I don't hear this distortion you talk about Frank, can you be more specific? What should we be measuring that we don't already?
I grant you early CD did sound grey and uninvolving, mainly for reasons already discussed here. I think in recent years, digital has come on a long way, very high quality sound is now more affordable than ever to the less than rich audiophiles among us. It still requires a bit of work though
davym, you've worked out a chain that "works", hence don't have that type of distortion as a significant issue. And at the recent audio show, the first one in many years for Sydney, the standard of digital was indeed much better than in earlier times - in part, because music servers were the mainstay. One particular room highlighted this beautifully - the standard was extremely good, but then the source was switched to CD to play a visitor's disk; an Oppo used purely for transport, feeding through to the single DAC in the room - SQ dropped off dramatically, the digital "unpleasantness" was there in spades.
That "unpleasantness" is a distortion - it's a corruption of the low level detail, some on the forum call it noise modulation. The difficulty is in the measuring, I feel the best approach is to do time domain comparisons; DiffMaker is one, highly flawed, attempt to develop such a tool, and there appears to be little interest in furthering that approach.
The "bit of work" is key - some expect the quality to emerge from the equipment, as is; my experience also is that plenty of fiddling is required, particularly in the area of minimising interference effects.
That "unpleasantness" is a distortion - it's a corruption of the low level detail, some on the forum call it noise modulation. The difficulty is in the measuring, I feel the best approach is to do time domain comparisons; DiffMaker is one, highly flawed, attempt to develop such a tool, and there appears to be little interest in furthering that approach.
The "bit of work" is key - some expect the quality to emerge from the equipment, as is; my experience also is that plenty of fiddling is required, particularly in the area of minimising interference effects.
Last edited:
I think CD has also had it's day if I'm honest, although a well done EAC rip to FLAC file does get past a lot of read error correction/guess work that CD players perform down stream. Bypassing windows mixer makes a big improvement to PC replay, if you cant hear it either your system or ears are not revealing enough information. The day is fast coming when physical medium is no longer wanted or needed by the masses, then CD will be where vinyl is now. I'm already over it
My recipe for anyone who wants high quality digital sound for the price of a semi respectable phono cartridge.
A PC netbook, quieter the better. A Samsung N140 250GB (used) and a suitable 2GB ram card can be got off ebay very cheaply. It's very quiet and plays music for up to 8hrs off it's battery.
Then follow this guide to turn it into a dedicated music server - Configuring Foobar2000 and Windows 7 for Digital Music Playback
I use ASIO4ALL v2 and also installed the ASIO foobar plugin. Be sure to install them correctly. If your windows volume slider still works you have NOT successfully bypassed windows mixer and you will wonder what the fuss is about. A lot of people make that mistake, I did!
Next plug the netbook into a decent USB DAC, I now use an HRT Music Streamer 3, partly because of it's very low power consumption because the netbook is often running on it's battery (no need for an external/upgrade power supply) but mainly because of it's excellent sound quality, for the price.
Plug that into your amp, play a decent FLAC file, be prepared to say good bye to your CD player.
A PC netbook, quieter the better. A Samsung N140 250GB (used) and a suitable 2GB ram card can be got off ebay very cheaply. It's very quiet and plays music for up to 8hrs off it's battery.
Then follow this guide to turn it into a dedicated music server - Configuring Foobar2000 and Windows 7 for Digital Music Playback
I use ASIO4ALL v2 and also installed the ASIO foobar plugin. Be sure to install them correctly. If your windows volume slider still works you have NOT successfully bypassed windows mixer and you will wonder what the fuss is about. A lot of people make that mistake, I did!
Next plug the netbook into a decent USB DAC, I now use an HRT Music Streamer 3, partly because of it's very low power consumption because the netbook is often running on it's battery (no need for an external/upgrade power supply) but mainly because of it's excellent sound quality, for the price.
Plug that into your amp, play a decent FLAC file, be prepared to say good bye to your CD player.
Been there done that.
And now you should say hello to a good USB to I2S converter placed directly near your DAC chip for best effect. With properly terminated twisted pair or coax.
Don't throw your CD's out just yet, you still have to live with the fact that NTFS corrupts all of your files eventually, until one day you are sitting there watching re runs of some tv show in VLC and it pops up with a box saying "this file is corrupted" despite playing the exact same file the day before.
My guide on how to rip to hdd and then do an md5 sum on the files: http://shadytreeprojects.com/index.php?title=How_I_rip_CDs_for_Archiving_to_my_Linux_Samba_Server
So unless you are doing MD5 hash calculations at the time that you've ripped your CD's to RAW (left justified PCM) format (using cdparanoia on linux) onto a extv4 formatted hard disk that is of the Western Digital brand (Seagate is too unreliable: http://shadytreeprojects.com/index.php?title=Main_Page#Reviews) then you are putting your CD collection in jeopardy.
Last edited:
Maybe it's lack of distortion and excessive dynamic range that is fatiguing?
One thing I've noticed with digital recordings is that they often have what I'd call an unnaturally low noise floor full of extraneous detail. In quite respectable classical music recordings you can hear sheet music rustling, musicians breathing and cars passing in the street outside. In pop music you can hear performers' feet tapping and so on. In my own digital recordings, before I had access to separate live and control rooms, I was plagued by fan and hard disk noise from the equipment.
An audiophile grade digital recording can also have dynamic range that, while it's more like the dynamic range of the original event, is a lot more than could have been achieved on vinyl. If you turn up a 24-bit orchestral recording so you can hear the quiet parts, a crescendo will slam you against the wall. Of course the original performance has even more punch, but the effect can be a bit overwhelming in your living room.
In the analog world, unwanted background noises would have been hidden by tape hiss and vinyl surface noise, and the gain would have been ridden on those crescendos. I can see how this could be perceived as easier to listen to for long periods and less fatiguing.
One thing I've noticed with digital recordings is that they often have what I'd call an unnaturally low noise floor full of extraneous detail. In quite respectable classical music recordings you can hear sheet music rustling, musicians breathing and cars passing in the street outside. In pop music you can hear performers' feet tapping and so on. In my own digital recordings, before I had access to separate live and control rooms, I was plagued by fan and hard disk noise from the equipment.
An audiophile grade digital recording can also have dynamic range that, while it's more like the dynamic range of the original event, is a lot more than could have been achieved on vinyl. If you turn up a 24-bit orchestral recording so you can hear the quiet parts, a crescendo will slam you against the wall. Of course the original performance has even more punch, but the effect can be a bit overwhelming in your living room.
In the analog world, unwanted background noises would have been hidden by tape hiss and vinyl surface noise, and the gain would have been ridden on those crescendos. I can see how this could be perceived as easier to listen to for long periods and less fatiguing.
Last edited:
I'm not one of those people. And certainly the scientific community in this field seems to agree with me.
They are not "band aids, to correct flaws inherent in the CD media and encoding."
They are proper engineering solutions to make sure the anti aliasing and reconstruction filters can be perfectly implemented without the need for extremely complicated and expencive analogue circuitry.
160 years of acoustical research.
If intermodulation products of ultrasonic frequencies exist, then a normal mic can record them and a redbook system can reproduce them.
Then you can point me to a research paper that clearly demonstrates that.
I only know of 1 such paper and they used very steep reconstruction filters and ear-splitting 120dB spl levels to just barely detect audible differences between redbook and hires. So the audible difference could just as easily be something else.
I have Pitunes and Pitunes2. Each has its own hard disk and one mirrors the other using rsync. The only issue I have is that for historical reasons
)) one of the disks is FAT32 and the other is ext4.
I have Pitunes and Pitunes2. Each has its own hard disk and one mirrors the other using rsync. The only issue I have is that for historical reasons
Nicely done but its no go for me. I need my hardware to be made up from widely available consumer grade components.
I run on a limited power budget aswell, as a result my server does much much more than just serve up music, when you think about it running multiple small micro controller computers is more inefficient than running a single computer with a single energy efficient cpu and ram. For microcomputers I would need multiple power supplies, adding cost and complexity.
It will soon be upgraded to a much more energy efficient cpu based around the BayTrail platform (7.5w TDP), a decent 'upgrade' from a power consumption point when your current server CPU is a 2120-T (35w TDP), the old server motherboard/cpu/ram will be given to family members, who will also be upgrading from a 160 watt (Core 2 Quad Q6600 at idle) computer down to the 2120-T with 35w.
Trickle down effect.
Its also difficult to use a 4TB or 6TB HDD on such a microcomputer. Or 6 of them.
Maintaining multiple sets of hardware is not only complex but clumsy and costly. If I had to buy a case for such a system it would take away maintenance money for my server.
Another advantage too that I have is that if a hardware fault occurs I only need to replace a hard drive, psu, cpu, ram, mobo.
Tell me what is the operational in-field repair manual for a broken wire on your device? For me its easy enough as I've got experience with whole x86 based computer systems going back 15+ years. It would show up also in logs. But with an arduino I would have to consult with google.
Last edited:
Needless to say I disagree violently with all of the above. Mainly because I have a lot more experience with ARM than x86.
The reason for 2 sets of hardware is simple, one hooked up to the stereo in the kitchen/workshop, the other in the living room. The fact that one is a backup of the other is an incidental bonus.
The only downside I found to the Raspberry Pi platform is that file uploads and syncing are a bit slow due to limited wifi speed. I tried to build a "pitunes3" using a surplus fanless Atom motherboard, but it didn't work out. The industrial BIOS on that board was so lousy that I couldn't get any OS to install.
Mainly because I have a lot more experience with ARM than x86.The reason for 2 sets of hardware is simple, one hooked up to the stereo in the kitchen/workshop, the other in the living room. The fact that one is a backup of the other is an incidental bonus.
The only downside I found to the Raspberry Pi platform is that file uploads and syncing are a bit slow due to limited wifi speed. I tried to build a "pitunes3" using a surplus fanless Atom motherboard, but it didn't work out. The industrial BIOS on that board was so lousy that I couldn't get any OS to install.
CD's are boxed up in the loft, I always keep them (if they are worth owning).
How about this - make WAV files and store them on a dedicated external drive which is rarely used, make FLAC copies for reply on your server then back them up on a second external drive.
Seems straight forward and if NTFS corrupts the files on your server you can simply replace them with fresh copy's. I haven't encountered this problem (yet) but I replace all the music on my server at least once a year anyway.
Last edited:
I've always been partial to WD drives, especially after all the Seagate 10 and 20 gig Seagate drives went bad the same week in my lab (Used to be a unix sysadmin). My backup scheme is to store all my CD rips on my Solaris server (file system is ZFS), backed up daily to a WD NAT box. I don't worry a lot about file corruption...
Low levels
But vinyl don't have any problem at low or very low level outputs only the noise.
The digital by a low level signals have no more steps for recreate the signal .
Is right ?
But vinyl don't have any problem at low or very low level outputs only the noise.
The digital by a low level signals have no more steps for recreate the signal .
Is right ?
Most digital audio systems are made to have only random noise remianing (called, dither) as the signal dissolves to nothing. The difference from analog is that a consumer digital playback system's noise floor can be made arbitrarily low, limited only be the self noise of the electronic parts. Said another way, digital has greater dynamic range. Dynamic range is the difference between the maiximum amplitude and the smallest resolvable amplitude. Analog systems do not have infinite resolution. Analog resolution is limited by noise, as it is for digital. If a given digital system has greater dynamic range than a given analog system, that means the digital system has greater resolution (or, steps, as you say).
However, I would agree that we haven't completely solved the problem of residual listening fatigue with digital, although, I do feel we are finally coming very close.
Last edited:
No. Digital audio isn't made up of discrete steps. A reconstruction filter reconstructs the input very very accurately.
If a digital system and a vinyl system have the same noise at the same level, both low level signals will be played back with the same precision. Assuming that both systems don't add any distortions. Witch we know is not true, vinyl adds quite a lot of distortion, where as the digital system does not. So in practice the digital system will reproduce the signal more accurate than the vinyl and have much less noise to.
A seldom talked about distortion, especially with low sample rate digital audio is envelope distortion. DSP engineers who develop envelope following algorithms usually require at least 20-30 times the sampling to data rate to avoid envelope distortion exceeding 1%.
This distortion only minimizes as the inverse first order of the sampling rate/passband ratio, and this significantly accounts for CD Redbook's poor HF and imaging performance.
This distortion only minimizes as the inverse first order of the sampling rate/passband ratio, and this significantly accounts for CD Redbook's poor HF and imaging performance.
I'm uncertain about what exactly you are refering to. I've not before heard of envelope distorion with respect to bandlimited PCM audio sampling and playback. There's no properly bandlimited signal which requires a sample rate more than 2x the signal spectrum. However, I'm always interested in learning something new about digital audio. Could you provide a link which discuss this issue. A quick Google search didn't appear to produce anything germane. Thanks.
Do you now and again check, fully, the integrity of that backup? A client I had faithfully did their tape backups, a daily ritual like cleaning your teeth - until one day a major glitch forced them to actually refer to the data on that backup ... uh, oh - a bug in the backing up software was effectively causing a corrupted backup to occur, had been happening for yonks - oh, dear!!
The trouble always is, that it's so easy to fall into a routine where you don't think any more about what's happening, what can go wrong - it's worthwhile, every now and again, to go through an exercise where you carefully think through, and check that every step is really happening as you think it is: a full integrity audit, IOW ...
That is the other advantage of my scheme, the backup gets integrity checked every time I listen to music in the living room.
I agree that envelope distortion is a thing, but I can't see any line of reasoning that explains how it could be audible as such. Our ears don't respond to envelopes, but to the sound waves themselves.
I agree that envelope distortion is a thing, but I can't see any line of reasoning that explains how it could be audible as such. Our ears don't respond to envelopes, but to the sound waves themselves.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.