Tascam 102i Review (Audio Interface)

[AnalogSteph]

Looks like a typical CS4272 job, and possibly a slightly older construction as well. The newer 2x2HR should give generally more encouraging results if Julian Krause's review is anything to go by, though headphone output performance remains a bit naff.

The 102i costing $399 must be related to other features like the S/MUX 8-channel digital input (is that ADAT?), DSP and MIDI I/O, not analog performance. A 2x2HR is $179.

 

[dualazmak]

The US-2x2HR seems to have marginally better THD+N measurements (based on company's specs), but the fundamental problems likely remain the same to make much of a difference
US-2x2HR | SPECIFICATIONS | TASCAM | International Website| | SPECIFICATIONS | TASCAM | International Website| | SPECIFICATIONS | TASCAM | International Website| | SPECIFICATIONS | TASCAM | International Website
SERIES 102i | SPECIFICATIONS | TASCAM | International Website| | SPECIFICATIONS | TASCAM | International Website

Just a reference for you and other people here, I use US-1x2HR for Fq response measurements in DAC's analog out level, in amplifiers' SP output level (together with nice High-to-Low converter AT-HLC150), and also in room sound at listening position with Behringer ECM8000 mic, as I shared recently in here through here.

 

[phoenixsong]

Just a reference for you and other people here, I use US-1x2HR for Fq response measurements in DAC's analog out level, in amplifiers' SP output level (together with nice High-to-Low converter AT-HLC150), and also in room sound at listening position with Behringer ECM8000 mic, as I shared recently in here through here.

Based off the company's specs the 1x2HR actually seems to slide a bit compared to the 2x2HR in performance

 

[Francis Vaughan]

Does this mean it's possible they just tapped one of the line outs for the headphones ?

Probably not. Indeed this output impedance is to be expected.

We have noted before, devices for studio or home studio use commonly place a resistor in series with the output. RME and Apogee use 10Ω on similar devices.

There is even standard. IEC 61938:1996 recommends 120 Ω for the output impedance of headphone amplifiers.
This appears to come from studio headphone distribution amplifiers, amplifiers designed to run multiple headphones at once.
In studio settings it is common to have everyone listening on headphones during tracking. Rather than a dedicated headphone amplifier per artist, they used (and often still use) a single amplifier fanned out to the outputs via resistors. Heapdhone manufacturers for studio work would be expected to balance their products frequency response for this output impedance. So, chicken and egg. There is an expectation that in studio settings, headphones used will be driven by a signifacant output impedance, and the headphone design will have been balanced for this, so studio headphone amplifiers provide this. Deliberatly.

Looking at a few studio headphone distribution amplifiers, PreSonus uses 51Ω across their range, Mackie uses 3Ω. The Mackie documentation explicitly shows the presence of resistors on the output. Behringer and Art don't list theirs.

The Simon headphone distribution box is still made - and indicates where all this comes from.

 

[AnalogSteph]

This IEC standard is pretty old hat by this point, I don't think anyone has been taking it very seriously in the last 20 years. The last headphones I can think of that were probably designed with this in mind (or even the 220-470 ohm jacks of typical integrated amps) are Beyerdynamic DT831 and DT931, and those are from what, the early-mid '90s or something?

In an interface like this, high output resistance is really just about keeping crossover distortion in the poor little opamp somewhat at bay (the 3rd gen Focusrite Scarletts are using an NJM8065, hardly a born headphone driver), as well as limiting the output current to keep the power supply section from crapping itself under load in bus-powered models (which the 102i is not).

BTW, the price difference between the 2x2HR and 102i is substantially less outlandish on this side of the pond, at 169€ vs. 269€ at Thomann. Not sure why the 102i is that expensive in the 'States.

 

[dfuller]

Looks like a typical CS4272 job

I'm continuously amazed $400 interfaces are still using CS4272s - the BLA Revolution uses one too. That's especially horrid considering at a lower price point Audient managed to cram in an AK5574 and a CS43198 into the iD14 Mk2, all while not sacrificing their mic preamp design which is certainly more costly to implement (being mostly discrete and all) than a basic op-amp based one.

 

[Andres]

To quote from the initial post conclusion "ADC however can't even clear the 16 bit hurdle with 4 volts in" - can anyone clarify if this means that the interface should perform better at a higher bit depth and sample rate?

 

[Moderate Dionysianism]

To quote from the initial post conclusion "ADC however can't even clear the 16 bit hurdle with 4 volts in" - can anyone clarify if this means that the interface should perform better at a higher bit depth?

Nope, it means that due to high distortion, the ADC is not able to capture the 16bit dynamic range (96dB).

 

[Andres]

Nope, it means that due to high distortion, the ADC is not able to capture the 16bit dynamic range (96dB).

Well, that doesn't sound too great

 

[AnalogSteph]

Nope, it means that due to high distortion, the ADC is not able to capture the 16bit dynamic range (96dB).

That's not accurate either. It merely means that calculated SINAD does not clear the maximum possible with 16 bits, which would be noise-limited to about 98 dB(A). Now SINAD is not the very most useful metric, as it invariably lumps together apples and oranges. It only tells you that you'll be good if it clears a certain hurdle, not that you won't if it doesn't.

Besides, distortion would be anything but constant with input level (it being mostly H3 here, I'd expect 2 dB less per level dB reduction), which is one major reason for SINAD vs. level graphs looking like they do.

Chances are you'll never hear the difference between an input with 106 dB worth of dynamic range and -92 dB worth of distortion and one with unmeasurable distortion levels at the same dynamic range. However, 92 dB of dynamic range but -106 dB worth of distortion would be substantially less useful in practice, as noise would be encroaching much sooner.

 

[JeffinMaine]

The channel balance seems to be off for the DAC part?

I just bought one of these despite the bad review, simply because it was a cheap ($180) way to get an ADAT input to use with existing outboard gear for recording in Cubase. I also need the DAC for monitoring, and I have noticed that on my 102i, the channel balance is off for both the balanced outputs and headphone jacks. The left channel is slightly louder, which is incredibly annoying. I think this review shows the same imbalance, and I'm wondering if maybe every unit has this same flaw?

Also, I'm assuming I would be correct that my old Apogee Rosetta AD would be better than the ADCs on this device? I'm positive that my Great River MP-2H preamps must be far better than those on the 102i, so if I keep the Tascam, I will only use the ADAT input.

I'm considering returning the unit, but if I run it into my existing mixer, I could fix the channel imbalance. The bummer is that the unit becomes basically useless as a portable device because I can't live with the headphone outputs being so imbalanced.

Any recommendations for an inexpensive usb audio interface that would provide ADAT input and decent DACs? I could potentially spend more if the recommended device was sure to provide good results.

 

[AnalogSteph]

Also, I'm assuming I would be correct that my old Apogee Rosetta AD would be better than the ADCs on this device?

The converter chips are in the same ballpark (CS4272 ADC specs are little worse than the CS5361's, and I presume that the converter design is much the same), but the Rosetta is likely to have a much more complex analog stage, it wasn't exactly cheap after all. For one thing, you can fine-tune analog stage gain for 0 dBFS level from +2 to +26 dBu, so it will handle even big studio levels with ease while being equally suited for consumer sources.

Any recommendations for an inexpensive usb audio interface that would provide ADAT input and decent DACs?

Audient iD14 (MkII or MkI) looks like a good candidate.

 

[Ifrit]

They need to keep portastudio reputation in modern days, apparently.

 

[JeffinMaine]

The converter chips are in the same ballpark (CS4272 ADC specs are little worse than the CS5361's, and I presume that the converter design is much the same), but the Rosetta is likely to have a much more complex analog stage, it wasn't exactly cheap after all. For one thing, you can fine-tune analog stage gain for 0 dBFS level from +2 to +26 dBu, so it will handle even big studio levels with ease while being equally suited for consumer sources.

Thanks for this analysis of the converters. This confirms that I only really need an interface with good dac and ADAT input.

I've realized that it's really uncomfortable while recording to not have the detailed meters of the apogee + the all important overage indicators that stay lit so you know if you ever went over. That alone is enough to make me stick with the rosetta.

Audient iD14 (MkII or MkI) looks like a good candidate.

Thanks! I will take a look. I wish MOTU made a cheaper interface that had an ADAT input. I like the look of the ultralite mk5's digital i/o section though I am a little concerned about the windows asio drivers - based purely on what I'm reading, it sounds like they give more attention to the stability of the mac drivers?

 

[Bergante]

I got the 208i model for one reason: with two optical ADAT inputs it can take 16 additional input channels at 44.1/48 kHz, 8 channels at 96 kHz.

In my case I am adding two Midas XL48 8 channel microphone preamplifiers. Very few audio interfaces (and I think none at this price point) offer so much I/O.
The only alternative I found was the RME Digiface USB which seems to be very rare now thanks to component shortages. And this one offers four mic preamps.

Another advantage I´ve seen is, you can defeat the 75 ohm terminator for the Wordclock input. Very few equipment offers that option.

So for me it´s more like a fancy ADAT network card rather than an audio interface proper.

Oh, another plus. It´s a fully class compliant device. Which means you can use it to record to an iPad (or, for instance, use it as a very compact
mixer using an iPad).

 

[teashea]

This is a review and detailed measurements of the Tascam 102i Audio Interface (DAC, ADC and headphone amplifier). It is on kind loan from a member and costs US $399.

I like the departure from usual designs by having the unit tilted up some:

View attachment 139990

Here is the back side (I put spacers under it to make it easier to take the picture):
View attachment 139991

The DAC output was surprisingly poor with stock class driver. I downloaded the image from Tascam which remedied that. Download speed was quite slow though. They need to host this someplace better. Once there, I lost all output. Had to figure out the cryptic routing UI to turn the computer output back on for each channel.

Note that my testing is focused on performance in specific areas. I am not testing functionality of these interfaces.

Tascam 102i DAC Measurements
As always, we start with our DAC output with volume adjusted to 4 volts out:

View attachment 139992

Not something to write home about but not broken either:

View attachment 139993

Letting the output swing out more doesn't get you better performance:

View attachment 139995

Dynamic range is similar:

View attachment 139994


IMD test shows surprisingly high level of noise when the signal is low:
View attachment 139996

Jitter suppression is good but there could be bad stuff buried in higher than wanted noise level:
View attachment 139997

The filter is unfortunately the typical, incorrect one:

View attachment 139998

While this is very common, I wish at least pro products implement this correctly with faster cut off.

Linearity could be better:
View attachment 139999

Last test was quite a puzzler at first:
View attachment 140000

Surely we don't have half a percent noise and distortion! Let's find out the cause:

View attachment 140001

This is the typical noise shaping we used to see in older DAC chips. Noise in the audible band is reduced by pushing it into ultrasonics. Native performance needs to be a lot better so this is not needed or at least not needed to this degree.

Finally multitone shows that higher frequencies produce more distortion:
View attachment 140002

Tascam 102i ADC Measurements
Here is our dashboard:

View attachment 140003

Pretty disappointing performance:

View attachment 140004

Distortion is the problem. Noise is much better:

View attachment 140005

Linearity is not great:

View attachment 140006

The high distortion seems to be frequency independent:

View attachment 140007

Tascam 102i Headphone Measurements
Here is our dashboard with volume adjusted for 4 volt out:

View attachment 140008

Not bad. Let's load it down to 300 ohm and see what we get:

View attachment 140009

Very strange response. Not sure what is going on. Maybe it has a dual stage buffer or something? Regardless, we have OK amount of power.

Switching to 33 ohm causes the power to drop way down:

View attachment 140010

Suspecting that it is due to high output impedance and I measured it at 85 ohm. This some of the highest I have ever measured and falls in the category of truly awful! You not only lose a lot of power but the frequency response of the headphone can change resulting in tonality variations.

Conclusions
Performance of the DAC in Tascam 102i is reasonable. ADC however can't even clear the 16 bit hurdle with 4 volts in. Headphone amplifier is just junk with high output impedance.

Based on performance measurements I perform, I can't recommend the Tascam 102i. For a current product, it simply is not very competitive.

-----------
As always, questions, comments, recommendations, etc. are welcome.

Man these interfaces take a ton of work to measure. So many subsystems. My fingers are hurting!!!

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Too bad that Tascam failed on this one. One would expect better.

 

[computer-audiophile]

Looks like a typical CS4272 job, and possibly a slightly older construction as well. The newer 2x2HR should give generally more encouraging results if Julian Krause's review is anything to go by, though headphone output performance remains a bit naff.

The 102i costing $399 must be related to other features like the S/MUX 8-channel digital input (is that ADAT?), DSP and MIDI I/O, not analog performance. A 2x2HR is $179.

I ordered this interface today because I want to activate my electric guitars a little more, and I need an interface for that. I don't have a practice-amp or headphone-amp for this purpose here anymore.

The main reason for choosing this model was the alleged Linux compatibility and the practical design of the connectors on the back with cinch, which makes it compatible to my HiFi-set. Besides, it doesn't cost much. Let's see if that works well. I'm also curious if the audio quality is good enough when I use it as a USB interface to listen to music. TASCAM is a brand I trusted a lot decades ago.

Measurements of this Unit I did not find yet.

 

[computer-audiophile]

I know Linux compatibility isn't important to many because they use other operating systems. This could be the reason why many manufacturers do not provide any information about the Linux compatibility of their devices, mainboards, etc. in their specifications. That's why I'm always grateful to find information about it on the net. It doesn't matter whether I want to build a barebone PC or just use a USB sound interface with Linux, as in the current example.

I can now say that the Tascam US-1x2 HR works very well. With Ubuntu 22.04.1 LTS, the USB device was recognized immediately and I was able to listen to music at my near-field listening position. The sound in CD resolution is pleasant and subjectively comparable to the previously connected Topping E30II.

I use Audacity as the recording software. I plugged in my electric guitar and was able to get started straight away with the default settings.

If I want to record the sound from PC, I need to use Pulse Audio (pavucontrol GUI) to select the appropriate channel to make it work in Audacity.

What I don't like about the device is the MINI-USB connector. This could be a reliability issue in the future.

 

[AnalogSteph]

What I don't like about the device is the MINI-USB connector. This could be a reliability issue in the future.

Actually, it's not a mini, it's type C plus Micro-B for aux power. Mind you, the point still stands, I'm not a huge fan of Micro-B either (seems to be kind of a weak point on any smartphone that has them). You're not likely to put all that many cycles on this one though, so it may well hold up OK.

 

[computer-audiophile]

Actually, it's not a mini, it's type C plus Micro-B for aux power. Mind you, the point still stands, I'm not a huge fan of Micro-B either (seems to be kind of a weak point on any smartphone that has them). You're not likely to put all that many cycles on this one though, so it may well hold up OK.

Yes, this is more precise. What I find amazing about my smartphone, which I often charge via this plug, is that it works so reliably.
With the TASCAM being a larger tabletop device, however, there is no need for such extensive miniaturization. The connectors on the front are rather robust.