Mac Mini to original Benchmark Dac1 and signal "conversion"?

[jay]

Chris, at the risk of writing the longest first post in your site's history, let me first say congrats on such a fantastic site. A reviewer buddy of mine linked me to you a month ago after he visited me on a business trip and saw my attempts to do a great-sounding computer-based front-end. I've been soaking up the info from everyone in these forums as well as your articles and reviews, and it's been absolutely fantastic, so thanks for your enthusiastic pursuit.

 

I'd been "very comfortable" with my set-up for over a year and a half now, having settled on a mac mini for a number of reasons, and having not had good reason (yet!) to stray from my original Benchmark Dac1 (non-USB version). My original goal was to make Pandora sound "great" and I definitely accomplished that and more with my current set-up (that's how I ended up with the mac mini rather than apple tv which was also a great front-end, but now feeding my HD tv in the basement). I've ripped most of my CDs to lossless at this point (any that I listen to) and have settled on iTunes as my software of choice to run the library (just to further the background leading up to my questions below). I had curbed my jitter concerns with what I had read about (and was hearing from) the Benchmark and was ok feeding the Benchmark from a custom-made (read Chinese-made, I'm sure) Ebay-bought glass mini toslink. I ran with this set-up for the last year plus with great results, feeling I had maxed it out well enough to just thoroughly enjoy it for awhile, which I did...until...

 

A recent speaker upgrade and the discovery of this site has caused me to revisit my set-up and start to experiment again, not because I wasn't happy or it didn't sound really good, but because I could hear new and better things again in my music. I started by tweaking midi-settings at the urging of a couple posts (one from you, I believe), and that was a great find as I believe the 96hz and 24-bit up/over sampling does give me a little noticeable improvement with 44.1hz stuff. Great, what else was I missing?

 

Then the questions started, with further experimenting with stuff I already owned but had long-ago pulled from the system yet never sold. Jitter reducers from Theta and Sonic Frontiers didn't improve anything (and in fact, actually made the sound worse in the case of the Sonic Frontiers unit). To test the 96hz signal I pulled out the trusty and often controversial(?) Behringer DEQ2496 and threw it in between the toslink mac mini signal, and because I had the cable sitting around, instead fed AES/EBU to the Benchmark.

 

To my surprise, I noticed an improvement I had not heard before, and almost by accident. The opening of Jeff Buckley's Hallelujah has a little guitar opening that had a much more noticeable attack and decay through the AES/EBU connection. Not revolutionary, but clearly better. I was bypassing all digital equalization from the Behringer, mind you (never having found that I needed it, or that it didn't hurt the sound), but somehow the signal conversion within it seemed to be giving the Benchmark a little more (or better?) signal to work with.

 

Is this even possible? Has anyone heard the same in a similar set-up? Now I strongly sit in the camp of "less must be better" when it comes to having stuff in the signal path between source and the speakers, but this finding caused me to pause and ask more questions. Why does adding a conversion seem to make the sound better? Is the AES/EBU input on the Benchmark just better than the toslink? Is the Behringer somehow further reducing jitter with the conversion? Something else? My imagination? I had read years ago when I first started that jitter was sometimes thought to have been reduced when a cable conversion was in the mix, but I was never able to hear a real difference (hence the Theta was removed from my system).

 

So that of course has made me pour through the rest of the forums to look for other cabling/signal-out options from the mac mini that might have further improvements, or at least replicate what I think I'm hearing. Since I can't send USB directly to my Benchmark, I've been curious to try Bel Canto's Link that's coming out shortly (I hope). My friend is very successfully using that to feed his MacPro to his Bel Canto Dac3. But then with USB I lose the 96hz "benefits" I think I heard with the midi settings, right?

 

Short of leaving the Behringer in the chain, does anyone have any other suggestions (short of wholly changing out the mac mini for an older mac pro running XP with the sound card everyone is raving about on these forums, or changing DACs to get USB directly to the DAC)? Spending thousands on a firewire solution? There's an M-Audio CO3 that appears to transfer 96hz that I still might be able to get my hands on for another toslink to AES/EBU conversion, but would it be as good/better than the Behringer?

 

I don't mind spending a couple hundred bucks to convert to an AES/EBU signal if that's the conclusion I reach, but more than that and I think I would either keep the Behringer in line just to do this conversion, or invest in a new DAC, or (god forbid) a new front-end.

 

Thoughts? Thanks folks!

 

[Ashley James]

It's important to understand that jitter is only an issue at the A to D and D to A stage and that the Benchmark is sold on its use of Sample Rate conversion to eliminate it.

 

It's also important to understand that any EQ or anything else done in the digital domain is purely mathematical and is not corrupting the sound, although it might be changing it.

 

Therefore you should treat all digital sources as being the same, you shouldn't worry about how you get your noughts and ones. Instead you should remember that minute level differences will always persuade you that louder is better and that it's terribly easy to imagine what doesn't exist. Analogue signal processing is where big differences can sometimes be made IMO.

 

Ashley

 

[jay]

Ashley,

 

Hugely appreciate the quick response and advice.

 

If I'm to understand your points correctly, you're very correctly calling out that I may be, in fact, not hearing a "real" difference, other than perhaps slight changes in volume. That's a super fair point. But I presume that would be due to what you would conclude to be some minute change in gain introduced by the DEQ even in pass-through mode, or again, my imagination and desire to hear something (anything) different, positive or negative as I might perceive that change to be. I totally agree in that possibility, and would even go as far to say it's highly likely. I'm going to get better ears than mine to try to validate what I think I'm hearing. But if I'm to understand your point further, and please forgive me if I'm too generally reading your comment here, that would also imply that you're not totally sold on some of the topics of this site that in fact, place some implied value on "better getting the 1s and 0s" to the DAC, in whatever means that entails?

 

I suspect you're actually correctly warning me to "beware of snake oil," another very good point. I recognize the overriding goals of this site and the majority of it's visitors is to create computer-based front-ends that satisfy the aggressive but very achievable needs of creating fantastic "audiophile" music. From reading your site (well done, btw), it seems clear to me that you completely support such ideas (hence your positive participation with this site), and perhaps even more importantly to me you support the full-on pursuit of such goals via very open thinking. I REALLY like your approach, as an aside. The systemic approach to audio design has insanely good merit in my head, and clearly your product line looks to solve real-world problems with totally open thinking, not embracing what may be traditional and now somewhat outdated concepts (like "high-cost," "boutique," "separates", etc.) to accomplish the same goals often at unrealistic prices for true music lovers. I was fortunate to be introduced to this line of thinking a few years ago by a rather seasoned sound engineer and equipment designer/builder long in the industry, but jaded by the business side of "audiophile." His most recent attempts to deliver phenomenal music were in fact a computer-based front end feeding an integrated DAC/pre/amp that he designed feeding his own speakers designed for the same purpose, not fed by expensive boutique wiring or a $5K transport. His point to me--don't meddle with his system, and he will promise very impressive sound and a total system price less than I must otherwise spend to get the same (or inferior) sound. Very eye-opening to me that such beautiful sound ignored a lot of traditional "equipment" industry "musts." Beautiful music for the masses without the endless pursuit of equipment experimentation--a wonderful concept. (Of course I had already dropped loads of cash on a system of individual components that I was basically very happy with, but had I visited him earlier in my pursuits, I probably would have bought one of his systems and saved some time and coin!). I'd say from your site that you may have gone a step further and done the same at even more affordable prices, which is highly commendable, so well-done. I hope to have the opportunity to hear your stuff.

 

I perhaps worded my original set of questions too generically. I'm actually hoping that I would find someone with the same or very similar set of gear I have and having experimented with the sames things I'm trying, if for no other reason that the fun side of experimenting for the elusive, albeit subjective, "better" sound. Having the Behringer back in line, I've actually spent some time in the last two days goofing with the room correcting EQ features, which kind of negates a bit of my line of questioning as now I'm purposefully using this item to mess with the sound, but regardless I am still interested to see if anyone has what they feel is a more advantageous way to get the mac mini to "sing" for me.

 

Thanks again for the feedback, Ashley, and please keep up your pursuit of digital excellence for the rest of us. I will admit that my first and only vinyl purchase was Michael Jackson's Thriller album as a third grader, so my appreciation for the analog realm (at least from a source perspective) is quite limited and therefore maybe I'm more willing to embrace the all-digital/all-at-my-fingertips approach to beautiful music than others, which is why I think I've found my safe-haven on this site!

 

[Ashley James]

Jay

 

Thank you very much for your understanding and for not being cross with me for being so blunt. You've understood what I said perfectly. IMO we should all concentrate on listening to the bits of a hi fi system that are known to make differences.

 

It's very hard to tell how good or bad things are when the differences are small and no one is there to help you. It's also possible to persuade yourself something not very good is okay, so if expert opinion (and I don't mean me, I mean experts) says that you can treat all digital sources as sounding the same, you should. Music is no different to anything else a computer does, so if everything else it's doing is being done correctly, then it's good. XP and Windows excepted.

 

You cannot EQ the effect of a room because your ears hear what comes from the speaker before the room takes effect, unless it has a big problem, in which case it won't really help. What you can do with EQ is to correct some of the more unpalatable recordings and where possible, you should do it by subtracting rather than adding so you don't eat up all your headroom.

 

As a general rule, too much bass needs a reduction in output in the 60-80 Hz region and too little or too bright a sound is best corrected by reducing output from between 1-3 kHz.

 

The iTunes equaliser is good to practice with.

 

Best of luck with your quest and thanks for again for your understanding.

 

Ashley

 

[mpmct]

What you can do with EQ is to correct some of the more unpalatable recordings and where possible, you should do it by subtracting rather than adding so you don't eat up all your headroom.

 

Ashley, can you explain this a little more?

I've been doing just the opposite with iTunes custom EQ settings --

compensating by adding, rather than by subtracting.

Was only my best guess as to which 'direction' would be best.

 

By the way, having experimented more with recordings that

were the most troubling to me, and making Presets for each,

and tweaking them repeatedly ... this is really something -- to

be able to have this sort of control on a micro level.

 

 

[Ashley James]

Sorry I've been out in the garage putting the gearbox back in my Bristol and I got cold and knackered so wasn't thinking clearly.

 

As you know iTunes allows you to preset EQ for each track and that's good, you can set it how you like without worry.

 

If you make big adjustments with the main equaliser, you'll use up all your headroom. iTunes does compensate, but it's best to start by reducing the outputs at different frequencies first and only boost them up if you have to.

 

If you push the gain up in the equaliser you may see iTunes pulling it down again to stop you overdriving anything.

 

The thing is to experiment till you get a feel for it.

 

Ash

 

[mpmct]

Thanks again Ashley!

 

[DanRubin]

Jay, thanks for sharing your story. One piece that surprised me was your finding that setting Audio Midi to 96hz and 24-bit gives you better sound. I have found the opposite to be true. And it makes sense to me that if any upsampling is to occur, better to have the DAC do it and not the Mac Mini. But I remain curious about this. What have others experienced?

 

Dan

 

[The Computer Audiophile]

Hi Dan - My experiences are the same as yours.

 

[jay]

So after going back and playing with the settings again for the last couple of hours, I think the difference for me is actually much less noticeable (not noticeable?) between 44.1 and 96hz, but instead the biggest difference comes in the 16-bit to 24-bit jump. 16-bit sounds very slightly "gritty" to me, at least in some of the voices (many of them male, and with higher and louder pitches). 24-bit sounds less fatiguing to me and more natural over the same sections. I had a musician friend blindly confirm this on my stereo a few days later with a bunch of music he's very familiar with. He was really surprised to hear any difference at all, actually. He came back a second day and played with the settings himself for about 30-minutes, and confirmed his opinion that 24-bit clearly sounded better on my system.

 

Now, this difference may be DAC-specific, hence why our individual experiences might differ. I'm using an original Benchmark (non-pre, non-USB)...

 

Benchmark's website wiki may support what I'm finding on the 24-bit setting.

http://extra.benchmarkmedia.com/wiki/index.php/ITunes-QuickTime_for_Mac_-_Setup_Guide

 

Section 2 seems to state their recommendation "for optimal audio quality" is setting midi out settings to 24-bit and setting the sample rate to match the source. However, much of the text references concerns w/ iTunes 6 vs. 7, so who knows how 8 is performing relative to their findings in 7.x. Hopefully not worse that 7.x. I updated to 8.x before I tried all these midi setup changes, and their wiki is not updated with any 8.x results.

 

Then later, in section 3, there is a slight contradiction on sample rate, as the author of that section recommends 96hz in 7.x (again, not updated with 8.x findings, so who knows)...

 

"Our measurements show that the iTunes 7.X SRC is so good that it should be inaudible (unless it is down sampling). An easy, all-inclusive setup is to set AudioMIDI to 96kHz before launching iTunes and keep AudioMIDI at 96kHz. This will lock iTunes at 96 kHz, which will upsample all sample-rates lower then 96 kHz. Setting the default to 96 kHz would allow playback of high-resolution 96 kHz 24-bit files without any processing, and would utilize the high-quality iTunes 7.X SRC only when playing low resolution 44.1 and 48 kHz files, or the rare 88.2 kHz file."

 

Now with 8, I definitely don't have to shut down iTunes before changes take effect (I can clearly hear a difference when the bit-rate is changed, and I see the difference in sample rate delivered on my DEQ2496's readout), but that's an aside not related to the sound quality.

 

Just for one more data point, I went back into some email archives to find one from a buddy of mine also using a mac, but a 12" macbook pro. Without optical out on that, he is feeding his Bel Canto Dac3 with BC's converter box (seen on this site)--USB to digi coax, and reports to me finding 48 hz to sound better than 44.1. He didn't mention bit-rates used. Again, who knows, leads me to believe it may not be Mac specific, but DAC specific.

 

For the record, while I have been very happy with my Benchmark and I believe they truly build a well-designed, well-built, well-tested product, I don't necessary take their word for everything in writing that comes from them...I will try their suggestions, but my conclusions don't always match. I do think it's system-specific in some cases (I won't try to argue their jitter graphs, however). As an example, their literature strongly recommends bypassing any pre-amp and using their volume control where possible. This produced worse results for me than going through some custom-made passive pre's I have. I can't tell you how many times I've tried to remove pre-amps from my signal path, but they're still there. I mention only to reiterate that I'm objective as to what I read, even for products I really like like the Benchmark.

 

One last point to that note, I'm hoping to get my nubs on my friend's BC DAC3 before the holidays while he reviews another model, so I'll repeat my experiment with a different DAC and see if I still prefer 24-bit.

 

If anyone else wants to loan me their DACs I'll be happy to test them as well!

 

[DanRubin]

Whew -- great response. I'm going to revisit all of this based on your experiences. Unfortunately, I don't yet have any USB DACs that support anything but 44/16, but I can experiment in Toslink. As I think about it, my findings may have predated iTunes 8, though I'm not certain of it.

 

Dan

 

[jay]

As I went to bed last night I thought of one more thing which I overlooked...

 

I don't think my Benchmark DAC does any upsampling/oversampling. That may be why Benchmark is suggesting the Mac do it. When you found 44/16 to sound better, was it on a DAC that does upsampling/oversampling itself (I presume so since that was kind of your question)? If so, that may explain it. It's highly likely that a good DAC could do a better job upsampling than mac software, to your point in your first response to my post. That may result in a better sound--hopefully the DAC manufacturers would certainly not upsample or oversample if they didn't think it helped the output. I know some have the option to turn it on and off. I think I also read once that Benchmark chose not to use that in their design, but I forget the reason or more likely never understood it anyway! Although kind of a head-scratcher that they would come out and say that you should upsample before sending the signal to them. Kind of implies they think a computer should sound better than a 44.1/16 cd from a good transport, doesn't it?

 

Interesting stuff!

 

[Beranga]

I know this is an old post but I had the same experience as the original poster using my Bryston BDA-1 as my DAC. Regarding the AES/EBU connction of the DEQ2496, I can notice an improvement when using an AES/EBU connection compared to Toslink (glass Optocoupler cable).

 

Since the original poster used a Benchmark DAC1 and I a Bryston BDA-1, I guess we can conclude that the improvement is due to the Behringer DEQ and not to the better/worse inputs of the DAC's as was initially believed. I also tried skipping the DEQ ang going straight to the BDA-1 with Toslink, but the Behringer's AES/EBU was still noticeable better.

 

As always, this is with my system, in my room and IMHO.