Tascam DA-3000
To my ears (using the Tascam only) 2xDSD (5.6MHz) recordings get me closest to the original vinyl. No doubt about it. So the best result could be to use a DSD-capable A/D converter, or have the A/D in the Tascam modified. It reminds me of the S/N mod to my Revox A77 that really lifted the quality, back in the 80ies.The Tascam is OK as it is, but when I am considering recording some hundred vinyl albums, I do want everything to optimal, since it is a lot of work, and should be (ahem) future-proof. |
Al: You do have a point, assuming that potential RF entering the Benchmark is somehow filtered differently so that it never gets into the digital bit stream. I did notice my 60 Hz hum levels were lower when using the Benchmark. Same gain setting on the Spectral, same interconnects. I attribute this to a better common mode rejection in the Benchmark balanced input amplifiers, which may have an impact on your point above. |
Fair enough, Dhl93449. Your post just above solidifies your case persuasively, as far as I'm concerned. My theory was an attempt to reconcile your sonic assessment of the Tascam with Jeff's (whose opinions command great respect in my book), while taking into account the nearly unique ultra-wide bandwidths of the Spectral products. One minor nit, though: I won't repeat that other post here, but basically the Spectral was coupled to the Benchmark which fed digital AES/EBU to the Tascam. If Al is right, the same RF would have been transmitted to the Benchmark and the sonic results would have been the same.My theory revolved around the assumption that the anti-aliasing filter in the Tascam was less than fully effective at RF frequencies, perhaps because the design is such that some RF frequencies might bypass the filter altogether via unintended/parasitic circuit paths. If so, the results would very possibly not have been the same with the Benchmark. Regards, -- Al |
Al: I do have oscilliscopes and RF spectrum analyzers and I can assure you there is no RF leaving my Spectral preamp. The Spectral also has a special protection monitoring circuit that is designed to pick up RF leaving the amp. That circuit is not or has ever been active. Plus, I am pretty certain that RF feeding a JC2 would not result in good sound through the rest of my system. Read my recent post on using the Benchmark ADC USB1. I won't repeat that other post here, but basically the Spectral was coupled to the Benchmark which fed digital AES/EBU to the Tascam. If Al is right, the same RF would have been transmitted to the Benchmark and the sonic results would have been the same. But they were not. There was a dramatic improvement in sound quality (and yes, even through the headphones), so I am not buying the mystery signal theory. Clearly, the Benchmark has a better designed and implemented analog front end than the Tascam, and one would expect it given the components used by Benchmark. Components DO matter, and there is a general consensus that some opamps are better than others, and that copious capacitor coupling with electrolytic caps does impact the sound. Why else do people pay $300 for a Mundorf coupling film capacitor? It is also reasonably agreed that an AD797 will sound better than a NE5532. There is no mystery here, and we don't need to make up theories of mystery RF signals to explain the observed differences. |
Dtc, if I'm not mistaken the phono stage Dhl used is built into his Spectral DMC-10 preamp. And he had indicated earlier in the thread that he was using the DMC-10's main outputs, not its tape outputs, to supply the signals to the Tascam. So the experiments that would seem to suggest themselves would be to either record from a different source altogether, with the Spectral disconnected and turned off, or perhaps to record from the Spectral's tape outputs. Regards, -- Al |
Al - your theory would be easy to test. All he has to do is turn off the Spectral pre and record from the Tascam, using the headphone jack as the monitor. Then, turn on the pre and listen to what was recorded. My guess, is he will get the same result. There is other evidence that the input section of the Tascam is not as good as that on other higher priced ADCs. But it would be an interesting and quick little experiment to do. |
FWIW, my suspicion is that the conflicting assessments of the Tascam recorder by Jeff (JWM) and Dhl93449 are both correct. And one look at the system description thread for Jeff's extraordinarily high quality system will make it abundantly clear that his much better assessment of the Tascam is not attributable to any lack of musical resolution in his system. So what could account for the differences in the two assessments? A possibility that occurs to me is that the ultra-wide bandwidth of Dhl's Spectral DMC-10 preamp resulted in spurious RF energy being introduced into the Tascam, which its anti-aliasing filter could not adequately reject. Perhaps because that energy simply bypassed the filter, coupling "around it" and into the A/D circuitry via stray capacitances, less than ideal circuit grounds, or other extraneous paths. Conceivably that spurious RF energy might have been picked up out of the air, from electrical or electronic devices within Dhl's home or even from radio stations. Keep in mind that there have been more than a few threads here in which people have reported actually hearing radio stations through their systems while listening to vinyl sources, while using components having vastly narrower bandwidths than the Spectral. The bottom line: If my theory is correct, while the poor results Dhl obtained with the Tascam reflect less than optimal design, the Tascam would provide significantly better results in most systems than it did in his application. Regards, -- Al |
JWM See the details under my post above on 11/8. Yes, I recorded a cut from an LP at 24/96K in WAV on the Tascam. Converted to FLAC on my PC, but also listened to the raw WAV file as well. Compared directly to my system playing the identical vinyl LP and three other versions ripped from CD to 16/44.1K FLAC. My playback system is Bryston BDP-1 to BDA-1 to Parasound JC2 for digital files; Phono is AT OC9/ii cart/SME IV arm/Technics 150 SLMKII turntable (modded with outboard PS)/Luxman MC transformer/Spectral DMC-10 phono preamp to Parasound JC2. Significant detail in the phono was missing in the recorded digital file. It was obvious on an A/B comparison. |
DTC: I think Benchmark only outputs 24/96K via USB. It will output 176 and 192k via the AES/EBU and digital outputs (including optical). Another reason not to go USB (and record to the Tascam), if you want higher than 96K. However, I was told 96K may be the optimum sample rate and higher rates may not sound as good or better. The other thing I discovered was the Benchmark has much more flexible input gain ranges. You can get up to 29 dB of gain, whereas the Tascam will give you a max of 12 dB. This was an issue due to the low output voltage of the Spectral with a MC cartridge, and the highest signal levels I could get were about -10 dBU at the peaks when recording an LP. The gain is also set in the Benchmark via analog devices like switches, resistors, and pots. The Tascam uses digital methods as you enter the gain numerically in 0.5 dB steps via the LCD screen. |
Sounds like you are making progress. I will say one reason I went with my Korg was to avoid usb. I think you can output to the usb and the Tascam at the same time. It might be interesting to do that and compare the wav files. I believe you can also do 24/192 to the Tascam if that is of interest. Not sure the usb supports that. Good stuff. Hope the Benchmark works. |
DTC: Spoke with an engineer at Benchmark today. Extremely knowledgeable and very straightforward with me. When asked about the analog front end he told me they did extensive listening tests and originally settled on the AD797, but switched to LME49860 which was even better. He agreed the performance of the analog front end was critical to the sound quality. They use one electrolytic coupling capacitor in the entire chain, and it is triple bypassed with film caps and NPO ceramics (which is unusual due the cost of NPOs). He mentioned that the 5532 is long outdated (fine in the 80's) but op amp development has left this part far behind. He agreed that recording to the Tascam would be the preferred method, because some PCs with Windows OS can mess with the sampling frequency (using software driver frequency converters) and you may not even know it. There may also be hardware issues with the USB ports re data reliability. Remember that connecting a USB cable from a computer to an audio device enables high frequency noise an entrance to your audio device (unless you optically couple like Ayre does). The Benchmark has a word clock output that can be hard wired to the Tascam word clock input to synchronize the clocks. The Ayre only has this on the uber expensive pro model. So it looks like I may order one from Audio Advisors because I could return it if I don't like it. Not so lucky with the Tascam as the seller wants a 15% restocking fee, so I may keep it just as a hardware recorder. Also, apologies to TOP TEN for high-jacking his thread. |
Music Direct and Benchmark themselves also have trial periods. I agree I would rather go straight to disk, but if you have a computer near by you can go direct. All you are sending is the bits, the timing is not an issue. For the price of the Tascam you can even buy a PC and then transfer to your main system over Ethernet. You could even use the PC as your music player if you do not have one already. I'll be interested in the results you get from the Benchmark. If you happen to be in the Boston area, you can try my Korg. |
DTC: I suspect the Ayre will work fine. I am leaning toward keeping the Tascam for simply converting digital from an Ayre QA-9 or maybe a Benchmark ADC1 to .wav files, because the variables of the computer conversions of USB data streams are eliminated. I think Audio Advisor carries Benchmark and maybe I can get one to try out. If it does not work, I'd be back to finding and (paying for) the Ayre. |
Rodge827 Unfortunately, the transformers only impact the output of the D/A stage, and have no impact on the input analog buffers (all those op amps) coupled to each other with electrolytic caps. Those buffers are between you signal source and what gets digitally recorded. I can see their argument about the digital hardware only recording without the need for a computer. I may keep it just for that, but still use an Ayre or Benchmark to run digital into the Tascam bypassing its input stages. Although its kind of expensive substitute for a computer these days. |
Steve over at Decware has a modded DA3000 (ZDSD). He had Jensen make special output transformers that are internally mounted. They are coupled to the balanced output stage and convert to unbalanced for output. This allows for the same voltage output as balanced with a pot to adjust the out voltage. There are some positive reviews of the mod on his forum. http://www.decware.com/newsite/ZDSD.html |
Bruce Brown from Puget Sound Studios did an interesting test of 5 A to D converters. This was a DSD comparison. He put the files blind on his website and people downloaded and listened. The thread is long, but interesting to skim through. He provided files from a Merging Horus, Grimm AD1, EMM Labs ADC8, Korg MR-2000(modded) and a Tascam DS-3000. There are some passing very positive comments on the Aye QA-9. There are lots of opinions, but the Merging and Grimm generally came out on top. The Tascam was generally on the bottom. Clearly in this test, price mattered. Bruce Brown Comparison |
Thanks for the update. I typically digitize at 24/192 but from your report, that is no going to make enough difference to make the Tascam viable. I know Chris at Busman modded one. Not sure if he can help. Did you look at the Korg MR-2000? No experience with the Benchmark and I don't read much about it anywhere. John Atkinson did a short review of it and compared it to the Ayre. Benchmark ADC1 Sounds like you may need a Horus but your wallet may not like it. |
Dtc: I got my DA-3000 today. It has a large number of op amps, either 5532s or 2114s (JRCs version of the 5532). No matter. Sound quality is mediocre at best. I did a test recording of Blind Faith "Had to cry today" because I have the UK vinyl LP and three different versions on CD (US Polydor 1986, German RSO 1986, and 2001 "deluxe edition"). The vinyl sounds better than any of the CD versions, so a faithful recording of the LP has a leg up. I recorded a 24 bit, 96K copy of the song direct from the outputs of the Spectral DMC-10 (what I listen to, not the tape monitor) and converted the .wav file to FLAC with dB Poweramp. I then A/B'd with all the CD versions and the LP playing directly. The copy recording made with the 3000 could not hold a candle to the LP. Period. Not even close. It was not even as good as most of the CD versions (although these are all different sounding from the LP). Symbols, in particular, were "mushed" and distorted compared to the vinyl. Actually quite disappointed. Swapping opamps is not really practical due to the large number. And what is worse, it appears all these amps are cap coupled with electrolytic caps. Without a schematic, modding is not even possible. For making copies of rare, vinyl not released on CD, or maybe versions were the CD is unlistenable, maybe the DA-3000 will work. But don't expect that vinyl air and magic to come through in the recording. And I also found out the DA-3000 will NOT record to USB drives, only CF or SD cards. It will not even function with USB HDs. It will only play back from a USB drive. So you have to record your files to an SD card and then convert it or transfer it to a USB drive on another computer. You can transfer files on the DA-3000 directly, but it very complex and slow. Better done on an outboard PC. Looks like I'll be looking for an Ayre. |
Ozzy - Have you tried Vinyl Studio for processing files? It includes a album lookup feature to get timings and names for album tracks, which makes it easy it pretty easy to split tracks. It does take a little time but, for me, it is better than sitting in front of the unit hitting a button. I spend that time working on another album in VS. |
DTC Did not look at the Korg 10R, but it does not appear to be significantly better than the Tascam 3000. It uses a different op amp (OPA1662) instead of the 5532, and has built in phono preamp (which I don't need or want, I'll use my Spectral DMC-10). Plus, it cannot record directly to digital media like a USB drive or memory card like the Tascam can. The Korg must be connected to a computer via a USB cable to save digital files. The Tascam is much more of a stand alone solution. I have not auditioned the OPA1662 vs the OPA627 so I cannot comment on that. |
Topten, Copying albums? Well if it was me using the Tascam, you could set the bit rate to 24 and the sample rate to 192Hz. While playing the albums you would have to sit there listening to when the track ends in order to place a track number to the digital recording. Otherwise, you would have one big long track. The Tascam does have a feature that can automatically place tracks when the sound level lowers to any amount you want to preset. This is supposedly places new tracks in between the songs. But, in practice it doesn't work that well. So, I do it manually like I stated above. Kind of time consuming, but when its done you will access to the tracks individually. |
DTC The QB-9 is the D/A. The QA-9 is the A/D. I have yet to see a QA-9 for sale. I went ahead and bought a TASCAM 3000 on fleabay for $720 new. For that price I can afford to try some op amp mods. I think I can remove the 5532s and replace with OPA627s on a Brown Dog adapter. For my purposes, only the input buffer amps need to subbed out. Once I have the digital file, it will be played back in my Bryston BDP/BDA combo which uses Class A discrete analog stages. If in the future a QA-9 becomes available, I'll consider it at that time, depending on the results I get with the Tascam. |
Dhl93449 - there is an Ayre QB-9 listed for sale here. The B is the successor to the Ayre with DSD, but opnly does 88 MHz DSD. There is a Korg MR2000 available here for $900. Busman does mods to those. They did mine and the recordings at 24/192 are very comparable to my turntable. The Korg records to a disk but than you can transfer to a PC via USB. That unit is no longer made, but was a "low end" professional unit. |
Topten: I realize that, and was responding to your question about the use of a "CD" recorder for capturing vinyl LPs. CD media is limited to 700 MB. By CD recorder I am assuming you mean recorders that directly write to CD media. Since most LPs are limited to < 20-25 minutes per side, you will only exceed this for a single 2 GB file for DSDIFF at the 5.6 MHz sampling rate! This is way overkill for LP recording, especially with IC opamp buffers in the input circuitry (they will limit the sound quality, not the digital capture conditions). In fact, with 2 GB, you can record a 26 min .WAV file at 24bit, 192KHz. Since the longest single, uninterrupted song on a LP is usually under 20 min, there is no need to worry about splicing files larger than 2 GB together. Also, those instructions are only for playback on the Tascam. I would convert the .WAV files to FLAC for playback on my music server and would never use the TASCAM for anything but the original A/D conversion. |
DhI93449, The largest file that this unit will create during record is 2GB. If a recording is about to exceed this size, a new file is automatically created, but the audio is processed so there is seamless continuity between the current and new file. On playback, in order to hear the audio seamlessly, please set the TRACK GAP mode setting to “GAPLESS”. (See “Setting gapless playback mode” on page 41.) Is this solution to the 2GB limitation: GAPLESS is selected during DSD file playback Depending on the status of the audio data, noise might occur at the division between tracks due to the nature of DSD files. If the noise is bothersome, set the track gap mode to NORMAL to avoid it. (Noise will not occur during the gapless playback of audio from tracks created using this unit’s auto track and track increment functions, for example.) |
topten: Also, MD (Sony mini-disc) is not a loss-less format. It is compressed, similar to MP-3. I also have mini-disc and would never equate this format with high end audio. So it would not surprise me that anything recorded to MD will sound flat and lacking in detail. I have many recordings ripped from CD to Sony MD, and they don't even compare to the CD quality originals. I did it because MD is convenient on a portable device were I did not want the larger sized CD. I think it is better than MP-3. |
topten: The TASCAM does not have the file size limitations of a "CD" recorder. You can build any size file if the media (USB thumb drive, HD, or memory card) is large enough (but will run into 2 GB file size partition issues since they use FAT file partitions). Neither does the Ayre QA-6, because you stream USB digital to a computer. Audio quality does become an issue, depending on how critical you are. Fremer did a review of the Ayre QA-6 recently in March 2015 along with some other A/D converters such as the PS Audio Nuwave, Lynx Hilo, and M2Tech Joplin: http://www.analogplanet.com/content/analog-digital-converter-files-identified-converters-profiled#wCSH81I66K8VFGXx.97 |
DTC The Ayre QA-9 you mention does have discrete (assumed class A) circuitry. It's easily double the cost of the Tascam and I am not sure if it's still available. Could not find any listed on 'Bay. The other issue is that the QA-9 requires a stand alone computer to formulate the final .WAV or .FLAC audio file, which is not that convenient. The TASCAM is a stand alone solution, which produces .WAV files on memory cards or USB media directly. Did discover that Brown Dog does make a SO8 to SO8 surface mount adapter to allow two OPA627 op amps to replace a 5532 dual. In my tape loop listening tests with the Spectral, I have found the OPA627 to be pretty close to the vinyl source, with a small amount of loss of SQ. For recording of vinyl for casual listening playback, this may suffice, assuming the BD adapters don't create stability problems with the 627s. |
Perhaps I'm not understanding your question but,... I own and use the Tascam 3000. I use a 32gb SD card with it and I can record many Hi-Rez files on it. Mostly from Blu Ray or DVD-A using the analog outs. When the SD is full or convenient I transfer the music to an external 2TB Hardrive. If your just copying CD's to a hard drive, the Tascam is not the way to go. As mentioned earlier you would want to cut the ripping time down to about 10 minutes or less per CD. |
