Er, um, John?
This time, just for a change, it wasn't Frank (fas42), it was me, I asked the question. FRank probably knows his answer to that, but I don't, and when I don't know the answer, I tend to ask those who I assume know it.
Thank you for the explanation. It'll take some tome for it to sink in, but eventually I'll get there.
Even more so for SY's example, slightly more complicated for me because I am not at all familiar with tubes. But I am persistent.
Currently waiting for Pavel's take and, of course, for George The Phono Man.
This time, just for a change, it wasn't Frank (fas42), it was me, I asked the question. FRank probably knows his answer to that, but I don't, and when I don't know the answer, I tend to ask those who I assume know it.
Thank you for the explanation. It'll take some tome for it to sink in, but eventually I'll get there.
Even more so for SY's example, slightly more complicated for me because I am not at all familiar with tubes. But I am persistent.
Currently waiting for Pavel's take and, of course, for George The Phono Man.
"Never to be recovered" is very far from the truth. If enough computing power is thrown at a problem the underlying 'truth' can always be extracted - in simple terms, if the mechanism that causes the distortion is understood then an intelligent "undoing", working only on the mangled version, will retrieve what's important. It may take several hours to "fix" a five minute sample, as an extreme example, but once sorted out it then remains sorted out.The most freaky part is that, apparently, the data on the medium already is time smeared, never to be recovered. Quite shocking, no idea if it's actually true.
Cheers,
But first we must find someone who is interested in doing so,
...I don't believe the issue is personal taste - the biggy for me, and has always been, is making playback robust - that is, no matter who unfavourable the overall environment is for decent SQ to be achieved, that in fact one can still manage at least an acceptable level ... over the years, this has been by far the biggest headache ...
Hi,
If it takes several hours to "fix" as you put it, a five minute sample then for all intents and purposes, that actually is "never".
More to the point, how did it end up there in such a state and is there no way of preventing this?
That 60dB would include the SUT which accounts for, I'm guessing here, 10 to 15dB of gain?
Cheers,
If it takes several hours to "fix" as you put it, a five minute sample then for all intents and purposes, that actually is "never".
More to the point, how did it end up there in such a state and is there no way of preventing this?
That 60dB would include the SUT which accounts for, I'm guessing here, 10 to 15dB of gain?
Cheers,
No, think noise reduction processes, that are used to "improve" old recordings - someone applies the algorithm once, correctly, and then there is a "refurbished" version of the recording, which can be released, sold, for everyone to play. Don't get caught in the trap of thinking that everything has to be done in real time, that is, while you're actually listening - far smarter, usually, is to do it beforehand - say, you have a precious LP, that is a mess: do your best encoding of that, conversion to digital, and then apply all the "fixits" to the digital capture; the "cleaned" digital is what you then play, from then on - if everything is done right then that should be far superior to just straight playback of the record ...
Last edited:
I am here at the High-End audio show at the Fortune Hotel in Bangkok. The best LP sound I have ever heard is from LUXMAN's room. I could live with that sound..... IF I never heard High Rez digital direct from masters/down loads. The upper freq distortion on loud passages was still unpleasant on LP and the 2H was always there to add a thickness to the sound. There is a comparative lack of 'see-thru' or clarity to the sound. But still.... best LP to date.... never heard it better. For me...... RIP. Heard the Weiss converter thru headphones.... extremely good.
THx-RNMarsh
THx-RNMarsh
Last edited:
Hi,
It has nothing to do with noise or imperfections of an analogue master. It is said to be intrinsic of the medium as it is:
A quote from the Rife paper I referred to earlier in this thread:
Granted it was written in 2002 but from what you wrote earlier and even allowing for caching on SSD or RAM and far more powerful CPUs than back then, it would still take an awful lot of time to correct (assuming it is at all possible) a small segment of the music to get started.
So, the question remains, what happens?
As far as I can tell nobody noticed any changes in the digital format when transferring data from say a HDD to a CD or DVD. If it would it could be easily spotted.
So this system works flawlessly for the IT guys but when the 1 and 0 strings contain music this does no longer work?
That's just not possible so, one more, what goes wrong and more ad hoc how and where does it go wrong?
This is assuming Mr. Rife is correct, something you seem to acknowledge.
Cheers,
It has nothing to do with noise or imperfections of an analogue master. It is said to be intrinsic of the medium as it is:
A quote from the Rife paper I referred to earlier in this thread:
Granted it was written in 2002 but from what you wrote earlier and even allowing for caching on SSD or RAM and far more powerful CPUs than back then, it would still take an awful lot of time to correct (assuming it is at all possible) a small segment of the music to get started.
So, the question remains, what happens?
As far as I can tell nobody noticed any changes in the digital format when transferring data from say a HDD to a CD or DVD. If it would it could be easily spotted.
So this system works flawlessly for the IT guys but when the 1 and 0 strings contain music this does no longer work?
That's just not possible so, one more, what goes wrong and more ad hoc how and where does it go wrong?
This is assuming Mr. Rife is correct, something you seem to acknowledge.
Cheers,
Time smearing, as Rife apparently is talking about it, is just phase shift; if one is paranoid about such - which I'm not - and you know you're going to put the source through a replay mechanism which introduces "excess" phase shift, then just pre-phase shift the material, in the "other direction". DSP allows you to do "anything" - so if you know the replay system is going to "distort" in a systematic way, then just apply "negative" distortion to the track - a "twist" is countered by a reverse twist applied beforehand.
While music remains digital everything is hunky dory - but at some stage prior to hitting the speaker drivers that data has to become analogue in nature - and that's when the poo hits the fan ...
While music remains digital everything is hunky dory - but at some stage prior to hitting the speaker drivers that data has to become analogue in nature - and that's when the poo hits the fan ...
Would that possibly be using Avid TT? I notice that Avid and Luxman sometimes go together, and in the last Sydney show an Avid TT was part of the only vinyl system that was worthwhile, to my ears ...
Hi,
This is assuming Mr. Rife is correct, something you seem to acknowledge.
Cheers,
I believe he is referring to what we call Group Delay. It can be severe and audible in the record/play processing.... esp where analog is used extensively.
THx-RNMarsh
I didnt pay attention and dont recall TT/arm/cart... didnt ask. I have not heard a TT/LP system in years and so was surprised how good it has become. But still a flawed medium IMO.
The new Sony push to make computer -digital downloads a part of mainstream audio is a positive move with lots of potential.
It all reminds me of analog video systems we used in the past vs digital video of today. I keep thinking... video cant get much better and then it does.... I had to buy a new Samsung curved screen UHD and UHD upconvert disc player because it looks sooooo good. I'll finally upgrade my 1080 plasma.
Didgital audio keeps getting better and has 'clearly' surpassed analog systems. We still need great analog power amps and speakers though. And the front end is still analog (mic etc). The best Current-Mode designs are still moving forward and so analog is not completely dead IMO. And, with memory being cheaper and cheaper, bandwidth and storage can now allow streaming without compression et al. We are in the middle of the second Golden Age of audio.
THx-RNMarsh
Last edited:
Hi,
It is a sort of group delay if you like but he's referring to CD as a source.
The article by Rife is entitled: Theory of Upsampled Digital Audio.
If you Google the title then it's the first one that pops up on the list.
Bottomline , according to Rife, is that the medium itself is hopelessly flawed.
Also, regarding your analogue adventure in BKK, although vinyl pressing quality in general has improved, I too hear quite a number of recordings and playback systems that suffer from blocking distortion.
Cartridges that mistrack high frequency content etc.
Much to my surprise, I incidentally came across some takes on You Tube the other day from a guy using an old Garrard and Decca Supergold cartridge that sounded way better than most of the recordings of high-end systems on You Tube I heard so far.
Allowing for the rather modest gear (a basic little Croft tube pre and a Michaelson TVP1) this came as a nice surprise to me.
Here are some of the guy's recordings:
stopi googli - YouTube
Cheers,
It is a sort of group delay if you like but he's referring to CD as a source.
The article by Rife is entitled: Theory of Upsampled Digital Audio.
If you Google the title then it's the first one that pops up on the list.
Bottomline , according to Rife, is that the medium itself is hopelessly flawed.
Also, regarding your analogue adventure in BKK, although vinyl pressing quality in general has improved, I too hear quite a number of recordings and playback systems that suffer from blocking distortion.
Cartridges that mistrack high frequency content etc.
Much to my surprise, I incidentally came across some takes on You Tube the other day from a guy using an old Garrard and Decca Supergold cartridge that sounded way better than most of the recordings of high-end systems on You Tube I heard so far.
Allowing for the rather modest gear (a basic little Croft tube pre and a Michaelson TVP1) this came as a nice surprise to me.
Here are some of the guy's recordings:
stopi googli - YouTube
Cheers,
A bit confusing at the moment (re dynamics and downstream IMD relationship). But to my ears, the only issue is indeed what I observe as "dynamics". Paralleling the chip, increasing supply voltage to slightly below 9V (with heatsink on top of the chips). And then, like you might have just mentioned, the downstream electronic (I/V, LPF, buffer) is really critical (soundwise) in a way that I don't understand what Physics is on the play. But you mentioned downstream IMD, which is confusing for me.
Assuming that you are right, about the usefulness of this LC, could you show how you made your LC?
Sorry, wrong ...
A nice reference point, Frank. As you say, way better than most of the efforts of pretentious systems that are uploaded. Still a system "sound" which favours certain styles of recording, problems with high frequencies at times - but the aspects of the sound which are got right certainly show a clean pair of heels to much of the YouTube material.Much to my surprise, I incidentally came across some takes on You Tube the other day from a guy using an old Garrard and Decca Supergold cartridge that sounded way better than most of the recordings of high-end systems on You Tube I heard so far.
Allowing for the rather modest gear (a basic little Croft tube pre and a Michaelson TVP1) this came as a nice surprise to me.
Here are some of the guy's recordings:
stopi googli - YouTube
Cheers,
Surprised that the Duke Ellington does not come across well ...
It was the brick wall filter in non-upsampling DACs that was often accused of being the cause of all kinds of nastiness: time smear (there you have it again), phase anomalies and what have you.
There's an interesting paper on all that by Doug Rife of DRA Labs. (Search for
Upsampling Theory Rev. 2.pdf.
The most freaky part is that, apparently, the data on the medium already is time smeared, never to be recovered. Quite shocking, no idea if it's actually true.
Cheers,
This is why its so easy to sell wacko digital ideas to the uninitiated. It seems few understand it and many want digital to be intrinsically bad and unrecoverable. Unfortunately what was written is not particularly defensible and not very correct. The math behind upsampling etc. is quite solid and works. Issues when they are there have more to do with implementation than the underlying math or technology.
One can fuss about pre-ringing etc. but having played with a setup to test for ringing it seems the content necessary to trigger that ring in any form is largely non-existent in actual audio recordings. Things like the ringing come from the math and can't be wished away. Dispensing with the reconstruction filter is asking for trouble. -3dB at 60 KHz on a non oversampling DAC means a lot of the folded energy is coming through. If the main band is -6 dBFS from 250 Hz to 3 KHz then you will also have an image at 85200 to 87950 with almost the same energy, as in -12 dBFS (-6 in the original and -6 from the low pass filter). Probably what has rescued most users is the limited frequency response of the subsequent stages. However that rolloff can be a very nonlinear stage and generate a lot of IM with in band audio. If you like the extra bite, life, texture etc. go for it.
The critical filter that must be present is the antialiasing filter on the input of the ADC. Skip that and you could have a lot of garbage in your recordings. Same for analog tape and ultrasonic beats with the bias. The better machines had filters to protect against this issue.
The phase shift in the low pass filters is not that major in the main audio band. Its only a problem with the very high frequency stuff.
A cutterhead is a whole pile of nonlinearities. With as much as 80 dB of feedback in the mid band. The open loop response of a cutterhead rolls off on either side of its resonance, usually in the mid band. The phase response of the system up to the disk can be pretty squirrely at the ends of the bands. The controversy about HF response and LF response also represents a lot of inconsistent phase shifts at the band ends. Same issues hold for analog tape. John Curl can attest to adding phase correction to a recording channel which was not a common feature of analog recorders, in fact only a few had it.
An unfiltered NOS DAC introduces many more spectral components than were present in the original. Therefore with more tones present (more energy, spread over a wider bandwidth) there will be more error noise caused by IMD. Its this error noise that we perceive as reduced dynamics. Eliminating the components >20kHz reduces the error noise hence improves the dynamics.
The first active component after the DAC has to handle a much wider bandwidth than audio. Lynn Olson made measurements suggesting a 20MHz bandwidth emanating from his PCM63. Given that its difficult to make something transparent which handles only the audio bandwidth, how much harder to get the I/V stage right if its active?
I posted up the schematic in this thread - http://www.diyaudio.com/forums/digi...-real-life-es9023-new-1543-a.html#post4069291
The inductors are off-the-shelf parts from Mouser (Fastron).
- Status
- Not open for further replies.