Are you able to tell which recorder is better?

Audacity Help Forum Windows
1

Hello

Using v 2.0.5 on win 7x64

I’m playing around w/ recording online audio streams. Thanks to u’r detailed instructions, I have Audacity & computer working correctly. This is not critical work, just fun stuff.

I have 2 puters, 1 online (OL), the other, my workstation (WS) is off-line. The WS has a better audio chip then the OL; ALC889 vs. ALC888.

I’ve recorded a couple of 50s oldies w/ Aud. on the OL. I next used a Loopback cable going to the WS & recorded the same songs (using it’s Aud.installation) via it’s audio chip. I’m trying to determine if it is worth the effort to use the cable & 2 computers.

I’ve formed an opinion, but w/o the ability to do blind tests, I don’t trust it. I don’t have the knowledge to look at the waveforms & determine which is better…they are different. I’ve exported both to 32 bit float.

Would someone please tell me what to upload here (waveform, chart or whatever) so it could be looked & you could make a judgement call.

Thanks for u’r time on this.

2

The most important thing when comparing two files/sources is matching the volumes. The louder one will usually give a better impression and “sound better”.

And, you can’t tell much by looking at waveforms. You can see clipping. You can see which one is louder. Sometimes you can see offset. But unless there is a major defect, you can’t usually tell which one sounds better.

The biggest difference between the two is likely to be noise. So, I recommend you match the volumes, and then listen to silence (if there is a silent part) at the beginning or end to see which one is quieter.

Streaming audio is generally more-compressed than MP3s from Amazon or AACs from iTunes. The compression might affect quality more than your setup.

but w/o the ability to do blind tests,

It’s not hard to do a [u]Scientific Blind ABX Test[/u], but it takes some time. Note that an ABX test doesn’t tell you which one is better. It only tells you if you can reliably hear a difference.

If there’s a difference you can actually hear, you get to choose the one that sounds better to you. In cases where you are comparing a compressed file (i.e. MP3) to the uncompressed original, if you hear a difference you can assume the original is “better”. But it’s still a judgment call, and you may prefer the MP3. …Some audiophiles prefer vinyl to CD, even though it’s technically inferior. :wink:

I’ve exported both to 32 bit float.

There’s no harm (other than big files), but there’s no need for it. You only need floating-point if you are doing something that boosts the peaks over 0dB. Audio DACs & ADCs are 16 or 24-bit integer, so they can’t go over 0dB during recording or playback.

The guys that do blind ABX tests will tell you that CD quality (16-bit, 44.khz) is good enough for just about anything, and better than human hearing.

Most Pro studios record at 24-bit/96kHz.

Audacity (and most audio editors & DAWs) use 32-bit floating-point for internal processing and temporary storage.

3

Thanks so much for u’r detailed answer. Of course, I was looking for the easy answer & of course, it always fails to arrive :wink:

The most important thing when comparing two files/sources is matching the volumes. The louder one will usually give a better impression and “sound better”.
And, you can’t tell much by looking at waveforms. You can see clipping. You can see which one is louder. Sometimes you can see offset. But unless there is a major defect, you can’t usually tell which one sounds better.

I match the volume in both recording & playback. Also, I record below the clipping point on both…so that part is the same.

Streaming audio is generally more-compressed than MP3s from Amazon or AACs from iTunes.

Interesting, I didn’t know that…I wish there was a way to determine what the compression is, but googleplay doesn’t seem to say.

The compression might affect quality more than your setup.

I suspected that. Since I export them to MP3>Extreme preset, I figured that if I had trouble telling the 32 bit waves apart, the exported file would surely be too close to call.

Doing the blind test is overkill in my case.

The biggest difference between the two is likely to be noise. So, I recommend you match the volumes, and then listen to silence (if there is a silent part) at the beginning or end to see which one is quieter.

OK, that’s what I’ll do. It seems the easiest way to see if there is a difference.

Thanks again.

4

Since you are recording streaming audio I would suggest trying using the WASAPI as your Audio host - see this page in the Manual: http://manual.audacityteam.org/o/man/device_toolbar.html
and select the “loopback” input device.

My understanding is that has less Windows “fiddling” with the signal and is able to more directly record the incoming digital stream (this page I found has some techy details http://wiki.xbmc.org/index.php?title=Windows_audio_APIs )

I use WASAPI loopback (on my W7 64-bit laptop) for recording web broadcasts od radio stations and get good results.

WC

5

A “loopback cable” carries the sound from the Stereo Line-Out of a soundcard of the Stereo Line-In of the same card. The on-line services think they’re playing to speakers or headphones and the recorder thinks it’s recording from an external cassette machine.

If you connected two different computers together, that’s just “a cable.”

Both of those connections are desperation methods. You only use those when you can’t get one of the internal systems to work. Any time you convert from digital to analog or back the process adds distortion and noise.

The best of the processes grabs the signal while it’s still digital and allows you to record that. The worst of them grabs the signal as an internal analog loopback. It incorporates all the distortion and noise of both converters. The only real benefit is that you don’t have to plug a cable in but the down side is you can’t adjust the speaker volume while you’re recording because the system is trying to record the speaker feed. Which one you have depends entirely on the soundcard and associated drivers.

Koz

6

If you are using “stereo mix” or “what U hear” of the built-in sound card (or almost any sound card) then it will be sending a copy of the output to the input after it has reached the output (after it has become analogue).

Almost all the solutions that capture the audio digitally without intermediate conversion to analogue are additional software (virtual sound driver) implementations.

Although I currently think WASAPI loopback is entirely digital, it captures all sounds destined for the sound card, not separate streams from separate applications. So WASAPI loopback will still capture system dings like “e-mail arrived”. You would need an application like SoundLeech to capture only the application that started playing the sound.

I don’t believe WASAPI loopback is bit-perfect (at least in Audacity) but if you want it to be “fiddled around with” as little as possible you should set the playback device to “Exclusive Mode” in Windows “Sound”.


Gale

7

Thanks everyone for the additional insights. It’s been an interesting exercise but the end result is anti-climatic.

First…the bottom line, at least in my case…use the simplest approach, as it doesn’t make a damn bit of difference as to what u hear. :astonished:

Using a brief period of quiet (the band stopped playing then started right up again) I recorded in Audacity, exported to 32 bit float wave & played the 3 samples back (VLC) on 2 different puters. They all sounded identical & played at the same volume. I had to turn up the speakers to an uncomfortable volume to hear anything.

  1. I used the cable method between the 2 different computers & their different sound chips described in my post.

  2. I couldn’t get Soundleech to work, so I installed (on the online puter) HiFi Cable Asio Bridge downloaded from
    http://vb-audio.pagesperso-orange.fr/Cable/index.htm.
    and described here
    http://sheesania.com/how-to-record-your-computers-sound-output-on-windows-with-virtual-audio-cable/#comment-4558

  3. I configured Audacity on the online puter, as

Audio Host> Win WASAPI
Output Device> Speakers (Realtek High Def.Audio)
Input Device> Speakers (Realtek High Def.Audio) (loopback)

I heard the music playing while recording. Software Playthrough was not checked.

I’d be interested in what anyone has to say about this. Was this the expected result? Could perhaps using a different streaming company, other music, etc, produce a end result which would indicate that one method is superior to another???

Oh…btw…I disabled Computer Sounds so any system sounds wouldn’t get recorded.