Vinyl Transfer
I am sure many of these links recently posted are for interest only (accidental stereo...how cool sounding!) and I love them, but as someone earlier said: beware of feature creep, so I would implore limiting the scope to a well defined target.
Having performed both I can assure you there is a big difference between restoring a recording and merely declicking a modern LP. A restoration requires many more features such as vertical vs lateral source selection, channel correlation, various eq settings for different media and label eq differences, de-clicking (to the extreme when piecing together a broken 78) de-hissing, and usually a lot of manual waveform stitching, as well as more I'm sure I've forgotten.
I would encourage limiting the project to minor automatic surface noise reduction of an otherwise high fidelity modern LP. This would limit the DSP functions to dealing with clicks and pops. This would be quite a great product in its own right, and it could of course be expanded by an enterprising individual later to cover more purposes if needed.
One very desirable function would be the identification and labeling of tracks as it seems the Sugar Cube has implemented. If code can be licensed from Gracenote (Music Recognition (Music) | Gracenote delivers music recognition technology and metadata) or Rovi (Rovi Music) for a reasonable cost, it would be great.
I mention this issue for two reasons, one personal and one related to my radio station: I would love to transfer more vinyl at home (2500 LPs) but in my current situation have multiple jobs and very little time to agonize over the housekeeping issues (tracking, labeling, metadata) associated with doing so.
Also, if this device could be actually integrated with decent software I could train students to use it and gradually transfer some of the 20,000 LPs in WXYC's library to forestall further degradation. The approaches used by many radio stations are barbaric, usually involving a crappy USB turntable and free student labor to track, label and save the files, complete with terrible spelin and zero additional metadata such as liner notes which I view as critical to anyone interested in the music. Even if done using decent equipment many of the time and metadata limitations remain.
Maybe this is outside the scope of what George and Scott were thinking when this thread was started, which brings me back to my initial statement regarding scope creep...
LOLOL...?
Cheers!
Howie
Howard Hoyt
CE - WXYC-FM 89.3
UNC CHapel Hill, NC
WXYC Chapel Hill, North Carolina - 89.3 FM
I am sure many of these links recently posted are for interest only (accidental stereo...how cool sounding!) and I love them, but as someone earlier said: beware of feature creep, so I would implore limiting the scope to a well defined target.
Having performed both I can assure you there is a big difference between restoring a recording and merely declicking a modern LP. A restoration requires many more features such as vertical vs lateral source selection, channel correlation, various eq settings for different media and label eq differences, de-clicking (to the extreme when piecing together a broken 78) de-hissing, and usually a lot of manual waveform stitching, as well as more I'm sure I've forgotten.
I would encourage limiting the project to minor automatic surface noise reduction of an otherwise high fidelity modern LP. This would limit the DSP functions to dealing with clicks and pops. This would be quite a great product in its own right, and it could of course be expanded by an enterprising individual later to cover more purposes if needed.
One very desirable function would be the identification and labeling of tracks as it seems the Sugar Cube has implemented. If code can be licensed from Gracenote (Music Recognition (Music) | Gracenote delivers music recognition technology and metadata) or Rovi (Rovi Music) for a reasonable cost, it would be great.
I mention this issue for two reasons, one personal and one related to my radio station: I would love to transfer more vinyl at home (2500 LPs) but in my current situation have multiple jobs and very little time to agonize over the housekeeping issues (tracking, labeling, metadata) associated with doing so.
Also, if this device could be actually integrated with decent software I could train students to use it and gradually transfer some of the 20,000 LPs in WXYC's library to forestall further degradation. The approaches used by many radio stations are barbaric, usually involving a crappy USB turntable and free student labor to track, label and save the files, complete with terrible spelin and zero additional metadata such as liner notes which I view as critical to anyone interested in the music. Even if done using decent equipment many of the time and metadata limitations remain.
Maybe this is outside the scope of what George and Scott were thinking when this thread was started, which brings me back to my initial statement regarding scope creep...
LOLOL...?Cheers!
Howie
Howard Hoyt
CE - WXYC-FM 89.3
UNC CHapel Hill, NC
WXYC Chapel Hill, North Carolina - 89.3 FM
I was just checking the purevinyl specs CHANNEL D - Pure Vinyl that doesn't do the recognition that sugar cube claims. If they can actually do it, then they have a USP.
I agree on the feature creep. But it is useful to explore what can be done and the limits of free or affordable software vs the tesla priced pro options.
I agree on the feature creep. But it is useful to explore what can be done and the limits of free or affordable software vs the tesla priced pro options.
Some may find this article on digitizing vinyl useful: Computer Audiophile - Guide to Converting Analog Vinyl To Digital Files Using Windows
A follow on article: Computer Audiophile - Guide to Converting Analog Vinyl to Digital Files Using Macintosh
While not strictly DIY, I hope it provides some insights to the software used, including Declick and Decrackle. A comment made by one of the posters suggested Vinyl Studio: Record LPs and Tapes to CD and MP3 on your PC or Macintosh which several folks feel is the best value for the money. I reviewed it in the second article linked above.
The fun bit is downloading the digitized samples in the first article to give a listen. Towards the end of the first article is a blind listening test on comparing a digitized, declicked and decrackled version of the LP to a CD-ROM rip of the same master. Can anyone tell which one is which?
All in good fun - enjoy your vinyl!
A follow on article: Computer Audiophile - Guide to Converting Analog Vinyl to Digital Files Using Macintosh
While not strictly DIY, I hope it provides some insights to the software used, including Declick and Decrackle. A comment made by one of the posters suggested Vinyl Studio: Record LPs and Tapes to CD and MP3 on your PC or Macintosh which several folks feel is the best value for the money. I reviewed it in the second article linked above.
The fun bit is downloading the digitized samples in the first article to give a listen. Towards the end of the first article is a blind listening test on comparing a digitized, declicked and decrackled version of the LP to a CD-ROM rip of the same master. Can anyone tell which one is which?
All in good fun - enjoy your vinyl!
Vinyl digitizing
Thanks for the links and all the work to write those articles, Mitch!
I think I've used most of the available software made between 1988 and 2005, hadn't had the time to look back into it since then, so I appreciate the share.
Cheers!
Howie
Howard Hoyt
CE - WXYC-FM 89.3
UNC Chapel Hill, NC
diyAudio server HTTPS page
Thanks for the links and all the work to write those articles, Mitch!
I think I've used most of the available software made between 1988 and 2005, hadn't had the time to look back into it since then, so I appreciate the share.
Cheers!
Howie
Howard Hoyt
CE - WXYC-FM 89.3
UNC Chapel Hill, NC
diyAudio server HTTPS page
I think this Audio Diffmaker could do something like that. It compensates for the two files having slightly different sample rates, so perhaps two analog plays (especially on the same TT) would be close enough time-wise for this thing to compensate/track them.
Actually it only generates a difference signal, so it's not directly applicable, but it shows that something like this is possible.
http://www.libinst.com/Audio DiffMaker.htm
1audio, why do you say it would not work with an outboard box?
Last edited:
I think its a typo. What I mean't was that you could not do it in real-time with a small DSP. Its conceivable that you could capture a track several times, possibly including in reverse and using correlation putt the content out of random noise etc. I would reserve efforts like this for only really important recordings.
Feature creep is the real enemy. If this is kept simple it becomes something that can be done pretty easily and quickly. If the hooks are left open in the hardware additional things can be added. What I outlined above could be used with a good DAC and may be a valid option to a conventional phono preamp and usable in real time.
It could be attached to a raspberry Pi and transmit wirelessly to multiple endpoints as one direction. Or to a PC for digital archiving. Or as I noted above used with microphones, tape heads or who knows what other transducers. As features get pushed in they can easily narrow the utility of the final product. or they are software tweaks to push in and hardware hooks already provided.
With a little help from Scott on the specifics of the algorithms this could turn around into prototypes in a few weeks with a dedicated team and $$. Right now its a sort of dream catcher collecting ideas but not getting clearer.
I think stand alone hardware that just works will be more useful long term than hardware that is dependent on someone elses software.
It could be attached to a raspberry Pi and transmit wirelessly to multiple endpoints as one direction. Or to a PC for digital archiving. Or as I noted above used with microphones, tape heads or who knows what other transducers. As features get pushed in they can easily narrow the utility of the final product. or they are software tweaks to push in and hardware hooks already provided.
With a little help from Scott on the specifics of the algorithms this could turn around into prototypes in a few weeks with a dedicated team and $$. Right now its a sort of dream catcher collecting ideas but not getting clearer.
I think stand alone hardware that just works will be more useful long term than hardware that is dependent on someone elses software.
The Izotope RX5 standard package is currently on sale at $249. That's in the territory of acceptable if it works as well as you say.
I should note at this point that my opinions should not be taken into account in the specification phase. My medium term goal is to get a flat pre into miniDSP with a parallel feed to a PC for ripping. Anything more than that is icing and cherry for me.
I should note at this point that my opinions should not be taken into account in the specification phase. My medium term goal is to get a flat pre into miniDSP with a parallel feed to a PC for ripping. Anything more than that is icing and cherry for me.
IMO no all-in-one box at a $300 price point is going to challenge software written by people whose job is to do this at the highest level for commercial returns (writing DSP software is certainly not mine). I think possibly unrealistic expectations can be as bad as feature creep. On the software side the point of my article was to show how small a subset of signal processing was needed for a simple LP equalization by IIR or FIR filter.
Every time I read Mark's comments I wonder what we would have if all the industry operated to his standards.
Demian
Was it the straight through ADC that didn’t work well or was it the DSP equalization that didn’t do?
I searched the net for a commercial product containing one of these chips and the only one I found was this
Brik Audio AP-32H: Qobuzism for this DAC with Headphone Amp and digitization of analog sources
At 370-400 Euros it is a very good price but I think they don’t make use of the equalization capabilities of the ES9112
Maybe this is outside the scope of what George and Scott were thinking when this thread was started, which brings me back to my initial statement regarding scope creep...
Howie, thank you for contributing here your valuable experience.
I don’t have a saying on how the thread develops. I am happy with watching the creek flow and throwing a leave in it from time to time . I enjoy the new info I read.
I understand the need of many for declicking, denoising and metadata retrieval but I am interested on the backbone of the idea: flat preamplifier, DSP RIAA equalization, how to best achieve it (from an amateur standpoint) and acoustic comparison with the traditional way of doing the LP rip (analog RIAA and digitization).
As for my needs, bill said it there:
Thank you Mitch for linking to your very good work.
As to your quiz
http://audio.computeraudiophile.com/2012/1029/Which-one-is-which.wav
I can’t discriminate one from the other (to be expected
).George
This thread is getting costly. In researching the restoration part I keep finding albums I want to buy and I probably have only a toe in the water on the labels who do this*. Latest find was West hill radio archives. I can't find any information about who they actually are, but they have some great previously unreleased stuff.
*Gramophone also are very bad for putting me onto stuff I need.
*Gramophone also are very bad for putting me onto stuff I need.
please look at new RME ADI 2 Pro offering https://www.rme-audio.de/en/products/adi_2-pro.php - this Is what I would like to use, If money are not object
Last edited:
I may have spoken too soon, but DirectX was never my strong point, but I have found another candidate for examining which is audiolava https://acondigital.com/store/. Not sure if this needs a parallel thread, but maybe we should get a reference file to test all these declickers and denoisers so we can compare and contrast? Most of them have a free trial.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.