Tascam DA-3000 vs Benchmark ADC1 USB


In a previous thread the subject about using a Tascam DA-3000 to archive vinyl was discussed. I had posted that my initial tests with the Tascam were far from satisfying. I speculated that I thought the issue was the analog circuitry in the front end of the Tascam.

Well I just received a Benchmark ADC1 USB and have run a few test recordings to compare with the Tascam.

I set the ADC1 up to feed digital AES/EBU 24bit/96KHz into the AES/EBU digital input on the Tascam. Clock on the Tascam was set to DI (digital input from the ADC1). Clock rate and bit depth were set to match the ADC1.

So I was effectively using the Benchmark for the basic analog to digital conversion, and the Tascam to convert the digital data stream into a WAV file saved to a 4 G SD card in the Tascam. This way I avoided any USB and computer related variables in building the WAV file of the recording.

The analog inputs to the Benchmark ADC1 were straight out (DC out) of the Spectral DMC-10 phono preamp. I used custom built single ended RCA to XLR cables. Surprisingly, I found hum levels were about 6 dB better that the same inputs into the Tascam directly.

I recorded some quick cuts from LPs I am pretty familiar with (Steely Dan Gaucho Babylon Sisters, John Klemmer's Touch, and Blind Faith's Had to Cry Today that I used initially). The recording levels were very easy to set as the Benchmark ADC1 has really nice analog front panel controls for gain. Setting up the Tascam to "Monitor" confirmed the digital levels and both units agreed with each other to within a dB or so.

What about the results?

I was very happy with the recordings made with the Benchmark. When A/B'd directly with the LP, the recorded 24/96K WAV was not identical, but pretty damn close. Much better than recordings made with the Tascam alone. The original LP was a tad bit smoother and very slightly more detailed, but if you were not A/B ing you might not notice the difference. What was important to me was that the recording maintained the space and 3d sound field of the LP, and not crush it into a plane like many CD recordings.

All in all not the cheapest solution, but still cheaper than the Ayre 9A product. Plus using the Tascam gives you a stand alone solution with no need to connect USB to a computer, but it's there if you want it.

I would say Benchmark was pretty true to what they said their product would do.
dhl93449
Interesting post. But you have not yet done, is confirmed that recording using the USB output is any way less then the set up using the Tasman. Can you do the next logical test and compare the wave file created with the tascam with one recorded directly on the computer using USB input? If they sound equally good, one could solely use the benchmark, which would be more simple and cheaper. Thanks.
Peter:

My issue with the Tascam was the quality of the analog front end. By cutting that out by using the Benchmark, the recordings have improved significantly. Plus, the original recordings were made with the digital conversion in the Tascam, so the only variables changed were the analog and A/D conversion in the Tascam vs the analog and A/D conversion in the Benchmark.

I don't have a computer set up to stream USB at this moment, as my PC server is too far away from my sound system to run a USB cable. I bought the Tascam for about $700 so it's probably a wash buying a laptop instead, plus I like the idea of not dealing with the noise and conversion issues of a standard PC. That being said, the Benchmark is specifically designed to stream USB to a computer, so I doubt there should be any issues, even though the tech I spoke with at Benchmark actually preferred the Benchmark/Tascam system if I was going to keep the Tascam.

If you are paying full retail for the Tascam and the Benchmark, then spending a little more for the Ayre 9A may make more sense (which is about $4K). I paid about $2500 for both, which is a considerable differential with the Ayre 9A.
Glad the Benchmark worked. I suspect you would get pretty much the same result going directly to the PC via USB since the digital signal going to the PC does not depend on highly accurate timing information. But I also like the idea of writing to a disk rather than tying up a PC to do the recording.

You observations support the idea that the Tascam may be OK for moderate level systems, but its shortcomings become apparent on a higher end, more revealing systems.

Thanks for posting your observations.

What are you using for processing the files? Have you tried Vinyl Studio? It has a free trial and is only $29 to buy. I find it easy to use and like that it will look up album information, so I do not have to hand enter it. It also does click removal, hiss, hum, etc. non-destructively, so you always have the original to go back to. As any software, it takes a little to get use to it, but the workflow is pretty straightforward, since it was designed for recording, not general purpose editing.

Have you compared 96/24 and 192/24? I find a large difference between 44 and 96, but only a very slight difference between 96 and 192. I use 192 mostly because disk space is cheap and this is hopefully the only time I will digitize these records.
Dhl93449, would the Ayre be able to stream USB to the Tascam the same way as the Bencmark.
Jwm - that should work, although the Benchmark has a word clock output and on the Ayre that is an option you have to add.
11-23-15: Dtc
You observations support the idea that the Tascam may be OK for moderate level systems, but its shortcomings become apparent on a higher end, more revealing systems.
Although not necessarily. See my post dated today in the other recent thread on this unit.

Regards,
-- Al
This is a good conversation. I'm thinking of buying the Benchmark and digitizing a bunch of albums. I'm not too worried about the USB direct to computer concerns, although I might consider inserting a galvanic USB isolator (purchased from Empirical Audio) to prevent elements from the computer from making their way to the ADC. But it then takes one to the question of which ADC to purchase, giving that one wants to only digitize LP's once. I was thinking the Benchmark, but the Ayre reviews sound a bit better - but just a whole lot more expensive. Any thoughts?
Also, does anyone know whether the optical out in the 1/8" headphone jack on a Mac Mini also serves as optical in?
DTC:

I am only using dB Poweramp which has limited file processing capabilities. Thanks for the suggestion for Vinyl Studio, as I would like software to "crop" lead in and lead out noise and song transitions. I would stay away from those pop and tick removers because they may be altering the sound data.

I have not directly compared 24/96K vs 24/192K, although I have read in some forums that 24/96K is the best sonically. This was confirmed by the Tech at Benchmark (sorry I cannot remember his name). The explanation had something to do with the anti-aliasing filters and digital processing...not my cup of tea. Plus, the 96K files are half the size so that is a plus when recording to flashcards. But to each his own and if you feel the 192K is more "accurate" then by all means, use it.

I do like the volume normalize function in dBpoweramp that lets you convert the file data to 32 bit floating point, perform all the math in floating point, then re-convert back to 24 bit.

Regarding writing to the Tascam vs PC, I am in the Bryston camp here. I have a BDP-1 (instead of a PC server) to play back my digital files and have not been disappointed with the sound quality. Their arguments re the use of linear power supplies and isolation from noise sources in a conventional computer make sense (speaking of RF Al). The Tascam uses similar techniques. Plus, the Benchmark tech actually confirmed that he prefers the Tascam/Benchmark combo vs a PC for those reasons.

Regarding the connections, the Benchmark has a word clock out and the Tascam has a corresponding word clock input. But you don't need them if you use AES/EBU. You have to set up the Tascam properly while the Benchmark is connected and powered on. Otherwise the Tascam will refuse the settings and tries to revert back to an internal clock.
JWM:

The Benchmark cannot stream USB to the Tascam, as the Tascam has no USB data INPUT. The USB inputs on the Tascam are for storage media only, and you cannot record to a USB drive, only play back from it (or transfer files to it).

To connect the Benchmark or the Ayre, you would connect the digital outputs such as coax SPDIF or AES/EBU XLR balanced. The Benchmark or Ayre has to connected and powered on for the Tascam to recognize the inputs and allow you to choose the clocks and sources correctly (which is done in menus in the Tascam).

You do not need a word clock connection between the Tascam and the Benchmark either if you use AES/EBU (or another digital input/output).
Peter S:

Regarding sound quality, I have always been partial to discrete component Class A circuitry. So I am partial to the Ayre.

But I could not find a used one for a reasonable price and $4K was a bit rich for my blood. Honestly, I originally thought I could mod the Tascam with better opamps and caps, but realized this would be difficult once I got the Tascam and opened it up. All the opamps are SOIC8 surface mount and difficult to access. Plus no schematic means a lot of guesswork.

The Benchmark was the next most reasonable choice (there may be others), and if you buy directly from Benchmark, they have a money back trial period if you are not satisfied. So you can try a few recordings and compare to your phono system to see if it is good enough. For me, I found it was, but I was fully prepared to return the unit back to Benchmark if the results were not satisfactory.

If that is the case for you, then something like the Ayre is necessary and you may have to spend the big bucks.
Does anyone have experience with the M2Tech Joplin-MK2 384/32 A/D converter? Seems like a contender to the Benchmark.
Dhl93449 - Vinyl Studio lets you set the end of a track and the beginning of the next one, cutting out the part in between. You set the end then move the green arrow to the start of the next track. Very easy to do. When you write out the files you can also add silence to the beginning and/or end of a track. It also does album lookup for track names and breaks, which a real time saver. Its main source for that is discogs, which is nice since CDs and records are sometimes different. My guess is dBpoweramp does lookups based on CDs.

VS also does normalization. It am pretty sure it does it in floating point, not sure what precision. But you can ask on their forum and Paul will be able to tell you.

With VS you can do corrections (like click removal and hiss removal) on individual sections of the track, so if you are worried about the process you can do it only on short sections that are particularly bad. You can see what the click removal does to the waveform, and you will see that often it does little except take out the obviously bad part. You can also control the amount of the hiss removed and the sensitivity of the click removal. You really should try them. In most cases I cannot hear any adverse effect on the sound, but the hiss and click removal makes many passages much more listenable. You can listen to the section with and without the correction to see how it sounds. I use headphones for this and although the sound on the PC is not as good as on the stereo, you should be able to tell if you are damaging the underlying music.

I know I sound like a VS salesman, but it really is a good program for a very low cost. Charlie Hansen at Ayre also recommends it, so I am in good company.

As to writing to disk or a PC, I am with you on that. That was one reason I went with the Korg.
Looking forward to trying VS once I get my Benchmark, or M2Tech, or Ayre...
DTC:

If VS can remove tics or pops "individually" without processing the entire track, that would be amazing. I have some early first pressings that have the one depressing loud tick or pop that ruins the whole track. Being able to remove that without affecting everything else would be a plus.

But how do you locate these? Do you listen while watching a waveform playback to identify their location?
Dhl93449 - You just listen to the track and identify the track time. You can then zoom in on that point. The click will be pretty obvious in the waveform. You can select a small region and have VS do the correction or there is a manual tool which lets you adjust the width of the correction. It shows you the original and the corrected waveform and lets you listen to either one. You do each channel separately. Again, I use headphones for doing this.

For more complicated corrections there are also patch tools (e.g. FFT) that analyze the surrounding waveform and fit a corrected waveform. This is usually a last resort, but can be good. for example, if you have a scratch that runs along the grove rather than perpendicular to it.

You should try doing click repair on a whole track at the lowest sensitivity. You can look at the waveform and you see clearly the little clicks and pops which may be barely audible but have distorted the waveform and the correction smooths them out. People are always concerned about changing the music, but the flaws have already done that. I would say the same thing about hiss. If you can hear hiss on a low volume section, why not try removing the hiss? Again, you can do that on the whole album, a single track, or on an individual section.

The nice thing about VS is that the corrections are non-destructive. That is, they are stored in a separate file and only applied on output. The original file is never changed.

People also use Isotope RX 5 as a correction tool, but that is $349. For me, VS does the job for a lot less money.
DTC

Could you clarify that point about the corrections?

Are you saying that at the end of the process you have two files, the original and a corrected version?

Or is VS needed to apply the corrections during playback?
The original files (e.g., wav files from the Tascam) are left unchanged and the output files (usually the individual tracks - I use flac) contain the corrected audio. The changes are maintained in separate files and are only applied to the output files when their are created. The VS website has documentation and screen shots of the process if you want to see the details.
I'm probably getting a Tascam DA-3000 to digitize my LPs.  What attracted me to it was this write up on the Decware site, where they sell the DA-3000 with an upgraded analog output stage:

“Although it can be hooked to a computer or your favorite transport or anything with a digital output... our favorite thing about it is the SD cards! This makes it possible to hear your sound files on a 100% hardware solution without being hooked up to a computer or the internet. Below is a list of the following variables this can eliminate:

    1) What type of computer you have and its effect on the sound?
    2) What operating system does it run on?
    3) How fast is the hard drive in it?
    4) How much memory is available for buffering?
    5) Which of the dozens of playback software solutions you have?
    6) Are they configured correctly?
    7) What plug-ins are installed?
    8) Do you have the proper and up to date software drivers?
    9) Are the sound adjustments on the computer configured right?
    10) What USB formats does your computer support?
    11) What kind of audiophile USB cable should you run?
    12) Is the computer running an aftermarket updated power supply?


“BONUS: YOU CAN RECORD DSD AUDIO WITH IT - If you're not into wasting money on special software and inferior hardware to record your albums or tapes, stay tuned because this machine is simply a fantastic digital recorder. It records directly to SD cards making it possible to get perfect hi-res bit perfect DSD and double DSD recordings of your analog treasures with results so true to the source you'll have trouble telling the two apart.”

The discussion in this thread re: Vinyl Studio confused me at first since as Decware states no special software is required for recording with the DA-3000.  But here we are talking about using VS to enhance the files created by the DA-3000, and the VS is not used during the recording process, correct?


But here we are talking about using VS to enhance the files created by the DA-3000, and the VS is not used during the recording process, correct?

Correct. You do not need VS to do the recording. VS will allow you to break the file into tracks and name them and also to eliminate the gaps between tracks. For DSD files you cannot do any correction (click removal, hum or hiss removal) without changing to PCM.

 I started with DSD but soon found that I wanted to do corrections on most of my albums and went to PCM to do that. At least with my DAC, I do not heard much difference between DSD and PCM. You should try both and see if you can hear the difference.

As a playback device I think the DA-3000 is somewhat limited. From the specs I believe it only takes 64 GB cards or flash drives. That can be limiting when each 5.8 MHz DSD album is over 4 GBs. If you can put a large external drive on it, then you might be OK, although the "user interface" is certainly limited. The usb interface is usb 1.1 according to the specs.
I won't be using the DA-3000 as a playback device since recently acquired a PS Audio DirectStream DAC with Bridge II that I stream to.  It sounds very good and is the reason I've decided to try digitizing my LPs.  I'm hoping they sound better as DSD files than PCM via the Directstream.  I'm sure many of my LPs will need some correction with VS, so based on what you've said, those will need to be PCM files.  I rip CDs with dBpoweramp which can convert file types.  Too bad it can't convert PCM to DSD.
Vinyl Studio will convert DSD to PCM but not PCM to DSD. JRiver can convert PCM to DSD.

I think the DirecStream converts PCM to DSD internally on the fly for playback.  I do not own what, but pretty sure that is what I read about its architecture. So I do not think you need to do any conversion yourself.
That's true the DirectStream does convert everything to DSD, so digitized LPs in PCM or DSD played through it might sound the same.  I put this question up on the PS Audio forum and someone said "The DA-3000 is superb. You probably know this, but Tascam offers free editing software that provides the ability to divide and concatenate DSD and PCM files, as well as conversion. It supports up to 11.2MHz DSD and 384kHz PCM."
From what I can tell, although I have not used it, the Tascam software is pretty limited.

It allows you to work on one file at a time and does not allow a full album file to be treated as an album. There is no tagging of album, artist, composer, track name, etc. To output tracks from an album, you need to select an area then write it out, typing in the track name if you want to use that as the file name. You repeat that for each track. That is a pretty laborious task compared to something like Vinyl Studio which looks up albums on the Internet and provides track timing and track names. The software does not appear to have a facility to find track breaks.

The software does not have any correction capability nor any normalization function.

It does not have support for file formats other than wav, dsf and dsdiff so, for example, flac is not available.

It does seem to convert PCM to DSD and vice versa. And, it appears to convert between different DSD sample rates. I cannot figure out from the documentation if it does so by converting to PCM or not. Perhaps someone of the PS Audio forum knows if DSD to DSD conversion goes through PCM. If not, that would be a nice addition to know about.

If it converts between dsd formats without using pcm as an intermediate, that would be very useful. Other than that, VS is much more versatile.

Can you check on whether dsd conversions go through pcm?
Vinyl Studio definitely appears to be the better product.

Do you mean check on whether dsd conversions go through pcm in the Directstream DAC or are you talking about the Tascam software?
I was referring to the Tascam software and thought you could check on the PS Audio forum. I do not participate there.

There is a need for some people to convert, e.g. 2xDSD to 1xDSD and a program that did that without converting to PCM as an intermediate would be nice to have.
Thanks. We will see if he knows. Otherwise I might call Tascam and ask them.
He did not know, and says Tascam is pretty responsive so perhaps emailing them will provide an answer.
Question sent to Tascam. There wording in their description is a little vague. Hopefully, they will be able to answer. A free utility that converted between different dsd sample rates without going through PCM would be a welcome addition.

I did download the software to an Windows 8.1 i3 system and it took a several minutes to display the waveform on a 6 minute track.

I was looking at the PS Audio NuWave Phono Converter as an alternative to the Tascam DA-3000.  There are a couple for sale on this site ($1100 and $1400).  It is a completely analog phono stage + an AD converter in one box.  My current phono preamp is a Heed Quasar which in my system sounds great.  Since I don't know anything about the NuWave phono, not sure if I really want to go there.  But anyway, reason for bringing it up, Paul McGowan of PS Audio, despite his product having the capability of recording in DSD, belives 24bit/96kHz PCM is plenty sufficient to digitally capture all the musical information found on a vinyl LP.  Since I think the DA-3000 is the way I'm going to go, this makes me think no need to be concerned with whether to record in DSD or PCM.  Just use PCM 24/96 for all LPs and then those that need correction with VS, do that and be done.

Quote from the Manual: We recommend for most users 96kHz and nothing higher. As previously explained, 96kHz provides full bandwidth for any LP or, for that matter, anything most people are likely to want to play. Using 96kHz and below engages the best sounding decimation filters in the NPC and provides a near-perfect zero group delay for PCM users. 96kHz gives full bandwidth to 48kHz, exceeding the threshiold of human hearing by twice (and vinyl LP’s don’t exceed 30kHz even under the best circumstances), and maintaining perfect phase relations within the audible band.
There is another potential option if you find 24Bbit/96K is sufficient. It is another Tascam product, but uses discrete Class A transistors on its analog input stage. I have not made a direct sonic comparison, but bought one of these (oh yeah, its a Tascam DR 60D MKII) to use as a mic pre-amp to digital converter for a sound re-enforcement system to record pink noise in a room, for playback to a spectrum analyzer. Tascam touts the linearity and low distortion of the discrete component mic preamps, but these can be gain adjusted to take low level line inputs as well. Might well work in a vinyl conversion set-up. Not as flexible as a 3000, but a heck of a lot cheaper with maybe better sonic performance.
There was a thread a while back where a guy used a Tascam DR-5 to record his vinyl. He was happy with the results, although I do not remember the setup.  Don't know what the input state looks like. As I remember he might even be trying to sell the unit.
I had a chance to do some more testing with the Tascam DR60DmkII.

I ran some pink noise spectrum tests on my speakers with a digitally recorded pink noise source file. I had a Shure SM81 mic connected to the mic pre-amp on the Tascam, and I needed the highest gain setting (there are three, LOW, MID, and HIGH) for pink noise at about 83 dB at the mic.
I then recorded 10 minute files with the mic pointed at different positions, for playback later to a spectrum analyzer.

This system was a lot cheaper than buying a dedicated stand alone sound measurement system. Most mic pre-amps (with phantom power) are more expensive than the Tascam. So you basically get the digital recorder for free.

The Tascam specs on the mic pre-amps indicate the input levels can be as high a 0 dBu (on the LOW gain setting), so there should not be an issue using a phono pre-amp with a balanced output, assuming it can drive the 2K input load in the Tascam. Only compromise I can see is the DR60DmkII only records at 24BIT/96K, unlike the DA-3000 which can record up to DSD frequencies. A further advantage to the little DR60D is the use of discrete components in the input pre-amps,

"The DR-60D's HDDA mic preamplifirers utilize discrete circuitry and premium parts approved after months of evaluation tests."  -- Tascam

Kind of wished they would have used those in the DA-3000.
dhl - Thanks for the info on the Tascam DR60DmkII.  Would you mind creating a recording of an LP with it and comparing the result to the straight Tascam DA-3000 w/o using the Benchmark for the analog front end?  Perhaps you still have one of your earlier test recordings of the DA-3000 by itself.  I was about to purchase the DA-3000 but now thinking the DR60 might be a better choice due to it's discrete analog input stage.  It's much less expensive and I plan on recording at 24BIT/96K.
jeff

Was wondering about that myself. I cannot make a direct comparison of two digital files in my system, but can make a comparison between a digital file made on the DR60 and the LP itself. I'll see if I can break some time free to do that test.
Hi Jeff,

I have a digital file that you can have access to of a Phillips recording of Mozart's 6 string quartets.  It is recorded on a Tascam DA-3000.  Send me a private message and I can give you the link.