struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm
Bit perfect. Does it matter?
bertdw
bertdw's picture
Offline
Last seen: 2 months 2 weeks ago
Joined: Sep 18 2007 - 5:41am

Very interesting experiment, Struts. Nice job.

Whatever happened to Reed-Solomon error correction? This was encoded into the CD or DVD itself, is that correct? Some bright engineer somewhere should be able to encode this into our wave and flac files so we don't have to worry about this anymore. I suppose it would make the files bigger, but with terrabyte hard drives getting cheaper every day, who cares?

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

No, I replayed the files in the PC; I didn't burn them to CD. Note that since I was editing the audio waveform itself, my changed sample would not have been recognized as an error if I had burned the file onto a CD so Reed Solomon would not have helped.

You are right of course, in theory all data errors should be caught by the CD drive's error correction code, however many CD-ROM drives are known to have firmware bugs (particularly in their handling of C2 error pointers) which means that errors can and do slip through. Here is a useful list of different types of CD-ROM drives showing among other things whether they correctly support C2 error pointers.

Using a ripping program that references AccurateRip will alert you to a potential problem, what I wanted to investigate here was the actual audible effects of such a problem. As such I wanted to simulate an uncorrected error, and the simplest way of doing that was to actually 'bake' the error into the file at the audio level (as opposed to the bit level).

If I haven't screwed up somewhere along the way my methodology simulates an uncaught data error as small as one single bit of audio.

bertdw
bertdw's picture
Offline
Last seen: 2 months 2 weeks ago
Joined: Sep 18 2007 - 5:41am


Quote:
No, I replayed the files in the PC; I didn't burn them to CD.

I knew that; sorry I didn't make myself clear. I meant to say that Reed-Solomon error correction was part of the data written on commercially produced CD's, and not merely something the player did after the fact. Sorry if I took your thread somewhat off-course.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

A fun experiment.

I don't think this sample error is equivalent to a reading bit error however. I think a sample error as you describe is a lot of wrong bits.

A single sample that is uncorrelated with the surrounding data is easily heard as a click.

Those of us that have tried to make straight edits in a recording without a cross fade know how easy these are to produce.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
I don't think this sample error is equivalent to a reading bit error however. I think a sample error as you describe is a lot of wrong bits.


Sorry, I was not trying to say they were equivalent I was trying to illustrate an extreme (albeit unlikely) boundary case. My point was that the very low level at which I found errors to still be audible means that an error as small as one bit could potentially be audible - if it happened to correspond to the MSB in that sample.

In practice I agree that errors affecting a single bit are probably extremely rare, the nature of the Compact Disc medium is such that errors are likely to affect areas of the disc spanning multiple frames (and therefore lots of bits), and this is exactly what CIRC is designed to cope with.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Got it. Sorry.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

It's been fascinating to see how my average rip times have gone up since setting dBpoweramp to use Ultra Secure mode. Unfortunately I haven't been noting down the results (maybe I'll start) but here are some rough-and-ready observations:

  • 'Clean' rips where the disc just rips first time with no rereads etc. are in the distinct minority, <25%?
  • Many discs I rip seem to be different pressings to those in AccurateRip. The CRCs don't match those in AccurateRip for any track on the disc (necessitating multiple rereads) but no errors are detected.
  • dBp seems to require 'Ultra-secure' rerips most of the time (>75%?) although in the vast majority of those cases the result subsequently agrees with AccurateRip.
  • Only in a very small number of cases (max 2-3%) am I completely unable to get a secure rip.

This would seem to indicate that read errors are pretty common and that error correction or error concealment are working hard to cover them up. Error correction should of course in theory be inaudible (and will be as long as the algorithm is bug-free and the system is working fast enough), however error concealment (interpolation) will tend to be less accurate the higher the frequency of the signal, to the point where at 22.05kHz it becomes completely useless. One possible effect could therefore be that poorer CD transports producing a higher incidence of read errors and relying more heavily on error concealment might exhibit poorer harmonic texture and imaging, both of which are reliant on high frequency information.

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

Thanks for reporting your experimental results and on your experience with ripping.

I kept track of my results as I ripped about 2000 CDs. I experimented with EAC, dBpoweramp and J. River media center for secure ripping. I didn't find any differences in results between these programs. I have used J. river media Center for ripping new CDs.

I had treated those CDs with care - no big scratches or other defects. About 95% of my CDs required no extra re-reads. (JRMC doesn't use the AccurateRip database or C2 pointers info.)

Another 3-4% required extra reads on one or more tracks but the tracks could be ripped with confidence. About 1% of my CDs had one or more tracks that could not be ripped with confidence by JRMC or EAC or dBpoweramp. In about half of those tracks, a burst mode rip produced a file with no obvious audible glitches.

I've been buying used CDs from Amazon marketplace for a couple of years. I am finding a slightly higher rate of problem tracks (perhaps 2-3%.) Recently, I had a problem track with 2 very audible glitches. Re-ripping with different drives, different speeds or different ripping software didn't help. I used Audacity to remove single sample spikes of the sort that you created. This resulting file was quite listenable with no obvious glitches.

Thanks again for your very useful posts.

Bill

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Thanks Bill. Couple of questions:

  • Which version(s) of dBp did you compare? Free, PowerPack or Reference? Note that the free version doesn't even support secure ripping, let alone 'ultra secure'.
  • When you say you didn't find any differences in results between JRMC, EAC and dBp, what do you mean? Subjectively (i.e. by listening alone) or by doing file comparisons of the resulting audio?
  • Do you know what method JRMC uses to produce secure rips? Is it purely statistical (i.e. rerips until some number of rips agree with each other)? Couldn't find this in a quick poke around on their website
  • Out of interest what was the basis of your choice of JRMC?
Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

All good questions.

> Which version(s) of dBp did you compare? Free, PowerPack
> or Reference?

I used the reference version of dBpoweramp.

> When you say you didn't find any differences in results
> between JRMC, EAC and dBp, what do you mean? Subjectively
> (i.e. by listening alone) or by doing file comparisons of
> the resulting audio?

All my conclusions are based on comparing audio streams in the music files using file compare techniques. I compared results from enough CDs in each category (1 - no extra re-reads, 2 - extra re-reads with success and 3 - failures to get a successful rip) to be confident of my understanding.

> Do you know what method JRMC uses to produce secure rips?
> Is it purely statistical (i.e. rerips until some number of
> rips agree with each other)?

JRMC does multiple reads and cache flushing looking for consensus. It starts ripping a CD at a moderate speed and speeds up if it gets good results. This effect is more pronounced on some drives than on others. It does not use C2 pointer info or the AccurateRip database. There are some other techniques for improving ripping accuracy; I don't know whether JRMC uses any other techniques. Matt (a JRMC developer) commented on their techniques 2-3 years ago on their forum.

> Out of interest what was the basis of your choice of JRMC?

About 2/3 of my collection is classical music where albums are not important and the Composer is. Another big chunk of my collection is Broadway show CDs and CD with Broadway standards songs where Composer also matters. I wanted to use both Composer, work name and performer information in browsing and selecting music. I wanted to have all those tags in iTunes like browser panes.

JRMC lets me use whatever tags I want. I have some galleries of screenshots of the views I use in JRMC (look in the Misc and MC screenshots galleries.)

screenshots

So JRMC seemed the best choice for browsing and selecting music from my Composer based collection. It also had built-in ASIO support which I wanted. The tag editing and utility features are a big plus and the browsing, sorting and searching features made the tag editing features much more efficient.

Why do I use JRMC to rip CDs?

Nobody has the right info in online tag databases for music where Composer is important. Even if both Composer and performer information is present, it is in the wrong place, is formatted in a useless way or names are spelled or placed inconsistently for different CDs. I entered most of the tag information by hand for genres where composer was important. It was just less trouble than re-arranging the tag data returned from an online database. The tag editing features in JRMC made that job much easier than it was when I used dBpoweramp. EAC was even more painful to use.

JRMC is by far the best tool for me. I looked for some months before finding it. The right tool for someone else might be a different set of programs.

Bill

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Thanks, Bill. Great info.

I understand your frustration with tagging as it relates to composers. It is an ongoing issue for many of us.

I gave up long ago and use the "artist" filed for composer and then note the performer, orchestra, etc. somewhere else - sometimes in the album name so I know what performance of a given requiem I am looking at, etc.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Thanks for the comprehensive answers Bill. You've certainly piqued my interest, might just have to test drive JRMC myself.

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm


Quote:

I understand your frustration with tagging as it relates to composers. It is an ongoing issue for many of us.

I gave up long ago and use the "artist" filed for composer and then note the performer, orchestra, etc. somewhere else - sometimes in the album name so I know what performance of a given requiem I am looking at, etc.

I'm pretty stubborn about doing things my way. When I started, I knew how I wanted to browse my collection. I looked for tools that would let me do things my way. Fortunately, I came upon J River MC. Now my music library is tagged the way I want it and I can browse the way I want to. Manual tagging for new CD purchases is quite manageable. I'm not frustrated now.

Here is my scheme for using tags:

Composer => Composer tag (Mozart)
Work Name => Album tag (Symphony No. 41)
(also in Work Name tag)
Track name => movement (1 - Allegro)
Performers => Artist tag (Fleisher_Szell_Cleveland Orchestra)
Genre => Genre Tag (Classical, jazz, Broadway Shows)
Sub genre => Sub_genre Tag
CD Version => Version Tag (Original Jacket set)
vocalists => Singers tag (Kelly, Gene;Reynolds, Debbie)

This scheme keeps each list I browse to a manageable length. Once I select a Composer, the other browser panes show only works and artists for that Composer. If I then select a work, only those artists performing that work are present in that browser pane. For a very prolific Compose like Haydn or Mozart, I can select a Sub_genre to shorten the number of works and artists.

I had to make some adjustments to my tagging scheme in the subset library I put on an iPod. I worked out the compromises before I purchased the iPod.

Bill

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Oh boy. Don't get me started on tagging!

Very interested in your approach Bill, I guess mine started out as almost the exact opposite, but after a painful journey I have ended up with something along the same lines as your scheme.

Initially I wasted a lot of time and suffered enormous frustration trying to use the tags in the way they were intended (in the largely misguided belief that this is how they would work best). However as the scars on my mighty audiophile derri

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Struts,

We have made the same painful journey.

My ALBUM tags are identical to yours: Bach: Toccata in D, Partita No. 2, English Suite No. 2 (Argerich)

I know composer, piece and performer at once.

However I elected to use ARTIST as defined (in popular music, the band or singer; for classical, the composer).

A personal oddity (one of many), I list classical composers by last name, first name (Bach, C.P.E.). I list popular music first name, last name (Miles Davis). I don't think of him as Davis, M. Listing popular performers by last name makes them harder to find.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
A personal oddity (one of many), I list classical composers by last name, first name (Bach, C.P.E.). I list popular music first name, last name (Miles Davis). I don't think of him as Davis, M. Listing popular performers by last name makes them harder to find.


Ha ha! Call it an impersonal oddity; I do the same. I don't really think of it as an oddity actually (or, rather it's an oddity I have been able to post-rationalize) as classical composers are generally known my their last names, and some of their given names are rarely mentioned at all (great Trivial Pursuit question, classical composers' given names). It helps that most of them are unambiguous, Bach and Strauss are the exceptions that spring to mind.

In my case this means it's the COMPOSER tags that are inconsistent between Classical and other genres, rather than ARTIST. It also means of course that in some rare cases where the composer is playing or conducting the same name occurs with two different representations in two tags. This disturbs one part of my brain greatly, however the other part wins most of the time

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

I'm headed out the door for an 11 day trip - mostly wildflower viewing.

xiph.org - Vorbis comments are quite flexible. You can use whatever tag name you want. I use the Flac format and put the tags I want in the files. J. River can use whatever tags I put in the files.

I looked at the Sonos system. I wasn't willing to live with the tag use limitations and the limit on number of tracks. You made a different choice.

I included links to screenshots of my J. River Media Center setup. In the classical music views, I have browser panes for Sub_genre, Composer, Work name, Artist (Performer) and Version. If I select "Sonata, Piano" from the sub genre column, I see what Composers and works I have for that sub genre. If I reset the sub_genre pane to "all" and select "Faure", I see all the works and artists I have for that composer. I can explore my library very flexibly. I don't have to remember what I own.

A small screen interface like the Sonos or an iPod usually displays a list of values for one tag at a time. You browse through a series of screens. You want the UI to support the tags you want to use AND you want the right order of browse screens. The iPod I use allows you to start browsing by Composer but it doesn't have steps for both Album and Artist.

Names: I too use last names for classical music. I remember composers by last names. Screen real estate is valuable so I use only the last name for composers like Beethoven, Mozart (W.A.), Haydn, Schubert and Dvorak.

All this comes down to a question of personal stubbornness. I just didn't want to settle for tools that required unsatisfactory compromises of me.

Bill

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Enjoy your trip Bill. This has been one of the more thought-provoking threads here in a good while. We can pick up when you're back.

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

I'm back from the second wildflower trip (with thousands of photos to go through). I'll restart the conversation about problems in using classical music with PC audio software.

We talked about fitting the Composer, Work Name and Performer information into the Album and Artist tags that most player software use. For portable players, small screen remote devices and 10 foot interfaces for TV use, there is another problem. The browsing interface on such devices usually provides a sequence of browsing steps. At each step, the player displays a list of values for one tag. The user selects a single value and then the player displays values for the next tag in the sequence. A player may provide several different sequences of browsing steps. My iPod provides these alternative browsing sequences:

Artist -> Album->Songs
Album->Songs
Songs
Genre->Artist->Album->Song
Composer->Album->Song

None of these sequences fits my need for a

Genre->Composer->Work Name->Performer->Version sequence. It's nice that Apple provides for browsing by Composer but they just didn't get it right.

Adding a mechanism to allow me to create my own browsing sequences would be fairly simple. I'm not holding my breath for Apple to implement such a feature. Does Sonos have such a feature now?

The Custom Scan and Custom Browse plugins for the SqueezeBox server populate the tag database the server uses and lets the user build the browse sequences he wants to use. However, Logitech keeps breaking the plugin interface and the plugin developer has discontinued development.

Bill

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Welcome back Bill.


Quote:
My iPod provides these alternative browsing sequences:

Artist -> Album->Songs
Album->Songs
Songs
Genre->Artist->Album->Song
Composer->Album->Song


Yes, the Sonos provides these exact same sequences which are clearly more suitable for performer-centric genres than composer-centric.


Quote:
None of these sequences fits my need for a

Genre->Composer->Work Name->Performer->Version sequence. It's nice that Apple provides for browsing by Composer but they just didn't get it right.


It's a shame because it would be really easy to implement this by simply building in a conditional step. The step following Genre would depend on the value of Genre. For, Classical, Choral, Chamber, Broadway etc. it would be 'Composer', whereas for Rock etc. it would be 'Album'.


Quote:
Adding a mechanism to allow me to create my own browsing sequences would be fairly simple. I'm not holding my breath for Apple to implement such a feature. Does Sonos have such a feature now?


No, Sonos' sequences are pre-defined and identical to the iPod's. I don't necessarily think one needs to go as far as user-programmable browsing sequences which while satifying the needs of a probably small proportion of power-users would risk adding complexity to the UI and cause customer-care issues. As far as I am concerned a simple context-sensitive browsing tree as described above would suffice.


Quote:
The Custom Scan and Custom Browse plugins for the SqueezeBox server populate the tag database the server uses and lets the user build the browse sequences he wants to use. However, Logitech keeps breaking the plugin interface and the plugin developer has discontinued development.


I am dismayed to hear this and grateful you raised it. I sincerely hope that Squeeze is not losing its mojo now after Sean's departure, but your observations are worrying. I will probably still try a Squeezebox Touch when it comes out (unfortunately the latest on that is that several big chains here have revised their projected availability dates to 8/1/10, let's hope it comes forward again), although I will consider it more of a speculative move now and 'proceed at caution'.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Duplicate post removed.

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm


Quote:


Quote:
None of these sequences fits my need for a

Genre->Composer->Work Name->Performer->Version sequence.


It's a shame because it would be really easy to implement this by simply building in a conditional step. The step following Genre would depend on the value of Genre. For, Classical, Choral, Chamber, Broadway etc. it would be 'Composer', whereas for Rock etc. it would be 'Album'.

I do something like that with the J. River Media Center player program. I have separate views for each of the Genres I specified. Each of those views has a set of browser panes for selecting music files and a set of tags to be shown for each music file shown in the list.

Once I've selected the Classical music view in JRMC, all the browser panes are visible on the screen. I can select tag values in whatever order suits me. On a small screen like a Sonos controller, you see one list of tag values at a time (like one browser pane). So a recipe for browsing has to specify the the tags that will be used for selection and the order in which they are used.


Quote:


Quote:
Adding a mechanism to allow me to create my own browsing sequences would be fairly simple.

I don't necessarily think one needs to go as far as user-programmable browsing sequences which while satifying the needs of a probably small proportion of power-users would risk adding complexity to the UI and cause customer-care issues. As far as I am concerned a simple context-sensitive browsing tree as described above would suffice.

I think that allowing users to design their own menus is more likely to produce a useful solution. If you leave the design of those context sensitive menus to the brain dead folks who designed most of the current set of players, they are going to get it wrong again.

The Custom Browse and Custom Scan Squeezebox server plugins came indirectly from discussions I had on the SlimDevices forum with another person. We proposed ideas that were written up in a set of bug reports by the other guy. Those ideas were never implemented in the server software itself. However, a user with the screen name Erland implemented them in plugins.

Erland was not a classical music listener himself. It took a lot of work to get him to understand the need for the full browse sequence. Allowing the user to build his own menus with the browsing sequences he wanted free the plugins from the limits of his understanding of the needs of a classical music listener.


Quote:
(me) However, Logitech keeps breaking the plugin interface and the plugin developer has discontinued development.

Logitech keeps shooting themselves in the foot by breaking useful plugins. They shot themselves in both feet with a large caliber weapon the long delay in releasing the Squeezebox Touch. The Touch has been praised for its sound quality and the tiny server running in the Touch is fresh functionality. I'll be interested to see whether the Touch takes off and interest in the Squeezebox architecture revives. That might encourage Erland to do some more coding. It is also possible that the server software developers (now mostly inside Logitech) will implement the equivalent of the Custom Browse and Custom Scan functionality in the server software itself.


Quote:

I sincerely hope that Squeeze is not losing its mojo now after Sean's departure, but your observations are worrying. I will probably still try a Squeezebox Touch when it comes out ...
although I will consider it more of a speculative move now and 'proceed at caution'.

The Touch seems to be out now.

I've followed the Squeezebox product for over 4 years now. It never quite gets there for me but maybe it will in the future. Smart hardware design that keeps getting better but I don't like the server software.

Bill

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am


Quote:
The Touch has been praised for its sound quality and the tiny server running in the Touch is fresh functionality.

Moving the server software to a stand alone device makes excellent sense. Now one can put together a server with just a Touch and a NAS. This should appeal to many.

BTW, the Touch is now shipping: Now Shipping

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
The Touch has been praised for its sound quality...


Where? I was looking for a review of an actual production unit in the audio press and couldn't find one. No customer reviews either, which is strange if it really is shipping. However...


Quote:
BTW, the Touch is now shipping: Now Shipping


...neither have I been able to find any evidence that the Touch is in fact now shipping. The top 5-or-so hits on Google Shopping (Amazon, Crutchfield, J&R, DigitalStar and WiredZone) are all showing no stock. To quote Larry Ellison, "The product isn't late, the annoucement was just early".

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm


Quote:

Quote:
The Touch has been praised for its sound quality...


Where? I was looking for a review of an actual production unit in the audio press and couldn't find one. No customer reviews either, which is strange if it really is shipping. However...

I was mostly referring to remarks made by John Swenson on the Audio Asylum PC Audio forum and The Logitech Squeezebox forums.

Here's one link to some comments by John. You can search for others:

http://db.audioasylum.com/cgi/m.mpl?foru...+swenson&r=

The "Buy Now" links on the Logitech site were not working just now either. Maybe glitches getting things set up. Maybe they can't handle a high volume of orders or ship a lot of product. Give them a week and we'll see. I'm a disinterested observer so I have lots of patience.

Bill

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
...neither have I been able to find any evidence that the Touch is in fact now shipping.

So, the first reseller in Sweden listed the Touch as in stock today and I placed an order. We'll see.

According to this post by John Swenson over at the Asylum the reason for the delay was Logitech used initial stocks to clear a pile of backorders. Not entirely unlikely given the slippage versus the original anticipated shipping date which was last December iirc.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Well, despite my earlier skepticism I can confirm that the Touch really is shipping. Mine arrived today and is feeding the bigrig with 24/88.2 'Studio Master' FLAC files of Linn Records' Messiah as we speak. Listening impressions to follow.

jazzfan
jazzfan's picture
Offline
Last seen: 5 years 3 weeks ago
Joined: Sep 8 2005 - 8:55am


Quote:
Well, despite my earlier skepticism I can confirm that the Touch really is shipping. Mine arrived today and is feeding the bigrig with 24/88.2 'Studio Master' FLAC files of Linn Records' Messiah as we speak. Listening impressions to follow.

Struts,

I also just got a Touch and so far i really like the sound. I have it connected to Squeezebox server via wired Ethernet since it's located very close to my router. I also have it connected to my computer room stereo via the digital optical to my Headroom Desktop Amp which has a Headroom Ultra DAC built in.

What DAC, if any, are you using with the Touch?

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Hi jf,

My DAC is a bit of a home brew based on Anagram Technologies Sonic2 and Timelock evaluation boards fed with a milspec Acopian power supply. It is the same chipset used in the Soulution 740 so it has some pedigree.

Sounds pretty good, although as with all such projects I know just what I would change if I were to do it again! We ran a SBT against the Weiss Minerva down at Audio Concept and the whole panel preferred my DAC which was extremely gratifying. Embarrassingly, even though I preferred it I still identified it incorrectly, thinking it was the Weiss I was preferring

Anyway, it is a useful improvement on the Boulder which makes it well worth the $3000-or-so it cost to make (most of which was the boards). Was a fun project too!

hurt1097
hurt1097's picture
Offline
Last seen: 3 years 11 months ago
Joined: Mar 17 2010 - 7:14am

I am using iTunes to build a computer based audio system. I am running a USB cable to a Music streamer and thence to a Yamaha RX 397 Stereo receiver. My KEF iQ1 speakers are bi-wired. Eventually I plan to upgrade this system. The computer is a newer Dell with the CD ROM built in to the flat screen panel. I already have a high end system (VPI, Audio Aero, Audio Research, Rogue, Vandersteen), so I am familiar with high end sound. However, I am new to the computer audio game so I would appreciate some advice.

The sound quality of my computer system is acceptable, but when I replay my music, which I recorded from CDs as WAV files I get occasional drop outs. These events are non repeatable. In other words when I immediately go back to the skip the music plays fine (until the next drop out somewhere else). This does not happen with iTunes radio. Should I be recording in Apple lossless? Should I replace iTunes with a better program?

Thanks,

Analog Al

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

> The sound quality of my computer system is acceptable, but
> when I replay my music, which I recorded from CDs as WAV
> files I get occasional drop outs. These events are non
> repeatable. In other words when I immediately go back to
> the skip the music plays fine (until the next drop out
> somewhere else). This does not happen with iTunes radio.

Your dropout problem is probably caused by the buffer at the DAC chip running dry. A bottleneck anywhere along the path from hard drive through iTunes to the Operating system audio stack to the USB hardware support in the Dell, across the USB cable to the DAC you are using can produce the symptoms you describe.

Solving your problem will probably not cost you anything. It may require some patience on your part and some willingness to do a few experiments to pinpoint the problem.

> I am using iTunes to build a computer based audio system.
> The computer is a newer Dell with the CD ROM built in to
> the flat screen panel.

Tell us more about the computer and the software on it.

What CPU, how much memory? Are the music files on a hard drive directly connected to the Dell? If so, is the hard drive internal to the Dell or is it connected via USB or Firewire? If the Dell is reading files from the hard drive via a network connection, is it a wired LAN or a Wifi connection?

What else is connected to the Dell via USB?

What operating system are you running? What other software is also running when you play music with iTunes and get dropouts? Do you have an anti-virus program running? Which one?

Is Windows performing auto-updates without your explicitly telling it to look or upgrades and/or install them? is Windows indexing the contents of the drive where you store your music files?

If you give me answers to these questions, I'll suggest ways to pinpoint the cause of your problem.

> Should I be recording in Apple lossless?

No. It probably won't help and changing stuff at random makes it harder to identify the actual cause of your problems.

> Should I replace iTunes with a better program?

No, not at this point. You will probably have the same problem with the next program you try.

Don't start flailing around guessing what the problem might be and changing your system.

Bill

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Hi Al,

Welcome to the forum. Listener's troubleshooting approach is spot on IMO, however there is one quick thing you can try first. The problem you describe sounds a lot like a DPC latency issue to me. Clicky here, have a good read then download and run the latency checker program. If my hunch is correct it should help you put your finger on the problem.

Good luck!

PS Your question would probably have sat better in a new thread, although now I look back I see we had successfully drifted quite a ways off topic already!

jazzfan
jazzfan's picture
Offline
Last seen: 5 years 3 weeks ago
Joined: Sep 8 2005 - 8:55am


Quote:
Hi jf,

My DAC is a bit of a home brew based on Anagram Technologies Sonic2 and Timelock evaluation boards fed with a milspec Acopian power supply. It is the same chipset used in the Soulution 740 so it has some pedigree.

Sounds pretty good, although as with all such projects I know just what I would change if I were to do it again! We ran a SBT against the Weiss Minerva down at Audio Concept and the whole panel preferred my DAC which was extremely gratifying. Embarrassingly, even though I preferred it I still identified it incorrectly, thinking it was the Weiss I was preferring

Anyway, it is a useful improvement on the Boulder which makes it well worth the $3000-or-so it cost to make (most of which was the boards). Was a fun project too!

Struts,

Doing a little math using the Lexicon mark-up formula ($600 -> $3500 then $3000 -> $17,500), where the $600 is a little high since Lexicon must likely pays somewhat less than $600 for the Oppo players, I would say that you have built the most expensive ($20,000 minimum list price) DAC currently available. Needless to say, you're sitting on a gold mine. Please let me know if you're looking for a US distributor.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Lol! You're not far wrong! The only two commercial products I am aware of based on these Anagram chipsets are the Soulution 740 (

jazzfan
jazzfan's picture
Offline
Last seen: 5 years 3 weeks ago
Joined: Sep 8 2005 - 8:55am


Quote:
Lol! You're not far wrong! The only two commercial products I am aware of based on these Anagram chipsets are the Soulution 740 (
struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
You're right Struts words do not suffice for these marvelous pieces of kit, as the Brits would call them. But I really think that your missing a big money making opportunity, heck you've even given them an ultra expensive sounding name, "Silvercore". I just love it!


It's Christof's company (the guy who made the cases). He is a certifiable genius and makes some drop dead gorgeous tube equipment, for more info see his website here. WARNING! There is some hard core tube porn on display - just check out his 1 meter long 'custom' preamp!

Who says size doesn't matter?

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am

Struts, both your DAC and Christof's other work are gorgeous!

hurt1097
hurt1097's picture
Offline
Last seen: 3 years 11 months ago
Joined: Mar 17 2010 - 7:14am

Thanks for the tip on the latency checker program. It was very helpful. I have reduced but have not eliminated the problem by downloading new device drivers. Some device drivers would not install because there were already newer ones on the computer. Another (Broadcom network adapterone said that there was no hardware associated with the program. I plan to uninstall that driver as soon as I can figure out if it's safe. I'm making progrss. Thanks for the help.

hurt1097
hurt1097's picture
Offline
Last seen: 3 years 11 months ago
Joined: Mar 17 2010 - 7:14am

As you probably noticed I tried a latency checker program suggested by one of the other readers. It helped but did not eliminate the issue.

My computer is a Dell XPS1 with an Intel Core Quad CPU @ 2.33 Ghz. It has 4GB RAM. Music files are on the internal hard drive. I have a Wifi connection (via a Linksys WCG200 installed by my cable company) but no LAN. The keyboard and mouse are wireless, which is the standard factory setup on this Dell model. (I have also had some

hurt1097
hurt1097's picture
Offline
Last seen: 3 years 11 months ago
Joined: Mar 17 2010 - 7:14am

PS. Thanks for the heads up on starting a new thread rather than breaking in on another conversation. I'll keep it in mind in future.

Al

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

> As you probably noticed I tried a latency checker program suggested by one of the other readers.

DPC latency checker is a good tool for seeing what's going on at one point in the audio path.

You appeared to have looked at the output and jumped to a conclusion that you needed to update your drivers.

How long did you let DPClat.exe run? Were you watching the DPClat display as you played music through iTunes? What did you see when you ran DPClat.exe?

> It helped but did not eliminate the issue.

So now you have changed the configuration of your system and you understand the behavior of your problem less well.

> I have a Wifi connection (via a Linksys WCG200 installed
> by my cable company) but no LAN.

Does your Dell communicate with the Linksys router via an ethernet cable or via WiFi?

> The keyboard and mouse are wireless, which is the standard
> factory setup on this Dell model. (I have also had some
>

RGibran
RGibran's picture
Offline
Last seen: 2 years 4 months ago
Joined: Oct 11 2005 - 5:50pm


Quote:
What DAC, if any, are you using with the Touch?

Guys, enjoy your new toy but at some point I hope you will connect the analog outs and give us your impressions of the Touch's onboard D/A conversion.

After all, we have Logitech forum users selling off their Lavrys', Benchmarks' and DacMagics after carefull comparisons. What's next...Transporters!

Your impressions greatly appreciated.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
You seem to lack the patience to follow a disciplined troubleshooting process. You also seem to view taking my advice as a last resort. I'll see how to react to this post and decide whether to spend any more time on this interaction.

Bill


Hey Bill,

Give Al a break, he did say he was a computer audio beginner. It's my fault for confusing the issue, at first glance it probably looked like you were proposing a difficult solution and I was proposing an easy one. In actual fact you were proposing a comprehensive and thorough broad-spectrum troubleshooting approach and I was taking a potshot based on a hunch. My advice was intended as a complement to yours, not an alternative. I realize I should have made that clearer now.

Anyway, let's see what Al comes back with and see if we can help him further.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm


Quote:
Guys, enjoy your new toy but at some point I hope you will connect the analog outs and give us your impressions of the Touch's onboard D/A conversion.


It would genuinely never have occurred to me but I'd be more than happy to. Just need to check if I have a pair of SE->Balanced interconnects. I'm sure I do somewhere...


Quote:
After all, we have Logitech forum users selling off their Lavrys', Benchmarks' and DacMagics after carefull comparisons. What's next...Transporters!


Transporters? Surely not! I thought they used the miracle DAC...


Quote:
Your impressions greatly appreciated.


Stay tuned.

jazzfan
jazzfan's picture
Offline
Last seen: 5 years 3 weeks ago
Joined: Sep 8 2005 - 8:55am


Quote:

Quote:
What DAC, if any, are you using with the Touch?

Guys, enjoy your new toy but at some point I hope you will connect the analog outs and give us your impressions of the Touch's onboard D/A conversion.

After all, we have Logitech forum users selling off their Lavrys', Benchmarks' and DacMagics after carefull comparisons. What's next...Transporters!

Your impressions greatly appreciated.

I'll try to hook up the Touch via the analog outputs and post some impressions.

Elk
Elk's picture
Offline
Last seen: 3 years 7 months ago
Joined: Dec 26 2006 - 6:32am


Quote:
After all, we have Logitech forum users selling off their Lavrys', Benchmarks' and DacMagics after carefull comparisons.

Seriously? For the $300.00 Touch?

Too bad one can only use it with Slim Server software.

struts
struts's picture
Offline
Last seen: 2 years 10 months ago
Joined: Feb 1 2007 - 12:02pm

Turns out I have two Balanced->SE cables but no SE->Balanced so testing of the Touch's analog outs has been confined to the headphone rig.

With a slight caveat for the fact that this is an initial impression based on one evening's listening I have to admit that I have as yet been unable to discern any differences whatsoever between the Touch's digital and analog outputs when feeding the Grace m902. Switching back-and-forth the levels, as luck would have it, are perfectly matched and as hard as I strain I find them completely indistinguishable In fact at one point (starting to doubt my own sanity slightly) I disconnected the digital and then the analog cables just to double-check that the Grace was actually switching between them!

This thing is good, for $300 it's very good.

hurt1097
hurt1097's picture
Offline
Last seen: 3 years 11 months ago
Joined: Mar 17 2010 - 7:14am

I appreciate the detailed advice you have provided. It is not patience that I lack, but time. I am a history professor dealing with various end-of-semester deadlines & commitments, including a stack of doctoral dissertations that must be read and commented on in the next couple of weeks. Consequently I was hoping that the "quick fix" would help me to reach a happy conclusion. In the process I learned (or at least observed) a few things that (along with your helpful advice) have helped me to understand the problem better.

Because I am a teacher I am also sympathetic to the frustration of teaching someone who won't be taught, or as you put it, who does't have the patience. I often get blind email inquiries from strangers wanting help. When I give it they seldom thank me. So, Listener, thank you, even if I do not immediately have the time to follow your good advice. When I do have the opportunity to follow your directions, perhaps in a couple of weeks or a month I may put up another post or ask another question. I have an additional reason for waiting that long. By then I should have an iMac so I am unwilling to invest time in the Dell if the new computer will solve the problem. (My acquisition of the iMac has nothing to do with the iTunes issue). My hope may be in vain, but at least by waiting I won't be wasting anyone else's time.

In the meantime, what I would like to know is simply this: is this a common problem? In the few articles that I have read in Stereophile I don't recall anything about dropouts, but I think most of the Stereophile writers use Macs.

Cheers to all,

RGibran
RGibran's picture
Offline
Last seen: 2 years 4 months ago
Joined: Oct 11 2005 - 5:50pm


Quote:
With a slight caveat for the fact that this is an initial impression based on one evening's listening I have to admit that I have as yet been unable to discern any differences whatsoever between the Touch's digital and analog outputs when feeding the Grace m902.

This thing is good, for $300 it's very good.

Wow! The Touch seems to be comfortable in the company of some very fine stand alone dacs. And a touch screen to boot!

Struts thanks for taking the time to give your much valued impressions. I'll order up!

Listener
Listener's picture
Offline
Last seen: 12 years 1 week ago
Joined: Jul 28 2006 - 12:45pm

> It is not patience that I lack, but time.

Thanks for taking the time to reply.

Bill

Pages

Log in or register to post comments
-->
  • X