wyn palmer
Addicted Member
Yes, the title's a bit tongue in cheek but recently I've acquired some new digital gear/software and I've been ripping vinyl and 15IPS 2 track tapes to digital and the results have been surprisingly good. As a result I wanted to share my experiences and to enquire as to what the equivalent experiences of various members may have been in this area.
Here's my main system.
Vinyl playback: VPI classic SE turntable with Miyajima Madake cartridge, Modified Hovland HP-100 preamp with custom 20:1 SUT. Modified ML Montis with added sub woofer + miniDSP, Sennheiser HD-800S headphones, Stax SR-L700 headphones, Rogue M180 power amps. Home made cables.
Digital rip: Hardware; Korg DS-DAC-10R, Topping D50, custom variable gain output buffer amp. Software; IZotope RX-7 elements, Korg Audiogate. All rips stored on a laptop and backed up. Playback via USB 2.0 through a central powered hub. Separate analog supply for the D50. Playback via JRiver and Foobar2000.
The process was not without problems. I took the signal to be digitized from the record output of the Hovland and fed it directly to the Korg- putting it in parallel with the feed to the recording inputs of my tape deck, then started digitizing at 192kHz/24bits.
At first the process seemed to be going well. I adjusted the gain on the recording software to peak the input at -6dBv and then ripped the Recut Records issue of " Beethoven Choral Fantasia" with Katchen/Gamba.
After I was finished I exported the file and processed the output using RX Elements- normalizing the gain to just below full scale for the max level and using the software to de-click and de-clip the rip.
The result was just terrible. It didn't even need an AB to know that what was digitized was a pale image of the original vinyl. This was very disappointing, and completely contrary to what was expected based on the Korg's reputation.
So I tried to "debug" the system.
I evaluated the Hovland preamp record output. I found that there were two issues.
The first was that even well below clipping the output seemed to be audibly distorted despite its low output impedance (about 350 ohms).
It turns out that the distortion gradually increases as the load drops from 100k to 10k and then rapidly as the load falls below 10k. The tape deck was presenting a 10k load, and the Korg a 47k load, so the output was being overloaded! So, the tape deck connection had to go. I don't record onto tape, so that was OK.
It helped, but there were still many clipping events reported by the software.
The second issue was the absolute signal level to the Korg.
The max output from the record output before clipping is +11dBv, but the max input before clipping into the Korg is +6dBv. So a 6dB attenuator was needed. I tried setting the Korg input gain to -6dB, but that did not help. I then tried a commercial in line 6dB attenuator. That too failed as it loaded the record output too much.
In the end I had to hack an RCA cable and add a 47kohm series resistor at the input of the ADC to pad the level down. I checked the bandwidth of the plugged in cable (>100kHz) and calculated the additional noise contribution (negligible...) and tried again.
I set the Korg line input gain permanently to 0dB and monitored the peak on Audiogate to see if any excursion to the clip level occurred- and none was seen.
Next I imported the file to Rx and went through the same process- this time no clips were detected and a very small number of clicks, so I had no reason to do any processing besides normalizing to set the max levels just below full scale.
The results were extremely good and I've now digitized a number of LPs (Janis Ian Breaking silence 45RPM, several Decca Beethoven, Mercury Respighi the Birds etc.) and several 15IPS tapes ( Dark side of the moon, Kind of Blue, Elgar cello concerto/Dupre etc.).
Brief ABX testing with the output levels matched between the vinyl/tape outputs and the DAC output was attempted with myself and a couple of audiophile friends and none of us could reliably tell which was which. More extensive listening awaits, but the result is extremely encouraging.
So, all in all, I would say that it was a very successful effort and I would definitely recommend it to those who would like to capture the sound of vinyl and tape but have the convenience of digital, and not just for portability reasons.
However, my experience also illustrates how easy it is to get the digitization process badly wrong.
Here's my main system.
Vinyl playback: VPI classic SE turntable with Miyajima Madake cartridge, Modified Hovland HP-100 preamp with custom 20:1 SUT. Modified ML Montis with added sub woofer + miniDSP, Sennheiser HD-800S headphones, Stax SR-L700 headphones, Rogue M180 power amps. Home made cables.
Digital rip: Hardware; Korg DS-DAC-10R, Topping D50, custom variable gain output buffer amp. Software; IZotope RX-7 elements, Korg Audiogate. All rips stored on a laptop and backed up. Playback via USB 2.0 through a central powered hub. Separate analog supply for the D50. Playback via JRiver and Foobar2000.
The process was not without problems. I took the signal to be digitized from the record output of the Hovland and fed it directly to the Korg- putting it in parallel with the feed to the recording inputs of my tape deck, then started digitizing at 192kHz/24bits.
At first the process seemed to be going well. I adjusted the gain on the recording software to peak the input at -6dBv and then ripped the Recut Records issue of " Beethoven Choral Fantasia" with Katchen/Gamba.
After I was finished I exported the file and processed the output using RX Elements- normalizing the gain to just below full scale for the max level and using the software to de-click and de-clip the rip.
The result was just terrible. It didn't even need an AB to know that what was digitized was a pale image of the original vinyl. This was very disappointing, and completely contrary to what was expected based on the Korg's reputation.
So I tried to "debug" the system.
I evaluated the Hovland preamp record output. I found that there were two issues.
The first was that even well below clipping the output seemed to be audibly distorted despite its low output impedance (about 350 ohms).
It turns out that the distortion gradually increases as the load drops from 100k to 10k and then rapidly as the load falls below 10k. The tape deck was presenting a 10k load, and the Korg a 47k load, so the output was being overloaded! So, the tape deck connection had to go. I don't record onto tape, so that was OK.
It helped, but there were still many clipping events reported by the software.
The second issue was the absolute signal level to the Korg.
The max output from the record output before clipping is +11dBv, but the max input before clipping into the Korg is +6dBv. So a 6dB attenuator was needed. I tried setting the Korg input gain to -6dB, but that did not help. I then tried a commercial in line 6dB attenuator. That too failed as it loaded the record output too much.
In the end I had to hack an RCA cable and add a 47kohm series resistor at the input of the ADC to pad the level down. I checked the bandwidth of the plugged in cable (>100kHz) and calculated the additional noise contribution (negligible...) and tried again.
I set the Korg line input gain permanently to 0dB and monitored the peak on Audiogate to see if any excursion to the clip level occurred- and none was seen.
Next I imported the file to Rx and went through the same process- this time no clips were detected and a very small number of clicks, so I had no reason to do any processing besides normalizing to set the max levels just below full scale.
The results were extremely good and I've now digitized a number of LPs (Janis Ian Breaking silence 45RPM, several Decca Beethoven, Mercury Respighi the Birds etc.) and several 15IPS tapes ( Dark side of the moon, Kind of Blue, Elgar cello concerto/Dupre etc.).
Brief ABX testing with the output levels matched between the vinyl/tape outputs and the DAC output was attempted with myself and a couple of audiophile friends and none of us could reliably tell which was which. More extensive listening awaits, but the result is extremely encouraging.
So, all in all, I would say that it was a very successful effort and I would definitely recommend it to those who would like to capture the sound of vinyl and tape but have the convenience of digital, and not just for portability reasons.
However, my experience also illustrates how easy it is to get the digitization process badly wrong.
Last edited:
