advertisement


Can you tell the difference between 96/24 and 44/16 and mp3?

I'm absolutely with you on the 'not losing sleep any more' over this bit rate thing. I bought into the 'hi res' fad a few years ago, bought loads of Linn and Hyperion studio masters (all classical) then realised one day that I simply couldn't hear the differences, even through Stax headphones. Being a piano tuner by trade I like to think I have keen hearing (though too many years of it could have taken their toll I suppose...) so I shrunk them all to 320 and have been happy ever since.
 
Julf said:
I mean that when you say "I can tell the lossy algorithms at work", how did you verify that you actually can?
Click to expand...

Same way as I can tell if one dish has more salt added than another of the same recipe.

To be more exact, (I work in data storage management, compression, deduplication, and replication btw, so am biased)(hence being barred from the other Ethernet cable thread)

MP3 compression physicality removes from the music certain parts based on psychoacoustic ideals it assumes we don’t need/won’t miss (and many won’t ) the music content is altered.

Flac compression uses deduplilaction reduction processes. By this is it looks for patterns or fixed length chunks of 1 and 0s that are the same, then replaces these with a smaller token marker, thereby reducing the footprint. When data is decoded all the tokenised bits are put back in, check summed then delivered to DAC (in case of audio). There is nothing taken away. The music is not altered.
Thereby WAVs And FLACS (and ALACS btw) are the same. Any sonics differences perceived or heard here, could be due how the code on board DAC or app handles the files.

Within LMS etc there are transcoding options.

I have these all set the same.

So fir example I have fav demo track Tom Petty ‘crawling back to you’

I can play it from source CD thru same system and into LMS that I have then ripped (dbpowrramp) I to aiff, wav then flac.then also to 320 and 256 mp3.

I can play 3 source streams concurrently then switch on fly. Each source stream route is identical from same NAS (same Ethernet cables, same router, same switch into one SBT then to same DAC then active speakers.

so Tom Petty “crawling back to you”

Source CD
WAV/FLAC/ALAC/MP3 320 256

I can play 3 concurrently (using 3 LmS servers into SBT/DAC). Then switch on the fly.

NAS to SBT to Benchmark HGC to Dyaudio Actives. Velodyne sub. It’s not a bad system.

This as as standardised test - from ripping SW to playback infrastructure commonality - that I have tried.

One track I know intimately played via different codecs on same system.

First three sound identical to Source CD - factoring in my ears and brain. If I play those three I cannot discern and forget which one is playing after too many switches.

MP3s are subtly different - sound increasingly lacking space depth ambiance and nuance as the bit rate decreases. Less enjoyable. I can tell instantly.

So that’s my evidence. Not scientific but I believe my ears on this one.

Ps no 24 bit available for this test. But I use Qobuz Sublime and can stream or download many of same tracks with 24/16 rates. I do this often and struggle to discern. Using above infrastructure and system.

As an aside Some Hd tracks are a complet con. For example Hawkwind Space Ritual I have CD, then bought 16 bit then 24 bit downloads from Quobiz. Not only do they sound no different to me - they were recorded analogue in first instance - the 16 bit and 24 bit files are the exact same size in total andbtravk by track. I should have complained.
 
guest said:
So that’s my evidence. Not scientific but I believe my ears on this one.
Click to expand...

Not scientific, but way better than the usual sighted listening. If you wanted to go further, I can definitely recommend the ABX plugin for foobar2000.

As an aside Some Hd tracks are a complet con. For example Hawkwind Space Ritual I have CD, then bought 16 bit then 24 bit downloads from Quobiz. Not only do they sound no different to me - they were recorded analogue in first instance - the 16 bit and 24 bit files are the exact same size in total andbtravk by track. I should have complained.
Click to expand...

Yes, I have also come across several HDtracks albums that are clearly upsampled.
 
I am a big Hawkwind fan, but I would never claim that their recordings were audiophile demo material. Nothing requiring better than 44.1/16
 
davidsrsb said:
I am a big Hawkwind fan, but I would never claim that their recordings were audiophile demo material. Nothing requiring better than 44.1/16
Click to expand...

Agree. That wasn’t best example. The Space Ritual cd mastering is also not the same as original vinyl I used to have. And my memories of Hawkwind could possibly have been going through a magic mushroom sndbweed filter all those years ago.

Other ‘more hi fi’ tracks I stream or downloads by Rush or Ryan Adams or Counting Crows are also Indiscernible 16 or 24.
 
Yes.

Not much experience of high res, but having ripped the same CD at 44/16 (lossless) and MP3 (320) it's quite obvious in a side-by-side.
 
There is a potential problem in creating something like an mp3/aac from material that is clipped (or *almost* clipped). Ditto for changing the sampling rate. People can easily understand that 'upsampling' may imply 'overs' in between the original samples. But there is less awareness that reducing the sample rate or removing some patterns can alter how the result clips.

So tests should use material that is clear of clipping to start with. Alas, I'm not sure that means every modern digital version would pass this requirement.
 
david ellwood said:
There’s a big difference in my system between playing flac and wav of the same recording.

On the fly decoding of the flac really messes up the rhythm compared with playing the straight wav.
Click to expand...
Same here. I have up on transcoding and converted my library to WAV, much better and totally free.
 
Julf said:
Played on the same gear? Sounds like a level difference.
Click to expand...
Of course, hardly a comparison otherwise. Same CDs ripped in both formats, played back through the same gear. Through both PC and even burning both formats to CD to compare that too.
 
Fretbuzz said:
Of course, hardly a comparison otherwise. Same CDs ripped in both formats, played back through the same gear. Through both PC and even burning both formats to CD to compare that too.
Click to expand...

So how did you verify the levels?
 
Totally agree with the superiority of flac with regard to metadata.

if wav wasn’t quite a lot better I wouldn’t entertain all the extra hard drive space.

If your hearing is insensitive to the difference you’ve saved yourself a lot of inconvenience.
 
david ellwood said:
Totally agree with the superiority of flac with regard to metadata.

if wav wasn’t quite a lot better I wouldn’t entertain all the extra hard drive space.

If your hearing is insensitive to the difference you’ve saved yourself a lot of inconvenience.
Click to expand...

If there is an audible difference, then either your player is really bad at decoding FLAC, or the FLAC files include ReplayGain information telling the player to play them at a lower volume.
 
I have a £3,600 cable that transfers 24bit into 48bit. Must admit I dont hear a difference but strangely my dog is looking more interested in music.
 
If someone really has a noise issue with a device decompressing (lossless) compressed FLAC, why not convert to uncompressed FLAC?

Shouldn't that be the best of both worlds? Useful metadata capabilities and no noisy number crunching?
 
Last edited:
You must log in or register to reply here.


advertisement


Back
Top