The quality of my rips

[The_I]

I am unsure whether this does fit exactly into this forum because I never got all the basics about audio-cd mechanics while working on a pc. The thing I’m worried about is that when I play and rip audio-cd’s I can only achieve 24-bit 44.100 khz. This can’t be perfect can it? I have a brand new Audigy 2 zs so I should be able to do way better, shouldn’t I? Yet when I playback audio-cd’s in winamp it only shows 44.100 khz, and in mediasource’s settings it says I am only able to rip in 41.100 khz waw. I have connected the cable supplied with the card from my LG dvd-drive to the port on the card, that shouldn’t be possible to do wrong, and isn’t that spdif and therefore digital?
It’s a bit important as I plan to start ripping 20-30 cds so I don’t want to do it wrong, I’m planning on using EAC and LAME rj3mix to create my mp3-library by the way.

 

[phsstpok]

You do know that audio CDs have only 44.1K samples/sec and 16-bit encoding. That's the Redbook CD Audio standard.

Don't know why your rips are showing 24-bit unless encoders convert the data to 24-bit. (I never really thought much about this). Even if they do convert to 24-bit, you can't achieve higher accuracy than the original CDs contain.

By the way, your sound card isn't normally used for ripping. You don't even need a sound card to do ripping. Normally, the digital data is read over the IDE bus/SCSI bus/USB port in data format directly from the CD drive and converted directly to the other digital formats, WAV, MP3 etc.

If you really want to you can record from the CD-in (the port for the standard, analog CD cable) of your sound card. However, all that will happen is the CDROM's inexpensive DAC will be used to convert the 16-bit data stream into analog. This will then be sampled by your sound card and converted back to digital.

I guess you can use the SPDIF-out available on some CDROMs/DVDROMS and feed it into your SPDIF-in on you sound card. I don't know the workings of how that data would be handled but remember, the source is still the 16-bit encoded, 44.1K samples/sec, audio CD.

<b>56K, slow and steady does not win the race on internet!</b>

 

[The_I]

Ok, now I can start ripping without second thoughts (when I have got the right cd-drive for it) your reply has helped me out of a few misunderstandings. Yet I remain confused about one thing: I’ve seen hifi cd-players that boasts a 24bit/96 kHz DAC, if cds are 16bit/44.1kHz how can this be?

 

[phsstpok]

I’ve seen hifi cd-players that boasts a 24bit/96 kHz DAC, if cds are 16bit/44.1kHz how can this be?]

That's a good question.

I can see receivers accepting 96 Khz. Also, Super Audio CD and DVD Audio would need the better resolution/sampling but those formats require DVD drives not CD players.

I don't know the answer but I would guess that it's to support MP3s. MP3s need not be sourced from 16-bit audio CDs. Any source could be used. Live recordings come to mind.

Do the CD players that you speak of support MP3 playback or are they Audio CD only players?

<b>56K, slow and steady does not win the race on internet!</b>

 

[The_I]

Ok, think I found the answer, surfed some of the sites of the cd-players I've seen with the 24bit/96khz dac, here's what I found on one of them:

'Normal CDs are recorded in 16bit/44Khz resolution with a theoretical dynamic range of 96dB. When played through Digit II, the high-frequency bitstream is intelligently interpolated on the basis of the last 64 input samples, then filtered and re-sampled at 24bit/192Khz resulting in a 3-5dB improvement in the dynamic range.'
I think they call it upsampling.

 

[jubei_woop]

its a way of getting more money from us. it wont sound any better, just slightly different from the sampling

<P ID="edit"><FONT SIZE=-1><EM>Edited by jubei_woop on 10/29/03 04:45 PM.</EM></FONT></P>

 

[phsstpok]

Hmmm, but that sounds artificial, increasing the dynamic range that wasn't in the original source, that is. That technique presupposes that the original was compressed, which may or may not be true.

All I can say is try it. See if you can notice a difference. Remember though, CD drives don't have this upsampling capability (AFAIK). You'll need a standalone player.

The question is how do you take advantage of the up-sampling? Expensive players may have SPDIF-out but is the increased resolution available?

If the higher resolution is only available via the DAC then you have to use the player in analog mode and record using your Audigy 2 ZS? This would mean one D to A process, one A to D process, and one digital compression process (MP3 encoding).

Or would it be better just to use a CDROM drive, extract the 16-bit data digitally (with the full accuracy of EAC), and then encode (simultaneously if you prefer)? This way everything is done in the digital domain.

Which method yields the best results? [I don't know]


<b>56K, slow and steady does not win the race on internet!</b>

 

[phsstpok]

its a way of getting more money from us.

That's my opinion, too.

<b>56K, slow and steady does not win the race on internet!</b>

 

[The_I]

Don't you think that the up sampling can be done by the soundcard? I am pretty sure that I get the sound purely digitally (via PCI-bus I figure as it doesn't seem my SPDIFF-connection to my DVD-drive works) I should be able to detect those 'samples'? I have noticed that sometimes I actually do play the sound in higher resolution, on some tracks on some CDs I can play high-definition audio, but that seems to be because the CD is coded that way, as far as I understand it up sampling is done on all CDs. And as far as I could gather from what I've read up sampling is done on all CDs in some way or another, in the 20bit mode mentioned it just uses some 'sample' and applies it to all of the track it seems. Anyway, it would be really nice if someone who knew exactly what is going on when you play a CD/HDCD/DVD-audio/SACD (that's how many there are now!).

What I've heard though is that anything with higher quality than CDs is not needed. On the site I read concerning mp3s (rj3mix.net, it's gone now) the author illustrates it by mentioning that the CD-medium was made to have 'headroom' so it recorded more than the human ear could hear and way more than the microphones could. According to him if you removed the other bottlenecks you should be able to hear the sound of the AIR CURRENTS inside the studio!
I have heard some DVD-audio samples supplied with the sound card and they DO sound good, but my suspicion is that they might sound as good in CD-quality, if anyone knows some test of this it would be interesting to read.

 

[phsstpok]

Never said I was expert. Feel free to disagree. I was just thinking about the logic of it. I mean, if you start with 16-bit and go to 24-bit then you are just extrapolating the data. You are creating something from nothing.

Maybe there are some clever algorithms that I don't know about which make the sound "better" but I don't see how.

Anyway for sound quality, I guess it's debatable but I certainly I have no evidence one way or the other whether 24-bit is noticably better than 16-bit, even if you have a true 24-bit source. I'm make no claims in this regard.

One aspect of sound quality I can challenge is noise and dynamic range. Even though CD players since 3rd generation (circa 1985) have claimed 100+ dB of dynamic range and S/N ratio this is irrelevent in a PC environment because PCs have fans. Even a so called silent fan at 20 dBA is going to ruin your S/N ratio and it will be worse if your PC happens to be closer to you than your speakers. Add to this the fact that the inside of a PC is electrically noisy. This probably ruins the theoretical S/N ratios of sound cards.

Moving on...

About getting digital sound from your DVD-drive, how are you doing the connection? I mean what port on your sound card are you using?

I'm also curious as to why you are trying to do recordings this way. Why would this be better than extracting the data via the IDE bus?

True, it's pure digital either way. However, if you are using EAC you have the advantage of the software controlling what the drive does. If there are read errors EAC can read the data over and over and take an average, the most common value (mode), or extrapolate if necessary.

To record via SPDIF the drive has to be in play mode, and is therefor at 1x speed. This is an extreme fault tolerant mode. The drive must keep playing no matter what (because you wouldn't want interruptions in the music). Any errors are quickly masked or extrapolated in hardware. The data can't be fixed. Your recordings will reflect this.

[edited for clarity]

<b>56K, slow and steady does not win the race on internet!</b><P ID="edit"><FONT SIZE=-1><EM>Edited by phsstpok on 11/02/03 01:52 PM.</EM></FONT></P>

 

[The_I]

I am going to use the IDE bus for rips, rest assured, and I’m wondering whether I’m not using it for playing as well – I believe I’ve selected a setting assuring digital playback, but I don’t believe the cable I have from my CD to the soundcard SPDIF-in is working (it can’t detect a signal. What signal will the card use for CD-sound if I remove the cable?
And I have to agree with you about being skeptical towards higher medium-qualities; the bottlenecks are probably somewhere else. Yet it would be nice to se a test or some such.

 

[phsstpok]

When you select digital playback the data is passed over the IDE bus just as if you were ripping. This form of playback does place a load on your CPU. The amount depends on the speed of your CPU and the sound card and its drivers.

If you select analog playback then you need have the 4-pin cable going from the CD-ROM/DVD-ROM to the port marked CD-ROM on your sound card. Since this is just an audio passthru feature there is no load on the CPU with analog playback.

I don't know how you would go about using the SPDIF connector on your DVD-ROM for playback. I would assume you feed the SPDIF-out on the DVD-ROM to the SPDIF-in on you sound card and somehow you set your player software accordingly. Just a guess on my part, since I've never heard of anyone doing this.


<b>56K, slow and steady does not win the race on internet!</b>

 

[The_I]

Well I figure I don't have to worry about quality then. As said I have a cable connected from my dvd-drive to the soundcard's spdif in, but I don't think the soundcard recognises it. Anyway, it's probably my drive that's not good enough, my cpu util averages at around 2% in playback, I think I can live with that.

 

[phsstpok]

Somehow I missed you mentioning you were using SPDIF-in. Sorry for asking/mentioning about it multiplie times.

I heard the SPDIF port on some CD-ROM drives isn't enabled (especially on drive that label it "reserved") but I didn't think it was an issue on DVD-ROM drives. I'd like to try it but I don't have a DVD-ROM drive.

2% CPU utilization with digital payback? That's great! I get about 3%-4% which is fine. The problem is I'm still using Windows 98SE. The WDM drivers for my cheapo sound card didn't seem to work. This means that I can't normally access digital playback.

PowerDVD gets around this issue somehow but I'd rather use MusicMatch Jukebox.

<b>56K, slow and steady does not win the race on internet!</b>

 

[umheint0]

Okay, lots going on in this forum, I'll try to attack it all.

We'll start with the first post. It doesnt' matter what CD player you use. All it does it take the raw digital signal, and transmit it to the CPU for processing. It doesn't actually do any decoding on it's own. The quality of the rip will depend on the codec you are using to make the file (ie: Blade, LAME, OGG, etc...), moreso than the actual bitrate.

For re-sampling now. Up-sampling theorizes what the missing samples will be, and using a complex algorithm fills in the gaps. You can upsample to any number you want, but it all depends on whether your player can handle it or not. In my judgement, it's difficult to add to a source what isn't there. It's completely artificial, though the quality of the codec doing the upsampling comes into play (once again).

Yes, the original CD tracks are compressed. Just an FYI, but the original tracks that are recorded at the studio use analog tapes a lot of the time. LPs themselves have about 5-10x the signal density of a CD. They give a richer sound because they aren't compressed, but aren't as clear because they are analog and are succeptible to wear.

Any other questions?

umheint0's phat setup --> <A HREF="http://home.cc.umanitoba.ca/~umheint0/system.html" target="_new">http://home.cc.umanitoba.ca/~umheint0/system.html</A><--

 

[The_I]

If you mean CD-drive by CD-player what I have heard disagrees with your post. Using EAC it should have some significance whether I use a CD/DVD-drive that has an internal cache, this should be a mayor drawback because, as I have understood it, it makes error-correction harder. There's also the issue of error-correction on the CD. These issues are probably not all that significant though, as I don't have a lot of old scratched CDs and therefore get rips with nearly no errors. Until now I've ripped my CDs to WAW and after that encoded them using LAME, I might start using LAME and EAC at the same time.

How do I resample? I would like to try it to check whether there's any hearable difference. Is it something I make my soundcard do during playback? is it something I have to make my ripper do? or is it something I should do with the mp3/waws?

I will make another enquiry elsewhere on the forum concerning compression-formats as I have about 25 CDs in mail that I would like to rip.

 

[umheint0]

Not really. I use EAC on all my CDs, even some badly scratched ones that don't work in a regular CD player. With all of my computer's drives (one CD, one DVD, to cd r/w's, and one dvd r/w), there is no problem, EAC is able to recover nearly the whole signal (only in some severe cases it can't). Besides any new CD drive has an internal cache. Ones without are so antiquated that they aren't worth mentioning.

Resampling involves taking the source track, using some audio program, and filling in the gaps between each sample. I don't know of any program offhand that does true upsampling to actually increase the quality of the source track, but you can upsample normally using any old program like Sound Forge.

I use LAME @ 320kbps, straight stereo, highest quality (slowest) encoding. They're larger files, but they sound the best, aside from raw WAV files. Lots of people like OGG, and I must say I'm impressed at the quality-size ratio, which is higher than MP3. However, this is at low bitrates, like sub-128kbps. There is also WMA, which isn't bad either. The problems with both are that they aren't widely supported, so you'd have to take a given MP3 file, decode it, then encode it into whatever format you want. Just get MP3s with at least 192kbps encoding and they sound great, far better than 128.

I stand by LAME @ 320. 10 megs for a 4-minute long track isn't bad, especially since I have 480gb of space to fill.

umheint0's phat setup --> <A HREF="http://home.cc.umanitoba.ca/~umheint0/system.html" target="_new">http://home.cc.umanitoba.ca/~umheint0/system.html</A><--

 

[TTZX]

Use the Creative MediaSource Organizer that came with the Audigy card. I used that program to rip my 170 CDs at 320Kbps and they sound really good. It also get all the CD information off the web for you and names all the songs and albums and puts them in folders according to the artist. BTW I'm using the spdif out from my CD drive to the Audigy.

 

[umheint0]

Well, considering it's your processor that does the encoding and not the sound card, that last part is irrelevant, unless you're encoding the CDs off the line in on our sound card, which would make the recordings far lower quality than if you just used the plain old data stream of the IDE channel.

Do you by any chance know what encoder Mediasource uses? There's nothing on the Creative site, and nothing anywhere else actually.

umheint0's phat setup --> <A HREF="http://home.cc.umanitoba.ca/~umheint0/system.html" target="_new">http://home.cc.umanitoba.ca/~umheint0/system.html</A><--