Well I find it 'interesting' that this GD takes such prominence in this application. If, as we seem to think, the object is to cut down on infra sound that does not contribute to the audible illusion, why would the GD of such a filter be audible at all?
I'm not saying GD is not audible in carefully controlled A-B tests, but that's no reason to worry about it.
For me it is a bit like absolute phase. In carefully controlled tests, some people can reliably hear a difference. But set up a stereo in reversed absolute phase without telling anybody and I bet nobody will find it out without special tests or measurements.
Jan
Last edited:
Well I would think if it was that simple, it would have been implemented in a $ 15,000 player that has been on the market for 10+ years. My bet, but cannot proof yet, is that it is partly airborne dust.
Electrostatic charge on a record should be very easy to fix in the player I believe.
Jan
Alan, it is a very strange conclusion, on my point of view. And I believe calibration tapes are not the good material to figure-out all the problems with magnetic tapes.
Take for granted that i used to calibrate my analog recording machines all the mornings (demagnetization of the heads, polarization, EQ).
Even in one day of wear with re-recordings, you can feel a loss of dynamic in your drums kit tracks, that we record first, most of the time.
(In France, we practiced a lot of re-recordings for recordings of popular singers with professional studio musicians. First rhythmic, then guitars, then soloist or choruses, then all the added decorative stuff like synthesizers, percussion, etc... and the voices alone at the end.)
When it was possible to use two synchronized multi tracks, i used to record bass + battery (and all other instruments from the first session)on one machine, and make a premix copy on them on the second one, that i used alone for the re-recordings. Back to the original at the mixing sessions.
Too, age has various effects on tapes, depending of the way they were manufactured and their brands. I have some old tapes in much better shape than others.
I can tell-you than ALL our DATs we used in our movie post production facility to stock our sound effects libraries quickly deteriorate with age (lot of digital errors, and even no read). In such a way that we had to hire somebody to groove them in urgency on CD and HD when we realized the problem.
Last edited:
What if you and Demian got together and used one of those distortion
analyzers ShibaSoku 729D or Panasonic 7722a as a front end for the premap filtering out the GD you want? It digitizes (ADC), categories, filters,
switches, measures, and can't you undigitize (DAC) by picking off
the proper relays and run it through a box DAC to the pre?
Esperado, I'd have to confirm what you are saying...I find Waltzingbears comment starteling,
however I am not saying that the research or observations aren't correct too.
Also I'm sure only buying and using ONE BRAND of TAPE, and only using it,
makes things last a lot longer on the best Machines and keeps noise to minimum.
Once different tapes are introduced with their different chemical compositions
take their toll on equipment, tape, recording and playback.
Last edited:
Bcarso, Brinely, Kindhornman,
We all know how these guys make money on the projects.
Set the scope, one review, one change, one final.
Everything else is an addition/revision/out of scope item
schedule slips...and round and round we go. Eisenhower
coined the turn Military-Industral Complex and feared
it with good reason. But maybe I'm Jaded.
We all know how these guys make money on the projects.
Set the scope, one review, one change, one final.
Everything else is an addition/revision/out of scope item
schedule slips...and round and round we go. Eisenhower
coined the turn Military-Industral Complex and feared
it with good reason. But maybe I'm Jaded.
The cartridge/arm mechanical resonance is a min phase natural resonant system, which will incur the same group delay as any min phase filter with that exact response, be it electronic or mechanical.
What you have in the plot is a composite of a number of filters, some are mechanical and some are electronic. So long as they are all min phase, which they probably are, it's possible to work backwards from the composite amplitude response in the plot to find the phase response, and thence to find the overall group delay. Know one, know them all.
Because it alters the arrival time of the 'thump' that you feel as it hits your chest, for example.
OK, but I don't have the expertise to do that. So basically my question is: with all those resonances and filters and what have you at the lower end of the band, would one more (the HP for the rumble) make a difference in the whole scheme of things here?
Is it sensible to reinforce one link in the chain if you have no control over the strength of the dozens of others that make up the chain?
Jan
This question of the resonance and inertia of the head+arm assembly in a LP turntable is very interesting. I was on this, with a lot of other questions 40 years ago.
Nowadays, to be honest, this question interested me as much as to try to improve the suspension of cart horses.
The only vinyls still in my listening room are those with an interesting cover artwork. Like this one:
(I cannot listen to Jethro Tull music more than 10 minutes without falling in a deep sleep dreaming I'm a heron ;-)
Nowadays, to be honest, this question interested me as much as to try to improve the suspension of cart horses.
The only vinyls still in my listening room are those with an interesting cover artwork. Like this one:
(I cannot listen to Jethro Tull music more than 10 minutes without falling in a deep sleep dreaming I'm a heron ;-)
Last edited:
Some are reading what they want out of the references and some of the references are sloppy in the use of GD vs. GD "distortion". One of Dick's references does discuss the fact that things like tapehead behavior at the low end and other mechanical issues are by nature minimum phase and have already messed things up. Same with arm cartridge resonance, plot the GD of a 10Hz 6dB peaking filter and see what it is at 20Hz.
EDIT - Missed the other answer.
Last edited:
We all know the sensation of subsonics, the thump in the chest is part of the total experience, and yes its arrival time matters.
yes, better have-it the same day ;-)
I remember the "Sensurround" effect during the "Earthquake" movie. Very low frequencies, uncorrelated with any sound in the audio sound track.
Last edited:
note: I specifically limited my comments to long term magnetic changes, on purpose.
the oft cited loss of initial "immediacy" is limited to probably only a few hours or a day, and no one to my knowledge has been able to quantify it as taking place in reality. I certainly have experienced the perception as well, but cannot say it was real or not.
Print thru is a potential problem, less so with back-coated tapes as the spacing is larger between layers. There is a way to deal with that called skimming by Studer and implemented in their A820 decks, it is a small amount of bias (say 5%, don't remember exact level) that is applied to the tape. Their research showed this to reduce only the surface magnetization which is where the print image resides and not change the deeper signal. But who is going to try that on their masters? Not me.
"Print thru" from the LP is a pressing problem primarily, complicated by the side timing (more land area, less chance). It can also be identified by jumping the needle one track vs tape print thru which will not be correlated with the previous track. Tape print thru timing will change with tape speed and position within the reel.
All of which gets us back to these are mechanical systems that have huge complications. Now, I love working on tape decks and cutting systems. I do it for a living. But I realize they are limited systems that only provide us with an approximation of what originally happened.
Alan
the oft cited loss of initial "immediacy" is limited to probably only a few hours or a day, and no one to my knowledge has been able to quantify it as taking place in reality. I certainly have experienced the perception as well, but cannot say it was real or not.
Print thru is a potential problem, less so with back-coated tapes as the spacing is larger between layers. There is a way to deal with that called skimming by Studer and implemented in their A820 decks, it is a small amount of bias (say 5%, don't remember exact level) that is applied to the tape. Their research showed this to reduce only the surface magnetization which is where the print image resides and not change the deeper signal. But who is going to try that on their masters? Not me.
"Print thru" from the LP is a pressing problem primarily, complicated by the side timing (more land area, less chance). It can also be identified by jumping the needle one track vs tape print thru which will not be correlated with the previous track. Tape print thru timing will change with tape speed and position within the reel.
All of which gets us back to these are mechanical systems that have huge complications. Now, I love working on tape decks and cutting systems. I do it for a living. But I realize they are limited systems that only provide us with an approximation of what originally happened.
Alan
Yes that's similar to what jcx proposed a while ago, certainly viable. I was thinking doing the 75us time constant would take a smaller cap and the component fetishists would be happier.
I had a thought last night (just a thought experiment), flip Brad's comment around. If you do the de-emphasis in analog don't the highs get short changed on bits? Digitizing straight in you have pre-emphasis of the highs and large oversampling of the lows, I would think a good thing?
that does seem plausible - the digital low pass filtering should generate "extra" effective bits a ways out on its attenuation slope from the "averaging"
but if the comparison is with a premium "24 bit " audio ADC you should easily get 18+ bits with current flagship monolithic audio delta sigma ADC
clearly there are a few ways to go: 2 different corner freq LP, shelving or full RIAA - I think the flat pre, all digital eq gives up some overload margin compared to some low pass in the preamp
but if the comparison is with a premium "24 bit " audio ADC you should easily get 18+ bits with current flagship monolithic audio delta sigma ADC
clearly there are a few ways to go: 2 different corner freq LP, shelving or full RIAA - I think the flat pre, all digital eq gives up some overload margin compared to some low pass in the preamp
I can see where this is going: optimal filtering for digitization. It begins to remind one of Keith and Pflash's high definition compatible digital. It will require a suitable name and abbreviation. At some point of course the filter will have more complexity than the inverse RIAA.
I'm not finding a lot of material on the real resolution with 24 bit ADCs, except for the super-slow-dog ones suitable for non-audio use. I thought someone in here would know off the top of her/his head. Do I recall that Grimm was working on an extravagant ADC with assured linearity and absence of missing codes? I might just have dreamt it.
EDIT: posted before seeing jcx post preceding.
Last edited:
- Status
- Not open for further replies.