Vinyl Asylum

Welcome Licorice Pizza (LP) lovers! Setup guides and Vinyl FAQ.

Return to Vinyl Asylum

Message Sort: Post Order or Asylum Reverse Threaded

Digitizing and archiving vinyl

96.248.86.126
Posted on June 25, 2016 at 12:43:58
Dave Garretson
Audiophile

Posts: 2448
Joined: June 14, 2005
 I'm torn about whether to go down this path, but it's certainly tempting to start shrinking my LP footprint sometime before my wife or daughter buries or cremates them with me. I'm interested in this only if it approaches all-analog performance at the top of the spectrum.

The alternatives I'm considering are (1)using my Pass XP-25 to front-end a Lynx HiLo AD converter with USB or SPDIF output to CPU, or (2) buy into the Channel D construct, which in its premier model replaces a conventional phono stage with its own device that performs RIAA equalization in the digital domain and works in conjunction with the Lynx HiLo for the AD conversion.

Channel D/Pure Vinyl's position seems to be that the benefits of digital equalization are an even greater determinant of a positive outcome than even the AD conversion. On the other hand, if this is so, why don't more TOTL phono stages offer digital EQ?

Since both approaches share a Lynx HiLo, I suppose the logical first step is to try one out with my current phono stage.

Any ideas or experiences here? Maybe a direct comparison between Channel D and a top conventional phono stage into the Lynx or other AD converters?

 

Hide full thread outline!
    ...
RE: Digitizing and archiving vinyl, posted on June 25, 2016 at 12:55:49
skriefal
Audiophile

Posts: 117
Location: SLC, Utah
Joined: June 29, 2007
 Phono stages with digital EQ would sell poorly. Most vinyl lovers want a pure analog path as much as possible. I suspect that most don't really know WHY they want it... but for most it probably comes down to the assumed "evils of digital", or somesuch.

 

I just record LPs from Mac C2500 Record Output to, posted on June 25, 2016 at 16:27:34
oldmkvi
Audiophile

Posts: 10583
Joined: April 12, 2002
 my Sony PCM D100, in 24/96 or DSD 64,
then copy the files to my Computer.
Works great for me.

 

RE: Digitizing and archiving vinyl, posted on June 25, 2016 at 16:30:18
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 I am getting good results with a HRT LineStreamer+ coming off my battery powered Camelot Lancelot Pro through a volume pot.

I am using Vinyl Studio, recording at 24/96,I am really happy with the results. Using the default declicker program I have been able to clean up records that were full of pops and clicks. Remember the early '70s when the oil crisis caused them to use some really crappy vinyl. You just couldn't get a good copy of King Crimson's Red. My buddy had two copies one with a decent side one, the other for side two. After returning numerous copies, I kept one that was so-so, now digitized it is quite, listenable.

Do a search there has been lots of discussion on this.

 

Nothing wrong with digital, posted on June 25, 2016 at 18:15:29
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 If anything, by all measurable metrics digital is by far superior. The reason analog sounds better in my estimation is that it is the only medium now a days that is designed specifically for two channel dedicated listening. The audio engineer isn't worried about limiting dynamics so it won't blow your ear drums out with headphones or compete with road noise in a car. The audio engineer has the luxury of just focusing on a stereo environment in a listening room etc... I have some vinyl that was clearly just the vinyl version of the digital release and it sounds terrible. Conversely, some of my ripped vinyl that I've listened to in the car sounds terrible.

I don't have any experience with the gear you mentioned so I can't comment. Just be sure the input impedance for the A/D converter is a decent match to your phono amp or whatever amplification system you use. I like the idea of the RIAA in the digital domain but have no experience with that either.

My only point in answering your post is to suggest that you go for it. If you put together a decent A/D system you will not be able to distinguish it from the real analog.

Don't get me wrong, I love vinyl. And an all analogue path is really appealing to me, but just because I think its cool, not that i think it sounds better. And I don't think you are going to get a better sound out of digital, but again that is only because digital audio engineers are focused on convenience/ear buds/cars/memory etc... If a record is engineered for vinyl it will certainly sound better than the same recording engineered to satisfy the lowest common denominator of all those other competing requirements that digital sees.

Anyway, just my thoughts, I'm hardly an expert.

Nate


You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

RE: Digitizing and archiving vinyl, posted on June 25, 2016 at 21:35:46
BruceS
Audiophile

Posts: 119
Location: Australia
Joined: March 18, 2004
 Of course with the channel D/Pure Vinyl approach you can have your choice - digital equalisation or the usual analogue.
I wonder if any one has done a comparison of both on the Pure Vinyl and would like to share their findings?
Thanks
Bruce

 

RE: Digitizing and archiving vinyl, posted on June 25, 2016 at 21:48:23
airheadair
Audiophile

Posts: 393
Location: California
Joined: October 18, 2010
 Has anyone heard of Sweet Vinyl? I think their original idea was to get the rights to digitize a lot of old classic vinyl recordings, and sell the high resolution track. But now I think they have released (or will release) an audio component allowing people to do it on their own.

I have no real information beyond this.

 

RE: Historical ramble..., posted on June 26, 2016 at 00:03:40
mr.bear
Audiophile

Posts: 4167
Joined: November 13, 2001
 You are somewhat oversimplifying the equation perhaps. You have to look at history a bit. In 1982, when I first heard any digital sound, it was EVIL! It took like 15 years before I spent any serious time/moolah on my first digital hi-fi equipment. It was enough to make a Bear feel a little... er... bitter. Im satisfied that good digital copies of my perennial favorite records sound about identical to the record (all things being as equal as possible.) But why add additional AD/DA conversions? A well implemented analog phono stage remains the state of the art IMHO. I don't think my Pass phono preamp can be beat, my feeling being that it was designed and *voiced* in a golden age where vinyl was the only source. All playback systems acquire a voice, some slight flavor of the sound chosen by the designer as result of component choices, circuit topology, etc etc. I used to think digital sound was terrible, and for a while it was. But I still keep my analog all-analog for simplicity. And It know it sounds good.

 

RE: Digitizing and archiving vinyl, posted on June 26, 2016 at 05:09:42
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 If your goal is to make digital recordings of vinyl that sound as close as possible to what you're hearing from your system, my advice would be to use your vinyl front-end as is rather than switching to a digital phono stage. If you haven't already bought the Lynx HiLo AD converter, you might consider a TASCAM DA-3000 digital recorder. That's what I use and it seems to be very accurate. It will record in PCM up to 24/192 as well as DSD(64) and DSD(128). It also can be used as an external DAC, although it doesn't have a USB DAC input. Therefore, I use an April Music Stello U3 converter when I want to connect my computer as a music server. However, it's not really necessary to use a computer music server because it can accept a USB flash drive directly containing both PCM and DSD files. So I use it as a stand-alone digital player, too.

 

Way too much work to be worth the result.., posted on June 26, 2016 at 08:34:06
Smelly_Socks
 If the only reason s to lower your LP footprint, I would say no way dude.
Unless you know have to move to a much smaller place in the foreseeable future.
I have a small one bedroom apartment and still have 4,000 LPs. I had the same size apartment and had 13,000.. Then I had to move!!
I kept 6,000.. All along two of the walls of my bedroom. (So the living room is not cluttered.)

But on the other hand, for selling LPs, now IS the time. I doubt the market could get much better than it is now.

 

Recording vinyl is definitely a lot of work..., posted on June 26, 2016 at 10:08:15
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 However, that's what makes it worth it to me.

Playing records is lot of work compared to playing a digital file. Therefore, if the two sound identical, it's well worth the trouble to record the record for the convenience of playing it over and over again in the future. The only caveat is whether or not the digital copy sounds like the LP. In my system they sound identical to me; therefore, I prefer to play digital copies rather than going to all the trouble of playing records.

To each his own!

Best regards,
John Elison

 

RE: Way too much work to be worth the result.., posted on June 26, 2016 at 10:47:38
Dave Garretson
Audiophile

Posts: 2448
Joined: June 14, 2005
 The boxes of 12,0000+ CDs and LPs going into my next move is greater in size and weight than all other possessions. Time to lighten up, provided that digitized vinyl done right gets close to a top analog rig. If so then I can imagine settling on around 1000 LPs and 0 CDs that are keepers as physical objects.

Time and money spent making the transition may be recouped by selling off CDs and vinyl at an organized, deliberate pace, rather than in bulk at the far horizon for pennies on the dollar. At a measured pace it might be a five-year project through Discogs, etc.

 

The most important word in your post is "if"... N/T, posted on June 26, 2016 at 11:44:04
musetap
Audiophile

Posts: 31879
Location: San Francisco
Joined: July 8, 2003
Contributor
  Since: January 28, 2004
 aa
"Once this was all Black Plasma and Imagination"-Michael McClure



 

RE: Recording vinyl is definitely a lot of work..., posted on June 26, 2016 at 12:02:20
Dave Garretson
Audiophile

Posts: 2448
Joined: June 14, 2005
 John, my point of departure toward digitizing has been the realization that through many iterations, my vinyl front end is about as good as it will ever be. The last thing I want is to discover that the effort of digitizing is later cancelled by a significantly better tonearm or cartridge. Did you feel any of that after transitioning to your ART7? BTW, I've lately discovered that the ART7 likes loading below 100R. Currently running 50R.

 

RE: Recording vinyl is definitely a lot of work..., posted on June 26, 2016 at 12:47:20
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 Hi Dave,

I still have digital copies from my Thorens TD-126 with my old MC-2000 and I enjoy them just as much as I enjoy any of my newer recordings from my SOTA / SME V / DL-S1 or ART7. I think any improvements you make from this point on will be small and inconsequential relative to making your digital recordings obsolete.

I'll try lowering the resistance load on my ART7. I'm using 100-ohms currently. I should also have a new Vacuum Tube Audio phono stage in my system within a day or two. I will use it with my CineMag 1131 SUT to replace my Yaqin MS-22B for my SL-1200.

I would love to own the XP-25 for its front-panel controls and two-turntable connectivity. I can't quite afford the price, though. Therefore, I'm hoping the Vacuum Tube Audio PH16 might sound as good or even better than my Pass Labs XOno. I'll let you know.

Thanks,
John Elison

 

RE: Digitizing and archiving vinyl, posted on June 26, 2016 at 13:07:28
cactuscowboy@bresnan.net
Audiophile

Posts: 7738
Location: Powell, Wyoming
Joined: July 23, 2007
 No comment on the equipment you're looking at as I have no experience with it.

I've done thousands of transfers over the past nine years, both for paying customers bringing their records & tapes to me and also for personal use.

My turntables, phono preamps and tape decks are connected to a M-Box Pro and that is fed into my PC. I use Goldwave audio editing software to create the WAV files. As needed, I'll use filters in Goldwave to reduce or eliminate tape hiss. I use "Click Repair" software to clean up transfers from vinyl & shellac discs. All files are CD compatible, PCM signed 16 bit stereo.

I've listened closely and compared my WAV files to original vinyl playback and cannot tell the difference. I've conducted blind tests with friends and family and they can't tell the difference. I've also done tests comparing higher resolution digital files to 16/44 and again, no discernable difference. There are advantages to higher resolution but as far as I'm concerned, a quality difference cannot be heard by normal humans. Of course we all know golden-eared audiophiles who'd argue otherwise.

Most important factors in getting good transfers: Excellent front end (sounds like you have that covered), clean records, quality ADC, correct levels settings and proper use of click & noise filters.

You're likely aware that it's a very time consuming process. Good luck with the project.

 

the assumed "evils of digital" ?, posted on June 26, 2016 at 13:51:17
richardl
Audiophile

Posts: 3555
Joined: September 5, 2002
 SMPS and hi-frequency oscillators? I do not want these in my phono stage. They are fine in my cd player or a digital source I guess since they don't always operate without them. Build a few things and deal with noise problems and you will know why anything with digital and phono in the same box is not too high end.

 

RE: Digitizing and archiving vinyl, posted on June 26, 2016 at 18:01:11
Jabs1542
Audiophile

Posts: 90
Location: Virginia
Joined: December 18, 2014
 I'm going to agree with John, and I have the Lynx Hilo. Remember that you went through a lot of trouble to get a good match between your cartridge and phono-pre, that is a special sound that you will only capture if you record through your existing chain.

I tried the digital phono-pre and it sounded dry and a bit edgy, sort of like poor PCM. My Koetsu > Cinemag SUTs > EAR 834P is a unique sound that I could not reproduce digitally. So I stayed with the analog front end and have recordings that reflect that exact same sound as my analog rig - pretty cool.

 

RE: Digitizing and archiving vinyl, posted on June 26, 2016 at 20:40:59
jllaudio
Audiophile

Posts: 859
Joined: May 27, 2002
Contributor
  Since: May 2, 2007
 I record from my VPI Classic II through a Cinemag 1131 Blue connected to a Graham Slee Reflex pre connected to a Wyred4Sound STP Pre with Stage II mods. The pre-amp is connected to a Standalone Sony CDRW. I take the resultant CD to my computer and up-convert to 192/24 flac files using a program called Cirlinca.

I store the files on a 6TB usb drive and two 4TB backups (too much work put in to not get paranoid on backups).

 

I have recorded LPs using, posted on June 27, 2016 at 00:51:46
EdAInWestOC
Audiophile

Posts: 6828
Location: Glen Burnie, MD USA
Joined: December 18, 2003
 An old Emu 0404 USB and I'm now using a TASCAM UH-7000 digital audio interface. The UH-7000 is an excellent solution if you are going to record 24/192 PCM to your computer. The UH-7000 has superb performance at 24/192 and has a better PCM only A/D converter than the DA-3000 (DA-3000 has a PCM-4202 and the UH-7000 uses a PCM-4220 A/D converter).

From what I have read, the TASCAM UH-7000's use of the PCM-4220 means that the UH-7000 performs better at 24/192 but the DA-3000 includes the ability to record DSD and the UH-7000 is a PCM only unit. I was not interested in recording DSD and was also not interested in an all in one recorder solution like the DA-3000.

The UH-7000 is very reasonably priced and an excellent unit. If you like I can send you sample recordings from the UH-7000 (if that helps your decision making process).

The recording chain that I'm currently using is:

1) Cartridge: Denon DL-103R (retipped with boron cantilever and line contact stylus)
2) Tonearm: OEM Rega RB-300 (Michell technoweight, Incognito wiring and Riggle VTAF)
3) Turntable: Thorens TD-126 IIB (NOS, recapped and recently put into service)
4) Phono Preamp: Liberty Audio B2B-1
5) Line Preamp: Audible Illusions L1 (NOS National 7DJ8 tubes)
6) ICs: Kimber Hero
7) Computer: Toshiba P75-A7200 17" Laptop (Intel i7-4700MQ Processor, 24GB DDR3L 1600MHz RAM, 750GB HD + 480GB SSD storage)
8) NAS: WD My Cloud EX2 w/4TB RAID 1 storage)
9) OS: MS Windows 10
10) DAW SW: Sony Soundforge Audio Studio

Ed




Life is analog...digital is just samples thereof

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 06:49:02
WntrMute2
Audiophile

Posts: 782
Location: Detroit
Joined: September 16, 2002
 I don't mean to be dense but don't you now have 4 copies of the same material? LP and CD as hard copies and 2 digital copies. Just seemail excessive. Not wrong of course.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 07:29:06
jllaudio
Audiophile

Posts: 859
Joined: May 27, 2002
Contributor
  Since: May 2, 2007
 Actually, it's two copies and one original. Copy's only take seconds, storage is cheap. One copy is attached to the same PC as the original. The other backup device is a NAS in another area.

If you ever have the misfortune of living in a service area of a power company like DTE, you can never have enough backups.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 08:06:34
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 I think that having fully balanced hardware when interfacing with the PC is the way to go. The Pass plus the Hilo I would imagine to be a pretty awesome pair.

I've really wanted to audition a Hilo and also a Mytek Stereo192 ADC.

You might consider adding good mic preamps to the chain. This would allow you to fine tune the gain for albums with different recording levels. I'm really loving my Millenia HV-35P's. They can be run off battery but I've never tried it or felt the need so far.

One thing I haven't added to my chain yet is a nice mastering eq for doing system correction in the analog domain. Problem is that many of the eq's I might buy, Millenia NSEQ-4 for example, cost more than a decent used car.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 08:30:31
WntrMute2
Audiophile

Posts: 782
Location: Detroit
Joined: September 16, 2002
 I live near Detroit. I have DTE as well. Seems like you have the LP, the CD and 2 backup digital copies. That makes 4 I think.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 10:01:30
jllaudio
Audiophile

Posts: 859
Joined: May 27, 2002
Contributor
  Since: May 2, 2007
 There is only 3 at 192/24. I never keep the CD.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 13:12:56
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 It would be a lot simpler to just get a good adc and go straight to your PC. You could bypass the preamp too, straight off the phono preamp.

Rather than upconvert, digitizing at 24/192 or 24/96 would actually record anything on the record that is above CDs 22k brickwall filters.
Save a couple of unnecessary steps, and give you a better quality recording.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 13:39:57
jllaudio
Audiophile

Posts: 859
Joined: May 27, 2002
Contributor
  Since: May 2, 2007
 Unless the adc is ethernet enabled it will be probably to far from my computer.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 15:53:53
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 How about getting a laptop and putting Vinyl Studio on it. You could then transfer the superior files to your main computer.

The way you are doing it is not giving you the best sound quality, and 24/192 is a waste of storage, by upsampling to it.

Laptops are very reasonably priced these days and if you are willing to record at 24/96 a HRT LineStreamer+ is about $300.

I am so happy with my vinyl transfers using LineStreamer+, Vinyl Studio, and a Toshiba laptop. So often I don't feel like getting up every 15-25 minutes to change the LP, I can program in a couple of hours of music and just kick back.
Another cheaper option would be to run long ICs from your phono preamp to the LineStreamer. How far away is it? Or there may be a way to do it with a converter from usb to Ethernet.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 16:14:08
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 What about the TASCAM UH-7000 that Ed recommended? Apparently the price has been cut in half from $800 to $400. This would allow copying at 24/192. It seems like a better deal to me than the HRT LineStreamer. What do you think?

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 18:24:01
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 No idea, but I did look at it, interesting.

It does only offer XLR inputs and outputs. I don't know if that would present a problem for jll, his listed phono preamp only has single ended outputs. Which is too bad as balanced cables are better suited for long runs. Maybe Ed can answer those questions.

I'm sure it is a better deal, at $400. Better suited, let's hope Ed answers that.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 19:07:33
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 >It does only offer XLR inputs and outputs.

Ah, you're right! I didn't look at it that closely.

My system is fully balanced presently, so it would work for me, but I already have a TASCAM DA-3000. When I switch to tubes, my system will be unbalanced. Fortunately, my TASCAM DA-3000 has both balanced and unbalanced I/O so I'm in good shape. Moreover, I really like the sound of DSD so I now do all my vinyl recording in DSD(128).

Thanks,
John Elison

 

Late to this party, posted on June 27, 2016 at 19:31:40
zelig
Audiophile

Posts: 117
Location: Connecticut
Joined: March 7, 2001



I had the same issue and over the past couple of years have digitised some 2400 lps, mainly rock and roll, blues and jazz and a small part of my classical collection which I will start on in the winter.

My front end was and is a SOTA Star with SME V, MB SL Glider, Dave's SUT, ARC PH3SE phone pre into a PCI card RME Hammerfell 9632 for the A to D. Ripped at 192/24 and the PCM files then processed into FLAC files using AlpineSoft VinylStudio software which has metadata integrated and pulled from Discogs. The project started with a different cart that unfortunately failed halfway through and was retipped by Soundsmith. In retrospect, I wished I had that cart through the entire process.

WAV files are stored on a Synology NAS with 4x4TB HDD in hybrid 5 RAID.

Your proposed setup 1 sounds good.

It took about 285 of 10+ hour man days for the 2400 lps buy YMMV. My experience after the great traverse is that for me, with a collection beginning in 1965 through about 1998, the following are massively critical and maybe more so than the brand of A to D converter:

Clean records which have been been treated for static
Perfect arm and cartridge alignment
Clean stylus tip
Proper loading of the cartridge
Proper setting of recording levels and avoiding overload due to mismatched cartrige/SUT (phono pre) and pre.

The equipment selection is the easy part.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 19:57:09
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 John, can you edit out pops, clicks, and the lead in, and noise at the end of the record?
I suppose at the end if you are right there, you can catch it. I use a Q-Up so I can do other stuff without the needle running on.

I was interested in the Decware version of your DA-3000, but it was over my budget and too limiting for what I wanted to do. I still have 5 crates of LPs waiting for me to digitize. If it would of had usb input/ouput I may have bought one, instead of the used Benchmark Dac2 HGC and the HRT LineStreamer+.

I was new to PC audio and never considered a usb to coax converter.

 

RE: Digitizing and archiving vinyl, posted on June 27, 2016 at 20:39:43
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 The problem with DSD is you can not edit it. However, you can split it into individual songs and you can crop each side. Therefore, you can cut off the beginning that contains the noise from the stylus dropping on the record and you can cut off the end so it's not important if the stylus rides the run-out grove for five minutes at then end.

The DA-3000 does accept a USB flash drive, which can be used to export recorded files from its SDHC memory card or its CF memory card as well as playing music files. I rarely use my computer for streaming anymore because I simply copy the music files I want to play onto a 64-GB USB flash drive and plug it directly into the DA-3000. Then the DA-3000 becomes a stand-alone digital music player.

All recording must be done with either SDHC memory cards, which is what I use, or else CF memory cards. My Toshiba notebook computer has a SDHC port so I can transfer recorded files to my computer by removing the SDHC card and plugging it into my computer, which is very fast and convenient. Then I use the free TASCAM Hi-Res Editor software to crop and split my recordings into individual tracks.

For my purposes, the DA-3000 is an excellent DSD recorder and DSD player.

Best regards,
John Elison

 

What if you don't care about 24/192?, posted on June 28, 2016 at 09:08:37
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 I'm not convinced we can hear the difference between 192 KHz and say 88.2KHz. In my estimation (which will get you a cup of coffee if you add $1.50) 44 KHz is probably fine and anything above it is just a waste of memory. Perhaps I'm wrong.

Anyway, you said the UH-7000 is better at 24/192. Am I wasting money on this product if I don't plan to go that high?

Thanks!

This is an interesting option.

Nate


You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

The higher sampling rate, posted on June 28, 2016 at 10:40:18
EdAInWestOC
Audiophile

Posts: 6828
Location: Glen Burnie, MD USA
Joined: December 18, 2003
 Has advantages. I put a link below that explains it better than I can.

In short the reason you have a higher sampling rate is to avoid missing relevant content that occurs between sampling windows. Musical content is anything but a predictable waveform and there is relevant information that can be missed when the sampling rate is a low as 44.1kHz.

If music was a sine wave then 44.1 KHz is capable of recording and playing back the sine wave content (up to approximately 20kHz). Unfortunately musical peaks happen wherever they happen and the regular sampling windows can miss information that cannot be reproduced.

The answer is to increase the sampling rate to try and capture all of the content and provide the DAC the data necessary to reconstruct the original musical waveform. Whether or not the music is reproduced faithfully is dependent on the musical content.

Sometimes CD resolution does a very good job of reproducing music. Where it starts to falter is when the musical waveform is complex. That can happen when there is lots of high frequency content and/or lots of out of phase information (like when its a live recording in a lively recording space).

There is also common sense involved. The idea is to record at the highest resolution you can and then you can master whatever you want from the high resolution master. When you have the 24/192 recording you can always make whatever lower resolution copy you want. If you start with a low resolution master you cannot recreate what you don't have.

Ed
Life is analog...digital is just samples thereof

 

RE: The higher sampling rate, posted on June 28, 2016 at 12:03:01
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 The reason I don't own a UH-7000 is this review. Apparently there is some higher levels of noise built into the design and limiting the achievable noise performance.

Not that I think this makes them unusable or anything but it seems interesting to me. Something I'd imagine potential buyers may want to be aware of.

 

RE: What if you don't care about 24/192?, posted on June 28, 2016 at 12:15:44
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 Some say that if you plan to do certain types of DSP on the files, the capture sample rate ought to be 5-10X the playback sample rate or you'll"suck the analog life out of it". By this measure 192kHz is maybe not quite barely adequate.

I'm no DSP expert and I'm not the one who said it. I have not done any testing to see if I can hear the difference. However this is from a forum where the participants make a living from the captures they create. Most there agree on this.

 

RE: The higher sampling rate, posted on June 28, 2016 at 12:22:52
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 I've copied vinyl at 16/44, 16/48, 24/96 and DSD(128). I liked DSD(128) best and I plan to keep copying vinyl in DSD from now on. I haven't tried 24/192, but I might try that, too.

 

RE: What if you don't care about 24/192?, posted on June 28, 2016 at 14:21:51
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 "However this is from a forum where the participants make a living from the captures they create."

Which forum is that?

 

RE: What if you don't care about 24/192?, posted on June 28, 2016 at 15:02:33
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 I don't intend on doing much processing. I don't even like to normalize because a pop or a loud noise could disrupt the volume on the rest of the track.

I understand that you lose information with the shape of the wave but we wont hear that information anyway. If you have a saw tooth wave for example that is at the upper limits our our hearing, we won't hear the "shape" of that wave because our ears wont detect the saw tooth. I could be wrong but I just can't imagine why we need much higher than what we can hear. I could by 88.2 or 96 just for CYA purposes but why 192?

I could buy the argument that significant processing of the signal may need the higher sample rate to cut down on processing error.

But like I said, I'm hardly an expert just a cynic.

Nate


You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

Noise spikes likely SMPS related , posted on June 28, 2016 at 15:14:34
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 The noise characteristics shown in the spectral analysis suggest SMPS interference being coupled into the analogue input either through the phono cable or the cables to the ADC or possibly an interaction with the phono stage power supply. I observed similar characteristics when playing with different phono stages. The ones using a SMPS exhibit the noise shown in the traces. My reference phono stage has a linear power supply and the noise spikes disappear. The use of Balanced/Starquad XLR improves things and limits the bandwidth of the noise. When using twin conductor shielded RCA cable to XLR (since the iFi only has unbalanced outputs), the noise bandwidth was wider.
My iFi Phono exhibited noise spikes around 20kHz. My MF M1-ViNL (unbalanced output) had spikes around 38kHz with the same wide spectral characteristic.

Noise spikes notwithstanding, they are shown to be at very low amplitude <-100dB so apart from being measurable, they won't be audible. With vinyl playback you would have to have a very state of the art rig to do much better than 75dB S/N for hum and noise (ref 0dBFS / 5cm/s). Typically you would be getting hum breakthrough at -60dB or higher (ref 0dBFS / 5cm/s).

In short, I wouldn't be too concerned about that review.
Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: What if you don't care about 24/192?, posted on June 28, 2016 at 15:40:31
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 Hi Nate

Here is a link to a slightly more rational white paper on the subject with a more detailed mathematical treatment which debunks the myth of needing higher than the Nyquist rate to reproduce a band-limited waveform. You are correct in your assumption that, mathematically, sampling at significantly above the (required) Nyquist frequency does not yield a more accurate representation of the band-limited waveform.
There are a number of factors that make the use of higher sampling rates above 44.1kHz desirable. The main one being implementing practical filter characteristics and reducing time smear. However, the "accuracy" of the band-limited 20kHz waveform will not be improved by sampling at 192kHz or higher (in theory). I say "in theory" because jitter affects the performance of the reconstruction filter and jitter will cause distortions in the waveform.
Of greater importance is the correct use of dither and noise shaping. My down-converted 16/44 files are barely distinguishable from the high resolution versions. There are subtle differences in spatial resolution and depth which I attribute to time smear of the filters rather than a limitation in the word length or sampling frequency. However, it could be argued the sampling frequency IS the reason due to the impulse response of the reconstruction filter...
FWIW, I have settled on 96kHz as the highest sample rate I use. As I use an external DAC, going higher in sample rate causes more issues in the transmission line/interface.

Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

For those into DSD, "double DSD" or higher..., posted on June 28, 2016 at 15:48:59
genungo
 ... (triple DSD! quadruple DSD! The higher the better!!!) is said to be the new Cat's meow - at least according to some over there on the Digital and Computer Audio forums.

 

RE: Digitizing and archiving vinyl, posted on June 28, 2016 at 15:51:17
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 Tascam provide free software for editing the files - you can convert between DSD and PCM. The software is very basic so if you want to declick etc you can convert DSD to PCM then edit in another program like iZotope RX/Advanced before using the Tascam software to go back to DSD.
Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

dither question, posted on June 29, 2016 at 06:43:50
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 Is a dither important for recording/digitizing? Perhaps I was mistaken but I thought a dither was important for processing, but not so much for acquisition. Also, Is the random error generated by the nature of vinyl playback an adequate dither?

Long question short, do you really need a dither if you are just recording?

Thanks!

Nate

You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

RE: Noise spikes likely SMPS related , posted on June 29, 2016 at 07:32:37
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 I don't disagree with any of that.

Many choose to live with the low level noises. Since it is easy to avoid it I choose to. At some level the noise interacts with the signal, can create imd artifacts, unnecessarily works power supply components etc. I have no idea if anyone can hear the difference but at least I rest well knowing it's not there.

BTW the newer MX-ViNL claims to have full balanced path. I'm eyballing that thing. It's noise specs look pretty good but it doesn't seem to have the overload capacity of some of its competitors. I have my own phono pre project going but it just keeps stretching in time so I may buy one until it's done to give me time to do it right.

 

RE: What if you don't care about 24/192?, posted on June 29, 2016 at 07:54:50
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 While I'll likely have great difficulty digging up the exact thread for you since I've been reading alot of this stuff lately, a search will show that there are very many, coincidentally usually owners of ADC's with max sample rate capability somewhere in the 96kHz range, with an aversion to doing any kind of sample rate conversion even if it is theoretically perfectly handled. I've seen a couple speculative posts suggesting working with extremely high sample rates may captures may be the magic bullet. I can't confirm any of this yet since my rig isn't up to my standards level which would allow me to begin my work of ripping things to make comparisons by. It may be that my current goal of very good 192kHz sample rips is still insufficient to achieve acceptable SRC to redbook for the car but at least I'll be able to finally run the experiment for my self.

 

RE: What if you don't care about 24/192?, posted on June 29, 2016 at 08:21:03
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 Slightly more rational than what? :)

It's ironic that I've posted the Lavry link myself recently in a similar discussion.

However, there is quite a bit out there explaining why some feel that higher than what may at first appear to be sufficient Nyquist rates can be useful in some circumstances. The explanation is not that Nyquist/Shannon were wrong but many who have made incorrect assumptions about real world converter hardware filtering. As I understand it, in order to overcome these real world limitations within reasonable budgets is where higher sample rates may start to make sense over more expensive filtering hardware etc when ultra low distortion is necessary.

As I've mentioned earlier, I'm no DSP expert so I'm probably the wrong guy to try and explain it all. I'd end up with some BS about imperfect brick wall filters, finite band limiting is impossible and some hand waving and be totally embarrassed but it is a fairly common topic via search.

I believe this linked white paper may start to hit on some of these subjects. I'll try to remember to ask my buddy at work about this stuff and see what he says. It's all very confusing for me.

Whatever the case it's easy to find many search hits suggesting even 96kHz sampled to 44.1 ruins quality even when "perfect" 64 bit algorithms are utilized.

I have not yet tried to verify any of it myself.

 

RE: dither question, posted on June 29, 2016 at 10:19:56
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 If you use a program designed for digitizing vinyl like Vinyl Studio you won't have to worry about adding dither.

I think for vinyl transfers, if you are going to use declicker software, then record in pcm. That is my opinion anyway.
When I looked into adc, I was limited by budget and knowledge. I wanted the option of 24/192 but was leary of the pro gear as it was made for microphones. While what I found at that time was. I found the adc shootout on Analog Planet, see link. I also was reading a lot about sample rates by Mark Walberg, and while I don't agree with his bits are bits, cables, and Jitterbugs don't matter, I do agree with his choice of 24/96.
The Tascam has the advantage of adjusting levels and if using XLR/RCA adapters aren't a problem, I would probably go that route today.

At the time I found the LineStreamer+ to be ideal for me. I had to add a volume pot inline between the phono preamp and the adc. That added two sets of.5m cables too. I did some listening tests and found tbat the alps pot and cables were transparent to my 61 year old ears.

The sound I'm getting is very good. I have no plans of getting rid of my vinyl, although I got started on the vinyl tranfers because I was given access to 6 large crates of records that had belonged to a friend who died and left his music collection to another friend.

Having the option of playing a bunch of records, without getting up every 15-20 minutes is really nice. Right now I programmed in 5 records and I am doing other stuff while listening to music I wouldn't have taken the time to play. I think you will find that option very enjoyable.
And the way I have it set up I can do transfers while watching TV. When I record I remove the leads from the phono preamp to the main preamp and plug in the ICs going to the volume pot.

When you get your adc try Vinyl Studio, I think they offer a free trial, and either way, it is like around $30 for the program. I bought that and JRiver, I was all set.

 

RE: The higher sampling rate, posted on June 29, 2016 at 10:30:36
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 The lack of processing tools is the only thing keeping me pcm at the moment. I feel like the click repair, normalizing and to a lesser extent in my system the offset removal are pretty important steps I would want to keep doing.

If you just listen to the raw files then maybe dsd is best.

 

input impedance and Mono question, posted on June 29, 2016 at 11:01:28
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 I have the linestreamer+ as well and am quite happy with it. I haven't heard files from higher end AD converters, I've been meaning to check out the ones Fremmer posted but haven't. One thing to consider, I noticed (and if memory serves Fremmer pointed this out as well) the ADC has a low input impedance at 5Kohm. My phono amp's manufacturer suggested much higher at closer to 60Kohm or above. So I bought the Boozhound Labs buffer kit (linked below) that uses high quality parts to provide an easy load to drive. I haven't used it yet, but I am very happy with Boozhound products. I've built 2 phono amps and the head amp and to my ears they are really great values! I want to add the pot as you did and haven't gotten around to figuring which one I want to use. Thanks for the suggestion.

So anyway, you may want to investigate if the output of your phono amp can handle a 5Kohm load, if not I recommend the boozhound even though I haven't actually used it yet.

As far as mono, do you think adding a mono switch or just combining the signal in the digital domain is ideal?

Thanks

Nate
You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

RE: input impedance and Mono question, posted on June 29, 2016 at 14:32:21
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 No idea on the mono, but I have transferred over 100 albums without any issues I am aware of impedance wise. I use a battery powered Camelot Lancelot Pro phono preamp.

Without the volume pot, I was clipping. Honestly between the pot and cables I have no idea what the impedance is. It works, it gives me good digital copies at 24/96. To improve on it would be costly. I got interested when PS Audio dropped the price of their phono preamp that includes adc, but I read some reviews that said it was lean sounding.

 

RE: dither question, posted on June 29, 2016 at 17:53:07
John Elison
Audiophile

Posts: 23900
Location: Central Kentucky
Joined: December 20, 2000
Contributor
  Since: January 29, 2004
 > Is the random error generated by the nature of vinyl playback an adequate dither?

Very good question! I have often theorized this to be the reason that my digital recordings of vinyl sound so accurate. It seems logical, anyway. ;-)

 

RE: dither question, posted on June 29, 2016 at 18:44:16
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 I guess my answer is in danger of being considered in the wrong forum.... but here goes!
Dither is essential to reduce the effects of quantisation noise and especially with bit-depth reduction (for example down converting a 24 bit word to 16 bit word).

As far as recording goes, the user does not have to add dither as a specific process in editing unless reducing the bit-depth, as it is an inherent component of the ADC process. Low bit ADCs utilising Delta-Sigma modulation require noise-shaping in combination with dither of an appropriate amount to de-correlate quantisation error from the signal.

When down converting a high res (say 24/96) down to 16/44 the use of noise-shaping and suitable level of dither enables a result that is very hard to distinguish from the original high res master unless one was specifically trying to "look for the faults".
My point being related to the claim that a high sampling rate and high bit-depth being "essential" to reproduce a 20kHz bandwidth depends more on the ear of the beholder than a fundamental limitation of the minimum sample rate of 44.1kHz.

Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: dither question, posted on June 29, 2016 at 19:01:52
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 Most if not all ADCs in common use are based around delta-sigma modulators in which case dither in combination with noise-shaping is an essential component within the chip to reduce the quantisation distortion depending on the number of bits used in the modulator. Due to RIAA eq, the HF noise is attenuated relative to LF rumble, and the LF noise will not be terribly effective at adding useful dither above a certain frequency depending on the frequency bandwidth required relative to your chosen sample rate.

The very first "digital" (DDD) recordings when CDs were first released were generally thought to be inferior to the ADD transfers. At the time, the tape hiss was thought to be supplying dither to the ADC which were multi-bit.

In principle the PCM4220 in the Tascam UH-7000 would yield superior results to the PCM4202 in the DA-3000 because it uses a 5-bit quantiser rather than a 1-bit quantiser (as in the PCM4202) so ideal levels of dither may be applied. Lipschitz wrote a paper suggesting that the use of DSD for archival recording was fundamentally flawed on this basis and that considerable levels of audible distortion were possible with the 1-bit modulator.

It is worthwhile noting that many commercial DSD recordings (such as those from Linn) are conventional PCM at 24/192 and then converted to DSD.
Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: Noise spikes likely SMPS related , posted on June 29, 2016 at 20:57:38
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 I honestly don't think there is enough information about the test system to determine if those noise spikes originate from the phono stage or introduced by the UH-7000. However, the SNR specification for the UH-7000 would tend to suggest it is not responsible by design although it is curious that it is noticeable after the unit has warmed up which suggests heat related component drift which might "tune into the noise".

The noise spikes will be masked by broadband noise from the vinyl so generally won't be visible on spectral analysis of an actual signal with content up to and beyond 20kHz. The ultrasonic noise will be of little consequence. DSD has way more ultrasonic noise! As to the level of IMD, the cartridge distortion dominates and will be adding 10%+ IMD above about 15kHz for wide bandwidth LOMC.

Yes, I noticed the same thing with the latest MF phono stages like the MX-VNYL! It is because they have opted to go for a 12V supply (or 5V microUSB for the LX-LPS). The overload margin is very low compared to the products of the past which is due to the 12VDC supply - Assuming they have done something similar to the V-LPS, they will just level shift the signal relative to the power supply rails for the active stage and basically have a reduced maximum output voltage. The original phono stages used to use 12VAC and have voltage multipliers to give the ±15V rails thus giving the 27dB or higher overload margins. I have the X-LPS and X-LPSv3 (both of which are heavily modified)plus the previous M1-ViNL which has balanced outputs (but not input). I wouldn't be surprised if it is the same circuit being rehashed in the MX-VYNL which MF have done in the past.

"I have my own phono pre project going but it just keeps stretching in time"

Haha me too! I bought my M1-ViNL for exactly the same reasons as you! There is always something else coming up...
What topology are you considering? Active/passive?
Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: Noise spikes likely SMPS related , posted on June 30, 2016 at 08:03:13
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 "I honestly don't think there is enough information about the test system"

For sure. Sometimes, due to circuit topology or whatever there is only so much which can humanly be done. It's the reason I'm moving away from single ended gear for ripping. My single ended ADC is dead quiet with nothing plugged in. Get the rest of the single ended, moving magnet version of the chain plugged in and power on the preamps and you see -105dB 60Hz poking above the rest of the noise floor while gain is set for reference track at -10dB of ADC full scale. Really fantastic 60Hz noise performance compared with the other single ended moving magnet systems I've compared with, true, but it bugs the hell out of me. Can't really blame the ADC other than having a circuit topology which would physically never allow me to completely achieve my goal of having no 60Hz noise show up in the noise floor while the rest of my system is plugged in and powered up.

Not having any noise spikes in my scans makes me smile more whether I heard them or not. For me it is worth it.

"What topology are you considering? Active/passive?"

The circuit I'm playing with has tons of paralleled Toshiba jfets in the front end. It's a direct coupled design biased in class AB. After many versions of trying other stuff I found that by going full active in my first stage I was able to get massive overload margin ~31dB with only +/-17VDC rails and also reduce the number of total stages required. It's all balanced and true dual mono including fully isolated dual power supplies.

Right now I'm still dealing with an instability caused by power supply sequencing issue during startup. If I could ever drag myself back to my PC to finish the power supply I'm real close to designing my boards.

I'm actually excited about it because I think it will work well. I've been very thorough simulating it, fine tuning the performance. My main problem now is the beautiful, warm, shiny sun outside is very distracting and draws me away from the keyboard.

How about your circuit?

 

"beautiful, warm, shiny sun ...", posted on June 30, 2016 at 08:52:38
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 HA! We have the inverse here in Texas. The burdensome, hot, oppressive sun drives me inside!

9 Months a year the weather here is like Coastal Southern California. 3 Months a year ... not so much. We have just begun those 3 months!

Nate


You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

RE: What if you don't care about 24/192?, posted on June 30, 2016 at 20:41:49
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 The Westcott pdf was the one Ed posted and what my comment was originally aimed at.

His main point is in relation to the "available" bandwidth relative to the theoretical Fs/2 bandwidth when implementing a practical system. However, whilst this is absolutely true, the concept of inadequate "accuracy" when applied to audio recording of vinyl (just to make sure this post is still considered Vinyl related!!) is somewhat misplaced especially in the context of the ADCs available for the audiophile. The Lavry paper provides the mathematical analysis to prove that this is simply not the case.

ADCs such as the PCM4202, 4220 etc utilise sigma-delta modulators running at a very high oversampling ratio dependent on sample rate (for example 128Fs for 44.1kHz, 32Fs at 192kHz) thus reducing the required order of the input buffer filter. The oversampling ratio is, of course, an inherent requirement given the limited number of bits in the quantiser in order to achieve the desired noise performance after the application of noise-shaping.
The PCM4220 datasheet recommends (at a minimum) a first order filter. Obviously for better anti-aliasing performance a higher order may be desirable. However, in my view, this does render the argument presented in the Westcott document somewhat irrelevant to the discussion on optimum sample rate for vinyl transcription. The output word at the desired Fs is the result of the decimation filter and therefore not subject (to the same extent) to the issues that Westcott asserts.

Therefore choosing Fs significantly greater than the required value for the signal offers no advantage beyond the reduction in ringing time for the reconstruction filter of an impulse response. The audibility of the filter impulse response depends on the cumulative effect of ALL filters from recording to the final analogue output stage.

As an extreme example, the Benchmark ADC1 runs the ADC at a sampling rate of 216kHz (IIRC) irrespective of the desired Fs. The desired output sample rate is then the result of passing the data through an SRC. Therefore the "accuracy" of the information is not dependent on the selected sample rate for recording, but more on the SRC algorithm...but that's another subject altogether!
You therefore only need to choose Fs based on the bandwidth requirements of your signal. In principle a high Fs would always seem best, but there are tradeoffs based on the clock accuracy and jitter which Westcott completely ignores.

On to the phono stage, I'm still trying to decide on my approach. To ensure good overload margin, I started off on the Single Gain stage/Active RIAA approach and I started off looking at DC servo based designs. Then I saw the elegantly simple Holman design presented in his paper to the AES - "New Factors in Phonograph Preamplifier Design" AES May 1976, Vol 24, Number 4.



Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: "beautiful, warm, shiny sun ...", posted on July 1, 2016 at 07:36:24
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 It's about perfect here now. We are just entering our 3 dry months. Wouldn't it be great if we trade weather at will?

 

RE: What if you don't care about 24/192?, posted on July 1, 2016 at 07:45:24
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 I actually don't disagree with with any of it. I just don't understand it well enough and can easily be convinced either way depending on whos stuff I read last.

Questions for you. I'm assuming you've also noted the fairly prevalent aversion to doing sample rate conversion by those doing vinyl captures.

How do you explain the very common perception that SRC (even the very theoretically perfect versions) is ruining the sound if it isn't due to numerical error introduced by the number crunching?

Do you think those claiming SRC ruins the sound, from say less than about 3X the playback sample rate, aren't hearing what they think they are hearing???

 

Sample rate conversion, posted on July 1, 2016 at 08:22:25
texanater
Audiophile

Posts: 1513
Location: Houston, TX
Joined: December 16, 2002
 I've always operated under the assumption that sample rate conversion done with integer divisors will introduce less error than non-integer divisors. By that I mean, If I plan to have a finished product of 44.1Khz I'll record at 88.2 rather than 96. Is this overly simplistic and wrong? Is it better to sample at the highest resolution possible regardless of sample rate conversion ratio?

I like to think of my self as a green belt in this kind of topic. I know just enough to get my ass kicked.

Nate


You can't cheat an honest man, never give a sucker an even break or smarten up a chump -- W.C. Fields

 

RE: What if you don't care about 24/192?, posted on July 1, 2016 at 08:57:41
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 Back when I was deciding which sample rate to use recording my lps I had a few brief conversations with the folks at Blue Coast. I was curious about their views on downsampling. Cookie said they could hear it. She felt the best recordings were recorded at the sample rate they would be released. I've not tried to hear the degradation myself but I have not downsampled (other than for the car). I'm sure her experience makes her better qualified than I.

I record at 24/96. I've tried 24/192 but did not hear an improvement. I refuse to give up editing therefore avoid DSD. I use RX to clean up the recordings and especially appreciate the compare feature. It makes hearing the before and after differences much easier.

I use a Tascam DA-3000. One reason for choosing it was sd/cf cards. I did not want usb. My computer and system are on a separate circuits.





 

RE: What if you don't care about 24/192?, posted on July 1, 2016 at 09:05:55
airheadair
Audiophile

Posts: 393
Location: California
Joined: October 18, 2010
 May I ask again if anyone here has heard of Sweet Vinyl, a company which devoted to transcribing vinyl to high-resolution digital?

 

sweet vinyl, posted on July 1, 2016 at 09:31:43
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 I've heard of it. Check Analog Planet for MF's comments. It's around $2,000. It allows recording, retrieving meta data, and declicking. I've no idea how it stacks up to say a Tascam, Vinyl Studio and RX which sell for less money.

 

Sweet vinyl price, posted on July 1, 2016 at 09:45:26
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 Here's a link to another article

 

RE: What if you don't care about 24/192?, posted on July 1, 2016 at 10:23:37
Ugly
Audiophile

Posts: 2912
Location: Des Moines, WA
Joined: August 22, 2006
 Several I've chatted with claim to hear a degradation when sample rate conversion has been used. Since getting high quality sample rates much above 192kHz seems to out of most peoples budgets I've never spoken to anyone to confirm whether higher sample rates, say 384kHz, allows for this type of math to occur without the same degradation.

Another clue to what might be happening may lie in the fact there seems to be a perception that recording with samples happening at integer multiples of the playback rate allows for the SRC to occur with much less damaging results.

Again, I'm not doing anything more than speculating here, but the number of users claiming poor results from src seems a bit to many to just dismiss them.

 

RE: What if you don't care about 24/192?, posted on July 1, 2016 at 10:47:37
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 "...recording with samples happening at integer multiples of the playback rate allows for the SRC to occur with much less damaging results."

I've read that was once the case - such as recording at 24/88 and downsampling to 14/44 - but with today's more powerful computers, not a problem. Again, no personal experience, just reading.

 

I use the pencil tool, posted on July 1, 2016 at 11:46:48
EdAInWestOC
Audiophile

Posts: 6828
Location: Glen Burnie, MD USA
Joined: December 18, 2003
 In Sound Forge to remove the worst pops/clicks from recordings. I zoom in on the waveform where the pencil tools is available, locate the pop/click and just draw the offending pop/click out of the recording (draw across the pop/click).

Its more tedious than applying a filter but it does not effect any other content, other than the pop or click in question. I have done this to a number of recordings and none of them have any negative side effects.

I have tried other methods and they all seem to have negative effects. The manual removal approach is more work but its like your momma told you. Nothing is free.

Ed
Life is analog...digital is just samples thereof

 

RE: I use the pencil tool, posted on July 1, 2016 at 15:00:14
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 Another RX option is to listen to what is being removed. You can hear when more than the click is removed and lower the setting. On a scale of 10, I normally use 1.4. If there is a stubborn click, I isolate the click (plus or minus hundredths of seconds) and use a higher setting.

50% of my listening is through a headphone system. I remove clicks I would not hear through my speakers but do with the headphone system.

Other than click removal, I use a 20 hz hi pass filter to remove the 9hz cart resonance. I wish there was a lower setting than 20 hz.

I did use a hum removal tool until I lowered 60 hz hum. Seeing the spectral data from my recordings has taught me a lot about my system. I had -60 db 60 hz hum. Now it's down to -81 db using a 64 db setting on my Sutherland with a Soundsmith Zephyr MIMC and -3 db setting on the Tascam.

Between tracks I sometimes use other tools. Other than that I think the less I do the better.

 

RE: Sample rate conversion, posted on July 1, 2016 at 23:33:43
Jeff Starr
Audiophile

Posts: 1574
Location: Milwaukee, Wisconsin
Joined: March 4, 2000
 When I got my LineStreamer+ I was going to record at 24/88 so it would be any even conversion to 16/44. I occasionally burn CDs for friends.
I asked over in the PC forum about 88 rather than 96 for my vinyl transfers. A few of the guys said with today's processers, and software it was best to record at the higher bit rate. It makes sense for me as I am not going to be listening to the CDs and none of my friends have decent equipment anyway. In fact I think my friend's wife probably just loads them on her phone as MP3s.

I have a policy though only non current music that is not available or if the artist is dead. I was telling her about how good the Sugar Stems were and she said well make me some copies of those. I told her no, they are a local band, and need the sales. That I would feel like we were stealing from them. She didn't get it, she got huffy and said fine then I just won't listen to it. She doesn't realize that music has a value.

We have two local bands that do power pop, the Sugar Stems and Testa Rosa. I listen to their music and think how can these bands not be nationally famous. Great lyrics with catchy tunes. Testa Rosa's last album is available on Bandcamp in 24/96 for like $8.

 

I usually use headphones, posted on July 2, 2016 at 00:15:17
EdAInWestOC
Audiophile

Posts: 6828
Location: Glen Burnie, MD USA
Joined: December 18, 2003
 I have a set of Stax Lambda Pros (full range estats) that have been excellent for the purpose. Unfortunately the driver for those is on the blink and needs to be repaired. Until then I'll use either Sennheiser HD 580, Sennheiser Momentums or HiFi Man HE 400S driven by a Schiit Lyr 2 amp for the purpose.

For the purpose of finding and editing clicks and pops the Stax were superb and the HiFi Man 400S are respectable. The Sennheiser are OK and I prefer the Momentums over the HD 580s.

I rarely record using my speakers. I have an full sized old office chair for headphone listening and recording. Once I record the LP sides, I split up the sides into tracks and then do whatever editing may be required.

I used to use Cakewalk Sonar but Sound Forge is easy to use and does a good job with all of the tasks. It took a little while to get familiar with Sound Forge, and its editing, but once you get the hang of it its very good with all the tasks.

Sound Forge Editing:

The pencil tool is used to draw across the click or pop or whatever you want to remove. Its obvious and easy to use. You have to drill down to a very close resolution to use the pencil tool (on the upper right corner of the screen shown). The cursor changes to a pencil and whatever you draw overwrites the waveform content being displayed. I highlighted a portion of the right channel waveform to show a candidate to be removed.

I start on the left hand side and draw a straight line across the center line (0db) and the waveform is replaced by the new straight line...and no pop.

I upgraded my DAW and added 16 GB, it now has 24 GB RAM. That speeded up everything nicely. Even better than that was when I added a second SSD drive for recording and virtual memory purposes. My laptop now screams and everything is done very quickly (even under MS Windows 10).

Ed
Life is analog...digital is just samples thereof

 

RE: What if you don't care about 24/192?, posted on July 2, 2016 at 04:28:06
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 You can edit DSD files using the free Tascam software or you can use it to convert formats (DSD to PCM) before editing in RX and then transferring back to DSD. It's a PITA, but at least it can be done.


Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: What if you don't care about 24/192?, posted on July 2, 2016 at 04:47:35
rrob
Audiophile

Posts: 762
Location: Kansas
Joined: February 7, 2010
 I do have the Tascam editor but had forgotten it does the DSD to PCM conversion. Thanks. I still prefer recording at my listening resolution. Korg had/has software that would convert and burn a cd from DSD files. My copy does not work anymore. This will give me another way of doing the same.

 

RE: What if you don't care about 24/192?, posted on July 4, 2016 at 17:55:06
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 Sorry, I missed the discussions on SRC degradation....
Are they referring to SRC in general terms or specifically software (or hardware) implementations?
In short, it is implementation dependent and SRC artefacts can be very audible - anyone reliant on the Windows audio driver will know first hand what a bad one can sound like! Hence the use of ASIO drivers to get bit-perfect recordings from a USB ADC.

If the SRC uses interpolation then there are potentially going to be "errors" between the (theoretically) desired sample and the interpolated sample. Anagram Technologies use curve fitting (cubic spline or somesuch) operating on several samples to calculate a new sample. They claim reduced amplitude errors.

The filter order and type will also affect the sound.

In principle, you should be able to get from one sample rate to another without degradation assuming perfect clock accuracy, perfect filter characteristics and you literally converted the source sample rate to analogue, then resampled at the new sample rate.
In practice, this is not possible and the audibility of the SRC process will depend very much on the method (for example interpolation) and filter design (number of taps used, linear phase or minimum phase etc).

As far as software SRC implementations go, the iZotope SRC is one of the best out there and is configurable as well.
You can choose where the filter rolloff occurs with respect to Fs/2 to allow a wider or narrower transition band, the filter size and also the filter phase characteristic (i.e Linear or minimum phase).

I am happy to sacrifice bandwidth and usually set my rolloff at 0.9 wrt Fs/2 to allow a lower filter order to minimise ringing artefacts, whilst achieving a stop band attenuation of >100dB. When I stuck to the default settings, I *thought* I heard a difference when comparing downsampled files using an integer value or non-integer value and preferred to record at 176.4 for safety. However, the DA-3000 ADC has a lot of ultrasonic noise above 50kHz and it seemed pointless to include the extra noise when my cartridges of choice (in combination with the bandwidth cut on the record) didn't have a significant response >50kHz. At which point, I decided to learn how to use the settings on the SRC more sensibly.

Now I don't hear any artefacts or obvious degradation when comparing my recordings to the original Master file beyond the normal subtle difference when comparing 16/44 and 24/192 files (both recorded at native Fs) for example.
Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: The higher sampling rate, posted on July 5, 2016 at 20:09:53
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 Hi Ed

I don't know if you read a comment I made in a different section of this thread regarding the Westcott paper, but I thought it would be worth drawing your attention to a couple of important points that actually make the concerns raised in that paper irrelevant to modern delta-sigma ADCs that most people on the forum would be using.

The first important point to note is that Westcott is specifically concerned with Nyquist Rate converters. However, almost without exception, the ADCs being utilised in the equipment being discussed for recording vinyl utilise oversampling (as a necessary requirement given the number of bits for the quantiser). In fact, the PCM4220 uses an oversampling ratio of 128 (8kHz
"If music was a sine wave then 44.1 KHz is capable of recording and playing back the sine wave content (up to approximately 20kHz)."

True - Fourier's theorem is that ANY waveform may be created from a linear combination of a suitable number (which may be infinite) of sin and cos functions. As long as the sample rate is high enough for all required harmonics to be coded, then you won't "miss" information that occurs between consecutive samples. The issue is over what you define to be the required or acceptable bandwidth relative to the original waveform. Phase and high frequency content (up to the cutoff defined by the anti-alias filter) will be reproduced exactly as recorded.

On the one hand, there is the more academic issue of bandwidth required to completely reproduce a given waveform shape. However, the question is what musically useful bandwidth >20kHz can you hear that affects your perception of realism. Whilst there is a lot of emphasis on the importance of sufficient HF content to reproduce ambience, in fact subharmonics and infrasound are just as important for evoking a range of emotions and perceptions.

If the waveform on the vinyl to be recorded were from one of the many "Digital" recordings that appeared from the late 70s onwards, then the bandwidth is 20kHz - with these recordings you will see from the Spectrum Analyser in Sound Forge that the spectral content above 20kHz is simply noise. With a Nyquist Rate converter, Fs = 44.1kHz will perfectly reproduce the waveform (as cut). Where differences will occur will be in the cumulative effect of ALL filters (both analogue and digital) from ADC to DAC which will affect impulse response and phase response. Therefore the audibility of the ADC will be to some extent dependent on the DAC design. The analogue filter to the ADC should be designed not to have the phase shift occur in the bandwidth of interest.

Anyway, the point I wanted to get across is that you don't need to set Fs any higher than dictated by the input waveform bandwidth to get a more "accurate" representation of your waveform (as in intersample information loss) as implied by Westcott. Secondly, even if you aren't convinced by this, rest assured the PCM4220 is sampling very much higher than Fs!

"The idea is to record at the highest resolution you can and then you can master whatever you want from the high resolution master."

If you want the highest quality, I would avoid the use of an SRC as you are introducing another filter characteristic into the chain of filters - use the Fs that you actually need.

These days, with streamers and USB DACs, just stick with the highest resolution you feel comfortable with and be done with it!
Having said that, if your copy of SF was bundled with iZotope's SRC, then that is about as good they come. However, it is worth experimenting away from the default settings to optimise the result.


Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: The higher sampling rate, posted on July 5, 2016 at 20:15:29
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 I don't know what happened, but half a paragraph got lost from that!!
This is the complete version of the paragraph in question:

In fact, the PCM4220 uses an oversampling ratio of 128 (8kHz
"If music was a sine wave then 44.1......"

Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

RE: The higher sampling rate, posted on July 5, 2016 at 20:17:35
flood2
Audiophile

Posts: 2558
Joined: January 11, 2011
 ...ok got it ... symbols of greater than and less than not understood by the forum software...

In fact, the PCM4220 uses an oversampling ratio of 128 (for Fs between 8kHz to 54kHz) which is equivalent to a sample rate of ~5.6MHz if you were to choose Fs at 44.1kHz. The output data at your desired Fs and word length is the result of a decimation filter which thus renders the arguments presented by Westcott completely moot. The analogue input band limiting filter (as a minimum) need only be first order filter.

"If music was a sine wave then 44.1 KHz is capable of recording and playing back the sine wave content (up to approximately 20kHz)."

Regards Anthony

"Beauty is Truth, Truth Beauty.." Keats

 

Page processed in 0.056 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]