| Author |
Thread |
|
Impulse_Response
Advanced Member
United States
733 posts
Joined: Jun, 2013
|
 Posted - 2014/09/08 : 05:07:04
At risk of looking like an idiot I decided to post this. This is something that has been bothering me for a while now and I have to find the answer. I don't mean to sound like a whiney bitch or an audiophile; I just want to know what's going on and if I'm being scammed.
I buy lots of tracks from Beatport, and a few from Junodownload and Trackitdown. I buy everything lossless when it's available. I've recently become very interested in digital audio and
sampling theory, so for fun I usually put the files in a spectrograph (I use Spek) to see what they look like and to confirm that they are, in fact, lossless files and not upconverts from mp3 or some other lossy format. I've noticed that on quite a few of my downloads, the spectrograms seem to show clear signs of lossy compression.
When I think of a good file I think of something like these (BTW you can see the track names at the top of the images. Get ready for an overwhelming number of images). They are clearly full lossless audio.
http://i.imgur.com/PhQBtPX.png
http://i.imgur.com/QvvNN7g.png
The main thing that I look for to spot compression is the steep "cliff," usually at 16kHz, though sometimes it's higher, like at 20kHz. The other is a really screwed up garbled or choppy appearance above that cliff. Those first two images show no cliffs and they cover the entire spectrum 0-22kHz (or close to it.)
I first noticed problems with some of my Nu Energy downloads, all of which are from Beatport and lossless. The earlier releases (~2002-2005) that I have from Nu Energy have a clear cliff
and garbled appearance.
http://i.imgur.com/GbKW6dD.png
http://i.imgur.com/mCLht6O.png
http://i.imgur.com/z1xGngm.png
http://i.imgur.com/PkL4CMn.png
However, later releases don't have this.
http://i.imgur.com/smCsasH.png
http://i.imgur.com/DiJILsd.png
http://i.imgur.com/OfgfysT.png
http://i.imgur.com/JRuQAcd.png
I noticed something similar with Relentless Vinyl. Earlier releases seem to show compression. . .
http://i.imgur.com/qT1pqr3.png
http://i.imgur.com/0sdtFbx.png
http://i.imgur.com/vMHrUHi.png
While later releases don't.
http://i.imgur.com/1s22EaQ.png
http://i.imgur.com/9o9MxZy.png
http://i.imgur.com/I3PGhU3.png
Finally, Warped Science / Underground Recordings. All but one of my files look suspicious.
http://i.imgur.com/fX7J8im.png
http://i.imgur.com/CedHfYK.png
http://i.imgur.com/AxdG9nJ.png
http://i.imgur.com/xuCyVsC.png
http://i.imgur.com/yVbtHpO.png
http://i.imgur.com/qI0chRd.png
http://i.imgur.com/SfK7lqS.png
. . . And the one good one.
http://i.imgur.com/xa2K1r8.png
Can someone explain what is going on here? First of all, am I a complete idiot and do I have no idea what I am talking about? Hopefully this isn't the case, as I researched this spectrogram
thing quite a lot.
If I'm right, then why would this happen? Did the producers use mp3 samples or synths that didn't go all the way up? Was something done in mixing or mastering that would cut out most of
everything above 16kHz? Did they maybe lose the original lossless files, upload an mp3 instead, and say "Let's just upconvert these and sell them as wavs to make a few extra pennies!"
Any insight would be appreciated.
As a side note, the digital downloads of the tracks from Underground Recordings sound horrible. Screwed up high end and near constant clipping throughout. Anyone know where to get good
digital copies of these?
__________________________________
Producers and record labels, please stop "loudness war" mastering everything. It sounds terrible.
Alert moderator 
|
silver
Admin
Japan
12,577 posts
Joined: Feb, 2001
894 hardcore releases
|
Posted - 2014/09/08 : 05:15:45
Lossless are as only good as the source, I would assume lossless in some cases is actually "the best version we have"
Alert moderator
|
Audio Warfare
Advanced Member
United Kingdom
3,047 posts
Joined: Mar, 2009
|
Posted - 2014/09/08 : 07:57:51
Some people roll off their masters at 20kHz and 30Hz so it could be that you're finding. Don't ask me why, seems pointless to me. I've heard people say it's to save headroom but it doesn't make a blind bit of difference I've found.
__________________________________
Listen to released and forthcoming Audio Warfare/Audio Weaponry tunes here:-
http://soundcloud.com/audio-warfare
Alert moderator
|
Samination
Advanced Member
Sweden
13,225 posts
Joined: Jul, 2004
195 hardcore releases
|
Posted - 2014/09/08 : 08:50:48
Considering NuEnrgy and Relentless are from the same place, it could be that the DATs wheren't the best (Kevin has had problems with missing and bad DATs, there's still a bunch of older NuNRG missing)
__________________________________
---------------------------------------------
Samination, Swedish Hardcore DJ
Happy, UK Hardcore, Freeform, Makina and Gabber
http://samination.se/
---------------------------------------------
Alert moderator
|
Future_Shock
Advanced Member
Australia
2,483 posts
Joined: Apr, 2007
|
Posted - 2014/09/08 : 11:06:59
Mastering for CD and vinyl are pretty different processes... It's a bit too elaborate for me to go into here, but you can research it quite easily.
A lot of older tracks, like silver said, would just be 'best version we have' and for some of them, the 'best we have' is a vinyl rip of from DATs in which case it would definitely NOT be mastered the same as a digital file and would show some of the the obvious differences you're seeing.
The other case is you don't know how it was mixed or mastered. Mastering engineers tend to know what they are doing hence why it is its own discipline, profession and has its own academia. Without having produced extensively or mastered extensively or have formal studio education, it is impossible to try and understand what they did or why.
Brief side note 1: it's basically ****ing pointless having anything in over 20Khz. You can't hear it. There is significant scientific evidence showing that humans, on average, tend to hear ~20hz - 20Khz. Again, this is pretty known scientific fact and anything you hear to the contrary is usually anecdotal. If you're been djing, producing in headphones or raving for years, chances are that it's quite a bit lower than 20 Khz.
Whilst any track that is old enough to come from DATs or is obviously a vinyl rip (due to limitations of old recordings) you can write off. You won't get a better quality. I've seen this myself buying MP3s from the next generation website itself.
However, if it's a relatively new track and you're seeing the same sharp rounding off at UNUSUAL frequencies (18-20Khz i would not call THAT strange because a well mixed a mastered track that would be one of the quietest frequencies anyway - but under that....) then you can assume you've been trumped, i think.
Another important side note is what you're listening to the track on. If you're listening on something like ipod headphones or worse LAPTOP SPEAKERS (shudder) - it's all in your head. They have a frequency range similar to that of listening to music through a dead fish. And your mind can play tricks on you with frequencies - badly. But it doesn't matter because you're listening on a shitty medium.
You know why people put up with 128 kbps MP3s? Because they can't ****ing tell the difference. With a well-encoded mp3 the average person either can't tell or doesn't deem the difference worth caring about. ****, even the people that can tell the difference between a 192 kbps mp3 and a 320 kbps mp3 are pretty ****ing rare.
Long and short of it is i wouldn't worry about it too much.
As a final side note, you could have run these through a spectrum analyser and had a much quicker time deciphering the results having not come from an audio background as they are much more plain to read.
__________________________________
New Future Shock Hardcore: https://soundcloud.com/futureshockgroup
Alert moderator
|
Samination
Advanced Member
Sweden
13,225 posts
Joined: Jul, 2004
195 hardcore releases
|
Posted - 2014/09/08 : 12:31:41
quote:
Originally posted by Andy_Influx:
You know why people put up with 128 kbps MP3s? Because they can't ****ing tell the difference. With a well-encoded mp3 the average person either can't tell or doesn't deem the difference worth caring about. ****, even the people that can tell the difference between a 192 kbps mp3 and a 320 kbps mp3 are pretty ****ing rare.
Ask any pirate, they will say that they can notice :P
Tbh, I only know of one song I could hear a clear difference in a high pitched vocal between 128 and 192kbps (mp3)
__________________________________
---------------------------------------------
Samination, Swedish Hardcore DJ
Happy, UK Hardcore, Freeform, Makina and Gabber
http://samination.se/
---------------------------------------------
Alert moderator
|
alstorm
Average Member
United Kingdom
190 posts
Joined: Dec, 2005
|
Posted - 2014/09/08 : 14:02:11
majority of our old releases, were uploaded as wav, the thing with older tracks is, hard drives, storage comes and goes, sometimes .. ive had to rip stuff off my old dj cds as the only remaining copy, although this should be wav, it could have been burnt as an mp3, and ripped as a wav, some tracks there is a chance all that was left alive is an mp3 that's HAD to be converted to be uploaded to labelworx etc, ... also sites do there own encoding too, as ive never converted a file to FLAC in my life, no guarantee what there conversion process is or what they have used as the source!
__________________________________
Al Storm
24/7 & Technique Recordings
www.247hardcore.co.uk
Alert moderator
|
Josephson_Junction
Advanced Member
United States
512 posts
Joined: Jan, 2012
|
Posted - 2014/09/08 : 14:27:39
What Andy_Influx and Al Storm said is spot-on. Aside from that, I presume that you're converting the files to FLAC yourself (considering that few digital retailers offer the format); The format is both lossless and compressed, so that's another thing that could factor into how the spectrograms appear.
I may as well ask: How are you going about converting the files? If you're on Windows, I'd recommend using the latest stable version of the FLAC encoder (1.3.0 at the time of writing this), which you can get here. It's limited to the command line, but usage is straightforward. For what it's worth, these are the arguments I typically use:
* -V: "Verify a correct encoding by decoding the output in parallel and comparing to the original"
* -8: Highest level of compression.
* --warnings-as-errors: Just a personal choice of mine, but I prefer to use this option on initial conversion runs since it's helped reveal major (but seemingly minor) issues with source files to me in the past.
Alert moderator
Edited by - Josephson_Junction on 2014/09/08 14:36:18 |
Impulse_Response
Advanced Member
United States
733 posts
Joined: Jun, 2013
|
Posted - 2014/09/08 : 18:18:38
Thanks for all the responses! Just to make sure I don't get the wrong idea across, I hold nothing against the artists or labels. I'm happy that they are making this music available and I enjoy listening to it.
I wondered about data loss, and that does make a lot of sense. If the best that still exists is an mp3, then I'll gladly take that. I can't hear a difference, and the only reasons I really want wavs are first that is feels good to have a "perfect" copy, and second for the best possible quality when I do my own mixes and go through recording -> mastering -> exporting -> any re-encoding that I do afterward to put it on my zune or whatever.
I guess I'll just stop worrying about this and keep buying music. :)
__________________________________
Producers and record labels, please stop "loudness war" mastering everything. It sounds terrible.
Alert moderator
|
Elliott
Advanced Member
United Kingdom
1,147 posts
Joined: May, 2012
|
Posted - 2014/09/08 : 18:25:28
quote:
Originally posted by Andy_Influx:
You know why people put up with 128 kbps MP3s? Because they can't ****ing tell the difference. With a well-encoded mp3 the average person either can't tell or doesn't deem the difference worth caring about
Proudly a member of this group.
__________________________________
old soundcloud
i gave up producing
Alert moderator
|
Impulse_Response
Advanced Member
United States
733 posts
Joined: Jun, 2013
|
Posted - 2014/09/08 : 18:45:30
Yes yes I know I know... I was just trying to figure this out. No reason to get angry at me, anyone.
__________________________________
Producers and record labels, please stop "loudness war" mastering everything. It sounds terrible.
Alert moderator
|
Captain Triceps
Advanced Member
United Kingdom
2,210 posts
Joined: Dec, 2011
|
Posted - 2014/09/08 : 18:48:17
quote:
Originally posted by Advather:
Yes yes I know I know... I was just trying to figure this out. No reason to get angry at me, anyone.
You total bastard!!!
__________________________________
Some of my remixes, original tracks and mixes here:
https://soundcloud.com/bradders-tracks-and-remix
https://soundcloud.com/bradders1982
https://soundcloud.com/paulbradley1982
Alert moderator
|
Future_Shock
Advanced Member
Australia
2,483 posts
Joined: Apr, 2007
|
Posted - 2014/09/09 : 07:24:20
I'm Australian... We have very colourful language...
....For ****'s sake :)
__________________________________
New Future Shock Hardcore: https://soundcloud.com/futureshockgroup
Alert moderator
|
Josephson_Junction
Advanced Member
United States
512 posts
Joined: Jan, 2012
|
Posted - 2014/09/10 : 17:58:46
quote:
Originally posted by Andy_Influx:
You know why people put up with 128 kbps MP3s? Because they can't ****ing tell the difference. With a well-encoded mp3 the average person either can't tell or doesn't deem the difference worth caring about. ****, even the people that can tell the difference between a 192 kbps mp3 and a 320 kbps mp3 are pretty ****ing rare.
How much of it is tolerance vs. actual inability to differentiate, though?
Alert moderator
|
|
Printer friendly
