Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: High resolution audio - lowest audio setup? (Read 6534 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

High resolution audio - lowest audio setup?

Hiyas, coz I'm quite disappointed that my audio gear otherwise can play from hires sources, but I can't make any difference of it, I'd like to have some idea where lays the perceptible threshold, and also where approximately is the true hi-res entry price level.
I realize that it may be the ears or bad mastering that prevents to hear any difference, but given that ears and record are sufficient, I'd like to have a thought how should the most minimal setup look like, which not only reads high resolution, but also is able to "write" it, ie. it can in real render the added resolution.
Can for example anybody post a sample of concrete setup where he's able to ABX 24bit PCM or DSD audio from standard resolution of the same source? (by standard I mean something like CD standard - 44100Hz/16b).

Re: High resolution audio - lowest audio setup?

Reply #1
44.1khz/16 bit is a pretty reasonable choice. Higher makes relatively little practical difference.


Re: High resolution audio - lowest audio setup?

Reply #3
Why 34 though? It surely depends on ears.
When I tested my ears range, the highest frequency that I could still confidently hear was about 18kHz. Therefore, to keep it, a sampling rate of ≥ 36 kHz would be needed.
Agree about 14 bits, for most practical use cases it could work.

Quote
if there was such a thing
Something close can be a thing — 32000 Hz (even a bit worse, but passable for many material) is a relatively common rate, and you can simply store only 14 bits in 16, most reasonable lossless encoders (incl. FLAC, WavPack) will deal with it efficiently.

(However, if one simply wants to minimize space usage in a portable player, it's still far more efficient to not mess with sample rates and bits and just encode the copy for the player in Opus.)
a fan of AutoEq + Meier Crossfeed

Re: High resolution audio - lowest audio setup?

Reply #4
Quote
Why 34 though?

16 khz is a widely used cut off in psychoacoustics. It is certainly as low as I would want to go.

Re: High resolution audio - lowest audio setup?

Reply #5
The ability to hear pure tones does not tramslate to being able to hear those frequencies in music.  Blame that pesky masking.

FWIW, 34 was chosen so that there would still be plenty of margin for filtering.

Re: High resolution audio - lowest audio setup?

Reply #6
Being able to really distinguish 16bit an 24bits music it only happened to me two times in a blind comparison, more correctly with only two songs of classical music that where direct recordings and also where specially dynamic and I'm not sure if it was due to the bit depth differences or the conversion... both song where level matched and in one I sensed if like the 24 bit version had more punch and in the other a near silent pasage in the 24bit version was totally silent in the 16bit version, and in the one that had more punch I guess the bit depth correctly because I knew that it was a direct recording.

But any other time that 16bit and 24bit sounded different was due to the audio path and the DAC, because in this situation while the 16bit and 24bit sound different a 16bit to 24bit converted file sound identical to the native 24bit audio version.


Re: High resolution audio - lowest audio setup?

Reply #8
We know nothing about the test or how it was performed, but I guess expressing doubt about the veracity of the claim without the usual shitstorm ought to be good enough.

Re: High resolution audio - lowest audio setup?

Reply #9
The ability to hear pure tones does not tramslate to being able to hear those frequencies in music.  Blame that pesky masking.

FWIW, 34 was chosen so that there would still be plenty of margin for filtering.
For some music/genres, there could be very loud tones in that area.
Off the top of my head, very loud HF tones (perhaps above 16kHz but I don't remember exactly) can be heard in "Perdition City" album by Ulver. (a musically good album by the way if you ask me).
I guess some other extreme/experimental/electronic music could have something like that too sometimes.
Sure, in most of the regular music the content above 16kHz would be extremely hard to hear if at all possible, but music is a broad term and it seems that signals with whatever characteristics may comprise music nowadays.
a fan of AutoEq + Meier Crossfeed

Re: High resolution audio - lowest audio setup?

Reply #10
Being able to really distinguish 16bit an 24bits music it only happened to me two times in a blind comparison, more correctly with only two songs of classical music that where direct recordings and also where specially dynamic and I'm not sure if it was due to the bit depth differences or the conversion... both song where level matched and in one I sensed if like the 24 bit version had more punch and in the other a near silent pasage in the 24bit version was totally silent in the 16bit version, and in the one that had more punch I guess the bit depth correctly because I knew that it was a direct recording.

But any other time that 16bit and 24bit sounded different was due to the audio path and the DAC, because in this situation while the 16bit and 24bit sound different a 16bit to 24bit converted file sound identical to the native 24bit audio version.
Your comparison was either not blind, or totally incorrect.
Bit depth conversion can't alter the amount of punch[es] at all, even in theory, even if done in worst possible way.
I'd suggest testing 24 bit record against itself converted to 16 bits. Not against some other record which could be made from something else entirely — that would make the test incorrect.
a fan of AutoEq + Meier Crossfeed

Re: High resolution audio - lowest audio setup?

Reply #11
Hiyas, coz I'm quite disappointed that my audio gear otherwise can play from hires sources
Please keep in mind that these high-res sources usually offer different material to normal 16/44.1 sources. (Usually a different master.) If you take those high-res files and convert them to 16/44.1, they will sound just as good. It's not the higher resolution that's giving the better audio quality, it's their higher quality source material.

So you don't need "high-res" equipment to reap the benefits of their higher quality source material. *IF* they actually have higher quality source material (someone here pointed out that this isn't always the case.)

Re: High resolution audio - lowest audio setup?

Reply #12
Off the top of my head, very loud HF tones (perhaps above 16kHz but I don't remember exactly) can be heard in "Perdition City" album by Ulver.
Show me.  Better yet, show me an ABX log indicating you can hear such an extreme mosquito tone in that piece of music when compared to a 16.5k low-passed version.

Quote
I guess some other extreme/experimental/electronic music could have something like that too sometimes.
I guess we're left with guessing.  But sure, you can't disprove a negative (or however the saying goes).

Quote
{music is a broad term and it seems that signals with whatever characteristics may comprise music nowadays.
Indeed. I guess it can be written for dogs and bats which will obviously require a higher sampling rate. ;)

Re: High resolution audio - lowest audio setup?

Reply #13
Off the top of my head, very loud HF tones (perhaps above 16kHz but I don't remember exactly) can be heard in "Perdition City" album by Ulver.
Show me.  Better yet, show me an ABX log indicating you can hear such an extreme mosquito tone in that piece of music when compared to a 16.5k low-passed version.

Alright, I just found the track which contains this (07 - "Dead City Centres"), and the frequency of that sound is ~18178 Hz, which will not be preserved with 32 or 34kHz sample rate.
I am quite confident that I will pass the test, so, shall I just apply a hard lowpass filter at 17000Hz, or resample to 32000Hz, or resample to 34000Hz for the sake of the test?
Also, post it here, or in a new thread in "Listening Tests" category?
a fan of AutoEq + Meier Crossfeed

Re: High resolution audio - lowest audio setup?

Reply #14
Show me.  Better yet, show me an ABX log indicating you can hear such an extreme mosquito tone in that piece of music when compared to a 16.5k low-passed version.
AFAIK, young teens and children are able to hear these more easily than adults. IMO, we shouldn't exclude kids. 44.1kHz sounds like the safest option for that?

Re: High resolution audio - lowest audio setup?

Reply #15
Listening tests. Low pass at 16.5k.  Check it to make sure there is no imaging and levels are identical in areas where there isn't any HF "musical" tones to be discarded.

Provide clips the critical part of the sample before and after low-passing that are 30 seconds or less.

Bonus points for providing a vinyl version that can withstand repeated plays.

Re: High resolution audio - lowest audio setup?

Reply #16
Listening tests. Low pass at 16.5k.  Check it to make sure there is no imaging and levels are identical in areas where there isn't any HF "musical" tones to be discarded.

Provide clips the critical part of the sample before and after low-passing that are 30 seconds or less.

Bonus points for providing a vinyl version that can withstand repeated plays.
https://hydrogenaud.io/index.php/topic,115880.new.html#new
I think I've done the lowpassing part correctly.
Spoiler (click to show/hide)
a fan of AutoEq + Meier Crossfeed

 

Re: High resolution audio - lowest audio setup?

Reply #17
Your comparison was either not blind, or totally incorrect.
Bit depth conversion can't alter the amount of punch[es] at all, even in theory, even if done in worst possible way.
I'd suggest testing 24 bit record against itself converted to 16 bits. Not against some other record which could be made from something else entirely — that would make the test incorrect.
It was not a personal comparison and was properly done, in it participated 20 person at 5 songs per day during two weeks with the same equipment, all the 16 bit versions where converted from the 24 bit masters with the level matched, also this both songs with anomalies where not from a commercial record but where direct orchestral records and nearly unprocessed (the only processing was the bit depth and sample rate conversion needed for the test) and for a clarification the where only 5 persons that heard the differences and not all heard the difference in both samples. This two outliers where special, we don't know the reason but there where suspicions that the one with more punch it was not level matched properly due to impulse behaviour in the resampler, in the other is that the near silent part was possibly below the noise level of 16 bit audio due to the conversion, an for the purpose of the comparison they where treated as statistical anomalies.

PD. I don't know if punch was the correct word as I'm not a native English speaker and the sensation that i had is hard to describe for me.