How much does audio interface really matter?

[Chelonian]

Upon post cup o Joe reflection, those numbers could be for the EIN* of the mic pre amps at some optimized gain but the noise OUTPUT will be vastly greater.

How could Julian Krause be measuring something other than the output?

 

[Deleted member 196982]

How could Julian Krause be measuring something other than the output?

I probably confused myself and others. Back to the JK measurements - he measures it, and in the chart I referenced in another of his videos, the very best measured noise is about -112dB (and you have to read the chart to see what that number means). Actual noise in typical use is a lot higher (smaller number), and is also in the chart. Maybe Dave can explain it, but I do pretty well trust that he is doing apples-apples types of comparisons.

review with comparison chart (pause the video when i starts and go full-screen)

how JK measures preamp noise

If you have rocks to throw - direct them at his videos

 

[ecc83]

Ok, I watched the video about the M2 and see I think that the noise figures were taken from a screen chart? They are, as suspected EIN figures and the reason they are better than the unity gain op amp is becuse they are A weighted. I got my figures from D Self's book and he does not give weighted numbers unless specifically mentioned.

Today I took delivery of a Fethead. Some anomalies have emerged, will report later.

Dave.

 

[Chelonian]

Thank you again! This is very helpful.

A 16-bit depth has a dynamic range of about 96 dB (which is nothing to sneeze at) but a 24-bit depth has a dynamic range of about 144 dB, which is pretty huge. It makes setting your gain to a setting of "loud enough to overcome the noise floor, but not loud enough to clip" almost a no-brainer. The wiggle room is huge, so it takes a lot of the stress out of setting your gain. That's pretty much the "trick" to gain staging: getting your signal level into an ideal range at each spot where you have control over it.

That's a superb point. I am trying to not only get good recording quality but make recording efficient and as few headaches as possible. And those likely go together, as if I keep having to fiddle with the gain and tire my brain and wear out my patience on that, I am likely to lose concentration somewhere in the performance/gain staging/mixing. I also love dynamics in music, such as a smokily fading away acoustic guitar bit followed by a big loud chord.

This write-up suggests that for mic-recorded music, 16 vs 24 won't matter much...but it also goes on to say that for direct in instruments, it very well might. I want to have the option to do everything.

Q. Do I really need 24-bit recording?

I'm a little skeptical that a Sound Blaster would be indistinguishable from an Apogee, but I'll reserve judgement until I've watched Ethan's content on the subject. I'd like to hear a null test between the two signals to see if they're truly identical or not.

He doesn't do a null test, unfortunately, but he does allow you to hear them both and makes some other comments suggestive that the sound card's A to D is rarely the weak link in the chain:

Audio Myths Workshop - YouTube

But after having used several entry-level interfaces from (almost) all of the major brands, I'd say to do yourself the favor of getting one that has dedicated ASIO drivers if you're recording on Windows. That'd rule out the UM2, UMC22, UMC202, and anything in the U-Phoria line without the "HD" in its model name.

What's the value of the ASIO drivers vs. the alternative? (is that ASIOFORALL?).

Thanks again!

 

[ecc83]

Re 24 bits and levels? My son regularly send me clips of classical guitar, mainly Bach transcriptions and they are often way down at -30 even -35dBfs. But his flat is very quiet most of the time and I can normalize to neg 20 and the tracks are still of very acceptable quality. (the 'waveforms' are mere blips!)

The Sound Blaster cards were always of very decent D2A quality I understand? They were of little use to recordists because of very high latency and were not designed for the MO of track building.

Dave.

 

[Eleanor Fudd]

I first started recording through a Presonus AudioBox. After a bit I got an RME Babyface. When I finished my first song, I was rather satisfied with it, the main non-vocal track being a soft and jangling rhythmic Stratocaster track with reverb. I had done a lot of messing on the Babyface, but when doing cleanup and going over notes, I discovered that that best and most important track had been done on the AudioBox. I thunk: go figure!