringing to the signal is also called "time smearing",
The frequency and duration of this ringing is dependant on the specific filter and on the sampling frequency.
In general one can say that the higher the sampling frequency, the less steeper the filters have to be to prevent aliasing, leading to less ringing and at the same time having a higher (ultrasonic) frequency.
That results technically in less time smearing.
WHAT IS NOT AT ALL PROVEN, is whether one can hear this time smearing.
On one hand there are interesting blind tests, telling that no difference can be heard between 176.4/24 PCM and DSD, others mention that no difference can be heard between 96/24 and 192/24 PCM and last but not least there are papers telling that with a top quality DAC, no difference can be heard between 44.1/16 and 192/24 PCM.
On the other hand, modern high end DAC's equiped with different filters to choose from, are told to be sounding better with an apodizing filter (i.e. no pre ringing) than with linear phase filters, but it is just the opinion of the tester of that audio magazine
Fact however is that 44.1/16 PCM's are already processed at the recording stage and will be suffering from time smearing, which can never be removed at playback time.
MQA tries to adress the streaming world, where it is highly unpractical to stream DXD with 352.8/24 PCM and that's why they claim to have come with a compression solution that sounds just as good as the original DXD recording.
To avoid the time smearing when going to 44.1/16 or 48/16, their algorithm is supposed to minimise / reduce or even eliminate the so called time smearing.
Very little is known about this algorithm, and yes it is very good to remain sceptical, but here is some information found on Wiki.
While the technology has received little comment in the general and mainstream press, it has been exalted by the audiophile and hi-fi press. Robert Harley, editor of The Absolute Sound has referred to it as "The most significant audio technology of my lifetime".[11] Editor John Atkinson writing in Stereophile magazine following the UK launch in December 2014 wrote "In almost 40 years of attending audio press events, only rarely have I come away feeling that I was present at the birth of a new world."[12]
Some critical but primarily speculative comments have been made in online forums such as the Computer Audiophile forum[13] and in audio magazine website comments, and a few writers have expressed concern in some areas. Over 80 detailed questions, some of which voiced these concerns, were submitted to the editors of the Computer Audiophile forum and subsequently addressed in detail by the creator of MQA, Bob Stuart, in an extended question-and-answer art
A Comprehensive Q&A With MQA's Bob Stuart
[URL]http://www.computeraudiophile.com/content/694-comprehensive-q-mqa-s-bob-stuart/[/URL]
I can only say that in an experiment that I performed, I could very well hear the difference between the sound of analoque recorded LP's played over an analoque chain of TT/Phono pre-amp / pre-amp / main amp / ESL63 versus
TT/phono pre-amp/24-192 AD/24-192 DA / pre amp / amp/ ESL63, being in favour of the first.
The AD/DA processed LP's sounded like CD's, very clean but less life.
Playing digital recorded LP's however did not show a difference in sound.
I know that this is absolutely no prove at all, but for me it is nothing but an indication that it COULD be that the process of digitizing plus filtering is acoustically not as transparant as it is supposed to be.
So maybe in the end, MQA might prove to be just a commercial gadget, but one should always have an open mind and accept that almost everything is possible.
So let's wait and see.
Hans
Hi Bill,
I could very well hear a difference, don't ask me for an explanation, I just could, but only with analogue recorded LP's and with my equipment, and nobody can convince me I was wrong
There are so many variables, that even a cable or a mains filter could theoretically make the difference, but I found it curios that digitally mastered LP's showed no acoustical difference.
Maybe, if there is enough enthousiasm, I could repeat the test with a larger group under well managed conditions.
But then again, it would be test nr X which could be put on the pile of other conflicting tests.
Hans
Hi Bill,
It started out of curiosity when testing the possible sound difference between LP and CD, with a CD produced by the magazine Stereoplay directly from the very same analogue LP's from the 50ties that I had.
I heard a large difference, but blamed it on the difference in Cart, TT and phono preamp that Stereoplay used.
But then I included the AD/DA in my LP chain, and now the sound between CD and LP were practically identical.
I did not expect this outcome at all.
Since I was very surprised, I asked my son to listen while switching the remote control for LP with and without coded.
He was not informed in advance and he had no idea what the technical difference was between remote 1 and remote 2.
Gain in both situations was the same within 0.1 dB.
He could see an LP running on a TT but that was all.
Nevertheless he came to the same conclusion as I did.
Maybe my body language or whatever else influenced him, but the differences were not subtle at all with an analogue mastered LP.
And again, a digitally mastered LP's did not show these differences.
So it is just a very personal non scientific experience, never meant to be a test to prove something as a general statement, but nevertheless I thought it to be worth mentioning.
Many people seem to accept sound differences between SS an Tubes or between op-amps to name a few, but this area seems to be slippery.
Hans
Member
Joined 2009
Paid Member
Look at the picture Hans, that's the pre-ringing you get from a real acoustic signal and a microphone (fairly good,+1, -3dB @ 500~20k). I stop reading when the examples shown violate Nyquist and the point is about real music. Where is the MILLI-second "time smear"?
Dave Wilson produced an LP (of the late Mr. Bongiorno's honky tonk piano) with one side direct to disk IIRC and the other side passed through a Soundstream AD/DA as the only added process. I ripped both at 24/96 and aligned them and at no point could I even get FFT's of sections to match better than +-3dB across the whole spectrum. They sounded identical to a very large degree.
Hi Scott,
In the MQA article a 2.5 millisecond time smear is displayed for CD and 0.5msec for 96Khz sampling rate. Those figures are dropping from the sky because this depends largely on the filters used.
Two examples showing time smear for reproducing an impulse are shown below for a typical D/A at 44.1Khz for a Fir filter and for an Apodizing filter. The first being smeared over 0.6 msec and the second over 2.6msec.
This already shows that time smear is very dependant on the type of filter being used.
Your sampling rate was 96Khz, but sampling does not cause time smear, filtering does. So how did you produce this figure.
Was any filtering at all involved in this process.
And when using a more relaxed 48Khz filter, this would almost produce no ripple, making time smear completely invisible in your graph for the ca 10Khz signal, confirming your measurement.
But I agree in general, there is a lot of bla, bla in the article.
Hans
This confirms that A/D-D/A is (technically) not a transparant process, but acoustically it is supposed to be.
The problem with the LP you describe is that there is time between switching from side A to side B, making comparing quite a lot more complex than switching a codec in and out in real time while the LP keeps running.
Hans
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.