Help me choose an interface please

[CherylJosie]

For home studio/bedroom solo recording and podcast where there's already a Zoom Q2N HD camcorder and an iPhone 11 pro available with several small mixers to choose from --

Any opinions on the Zoom interface:

UAC-232 Audio Interface

32-Bit Audio Interface

zoomcorp.com

versus the Presonus:

PreSonus® AudioBox USB® 96 25th Anniversary | Interfaces

PreSonus® AudioBox® USB® 96K 25th Anniversary, Black

www.presonus.com

versus the Focusrite:

Scarlett 2i2 | Focusrite

focusrite.com

versus the ...?

Although the Zoom says it has 80KHz bandwidth in 192K mode, I'm finding that difficult to believe. First off, there's no conceivable reason for that bandwidth in an audio interface that isn't intended for scientific or industrial instrumentation. Secondly, increased bandwidth is not my understanding of how high definition audio works.I don't see much point in belaboring 'the sound' since I expect all three to sound substantially similar.

What I'm after is:

1) ease of connectivity to a FOSS DAW even if there are no Linux drivers available from the manufacturer, with Windows as an undesirable backup/workaround if all else fails
2) reliability
3) durability
4) cost easy on the budget
5) additional hardware (or maybe software) features that vary between models

Is there any substantial difference between these various devices on the market, or should I just go with the least expensive model available? The price differential seems to be considerable across the offerings. Why?

Thanks

 

[ecc83]

If you need Linux capability look at the Native Instruments KA6 (maybe the smaller ones work with Linux but don't know) I know for an absolute fact that they work with Linux very well because I loaned mine to a guy for a month and he bought one on the strength.

Dave.

 

[Papanate]

Although the Zoom says it has 80KHz bandwidth in 192K mode, I'm finding that difficult to believe.

It doesn't show that - what the documents say is:

• Video resolutions and frame rates: HD 1080p / 30 fps, HD 1080p / 24 fps, HD 720p / 30 fps , HD 720p / 24 fps,
• Audio formats:WAV (24-bit/96 kHz, 24-bit/48 kHz, 16-bit/44.1 kHz)

First off, there's no conceivable reason for that bandwidth in an audio interface that isn't intended for scientific or industrial instrumentation. Secondly, increased bandwidth is not my understanding of how high definition audio works.I don't see much point in belaboring 'the sound' since I expect all three to sound substantially similar.

You are chasing Windmills here -

Is there any substantial difference between these various devices on the market, or should I just go with the least expensive model available? The price differential seems to be considerable across the offerings. Why?

The Zoom and the Scarlett are quality units - the nod goes to Scarlett due to the interface - I wouldn't touch a Presonus with a 50 foot pole.

 

[TalismanRich]

If you use Nyquist theory, then the result of a 192kHz sample rate is the ability to record up to 80kHz before aliasing comes into play. This means you might be able to use a less invasive filter, although digital filtering has improved drastically over the original brickwall analog filters. While you may be able to record ultrasonics, you really don't need to worry about it since very few microphones maintain output much above 20k. It's possible that you could run into issues if you are recording things like synths, where you could get artifacts above 20k, which can cause audible issues from IM distortion. The biggest "innovation" of the Zoom is using 32bit floating point. There's debate as to whether it gives any significant advantages as the electronics are going to be a limiting factor in the overall noise level. Does having 1500dB of dynamic range mean anything when your preamps and associated electronics only have 125 or 130dB? I wonder also if it means a possible increase in latency, since you are now processing either 25 or 50% more data per sample. A 256 bit buffer doesn't really care if represents 8 32bit samples or 10 24bit samples or 16 16bit samples, but your audio program certainly will.

For the most part, I've found that unless you abuse most equipment, durability isn't a major issue, and reliability is usually good for most major brands. As for added features that's probably not going to be a big differentiator in this class of devices. They are basic units, 2 inputs, 2 outputs, USB and maybe midi. You need to spend more to get things like internal DSP processing. Now you're looking more at the UA Volt series, the Tascam 208 or RME Fireface as examples with DSP EQ, compression and reverb available internally, not via the DAW. This can offload processing from the computer.

That aside, Linux support will probably be more available for older interfaces. The UAC232 is pretty new, so it's doubtful that many users have tried configuring it with ALSA/Jack. The Scarlett is proven to work with Linux, as is the Presonus. The Zoom and Presonus include Midi, the Scarlett does not.

Is there a major reason to restrict your DAW selection to FOSS? A program like Reaper isnt "Free" or Open Source, but it's certainly cheap enough and has excellent support. Cakewalk by Bandlab is free, but isn't Open Source.

 

[Papanate]

If you use Nyquist theory, then the result of a 192kHz sample rate is the ability to record up to 80kHz before aliasing comes into play.

The unit doesn't go to 196kHz - so it is a moot point.

 

[rob aylestone]

Oddly - quite a few interfaces have a pretty decent performance, way up into what is technically RF, not audio frequencies. Many of the modern class D digital amps can do the same. It used to be technically impossible, but things have moved on quickly. Pointless I suppose - but they could filter it off, but perhaps leaving it is easier, and also becomes sales claims!

 

[TalismanRich]

The unit doesn't go to 196kHz - so it is a moot point.

It doesn't show that - what the documents say is:

• Video resolutions and frame rates: HD 1080p / 30 fps, HD 1080p / 24 fps, HD 720p / 30 fps , HD 720p / 24 fps,
• Audio formats:WAV (24-bit/96 kHz, 24-bit/48 kHz, 16-bit/44.1 kH

There is no video resolution involved.

You must be looking at a video recorder, not the new Zoom audio interface. It looks like they are using the interface portion of the F8 field recorders, going with the 32 bit audio.

Spec on the Zoom UAC232:

 

[ecc83]

In my 60+ years of interest in all things Audio there has never been any peer reviewed data that suggests an audio 'system' needs a bandwidth much in excess of 30kHz (3dB point) There are MANY situations where the extended response of say a mic pre amp or mixer has resulted in the reception of the local taxis or MW radio.

The world contains many devices that operate over a range of frequencies but to mention just one? Active car suspension, you do not build it to cope with signals from the pea gravel in tarmac! Only the audio industry wastes time, money and resources on amplifying things only bats can hear. How many people here have ever been bothered by the fact that FM radio chops off at 15kHz...EVEN when you COULD hear 16kHz? How many were bothered by the 19kHz pilot tone or the 15.625kHz "scream" of analogue telly? (I did get complaints from a few ladies about the latter)

"The wider you open the window the more the **** flies in"

Dave.

 

[jamesperrett]

Dan Lavry wrote a good document on why the 192kHz sampling rate was actually technically inferior to 96kHz around 20 years ago. Basically, there is no useful audible information beyond about 22kHz (young kids can hear up there although not older people) so there is little point in going too high because you are only letting more noise and interference through.

https://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf

There is a school of thought that gentler filters sound better so 96kHz could be justified but not 192kHz. There is also the question of whether you want to preserve musical signals above the maximum audible frequency - acoustic percussion or even analogue synths and drum machines produce a fair bit of energy that can't be heard by humans so it may be interesting for a research paper but I'm not sure that it is needed for normal audio recording.

 

[TalismanRich]

In my 60+ years of interest in all things Audio there has never been any peer reviewed data that suggests an audio 'system' needs a bandwidth much in excess of 30kHz (3dB point) There are MANY situations where the extended response of say a mic pre amp or mixer has resulted in the reception of the local taxis or MW radio.

The world contains many devices that operate over a range of frequencies but to mention just one? Active car suspension, you do not build it to cope with signals from the pea gravel in tarmac! Only the audio industry wastes time, money and resources on amplifying things only bats can hear. How many people here have ever been bothered by the fact that FM radio chops off at 15kHz...EVEN when you COULD hear 16kHz? How many were bothered by the 19kHz pilot tone or the 15.625kHz "scream" of analogue telly? (I did get complaints from a few ladies about the latter)

"The wider you open the window the more the **** flies in"

Dave.

These days I can't heard much above about 10K, but years ago, the flyback transformer's 15K squeal from the TV used to drive me nuts!. If I smacked my 27" Sony just right, it would get quiet for a while. So, now we get digital TV, without that transformer squeal, and I can't hear it anymore anyway!

There's no justice in this world!!

 

[ecc83]

Rich, when TVs were made with steel chassis and 3/8" ply cabinets we did not get complaints from the distaff sides. Then they were made with a light steel frame carrying several PCBs and all contained in a sub 3mm plastic case. Acoustic line power was barely phased in its transmission.

James, there is at least one very well respected active monitor that uses 48kHz for its DSP...PRECISELY to keep ultrasonics out of the system.
If drum machines,synths produce **** beyond 22kHz that is just bad design. Have they never heard of "Sallen&Key" ?

I do understand that very high energies at high (audible) frequencies can cause problems at lower sampling rates but the solution is as simple as 'drop the overall recording level'? That is my understanding but I am well known here as a digital dunce! Anyway, there still seems to be no case for 100kHz pre amp bandwidths.

Maybe one reason for the continuing love for valves and transformers is their naturally curtailed HF response?

If there is a large, one off musical event (or say the coronation, not that I personally gave a *&*!) With Sir john Blimpy conducting the once only performance of Fartfinger's 3 hour dirge then yes. Record it at 192kHz AND in Ambi and every other format you can lay your hands on to preserve THE best recording you can get for future generations.
For most stuff 24 bits and 44.1kHz is easily good enough IMHO.

Dave.

 

[rob aylestone]

Dave Rat did an amp test recently and discovered it was flat to 30K, but had significant output at 200K!!! Why filter it off seems to be the design philosophy.

 

[ecc83]

I did not see that 200kHz reading Rob but then I must confess I found the guy hard to follow! Do you have a time mark?
My view is that yes, they should filter out everything above about 40kHz, having a significant 'RF' current flowing through speaker cables is bloody irresponsible. Yes, other stuff SHOULD be proofed against it but there are limits to how much input filtering one can do from both a performance and economic point of view.

The finding that the amp with the high 'peak' power output was louder was interesting to see but no surprise. Forgive an old Bottle Jockey but this is "valves" again! Valve amps are almost never 'regulated'. That would be very expensive to do, waste gobs of heat and make the amp worse! It is the ability of valve PAs to deliver high, short term peaks that gives them their "punch". Some say the HT "sag" contributes to the "vibe" and "tone". Not so sure about that one!

Dave.

 

[jbuckey]

My first interface was a Presonus Audiobox and I hated it. I switched to Focusrite and haven’t looked back. I’ve had the 2i2, 2i4 and now use the 18i20 in my studio. The only complaint I have is the Focusrite Control software takes some getting used to if you use multiple inputs.

I’ve also used a Tascam 1600 interface and it was top notch too.

 

[TalismanRich]

One reason for using a higher sample rate is the effect on latency. Some driver software will limit how low you can set the buffer. A 256 bit buffer at 44.1K has a minimum of 5.8ms one way or 11.6 round trip minimum, regardless of the efficiency of the driver. Minimum latency is simply buffer size/sample rate. That doesn't take into account any processing that your computer has to do. If you keep the same buffer, but double the sample rate, you effectively halve the minimum latency, just as if you had gone to a 128 bit buffer. If your computer is fast enough to handle a lower buffer, that's fine, but if you can drop the buffer AND double the sample rate without any overruns, they you can significantly drop your latency.

I will say that I seem to hear a bit of a difference between my recordings at 88K vs 44K. It's not in frequency but it seems that the ambience or space is different. I can't explain it, but the higher bit rate seems more "spacious". Therefore I do my recordings at 24/88.2. I'm sure anything I might sense is lost when I render it down to 44K, but it certainly doesn't hurt running at the higher rate.

RE: the amp going to 200kHz, it's not unheard of. My 40+ yr old Bryston 2B was spec'd at 1-100kHz for the -3dB points.

 

[ecc83]

That is fine Rich, If you have a preference for the higher sampling rate that is of course your business. I have no issue with sampling rates per se, just gear with bandwidths up into the RF spectrum.

The Bryston amp will not be the only amplifier to have a ~100kHz 3dB point (IIRC the venerable Mullard 20-20 had a closed loop bandwidth close to that.) That does not matter so much SO LONG AS you don't FEED it RF! Plus they are 60W and 20W amps respectively. The amplifiers in that YT are class D and 400W+ and so could well be generating that 200kHz themselves.

Dave.

 

[jamesperrett]

I will say that I seem to hear a bit of a difference between my recordings at 88K vs 44K. It's not in frequency but it seems that the ambience or space is different. I can't explain it, but the higher bit rate seems more "spacious". Therefore I do my recordings at 24/88.2. I'm sure anything I might sense is lost when I render it down to 44K, but it certainly doesn't hurt running at the higher rate.

RE: the amp going to 200kHz, it's not unheard of. My 40+ yr old Bryston 2B was spec'd at 1-100kHz for the -3dB points.

If you are using any plug-ins or processing then they may change their behaviour at high sample rates. I've certainly noticed that some plug-ins, particularly compressors, sound better at high sample rates - even if the input and output are resampled from/to 44.1kHz. That's why some plug-ins offer an oversampling feature.

 

[jamesperrett]

James, there is at least one very well respected active monitor that uses 48kHz for its DSP...PRECISELY to keep ultrasonics out of the system.
If drum machines,synths produce **** beyond 22kHz that is just bad design. Have they never heard of "Sallen&Key" ?

I think I know the monitors you are talking about and they have a great reputation.

The drum machines and synths that I'm talking about were designed before digital recording became prevalent (it was around but strictly the province of classical music) so there was no impetus to limit bandwidth. Analogue recording is naturally band limited but I regularly see signals from old analogue studio tapes out to 30kHz and possibly beyond when these synths are involved.

 

[CherylJosie]

What a great response to my post. You guys are the best.

Well the voting seems to favor Focusrite. I had assumed as much since it seems to be what everyone is buying and the box is metal. Thanks for your input everyone. I'll get that one.

Regarding the Zoom, the reason I'm not believing the 80KHz spec is because the higher sampling rates are usually implemented specifically to simplify (shorten) the anti-alias filter and reduce ringing with a more gradual rolloff as compared to a brick wall filter with a narrower transition region between the pass band and the stop band. I'm willing to be convinced otherwise, but until I see a test bench result on that interface, I'm retaining my skepticism about their published specification.

Thanks also for the walk down memory lane. Although I'm not as 'into it' as you guys, I do understand (most of) what you discussed. I was especially interested to learn that older drum synths generate ultrasonics. I'm guessing those are analog drum synths since digital synths all have Nyquist reconstruction filters on the D/A output.

I tend to agree on that amp generating 200KHz. If it's there at all, it's probably because its output filter is poorly designed and ringing. Maybe I'll look into it to satisfy my curiosity. Thanks for the link.

I'll keep in mind the other recommendations too on options I hadn't listed. I'll do some reading before I order, just in case there's something I think may be of value. The FOSS is because I'm on a fixed and extremely limited disability pension, so I've got to make every penny count. Open source isn't the requirement; free is.

Happy 4th all. Enjoy your barbecue. With this heat wave you can probably fry your steaks right on the patio concrete.

 

[TalismanRich]

A bit of FYI, here is Julian Krause's review of the UAC-232. If you go to 12:50, he has the graph of the line input response and it is indeed flat out to 80kHz. But as always, there are good and bad points about any device, and there are some things that might be considered questionable choices in the design like not having any input attenuation. That might be fine for using in 32 bit mode, but if you go to 24 bit, it might well be an issue..



As for the steaks.... I've got 4 NY Strips in the fridge, just waiting to be tossed on the grill (it's not going to be abnormally hot here in the Midwest US).