Hey Craig, who at Tascam do you bug/whine to with new ideas?

[bookumdano4]

Since usb3 allows for massive pure power that usb2 doesn't, I'd like to see a new usb3 bus-powered audio/midi interface from Tascam with...1.. pad switches and 2.... at LEAST a 1watt power circuit for the headphones. 5watts would be better, but 1watt will do. Since the usb3 spec allows for ungodly power transfer... what is it... 100watts or something?.... it's time imo for riddance of the anemic mw headphone circuits. I need more volume in my phones!!!! My 7506s have great sensitivity so I just need more watts. Or is that "more whats?"

 

Yeah, my laptop's battery will zap out a bit faster, but I can live with it. The uh7000 headphone circuit sounds SO nice and loud. Now I just need that in a bus-powered single box.

 

Maybe a big brother to the us366... since I mostly like the I/o setup on that little box... although I'd personally dump the dsp stuff.

 

As an aside, I also want a reintroduction of the 1965 or whatever Firebird electric 12 string. Best elec 12 I ever tracked in my life in the early 1970s. I still solo that guitar on my multitracks just to figure out how I'm EVER gonna get that sound again. But don't put a Tascam name on that one please.

 

Thanks Craig. If I don't see this stuff under the Christmas tree next week, I'll look for it at Namm in four weeks.

[MikeRivers]

I hate it when a company comes out with a new product that I'd like to use, but I have to get a new computer in order to use it. And if all I needed was more headphone driving power, I'd rather deal with a headphone amplifier in another box than to re-build a working Windows XP computer system around a new computer (which will probably have Windows 10) just to get more current out of the USB hole.

 

Focusrite has been putting pad switches on their interfaces for quite some time now, though they're software controlled. I rarely record anything where I don't need pretty close to full gain, so personally I don't need pads for the mic preamps.

 

Headphone amplifiers are getting better, but unless you re-purpose a power amplifier as a headphone amplifier, I don't think that even the high class ones from SPL or Grace can put out much more than about 1 watt. (I just looked - the Grace M903 DAC/Headphone Amp puts out 1.3 watts into 50 ohms. I think the Sony 7506 phones are around 60 ohms).

 

When you see a headphone output spec these days, it's getting more likely that you'll see it as volts, not watts, for the sake of "mobile devices." They make the source impedance extremely low, and rely on low impedance (15 ohms or lower) earphones, to get reasonable power from the typical 3 volt power supply in those devices. And with semi-pro gear, the tendency is to design around the same components that are sold by the gazillions to the phone manufacturers, because it's convenient and they usually work pretty well.

[bookumdano4]

The Grace amps won't work as those are ac. I'm ready for a nice, high powered headphone amp (1 to 5 watts) on bus-power for when I travel.

 

Physical Focusrite pads are cool but the only bus powered box they do that on (i2x4 or whatever it is) doesn't have much in the way of I/o. I like the little us366 because the 6 ins or 6 outs are really useful for me, especially since some of them are optical and play nice with Cubase. But I'd like a couple of physical pad switches.

 

There's some $90 Behringer interface that has physical pads and a lot of I/o but I haven't followed up on the scant info I've heard about it so far... no doubt though that the headphone circuit is anemic ... like they all are so far for my purposes. Except ac powered stuff.

 

Pretty much every single brand of every single bus-powered interface is currently at 10 or 18 milliwatts on the headphone circuit. Which with usb2, I understand. Like trying to sample a grand piano into an old Ensoniq Mirage. You gotta make everything work at less than 2.5 or 5 volts (or whatever usb2 is limited to).

 

It would be cool to have a usb-powered circuit that would maybe deliver 1 or 2 watts into the headphone circuit when usb3 is the connection. Or still stay at the current 18mw when the interface senses it's only getting 5 volts of power from an older usb2 port.

 

On the other hand, I don't claim to know what I'm talking about.......

 

other than I WANT at least 1-5 watts on my headphone circuit on (preferable) one connected interface when I'm out and about.

[Anderton]

I looked at specs for headphone amps and curiously, hardly any showed specs...even something like Aphex's HeadPod 4, which is all about delivering lots of output, doesn't show the wattage. Curiously, a lot of interfaces do show the wattage. There are some headphone amps for audiophiles that go up to 6 watts, but they're super-pricey and it's expected you're using them at well below the max wattage.

 

Interestingly the UH-7000 power which you find loud enough is 45 mW per channel. The new US-20x20 USB 3 interface is AC-powered, it has class A preamps so it draws 20 Watts and does 70 mW per channel into 32 ohms. If I turned them up all the way, I think my ears would bleed...then again I'm using KRK KNS-8400 headphones, which are 36 ohms. You'd find them a lot louder than Sony phones, which are 60 ohms.

[MikeRivers]

The Grace amps won't work as those are ac. I'm ready for a nice' date=' high powered headphone amp (1 to 5 watts) on bus-power for when I travel.[/quote']

 

I only used that as an example, though I didn't realize from your initial post that you were looking for something independent of mains power, and a full out multi-channel interface.

 

Pretty much every single brand of every single bus-powered interface is currently at 10 or 18 milliwatts on the headphone circuit. Which with usb2, I understand. Like trying to sample a grand piano into an old Ensoniq Mirage. You gotta make everything work at less than 2.5 or 5 volts (or whatever usb2 is limited to).

 

This isn't a USB 2.0 limitation, it's a what-the-manufacturer-thinks-you-want-to-use-it-for limitation. A USB 2.0 port can provide up to 500 mA at 5 volts. That's 2.5 watts. But the manufacturer figures you're going to use it with a low power device and they don't want you to be using all of your battery to power one external interface.

 

Understand, too, that if your laptop computer has three USB sockets, they're most likely on an internal hub connected to a single port. Unless it's a powered hub (something I've never heard of, probably because they want to save your battery), you can't get 500 mA out of each of the three sockets, it's 500 mA total.

 

On the other hand, I don't claim to know what I'm talking about.......

other than I WANT at least 1-5 watts on my headphone circuit on (preferable) one connected interface when I'm out and about.

 

You may think you know what you want, but you may not know why you think you need it. There's an article on my web site that talks about headphone amplifiers. Think about your headphones, too, understand their specification for sensitivity, and then consider whether you really need 5 watts, or even 1 watt.

 

You'll be amused to know that there's a new requirement for the CE certification that headphone outputs have to warn you when they think you're headphones are getting too loud. They don't know what kind of headphones you're using, so I guess they just pick some nominal sensitivity and impedance. So, whatever the gadget's rated headphone power output is as long as it's above a few mousefarts, don't be surprised if, on your next gadget with a headphone output, it pops up a warning that getting any louder might cause hearing damage.

 

[bookumdano4]

I'll try the krks. I've been locked into the Sonys for so long that I never bother to check others. As to usb2 spec, I figure that if manufacturers put a 1 watt amp in the box, it would start scrunching what they can do with the remaining bus power.... like enough juice for phantom power, little leds etc. Probably why we also don't see 24i/0 interfaces with phantom power on each of the 24 pres and bus powered.

 

I just figure usb3 solves all that with the bus power spec now raised to 100watts or whatever. But as you say Mike, the manufacturer decides what to put in there.

 

My 1watt+ ref is based on what I get in the studio. My old 3700 console produces 2watts at the headphone circuit and my ancient rackmount Model 1s produce 1 watt at the back headphone circuits. Those are so nice when I use them. My various laptop interfaces are at about 18mw.... I actually had to call NI to find out their rating on one of their boxes and it took them some time to figure it out.

 

When I listened/auditioned the uh-7000, the first thing I did was route audio to test its headphone circuit....ah.... nice. So for my ears, even 45mw+ would be nice I guess

 

But I will definitely try the krk once I find a pair. SoCal GC and Sam Ash don't have them at the moment. Haven't tried the Hollywood stores yet.

 

By the way, my phone does warn me when I crank it up to 11 to listen to certain mixes. I suppose its at about 10mw.

 

[bookumdano4]

The new US-20x20 USB 3 interface is AC-powered, it has class A preamps so it draws 20 Watts and does 70 mW per channel into 32 ohms.

 

Just out of curiosity from a purely spec perspective, do you suppose the 20x20 could have been designed to be bus-powered on a strictly usb3 port? Drawing 20watts over a usb3 port is a cinch based on the spec... it seems.

 

But alas....it has to be ac powered instead in order to also be useable on older and more plentiful usb2 ports. That's the sense I get on the ac design.

 

I sense that marketing analysis isn't going to reveal the need for usb3-only version interfaces any time soon. No matter how many goodies we users could pick up.

 

 

[MikeRivers]

 

Just out of curiosity from a purely spec perspective, do you suppose the 20x20 could have been designed to be bus-powered on a strictly usb3 port? Drawing 20watts over a usb3 port is a cinch based on the spec... it seems.

 

Just because the spec allows for much more power doesn't mean that they're going to be building battery powered laptop computers that can supply that much power very soon. You might find yourself carrying around something as heavy as a car battery. I think the "full powered" USB ports will be reserverd for computers that are plugged into mains power.

 

I sense that marketing analysis isn't going to reveal the need for usb3-only version interfaces any time soon. No matter how many goodies we users could pick up.

 

I've seen a couple of USB3-only interfaces already, but it's not because of power, it's because of data I/O transfer speed. While it's nonsense to think that USB 3 speed will improve the performance of a piddly little 2- or 4-channel interface (which is what I've seen in this category so far), the manufacturers read the low latency hype presented by the folks who make Thunderbolt (only) interfaces and with USB 3, they can now keep up without restricting their customer base to Mac and a very few PC users.

 

I have a quote stuck over my desk that's appropriate here: "Hype hides alternatives. It supports the notion that problems have single solutions. Worse, it ignores inevitable trade-offs and the doctrine of unintended consequences."

 

 

 

 

[Anderton]

 

Just out of curiosity from a purely spec perspective, do you suppose the 20x20 could have been designed to be bus-powered on a strictly usb3 port? Drawing 20watts over a usb3 port is a cinch based on the spec... it seems.

 

USB 2 provides up to 500 mA at 4.4 to 5.25V, while USB 3 can provide 900 mA at the same voltage range. So the most wattage you could get from a USB 3 port is 4.725W max - not enough for your 5W headphone amp. A dedicated USB charging port can deliver 1,8 amps.

 

I sense that marketing analysis isn't going to reveal the need for usb3-only version interfaces any time soon. No matter how many goodies we users could pick up.

 

It depends on how much "feel" matters to you with virtual instruments and amp sims. I've been testing out the TASCAM US-20x20 and am getting 9 ms round-trip latency (1.5 ms sample buffer) with 64 samples at 44.1 kHz on pretty complex projects. Playing guitar through an amp sim is a very different experience, I really don't feel much difference compared to the 3 ms or so you get with Thunderbolt. OTOH if all you do is track bands and mix, USB 3.0 isn't necessary.

[MikeRivers]

 

I've been testing out the TASCAM US-20x20 and am getting 9 ms round-trip latency (1.5 ms sample buffer) with 64 samples at 44.1 kHz on pretty complex projects. Playing guitar through an amp sim is a very different experience, I really don't feel much difference compared to the 3 ms or so you get with Thunderbolt.

 

That's really a big difference. I would have guessed that most of the 9 ms was how long it takes for the simulator to generate the output from the input. But if moving from USB 2 to Thunderbolt really reduces 9 ms latency to 3 ms, that's impressive. I wonder if some of that 9 ms is a fixed buffer as part of the USB driver (at the data interface itself, not the usual A/D or D/A sample buffer). Some drivers incorporate such a buffer without telling you about it, to assure that you won't complain to Tech Support that your interface is skipping and hiccupping. Others let you turn it off and see if it works that way.

 

 

[Anderton]

 

But if moving from USB 2 to Thunderbolt really reduces 9 ms latency to 3 ms, that's impressive.

 

Actually I'm talking about moving from USB 3 (with 64 samples buffers) to Thunderbolt. Due to the complexity of the project, I have to use 128 sample buffers with USB 2 so that increases latency considerably.

 

I wonder if some of that 9 ms is a fixed buffer as part of the USB driver (at the data interface itself, not the usual A/D or D/A sample buffer). Some drivers incorporate such a buffer without telling you about it, to assure that you won't complain to Tech Support that your interface is skipping and hiccupping. Others let you turn it off and see if it works that way.

 

I'm pretty sure that's the case, because the time contributed by the sample buffers by themselves is 1.5 ms. Regardless, the difference between 3 ms and 9 ms is not really noticeable when you're playing in real time. Typically monitor speakers are 4 ft away from your ears, so if you're monitoring on headphones, with Thunderbolt you have less latency than listening to monitor speakers and with USB 3, 5 additional milliseconds.

 

9 ms may be something drummers would notice, because they're sitting in the middle of their kit and get maybe 2 milliseconds between the time they hit a snare and they hear it. With guitar players typically being several feet away from an amp, we're already used to at least 6-10 ms of latency. So 9 ms isn't that big a deal, comparatively speaking. Given that Thunderbolt on Windows just isn't happening, USB 3 with Windows 10 (definitely snappier, I'm becoming a fan) is a good alternative.

 

 

[bookumdano4]

Okay, so I guess what I want for Christmas is a set of KRK phones, the 6i/o us-366"B" which includes two physical 10-20db pad switches for inputs 1-2, and internal sensing of a usb3 port/power which then jacks up the headphone output range to 50mw where appropriate voltage is present. All bus-powered. And dump or upgrade the dsp stuff with the usb3 sensing. Japan should be able to whip that up in the next two or three days, right?

 

And I want that Firebird.

 

As to Thunderbolt, I get the impression MS is within weeks or month max of official support on Windows10... certainly by Namm where MS is beginning to have a physical presence in the form of Pete whats-his-name... who partners with all the interface etc guys in a MS capacity. My money is on usb3 though for the moment in the form of actual audio interfaces for the masses longterm. Or the Ethernet stuff for higher end. Or of course, I can easily be way off on the intuition.

[MikeRivers]

Okay' date=' so I guess what I want for Christmas is a set of KRK phones, the 6i/o us-366"B" which includes two physical 10-20db pad switches for inputs 1-2, and internal sensing of a usb3 port/power which then jacks up the headphone output range to 50mw where appropriate voltage is present. All bus-powered. And dump or upgrade the dsp stuff with the usb3 sensing. Japan should be able to whip that up in the next two or three days, right?[/quote']

 

Dreamer! No matter how many times a company says "We listen to our users" they still pick a price point and stuff everything they can into it at that price. You can't convince them to make one that costs $50 more just because you're willing to pay for it.

 

As to Thunderbolt, I get the impression MS is within weeks or month max of official support on Windows10... certainly by Namm where MS is beginning to have a physical presence in the form of Pete whats-his-name... who partners with all the interface etc guys in a MS capacity.

 

The problem in the Windows world isn't just the operating system, it's the hardware. And that's the case with Thunderbolt, too. It's convenient that all new Windows computers seem to be coming with USB3 ports now, and have for most of this year, but there are a lot of perfectly servicable computers that are 3 years old (I have a couple that are 10 or more years old) that don't have USB3, and most of them don't have a slot for a USB3 accessory card (which probably won't fullly support USB3 anyway). If the interface isn't backward compatible with USB2, they can't sell it to a user who isn't ready to get a new computer and housebreak it.

 

My money is on usb3 though for the moment in the form of actual audio interfaces for the masses longterm. Or the Ethernet stuff for higher end. Or of course, I can easily be way off on the intuition.

 

I don't know why there isn't more low end stuff coming along that uses AVB. Every computer made since about 1995 has an Ethernet port, and there's a free "virutal sound card" driver. It's a little fussier to set up than Dante, but for a simple system - one interface connected to one computer - it's no harder than installing a driver and plugging in a USB cable.

[MikeRivers]

 

Actually I'm talking about moving from USB 3 (with 64 samples buffers) to Thunderbolt. Due to the complexity of the project, I have to use 128 sample buffers with USB 2 so that increases latency considerably.

 

I'd like to play with Thunderbolt, but I don't have a computer with a Thunderbot port and don't anticipate getting one in the near future. I'd like to play with AVB, though. MOTU has a USB/AVB interface or two. I should hit them up for one for a review.

 

I'm pretty sure that's the case [the USB port buffer], because the time contributed by the sample buffers by themselves is 1.5 ms. Regardless, the difference between 3 ms and 9 ms is not really noticeable when you're playing in real time. Typically monitor speakers are 4 ft away from your ears, so if you're monitoring on headphones, with Thunderbolt you have less latency than listening to monitor speakers and with USB 3, 5 additional milliseconds.

 

I'll take your word for it about latency when playing. Usually when I'm reviewing an interface with a USB port, I'll load up a bass or piano plug-in, connect a keyboard, and verify that it works. I don't notice latency, but then I'm not trying to follow a part. I do remember measuring it once, by connecting the analog output of the keyboard (assuming that latency between pressing a key and getting a sound is negligable) to one channel and the monitor output of the virtual instrument to another channel and measuring the difference between them. I don't remember what I measured, but it was around 10 ms. This was probably with USB2 on a Windows XP Pentium 4 running Reaper.

 

The latency that bothers me is when I'm listening to my own voice on headphones. Any more than about 0.5 ms will give you comb filtering unless the headphone volume is loud enough to swamp out the direct sound through your head (which I suspect is why most people have never noticed it). Direct analog monitoring is the best solution there. Even "a zero latency DSP monitor mixer" in the interface hardware is often that only for large values of zero. It's getting better though. I've measured a couple of tenths of a millisecond on the last couple of Focusrite interfaces I've had in here.

 

I don't think I've ever heard of a guitar player giving a thought to "through the air" latency even though it can easily be on the order of 10 ms. My Casio MIDI guitar is a different story, however. I never got the hang of anticipating its delay when trying to play along with someone else, or a recorded track. But that guitar is older than any of my old computers.

[bookumdano4]

 

Dreamer! No matter how many times a company says "We listen to our users" they still pick a price point and stuff everything they can into it at that price. You can't convince them to make one that costs $50 more just because you're willing to pay for it.

 

 

Well in that case, I forgot to ask for the mic pre input controls to offer 80db of clean gain