Looking for an ADC for recording from a turntable (again)

[KITR]

Hi everybody,

I know it's been asked about trillion times already, but never actually answered to my complete satisfaction.

What I am looking for is a cleanest (not necessarily the most expensive) way possible to record something from a turntable. My turntable is an old Nakamichi Dragon-CT, so it's got an analog output of course. Right now, I use Sound Blaster X-fi HD, but it's phono preamp is horrible, it introduces lots of noise and is not particularly clean. I need either some ADC with a good preamp or an ADC with good line in interface, that would be fed through a regular preamp. I am inclined towards the latter, and my eyes are on EVGA nu audio pro, which is not crazy expensive, but according to measurements, it's one of the better ones.

Does anybody here have some experience with either a good ADC with an integrated preamp or with the EVGA card?

 

[MusicNBeer]

I use a cheap Asus Xonar DSX which works fabulous for vinyl. It's got 90dB sinad in my PC, THD approx 0.001, and flat response to 20K.

I record at 44.1Ksps 24 bit. After, I high pass filter to remove rumble, amplitude adjust to full scale peak, and save.

You do need to ensure recording levels into card are not clipping input.

 

[Vincent Kars]

with the EVGA card?

https://www.audiosciencereview.com/...u-audio-pro-review-internal-sound-card.19728/

 

[AnalogSteph]

OP, what kind of cartridge are you using on the turntable? I would advise going the phonopre route, which would allow you to take care of cartridges with special needs in terms of termination as well. The X-Fi HD at line level should be quite sufficient as-is; I would recommend 24/96 recording (this may be a bit tricky to set up with all of driver, OS device and recording software settings potentially having to be aligned, but it is a sort of familiar dance at this point).

I record at 44.1Ksps 24 bit. After, I high pass filter to remove rumble, amplitude adjust to full scale peak, and save.

Just as a side note, I wouldn't necessarily recommend recording straight to 44.1 kHz with a WM8776 (as used by the DSX) or other consumer-level ADC. While anti-alias filter stopband starts at 0.546 fs, stopband rejection of >65 dB is what they call a bit naff, not the best match for an analog source with a reputation for spitting out no small amount of ultrasonics. I would go to the extra trouble of recording at 24/96, normalize at that stage (note: I wouldn't recommend exactly 0dBFS peak but rather -0.5 or -1 dBFS), and then resample to 44.1 for further processing. (Audacity has had a SoX VHQ level resampler available for quite some time now.)

In the Xonars, controlling the recording side settings is quite straightforward, you just have to set up the recording device and software. (You can use Audacity in WASAPI I/O mode to check whether everything is set up right, it won't let you record or play anything if there's a mismatch in sample rate. Plus, that's the only way to get 24-bit recording on Windows anyway.)
The playback side is a bit trickier, I would recommend using XonarSwitch to manage settings. I made a "recording" preset that sets everything to my recording sample rate. I found that upsampling in the driver would work fine (i.e. system sample rate < card sample rate) while downsampling gives super crappy results, at least using a D1 or D2.

 

[MusicNBeer]

Just as a side note, I wouldn't necessarily recommend recording straight to 44.1 kHz with a WM8776 (as used by the DSX) or other consumer-level ADC. While anti-alias filter stopband starts at 0.546 fs, stopband rejection of >65 dB is what they call a bit naff, not the best match for an analog source with a reputation for spitting out no small amount of ultrasonics. I would go to the extra trouble of recording at 24/96, normalize at that stage (note: I wouldn't recommend exactly 0dBFS peak but rather -0.5 or -1 dBFS), and then resample to 44.1 for further processing. (Audacity has had a SoX VHQ level resampler available for quite some time now.)

My looptest results are not confirming that. I generate a 96Ksps signal using Audacity with a tone at 25KHz at about -7 dBFs. This signal is outputted via USB to external DAC and looped back to Xonar line in. When recording at 44.1Ksps on the Xonar, the tone is attenuated by greater than 100 dB (there's no image folded back greater than -110 dBFs). When I record at 96Ksps, the tone is present.

 

[KITR]

Thank for the replies.

AnalogSteph
I'm using Yamaha MC-1000. Well, I wanted to avoid using a preamp. But I'll try it anyway together with the Sound Blaster, maybe it will be the best possible solution.


Vincent Kars
Thanks for the link, but that's what made me look into that card in the first place. I reckoned that somebody would have a real life experience with the card and the turntable recording.

 

[AnalogSteph]

@AnalogSteph
I'm using Yamaha MC-1000. Well, I wanted to avoid using a preamp. But I'll try it anyway together with the Sound Blaster, maybe it will be the best possible solution.

That's one of those "well, duh" moments. Like the name suggests, the MC-1000 is an MC cartridge, and apparently not overly high output at that. Why did you expect the phonopre in the X-Fi HD to be suited for them in the first place? It's basically just a nice bonus feature to begin with, asking for more than a basic MM job is a bit much. Right now you are lacking about 20 dB of gain, and input voltage noise probably is too high by just about as much.

You would need an MC prepre at the very least, or a whole external MM/MC phonopre (most of which tend to cost more than the whole soundcard). A hi-fi preamp or integrated amp with MC inputs could be made to work as well via its recording outputs.

My looptest results are not confirming that. I generate a 96Ksps signal using Audacity with a tone at 25KHz at about -7 dBFs. This signal is outputted via USB to external DAC and looped back to Xonar line in. When recording at 44.1Ksps on the Xonar, the tone is attenuated by greater than 100 dB (there's no image folded back greater than -110 dBFs). When I record at 96Ksps, the tone is present.

Huh. Interesting. I can only guess that the 25 kHz tone (0.56689 fs) may be "slipping through the cracks" and just happen to be ending up near a response null of the filter. A test with a more wideband signal like white noise may clear this up. I don't see why the WM8776 would be deviating from the ADC filter response printed in its datasheet. According to that, the worst filter rejection would be at about 1.95 fs (~86 kHz), followed by ~1.65 fs (~73 kHz).

 

[apson]

I purchased the MOTU M4 specifically for vinyl ripping based on the review here on ASR. My second choice was the RME ADI-2 Pro which was considerably more expensive. I've found, not surprisingly, that turntable > cart > pre-amp are the critical elements of this process and I would at the very least get those elements where you want them before introducing an ADC into the mix. In my case it's a Rega P3 > Elys 2 > Rega Brio. I actually ripped a fair amount of my collection ~100 records before deciding to upgrade my cart which hasn't come yet. I do everything into the MOTU at 96k 24-bit using Audacity. I put a fair amount of thought into the logistics from there on as well. First I peak normalize them in ffmpeg but that adjustment is not done per side, but per master/release. It's a two step process in which I determine the loudest of the sides and apply gain to each so that the loudest side peaks at ~0. My Rega Brio doesn't push the MOTU very hard; but utilizing 24 bits with vinyl is probably silly at best anyway. I just wanted archival preservation so every pop and click is included. Again, this is peak leveling, I specifically didn't want the dynamics messed with. From there I go to FLAC as well as VBR AAC for on the go access. For the AAC I use qaac via command line via ubuntu with wine (which was a nightmare and made me wish I had at least one Windows machine laying around). I have a Dockerfile to run qaac anywhere if anyone wants it.

I also experimented with the HifiBerry's ADC (both the pro and "not pro" utilize the TI PCM1802). I'm going to give the HifiBerry a better shake at this competition soon. It was done with my U-turn Audio Orbit with a built in pre which is just terrible science for comparison purposes. The other interface I have is a Focusrite Scarlett Solo gen 2 -- I haven't tried that with vinyl yet but I'm fairly sure I can get the two inputs to work.

Any of the phono-pre + adc combo units I found always worried me because I felt like it was so application-specific that it must be a gimmick. Yet, I am currently trying to find a designer to help me market a possibly vinyl-optimized usb class compliant audio interface. It's sort of a non-profit project aimed at archiving the world's vinyl. Long story...

I have files I can share with you. I'm not sure it's okay to post them here though it should be well within fair use. If the moderators approve I can attach examples to this thread just tell me which genre you're after.

Here's my qaac command for reference.

qaac --tvbr 127 --rate 44100 --quality 2 -o led_zeppelin_ii_a.m4a led_zeppelin_ii_a.wav

 

[KITR]

I purchased the MOTU M4 specifically for vinyl ripping based on the review here on ASR. My second choice was the RME ADI-2 Pro which was considerably more expensive. I've found, not surprisingly, that turntable > cart > pre-amp are the critical elements of this process and I would at the very least get those elements where you want them before introducing an ADC into the mix. In my case it's a Rega P3 > Elys 2 > Rega Brio. I actually ripped a fair amount of my collection ~100 records before deciding to upgrade my cart which hasn't come yet. I do everything into the MOTU at 96k 24-bit using Audacity. I put a fair amount of thought into the logistics from there on as well. First I peak normalize them in ffmpeg but that adjustment is not done per side, but per master/release. It's a two step process in which I determine the loudest of the sides and apply gain to each so that the loudest side peaks at ~0. My Rega Brio doesn't push the MOTU very hard; but utilizing 24 bits with vinyl is probably silly at best anyway. I just wanted archival preservation so every pop and click is included. Again, this is peak leveling, I specifically didn't want the dynamics messed with. From there I go to FLAC as well as VBR AAC for on the go access. For the AAC I use qaac via command line via ubuntu with wine (which was a nightmare and made me wish I had at least one Windows machine laying around). I have a Dockerfile to run qaac anywhere if anyone wants it.

I also experimented with the HifiBerry's ADC (both the pro and "not pro" utilize the TI PCM1802). I'm going to give the HifiBerry a better shake at this competition soon. It was done with my U-turn Audio Orbit with a built in pre which is just terrible science for comparison purposes. The other interface I have is a Focusrite Scarlett Solo gen 2 -- I haven't tried that with vinyl yet but I'm fairly sure I can get the two inputs to work.

Any of the phono-pre + adc combo units I found always worried me because I felt like it was so application-specific that it must be a gimmick. Yet, I am currently trying to find a designer to help me market a possibly vinyl-optimized usb class compliant audio interface. It's sort of a non-profit project aimed at archiving the world's vinyl. Long story...

I have files I can share with you. I'm not sure it's okay to post them here though it should be well within fair use. If the moderators approve I can attach examples to this thread just tell me which genre you're after.

Here's my qaac command for reference.

qaac --tvbr 127 --rate 44100 --quality 2 -o led_zeppelin_ii_a.m4a led_zeppelin_ii_a.wav

I was actually thinking about getting MOTU M4 as well, then I found out about EVGA nu audio pro and it seemed like a better choice. That's why I wanted to hear somebody else's opinion on this card.

BTW, I'd love to hear some sample recording. Do you have some kind of jazz record?

That's one of those "well, duh" moments. Like the name suggests, the MC-1000 is an MC cartridge, and apparently not overly high output at that. Why did you expect the phonopre in the X-Fi HD to be suited for them in the first place? It's basically just a nice bonus feature to begin with, asking for more than a basic MM job is a bit much. Right now you are lacking about 20 dB of gain, and input voltage noise probably is too high by just about as much.

You would need an MC prepre at the very least, or a whole external MM/MC phonopre (most of which tend to cost more than the whole soundcard). A hi-fi preamp or integrated amp with MC inputs could be made to work as well via its recording outputs.

I know, that's why I wanted somebody else's opinion on EVGA card.

 

[LTig]

I purchased the MOTU M4 specifically for vinyl ripping based on the review here on ASR. My second choice was the RME ADI-2 Pro which was considerably more expensive.

The sensitiviy of the RME is good for line level signals only, you'd need a preamp even for MM because for MM the input impedance is too low (9 kOhm versuse 47 kOhm required). OTOH I once recorded direct from ym VdfH MC1 Special MC pickup with 0.65 mV output and while youicould hear the noise in pauses it was astonishingly free of noise when music played.

 

[apson]

The sensitiviy of the RME is good for lone level signals only, you'd need a preamp even for MM because for MM the input impedance is too low (9 kOhm versuse 47 kOhm required). OTOH I once recorded direct from ym VdfH MC1 Special MC pickup with 0.65 mV output and while youicould hear the noise in pauses it was astonishingly free of noise when music played.

Yes, I am using the line level output of my Rega Brio phono preamp into the MOTU M4's rear panel "Line Input" 3-4 (though manual states that 1-2 behave the same way when the gain knobs are set to 0?). I never bothered trying any of the ADCs with a built-in phono preamp because I just assumed expenses were spared on the preamp. It's a marketing mine field out there and this forum is the only place I feel safe .

BTW, I'd love to hear some sample recording. Do you have some kind of jazz record?

I have Miles Davis Kind of Blue side A if that helps. Sending you link because I'm still not entirely sure if ASR knows it's covered under fair use (by the name alone, even). If moderators say it's okay to post the link to dropbox I will. Just so you know it's exactly this: Rega P3 > Elys 2 > Rega Brio preamp > MOTU M4 (line in on back/3-4) to PCM 96k/24 > qaac -tvbr 127 -quality 2 and downsampled by qaac to 44.1 for maximum portability.

Which ADC is in the EVGA? The CS5346 is mentioned in the EVGA NU Audio Pro review. I'm still searching for a designer to hire to help me create a vinyl-friendly usb compliant/xu208 ADC. My goal now is to get a high quality reference LP and one of these QA401/2 and start making something ideal for vinyl. I just want to sell them for cost.

 

[KITR]

I have Miles Davis Kind of Blue side A if that helps. Sending you link because I'm still not entirely sure if ASR knows it's covered under fair use (by the name alone, even). If moderators say it's okay to post the link to dropbox I will. Just so you know it's exactly this: Rega P3 > Elys 2 > Rega Brio preamp > MOTU M4 (line in on back/3-4) to PCM 96k/24 > qaac -tvbr 127 -quality 2 and downsampled by qaac to 44.1 for maximum portability.

Which ADC is in the EVGA? The CS5346 is mentioned in the EVGA NU Audio Pro review. I'm still searching for a designer to hire to help me create a vinyl-friendly usb compliant/xu208 ADC. My goal now is to get a high quality reference LP and one of these QA401/2 and start making something ideal for vinyl. I just want to sell them for cost.

Cheers for the file, it sounds quite adequate. I reckon the MOTU would be enough for my needs.

 

[Oukkidoukki]

Rme adi pro is about worlds best adc.......go pro or regret later.....

 

[Oukkidoukki]

Rock solid pre amp.....vovox sonorus cables from switzerland.....rme adi pro......recipe for success

 

[apson]

Rme adi pro is about worlds best adc.......go pro or regret later.....

I owed it to myself to try the US $230 MOTU M4 first The ADI Pro is US $2000 and contains a highly performant DAC that I do not need.

Cheers for the file, it sounds quite adequate. I reckon the MOTU would be enough for my needs.

Sure thing! If anyone else wants to hear it let me know.

 

[Oukkidoukki]

Okey

 

[LTig]

Yes, I am using the line level output of my Rega Brio phono preamp into the MOTU M4's rear panel "Line Input" 3-4 (though manual states that 1-2 behave the same way when the gain knobs are set to 0?). I never bothered trying any of the ADCs with a built-in phono preamp because I just assumed expenses were spared on the preamp. It's a marketing mine field out there and this forum is the only place I feel safe .

If you have an MC system you could connect it directly to the MIC input of the M4 and do the RIAA deemphasis by software.

 

[apson]

If you have an MC system you could connect it directly to the MIC input of the M4 and do the RIAA deemphasis by software.

I read about this a while back but now you're making me curious. My Elys 2 cart is MM but the new cart coming soon (a Hana SL) is MC.

In that case it really makes me wonder about the sea of other pre-amps I've worked with over the years. One that comes to mind I think might work well for vinyl is the Sytek MPX-4. I'll give this a shot, thanks.

 

[AnalogSteph]

A very low-noise mic pre is definitely in the right ballpark at least. If you've got any interface that Julian Krause found to reach an EIN of around -131 dBu(A) @ 150 ohms, those would be well worth a shot (e.g. M-Audio AIR 192 | 4, Tascam UR-2x2HR, SSL 2, Audient iD4/iD14 MkII). Should get you about on par with a run of the mill MC input... a little over 1 nV/√(Hz) of input voltage noise density.

MC cartridges are actually (theoretically) capable of delivering higher SNR than their MM cousins, but at between 8 and 20-something ohms typical, they are even lower impedance than the lowest-impedance dynamic mics I know of, oddballs of around 50 ohms (mostly historic, I found like a single currently available model like that - the vast majority are 150 ohms up). Getting the last bit of SNR out of them requires seriously low input voltage noise, and mic preamps generally just don't bother going that low as it's simply not necessary but would substantially increase current consumption and either increase parts count or reduce parts availability / choice. A majority of very low rbb', highish beta transistors have been EOL'd in the last decade or so.

For reference, if you want an input that's no noisier than a 10 ohm resistor (meaning that you're still looking at a noise figure of over 3 dB with an 8 ohm cartridge... plus cable resistance), you'd be looking at -130.6 dBu unweighted @ 20 kHz with a 150 ohm source, that's about -132.9 dBu(A). (Input noise density about 0.4 nV/√(Hz).) An Earthworks ZDT preamp gets somewhat close but would still be about 0.3 dB short... doesn't sound much but keep in mind this is added to a 150 ohm source - 0.6 nV/√(Hz) is about 22 ohm equivalent. The Sytek MPX-4Aii's EIN spec of -134 dBu (50 ohms) works out to pretty much the same (23 ohms), so another good candidate indeed.

 

[rkay5]

Mytek made some really good ADC's the 192 or 96. Also the Brooklyn ADC it has USB, AES,S/PDIF, and BNC for DSD. I had the Mytek ADC192 that I used for recording vinyl and it made some great sounding recordings but I wanted more so I old it.
And got the Tascam DA3000 that I still have. But what I use now is RME ADI-2 Pro, it's the best that I've found that I can afford.