Sure. It is the ITU BS1116: https://www.itu.int/rec/R-REC-BS.1116-3-201502-I/enThere is a long document on perceptual testing for small differences by one of the European standards committees but I lost the link to it. Maybe someone can find it.
Sure. It is the ITU BS1116: https://www.itu.int/rec/R-REC-BS.1116-3-201502-I/enThere is a long document on perceptual testing for small differences by one of the European standards committees but I lost the link to it. Maybe someone can find it.
Sure. It is the ITU BS1116: https://www.itu.int/rec/R-REC-BS.1116-3-201502-I/en
It absolutely is. It should be read over and over again by anyone wanting to perform blind tests. When I was at microsoft we would commision third-party tests of our codec versus the other using BS1116 as the guideline.Thanks, I think this is a good model for a scientific comparison.
We're deep into conference season and crazy R/D, so I'll have to bow out for a while. If anyone is at NAMM, next week, please stop by and say "hi" and see our advanced new product we'll be showing for the first time. Hint: Ethernet audio, redundant AC/DC powering, app-based control, heroic miniaturization, industrial-grade environmentals (>65C), hospital-grade MTBF.
Pro audio is fully Ethernet, nowadays.Looks like audiophile Ethernet audio is going to be the new trend for this year’s audiophoolery. When will we see audiophool wireless routers cause you know wired networking is getting obsolete these days
Looks like audiophile Ethernet audio is going to be the new trend for this year’s audiophoolery. When will we see audiophool wireless routers cause you know wired networking is getting obsolete these days
So can you distinguish these in an ABX test?
foo_abx 2.0.2 report
foobar2000 v1.4.8
2020-01-09 08:03:24
File A: boston11.wav
SHA1: 5ae8849c8ce3bda8321ffb3f16ad9e797bf736f3
File B: boston22.wav
SHA1: 4955964da2691ab762acf72f25416a7431748109
Output:
ASIO : Creative Sound Blaster ASIO
Crossfading: NO
08:03:24 : Test started.
08:03:50 : 01/01
08:04:16 : 02/02
08:04:29 : 02/03
08:04:47 : 03/04
08:05:05 : 04/05
08:05:38 : 05/06
08:05:58 : 06/07
08:06:18 : 07/08
08:06:30 : 07/09
08:06:42 : 07/10
08:06:42 : Test finished.
----------
Total: 7/10
Probability that you were guessing: 17.2%
-- signature --
be1c42344ce9725cf5ba40ebd94fe706a6da2662
foo_abx 2.0.2 report
foobar2000 v1.4.8
2020-01-09 21:24:48
File A: boston11.wav
SHA1: 5ae8849c8ce3bda8321ffb3f16ad9e797bf736f3
File B: boston22.wav
SHA1: 4955964da2691ab762acf72f25416a7431748109
Output:
ASIO : Creative Sound Blaster ASIO
Crossfading: NO
21:24:48 : Test started.
21:25:09 : 01/01
21:25:19 : 01/02
21:25:29 : 01/03
21:25:49 : 02/04
21:26:01 : 03/05
21:26:20 : 03/06
21:26:34 : 04/07
21:26:43 : 05/08
21:26:55 : 06/09
21:27:13 : 07/10
21:27:13 : Test finished.
----------
Total: 7/10
Probability that you were guessing: 17.2%
-- signature --
b6514321df42df307fdfd400195430bbd81423f3
Pro audio is fully Ethernet, nowadays.
Dante and the like are real pro-level networking technologies.
16 channels x 24 bits x 192kHz, up and down with a single cable...
That's NO BS !
After runing a "Training mode" of Foobar abx for some time, I get this:
Code:foo_abx 2.0.2 report foobar2000 v1.4.8 2020-01-09 08:03:24 File A: boston11.wav SHA1: 5ae8849c8ce3bda8321ffb3f16ad9e797bf736f3 File B: boston22.wav SHA1: 4955964da2691ab762acf72f25416a7431748109 Output: ASIO : Creative Sound Blaster ASIO Crossfading: NO 08:03:24 : Test started. 08:03:50 : 01/01 08:04:16 : 02/02 08:04:29 : 02/03 08:04:47 : 03/04 08:05:05 : 04/05 08:05:38 : 05/06 08:05:58 : 06/07 08:06:18 : 07/08 08:06:30 : 07/09 08:06:42 : 07/10 08:06:42 : Test finished. ---------- Total: 7/10 Probability that you were guessing: 17.2% -- signature -- be1c42344ce9725cf5ba40ebd94fe706a6da2662
Code:foo_abx 2.0.2 report foobar2000 v1.4.8 2020-01-09 21:24:48 File A: boston11.wav SHA1: 5ae8849c8ce3bda8321ffb3f16ad9e797bf736f3 File B: boston22.wav SHA1: 4955964da2691ab762acf72f25416a7431748109 Output: ASIO : Creative Sound Blaster ASIO Crossfading: NO 21:24:48 : Test started. 21:25:09 : 01/01 21:25:19 : 01/02 21:25:29 : 01/03 21:25:49 : 02/04 21:26:01 : 03/05 21:26:20 : 03/06 21:26:34 : 04/07 21:26:43 : 05/08 21:26:55 : 06/09 21:27:13 : 07/10 21:27:13 : Test finished. ---------- Total: 7/10 Probability that you were guessing: 17.2% -- signature -- b6514321df42df307fdfd400195430bbd81423f3
Was this an answer for me, sorry there's a lot going on in this thread. Lol. Thanks if it is, so are you saying that if the DAC hits the magic number at 4v it will still be totally transparent at say .2v? I can't get my stupid head round that. I did ask John Westlake about digital volume when he was designing the mdac, he said yes there will be an impact in signal quality even with his 32bit volume when it's at low level. I won't go into my subjective findings here, so I think your saying if it measures that good at full output it should remain totally transparent at any level?Consider that most DACs (and certainly amps) produce the lowest distortion below 0dB FS as shown in the plot below
So no need to worry about that. The seemingly higher distortion levels aren't caused by a higher distortion. It is the noise floor creeping relatively closer to the signal amplitude as the noise floor is constant.
Don't get hung up on distortion at 1khz and full output voltage and noise levels relative to it.
Worry about actual noise levels and transducers instead.
SINAD is just a generated number under specific conditions.
Sometimes you do, sometimes you don't. Some DACs start to saturate and raise their distortion at full level.Hi folks, probably not the best place to pop in a first post but I could do with a little help understanding SINAD. So we get the highest figure at full output?
At lower output levels the distortion products (depending on the used circuit/components) is generally also lower or drown in the noise floor.
So when the noise floor is say 110dB down opposite 0dBFS then at -40dB FS, due to digital attenuation, the resolution will still be quite the same but the noise floor will be 40dB higher so 70dB (40 dB attenuation is a LOT).
Assuming you will get the DAC-10 (eying it usually means buying it) means 0dBFS = 2V as the 0.7V input efficiency does not appear to be an XLR input but rather RCA.
So full power is -9dB in your DAC. This means that assuming 115dB S/N ratio of the DAC you will end up with a S/N ratio of 105dB.
This is (depending on speaker efficiency) an inaudible level unless you put your ear against the tweeter perhaps.
This also means that the audio signal will always be transparent when digital volume is used as all distortion products most likely will drop below the noise floor which you can't hear.
Sometimes you do, sometimes you don't. Some DACs start to saturate and raise their distortion at full level.
PS audio DAC had a design flaw (obviously) and its a good thing they offered to fix it for free.
There isn't much difference in performance between (properly executed) digital volume control and that of the D10.
Resolution will be the same and noise most likely will be dominated by the background noise of your power amps.
Ultimately that and your hearing threshold will determine the resolution that can be had.
And the limitations of the hearing are worse than that of the DAC