• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Review and Measurements of Teac HA-501 Headphone Amp

Tks

Addicted to Fun and Learning
Joined
Apr 1, 2019
Messages
948
Likes
991
Location
NYC
#41
I don't listen to music loud. For my continuous listening, the levels are quite low. It is usually -20 to -40 dB on DX3 Pro. I do occasionally turn up the volume on a track or two but that is not my normal listening.

These are very difficult tests to pass. They require extreme attention to what could be distorted and focusing on them after some trial and error. Once there though, you can see extremely good success rate.

Bottom line: I have training that allows me to listen through the music and focus on different aspects of it. If such issues are at very high frequencies, then yes, you can't rely on me. Otherwise, my skill levels are well above average not because I was born with it, but because I have spent hundreds of hours testing for artifacts in audio in controlled listening tests (where one would know if the answer is right or wrong).


No offense. I think it is best to ask me about such things rather than reading the tea leaves. :)

For many audiophiles that are as old as me and older, it is good news that one can be a critical listener, able to hear small impairments well above general public despite losing high frequency hearing.
Few things, -20db would be torture for me. -60db to -50db is what I normally listen to on headphones/IEMs. On speakers, I could approach -40db and it not be an issue depending on seating positions and such.

The tests that are difficult to pass, pertaining to 128kbps vs uncompressed is an overstatement (if that was the tests you are referring to). creating a diff of uncompressed vs 128kbps reveals essentially so much chopped out data, you can basically easily identify a song and even hear most lyrics even if it's your first time listening to it.

On the topic of training, I spoke favorably for your stance on that metric, you've got the experience especially after training yourself on what to watch out for, allowing you better ability to pick out undesired portions in audio much easier than many of us.

As for asking instead of reading tea leaves, asking what exactly? Age? Also, what tea leaves? Nothing about my statements are false (aside from the endless discussion as to what could be technically "too loud"), and I'm not reading into anything without deduction.

And finally, there was never doubt you could not pickout artifacts at your age, but as I said prior - it has to do with the extensive training. Music would be like gymnastics otherwise - once a certain ages hits, audio engineers and mastering folks would be out of the job if the case to the contrary was true, which of course it isn't as you know (and live as).
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
23,474
Likes
40,466
Location
Seattle Area
#42
Few things, -20db would be torture for me. -60db to -50db is what I normally listen to on headphones/IEMs.
On what music? There is plenty of music recorded well below peak. Even bass heavy tracks like Yelo have low levels of recording requiring much more gain. In addition, this is all headphone dependent.
 

pkane

Major Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
1,096
Likes
1,458
Location
North-East
#43
Ah, try 4 volt out, 4 volt in as I tested and let's see if the results are similar.
Aha! That makes a difference ;) I can't get to 4.0v out of my DAC, 3.5v is what I get at 0dbFS. I set HA-501 input/output to 3.5v. Now this is much closer to your result, THD+N of around -85dB. The original output of 1.9v was for my normal listening level with HD650s.

hpa2.png
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
23,474
Likes
40,466
Location
Seattle Area
#44
The tests that are difficult to pass, pertaining to 128kbps vs uncompressed is an overstatement (if that was the tests you are referring to). creating a diff of uncompressed vs 128kbps reveals essentially so much chopped out data, you can basically easily identify a song and even hear most lyrics even if it's your first time listening to it.
The tests I post were at 320 kbps. They are no walk in the park.

As to 128 kbps, make sure you run the blind. We had third-party labs perform large scale tests for us and very few people can tell the difference even at 128 kbps if the content is picked correctly.
 

LTig

Major Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
1,180
Likes
1,656
Location
Europe
#45
Few things, -20db would be torture for me. -60db to -50db is what I normally listen to on headphones/IEMs. On speakers, I could approach -40db and it not be an issue depending on seating positions and such.
You reach sufficient sound levels with -40 dB volume? Do you have highly sensitive speakers (>= 100 dB SPL at 1W/1m)? If not I cannot believe that your amplifier or speakers are able to handle 0 dB (that would be 10000 times the power at -40 dB) - meaning the power amplifier is probably not good matched to its feeding source.
 

Tks

Addicted to Fun and Learning
Joined
Apr 1, 2019
Messages
948
Likes
991
Location
NYC
#46
On what music? There is plenty of music recorded well below peak. Even bass heavy tracks like Yelo have low levels of recording requiring much more gain. In addition, this is all headphone dependent.
Various, anything lower than like 10 or 9 rating of Dynamic Range easily will blow my ears to oblivion after a minute of the song.

Obviously it's headphone dependent, (mine being LCD2C's). I do have a pair of HD6XX's my brother uses most of the time I could test and see if it still holds true (doubt it does, and should approach your db metrics, with them being such high impedance and whatnot).

But regardless, knowing this, I don't see why even raise the point itself in that case if our listening gear wasn't going to be a factor of self evident contention.
 

Tks

Addicted to Fun and Learning
Joined
Apr 1, 2019
Messages
948
Likes
991
Location
NYC
#47
You reach sufficient sound levels with -40 dB volume? Do you have highly sensitive speakers (>= 100 dB SPL at 1W/1m)? If not I cannot believe that your amplifier or speakers are able to handle 0 dB (that would be 10000 times the power at -40 dB) - meaning the power amplifier is probably not good matched to its feeding source.
Speakers I'm not sure, don't listen to them much (they're at the living room TV, I don't use speakers for my desktop). As for headphones (LCD2C's and HD6XX's), amp source being a Magni 3 on low gain (till the THX 789 arrives).
 

Tks

Addicted to Fun and Learning
Joined
Apr 1, 2019
Messages
948
Likes
991
Location
NYC
#48
The tests I post were at 320 kbps. They are no walk in the park.

As to 128 kbps, make sure you run the blind. We had third-party labs perform large scale tests for us and very few people can tell the difference even at 128 kbps if the content is picked correctly.
Yeah I failed ABX testing some fast paced tracks that have electronic portions with shit Dynamic Range. But if it's music I've heard prior and had good chance to sample both 128kbps and FLAC for example, the rate of success is essentially 100% easily blind tested (Though I don't know the arguments against having to sample each source prior to picking them out all in one go after that, and how that invalidates blind testing). I also used foobar for these tests a while back just before joining the forum here.

EDIT: As for 320kbps I have a defeatist fallacy going on within me where I'm convinced I don't think I can pick those out between FLAC and MP3. I have had success rates of 70-85% aproximately, but honestly believe those were pure luck. Especially after seeing diff files between FLAC and MP3 320kbps.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,887
Likes
4,862
Location
Riverview FL
#49
Doing the current measurements and then moving on to subjective listening tests will inevitably introduce "affirmation bias" in the listening tests. I seem to detect that in some of the comments of the tests
Good catch.

But listening first could lead to "affirmation bias" when subsequently editorializing upon the measurements.

Back to square 1.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,887
Likes
4,862
Location
Riverview FL
#50
You reach sufficient sound levels with -40 dB volume? Do you have highly sensitive speakers (>= 100 dB SPL at 1W/1m)? If not I cannot believe that your amplifier or speakers are able to handle 0 dB (that would be 10000 times the power at -40 dB) - meaning the power amplifier is probably not good matched to its feeding source.
Experiment:

Listen to Joe Biden tell story (TV).

Attenuate the DSP while advancing the volume on the preamp to "wide open".

-45dB turns out to be a comfortable TV talk level here.

1556571257168.png
 
Last edited:

LTig

Major Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
1,180
Likes
1,656
Location
Europe
#51
Experiment:

Listen to Joe Biden tell story (TV).

Attenuate the DSP while advancing the volume on the preamp to "wide open".

-45dB turns out to be a comfortable TV talk level here.
I fear I don't understand much here - and Joe Biden does not tell stories in our TV channels.:p

In my chain (Classe Sigma SSP -> K&H O300D) the volume goes from -93 to +14 dB, with 18 Vrms maximum level at the balanced output (ergo 3.6Vrms at 0 dB). The sensitivity of the O300D is +6 dBu = 3.8 Vrms for full output of ~112 dB SPL, so it's properly matched and I have almost 14 dB reserve for digital tracks with very low level (seldom these days).

I reach sufficiently loud levels around -20 dB for compressed rock, around -10 dB for classics, and only Dire Straits Ride across the river on Brothers in Arms (first edition CD, phonogram/vertigo 824 499-2) requires +5 dB for full loudness. When I listen really loud the SPL meter on my smartphone shows average levels around 85 to 90 dB ( I calibrated the app with pink noise according to an analog instrument by Rhode & Schwarz which I could borrow).
 
Last edited:

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,437
Likes
6,475
Location
Gold Coast, Queensland, Australia
#53
Based on the powers measured by Amir I conclude that the amplifier runs with a supply of +/-15 V (215 mW at 300 Ohm require 22.6Vpp).
A logical conclusion and one I also considered. If that's the case however, the twin rail regulators on the other side of the case (next to the transformer) (we assume as 7815/7915 or equivalent) would be marginal running two channels and would get awfully hot- much more than those pcb mount sinks would be able to deal with IMO. That is, if it is Class A.

Are you sure that 8920 is only doing buffering? Or it's the gain stage or it's differential amplifier to sum the xlr input? I personally think half of the 8920 does summing and the other half provides gain. And the soic8 opamp provides feedback for the discrete output stage.
There's more to it than that, and it's pretty much impossible to work out exactly what they are doing without one in my hand.

This pic shows a lot more but is low resolution (can't zoom in). This PCB is also different to the PCB I posted earlier- note the electros immediately above the 8920, the row of muting transistors/diff pairs and the lack of a hole. It may be a different product or an earlier iteration.

HA-501 pcb.JPG


The 8920 is likely the diff front end.
The SMD SOIC-8 is likely the DC servo they speak of. (one for each channel)
The SOIC-8 in the middle is likely a delay/dc protector/comparator that indirectly drives the muting. Obviously the damping resistors and 5x DPDT relays are at the bottom of this pic.

That's my take based on a few grainy pics and not much else. Maybe Amir can gut the thing and settle it for us?
 
Last edited:

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,437
Likes
6,475
Location
Gold Coast, Queensland, Australia
#54
Doing the current measurements and then moving on to subjective listening tests will inevitably introduce "affirmation bias" in the listening tests. I seem to detect that in some of the comments of the tests.
But then you have the opposite too. Where you 'like' something for esoteric/nebulous reasons (subjective) and seek to dismiss test results where the issues are arbitrarily deemed below the threshold of audibility and irrelevant to real listening enjoyment.

I detect that too in some of the testing on ASR.

It's impossible for one person to be completely impartial. I think that's why they have subjective reviewers and objective reviewers that compare notes at the end. One seeks to explain what he/she hears crosschecked with the measured data and the other gets to see whether the listener heard the technical anomalies they found.
 

Dro

Member
Joined
Apr 8, 2018
Messages
99
Likes
49
#55
Being able to identify high quality lossy compression vs lossless can be indicative of hearing damage. The psychoacoustic models used by lossy compression rely on healthy hearing.
 

GGroch

Senior Member
Forum Donor
Joined
Apr 7, 2018
Messages
439
Likes
553
Location
Denver, Colorado
#56
Being able to identify high quality lossy compression vs lossless can be indicative of hearing damage. The psychoacoustic models used by lossy compression rely on healthy hearing.
That's interesting. Do you have any documentation for this seemingly counterintuitive claim?
 

Dro

Member
Joined
Apr 8, 2018
Messages
99
Likes
49
#57
That's interesting. Do you have any documentation for this seemingly counterintuitive claim?
I recall a very old MP3 encoder listening test in which the candidate with hearing damage performed the best, but I don't have that handy.

I would not call it counterintuitive though. Lossy compression relies on auditory masking which normally prevents you from hearing one tone if you hear another tone. The masked tone can then be discarded to save data. If you have hearing damage and cannot hear the masking tone, you will notice the normally masked tone is missing.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
23,474
Likes
40,466
Location
Seattle Area
#58
Have you tested the Teac against its rated specifications? 1 volt in, 1 volt out?
You can see that in the power graphs:



You just have to work backward to convert the watts into voltages.

For my dashboard, I tested its competitors at 5.8 volt so going to 4 volt is actually generous to it.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
23,474
Likes
40,466
Location
Seattle Area
#59
Being able to identify high quality lossy compression vs lossless can be indicative of hearing damage. The psychoacoustic models used by lossy compression rely on healthy hearing.
No. Perceptual masking works whether your hearing is good or not. Ditto for quantization noise floor. The brain wants to discard information it thinks is of no value. It doesn't become damaged because your inner ear nerves do. :)

What training does is to enable you to break these barriers to some extent. That is why I can hear these distortions and others cannot.

As it happens, my threshold of hearing is better than average population in my last audiologist test (a couple of years ago).

Hearing loss does lower one's ability to hear impairment, not improve it.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
23,474
Likes
40,466
Location
Seattle Area
#60
If you have hearing damage and cannot hear the masking tone, you will notice the normally masked tone is missing.
The masker is at lower frequency than what is being masked. https://www.audiosciencereview.com/forum/index.php?threads/audibility-of-small-distortions.67/



If you don't hear the masker (gray) due to age related nerve damage, then you are not going to hear frequencies after that either.

The masking effect is much weaker in reverse (i.e. masking the frequencies before) so that is not going to be a likely scenario either.

Regardless, my issue is age related nerve damage which impacts ability to hear higher frequencies. Our hearing thresholds are the best at mid frequencies as you see in the black line. So with respect to hearing impairments in that critical area, nothing is changed.
 
Top Bottom