This is a high resolution listening test.
Simply play the file and count the number of times you hear: "This is a High Resolution Listening test". Please report the number. No mysteries, no double blind, no guessing. Takes about a minute. Only two simple requests: (1) Set your volume as you are accustomed to listen, and don't touch the volume control while you play the file.
The more times you hear the phase the higher resolution and better ears it takes to pull that off.
The 24/96 FLAC file is too large (44 MB) for posting on this forum, but you can download it here:
https://lookaside.fbsbx.com/file/high%20resolution%20listening%20test%20audacity%20HD.flac?token=AWzKGdkJR4_ddmmAaHV8ZdEXlBFlrCDyJ7jer2v9pOJqV2VHI556PeIPKOYaexq_SbAS6z5ndXzHPRzmA-baup4WaRBOGP_qE9a1twNvDcbtRstNoEMDBCm0QtLyaYb9as18-NjIXBZO7QBbbYpSYqlezFsKiSSML2VxCKYRcSvLuvto9QEoxFdOyvZabF6vXfytFmIHPUVRlvxrS5NjE4a9
7x
Six. I hear some mumbling on the seventh time, but wouldn't have known what it was on its own.
P.S. Is the speaker drunk, or is my DSP messed up?
The 8th pass is only silent something. The 7th one is already hard to understand on its own i guess when not listened as the same as before in a row.
Why this is a high resolution listening test i don't get yet. Maybe bit depth?
Six. I hear some mumbling on the seventh time, but wouldn't have known what it was on its own.
P.S. Is the speaker drunk, or is my DSP messed up?
Neither, but this was before his first cup of coffee. :-)
6x (not sure/almost 7x) with studio speakers
7x (not sure/almost 8x) with open over-ear headphones
8x with in-ear phones.
P.S. All tests were done at my usual loudness.
Yep, 8.
6 :(
9. The ninth time is unintelligible but certainly I hear something (if that counts). Speakers, quite room. ~75dB-C
My main output is resampled at 44.1kHz for DSP purposes.
7 x
I hear some mumblings for the 8th time but cannot realistically distinguish the full phrase.
The more times you hear the phase the higher resolution and better ears it takes to pull that off.
Numerically, take the count of number of times you hear the phrase, subtract two (2), and multiply by ten (10) That is your estimated dynmaic range in dB.
For example if I do the test I get 9. So I take 9 subtract 2 leaving 7. Multiply 7 by 10 obtaining 70. That is the dynamic range that I am hearing: 70 dB.
90 dB or 11 samples is the equivalent of CD quality. If you run all samples that is 14, which corresponds to 120 dB or about 20 bits. The sample has a crest factor of about 18 dB which is typical of pretty dynamic music.
Highly compressed rock might be 10 dB easier to hear than my vocal sample.
IOW if the sample were highly compressed rock, you'd probably get one more sample right.
7x
I can still hear it very faintly at the 8th time
I can't really make out the 7th, and 8th is too faint to really make over the background noise of my room.
Headphones, usual volume level. I usually keep the player 50% volume, though, which would have made it even harder to hear the 7th.
Hey, no wonder I can't really tell the difference between PCM and ADPCM style lossy in most cases.
7.
My volume is set so that heavily compressed rock is comfortably loud, if I raise for classical music I can hear some words of 8.
At the volume I normally listen to, and the noise of my PC a few feet away, with volume set roughly where it would be listening to a podcast:
Speakers (Polk Hampden): first 4 can still make out the words, on the 5th I only hear what I can already expect would be the noise of the sentence but can't make out the words.
Headphones (Sennheiser HD-25-II): first 5 can make out the words, very faintly can hear 6.
I normally listen to music at a different volume level however so was curious what the results would be after applying ReplayGain to the track and leaving the volume level at what I'd listen to music at.
Speakers: first 4, 5th only like a couple words. 5th in this case was oddly (marginally) clearer than the headphone result below despite the ambient noise.
Headphones: first 4, 5th only like a couple words.
Edit:
The 24/96 FLAC file is too large (44 MB) for posting on this forum, but you can download it here:
It's 4.84MB when I downloaded it, was it the correct file?
The more times you hear the phase the higher resolution and better ears it takes to pull that off.
Numerically, take the count of number of times you hear the phrase, subtract two (2), and multiply by ten (10) That is your estimated dynmaic range in dB.
For example if I do the test I get 9. So I take 9 subtract 2 leaving 7. Multiply 7 by 10 obtaining 70. That is the dynamic range that I am hearing: 70 dB.
90 dB or 11 samples is the equivalent of CD quality. If you run all samples that is 14, which corresponds to 120 dB or about 20 bits. The sample has a crest factor of about 18 dB which is typical of pretty dynamic music.
Highly compressed rock might be 10 dB easier to hear than my vocal sample.
IOW if the sample were highly compressed rock, you'd probably get one more sample right.
Anecdotal, but...
This agrees completely with my measurements of my current noise floor. Also, seems to show that 24bit is TOTAL overkill for playback since there's no why in hell I would hear to sample number 11.
It's 4.84MB when I downloaded it, was it the correct file?
Yes. the 45 megabyte number is in error. Thank you for your attention to detail. I was reading the wrong column and/or rwo in the file list on my computer. The wave file version is 45 MB. Being highly redundant, it compresses splendidly!
Numerically, take the count of number of times you hear the phrase, subtract two (2), and multiply by ten (10) That is your estimated dynmaic range in dB.
What if someone only hears 2, or even worse 1. Is there a negative dynamic range? ;)
Numerically, take the count of number of times you hear the phrase, subtract two (2), and multiply by ten (10) That is your estimated dynmaic range in dB.
What if someone only hears 2, or even worse 1. Is there a negative dynamic range? ;)
One principle of designing tests is making them self-checking. Certain answers can be safely interpreted as: "Throw away this trial".
8. There is quite a bit of music I play slightly louder, I tried that too. Then the first two are slightly uncomfortably loud, but I mostly hear the sibilants at the ninth.
Afterwards I came to think that I usually play back with RG as scanned by fb2k, but that reports an RG figure of -0.71. Shouldn't matter.
I cannot download this. It tells me to login to facebook (which I will never do.)
Can't you just attach the file, since others said it's just 4.5MB?
I cannot download this. It tells me to login to facebook (which I will never do.)
Can't you just attach the file, since others said it's just 4.5MB?
I uploaded test file to a more neutral location:
https://www.dropbox.com/s/n50g7b6hs14x8dj/high%20resolution%20listening%20test%20audacity%20HD.flac?dl=0
It is also an attachment to this post.
8x.
The 9x time is almost completely unintelligble yet I can hear that something is happening so I'm not counting it. Tested with headphones.
8. There is quite a bit of music I play slightly louder, I tried that too. Then the first two are slightly uncomfortably loud, but I mostly hear the sibilants at the ninth.
Afterwards I came to think that I usually play back with RG as scanned by fb2k, but that reports an RG figure of -0.71. Shouldn't matter.
The same for me if the sound is loud, turning my chair into vibrator, but there is still some power reserve for half of the tracks.
Minus one (total 7) if I set music loudness to background (still quite loud) level.
If I listen to music at night mode (with EBU R128 Compressor DSP enabled), then again 8. ;-)
7x with open back headphones
I get 8. Laptop>Focusrite Scarlett 2i2>Sennheiser HD 280 Pro's. (64 ohm version.)
Unfortunately, we live a stone's throw away from a major highway. Highway noise masks anything beyond 8.
Strange, I also get a higher score ( 8 ) with my Sennheiser HD280 at a lower volume than my Shure SRH940 ( 7 ). I think it might be because of the better isolation. My room is a bit noisy.
7 times.
Though I did a Replaygain scan (foobar 1.4 beta 7) and applied that, I also have the RG gain set to -5 (so foobar playback adheres to EBU R 128 standard of -23 LUFS.
In addition I got EqualizerAPO with a rather big shelf filter (18 dB, my headphones/audio out lack bass/power), so stuff above 200hz or so is basically -18dB in addition.
My PCs volume is at 100% which in many cases are a tad too low IMO if I listen to replaygain tagged (or -23 LUFS target) audio. I'm guessing I might have heard 8 times or 9 times if my volume wasn't maxed.
According to the math 9 times would equal 70dB dynamic range? Myself I prefer to target 83dB SPL (which is movie loud) if possible (which would roughly be 10 times in this test?). Movies are among the most dynamic things out there and 83dB dynamic range uses less than 14 bits, if one ignores the 16th bit on the basis of random noise/noise floor, you still got 1 bit "in reserve" for a movie. And if TruePeak is supposed to be -1dBFS (actually it's -2 now in EBU R 128 isn't it?) that's the entire 16bit range.
83dBSPL is "safe" you might get tired ears but your hearing shouldn't be damaged even if you listened 8 hours straight at that level. When I listen to music I tend to end up with a lower level than that (music is more compressed dynamically and gets fatiguing much faster so I do not like to listen to that at full volume always).
I kinda like you started at the loud end as that forces people to set their audio level to "normal" at the start.
Could you do another test (same voice sample) but with some music, just find some Creative Commons licensed music that has some variation but during the test duration it's pretty consistent in it's loudness (make sure it's -23 LUFS maybe?)
It would be interesting to se how many times people can hear you then. I'm guessing that based on my 7 times I'd probably have your voice vanish into the music at around 4 or 5 times or possibly even 3 it's hard to predict.
8 I could understand. 9th could hear something, but any words I thought I heard were probably because I knew what they would be.
Galaxy S5 phone using Poweramp player > CCA in Onkyo Tx 8270 receiver > Sennheiser RS180 headphones.
Since it is my test, I should report my results. :-)
I was able to hear the test samples through the ninth (9) samples.
The test file was played on the computer it was developed on - a Windows 10 system with an audio device obtained from eBay that called itself "USB Mini Portable DAC Decoder amp HIFI Fever Sound Card SA9023A + ES9018K2M" The headphones were AT ATH M50s. The headphone amp was a Topping NX1 (original model) A technical test on the audio interface is attached.
I might have listened a little bit louder than usual, since I turn the volume up and down on my headphone amp depending on the level of the music. Anyway, the first time I could hear nine and I sensed ever so slightly number 10 as well. When I turned the volume up a little bit, where the first two were still listenable, I could hear number ten a little bit more clearly, but it was still very, very faint.
I then tried to crank the volume to max on my headphone amp, and I could hear 11, although number 11 was very, very faint. Number 12 I sensed ever so slightly, but I couldn't make out anything. Number 12 was the last one I could hear.
I then repeated at normal volume, where a very, very faint number 10 was again the last one I could hear.
I've never considered myself to have a super hearing, but I have noticed that I sometimes hear very faint sounds, like a telephone ringing in the flat downstairs or a very faint hum from electronics, when people around me don't hear anything. I do think my ears are sensitive in the sense that I have to wear ear plugs in the cinema, at the gym, sometimes even in the street in certain noisy cities around the world. But I don't claim that I can hear jitter ;-).
My wife has tried, with the same kit (S5 > CCA > RS180) as me, and also heard 8 but not sure of 9. We are both in early 70s.
Both know, her through proper hearing tests and me via a Google app, we have very reduced frequency range hearing. Mine is below 11Khz.
This is a cool and fun test. I just ran this by chance, daytime, my girlfriend doing some noises closeby.
First time I managed 8 times: iMac > RME ADI-2 Pro > Oppo PM-3
I wanted to do better, so I swapped the headphones with custom-fit IEM. I could hear the "high resolution" of the 9th time, but not the complete sentence. I forced myself to get past the first two parts by going beyond comfortable listening and I could hear the 10th time as noise - if at all. So I guess 9 is the absolute limit for headphones or private listening, IMO.
I thought I'd have another go to see how much the DAC can handle, so I slowly increased volume as the track played. I easily managed 13 times at max volume and was able to understand the first word of the 14th time. Since I was using high gain, the noise floor made it difficult to understand the rest. But of course this test is irrelevant because I would be deaf in a normal listening test.
I guess it's impossible to hear the full dynamic range of a CD. I chose my username well...
8 times. hd650 headphones, plugged into a xonar stx soundcard that's set to a high gain, windows 10 playback volume at 50, relatively quiet room.
6 times with speakers, 7 times with headphones.
Not sure what this test is about though. The results are the same with 44.1kHz and 22.05kHz and 16-bit. It doesn't sound different. Theoretically, 24-bit could allow very silent audio to be clearer, but I don't think higher sample rate will offer any advantages whatsoever.
I don't do Facebook, so I downloaded it from Dropbox and also tried the attachment to your post.
In both cases, my results were as follows:
Decoding failure at 0:00.085 (Unsupported format or corrupted file)
Anyone else get this error?
Decoding failure at 0:00.085 (Unsupported format or corrupted file)
Anyone else get this error?
Nope. The file is fine:
flac -t "high resolution listening test audacity HD.flac"
flac 1.3.2
Copyright (C) 2000-2009 Josh Coalson, 2011-2016 Xiph.Org Foundation
flac comes with ABSOLUTELY NO WARRANTY. This is free software, and you are
welcome to redistribute it under certain conditions. Type `flac' for details.
high resolution listening test audacity HD.flac: ok
md5sum "high resolution listening test audacity HD.flac"
1de01dd315f243e4a53e98c4ad0be3d1 high resolution listening test audacity HD.flac
Anyone else get this error?
Nope. The file is fine:
OK, problem solved. Metaflac shows that Arny's file was encoded with libFLAC 1.2.1, and I was using an old 1.1.1 FLAC decoder.
Apologies - I should have checked that before my previous post. Downloaded a new FLAC decoder and all is well.
Results, played through my Lenovo T420 laptop's (presumably not very good) headphone output in a typical domestic study:
- Could hear up to #8 using Sennheiser HD435s. (Open-back headphones).
- Then tried Shure SE215 IEMs and could just hear #9.
Speakers: 7 at normal volume (for pop/rock), at elevated volume I can still faintly pick up 8.
No surprises there, I've always been a pretty quiet listener (estimated little over 60 dB SPL on the PC, somewhere in the 50s in bed). The PC is running with a folded-up blanket thrown over the front right now for different reasons (microphone) - being a BTX setup with intake fans in the front, this cut down on noise quite a bit, in addition to a somewhat involuntary all-SSD upgrade (my data drive packed it in a few weeks ago, and even though it had never been loud, I thought the computer's sonic profile was noticeably more agreeable without). It's a quiet Sunday evening here, and my room is not too loud in general.