Naim DAC vs Transporter

Posted by: AMA on 16 February 2010

Now I have finished the first bunch of tests over Naim DAC.

The setup was Logitech Transporter/DigitalCable/DAC/XPS/282/2*HC/250.2/ProAc D28.
Digital Cable was Klotz RAC coaxial and VDH Optocoupler optical.
Klotz was a tad better than optical. Better clarity -- very small effect.
I'm not sure this was because of the cable difference or
because of the difference between Transporter RCA/TOS output
or because of the difference between DAC RCA/TOS input.
The rest of the listening was performed with coaxial.

Brief summary of contenders.

I will not go deep into A/B test against CD5X because TP->DAC combo was WAY ahead of it in all aspects.
The actual test was against Logitech Transporter through analogue output and TP->PS Audio DLIII. Both used 2*RCA->DIN cable.

Logitech Transporter is one of the best sounding sources I have ever tried -- in Red Book it comes very close to KDS -- although the KDS owners will consider the gap as HUGE and put TP to ADS level or even lower Smile
In terms jitter this is one of the best digital transports on the market. I'm not sure if further reduction of jitter on Red Book can be audible.
TP shows fantastic resolution and 3D imaging.
The problem is that I don't like TP analogue output. It's a bit slow and lean and analytical -- it lacks a PRAT of Naim CDPs (like CD5X).
I recall that I liked KDS more than TP just because KDS was more energetic. KDS is also smoother than TP - I think because of the transformer coupled outputs.

PS Audio DLIII is a very good DAC -- same league as Lavry and TP but owns idiosyncratic sound presentation.
Not as neutral as TP and a bit forward (human voices fire up a bit "in your face"). But very soft and energetic -- closer to Naim.
It does not use re-clocking and needs a low jitter input to show it's best. When pairing with TP you get a very good resolution.

After several months of continuous trials I have found myself of listening TP->DLIII almost all the time which means that I belong to those of us who can compromise (with a deep sorrow of course) with a sound resolution in favor of tuneful sound presentation.

Both TP and DLIII were the benchmarks for Naim DAC in different aspects: high resolution and imaging of TP and energy of DLIII.

I was the first user of the long-awaited demo piece and had a chance to unpack the new box.
Honestly -- after those shipment delays and poor explanations and gossips around for almost 6 months I expected from Naim to pacify customers frustration with a bonus bottle of whisky in the box Smile

I would avoid a tedious description of the first days of auditioning -- I only mention that sound was kept changing for all 4 days while DAC stayed with me.

I started tests with a bare DAC, original IC and stock power cord and Red Book CDs ripped on a NAS drive.
Many of the test records were XRCD and XRCD24 which offer a superior dynamic range and very low jitter.

First of all -- it betters TP-> DLIII in resolution and on par with TP through analogue.
I was prepared to hear a vast superiority of Naim DAC in resolution due to re-clocking circuitry -- but it didn't happen.
Possibly because TP as a digital transport suggests a very low jitter bitstream so that there is no space for improvement.
Possibly because the DAC was not fully burnt-in.
Second -- DAC soundstage is very deep and wide and the imaging is holographic -- better than TP and DLIII (not much).
Third -- it's a very typical Naim sound. Very neutral, very fast and very euphonic. Transients are crystal clear.
The downside is that DAC is a bit edgy -- more aggressive than CD5X, closer to CDX2.
Overall it was definitely far from the culture shock after the first listening session of CDS3...

Adding a PL and HilIne was more audible than when you hook them on CD5X. HiLine brought much more than PL.
The sound became more open and detailed -- but improvement revealed more edginess.

The last step was adding XPS. I didn't hear much of the changes in resolution. But the analogue section suffered a serious transformation -- it smoothed the sound substantially.
The background became blackish like TP. Bass became even tighter. Very analogue-like. This was much better than CDX2/XPS and much better than both TP and DLIII.
PL on XPS is more audible than on DAC.
The TP/DAC/XPS/PL/HL setup is definitely close to the best sources I have ever tried: CDS3, ARC CD5 and Klimax DS -- would be interesting to do A/B test.
It's also interesting to test TP through BNC against CDX2-2 and HDX through BNC. I will not be surprised to see TP to be a winner, but there is only way to know it for sure ...

I have some difficult tracks where an instrument or a voice stands out "in your face" or dominates over the rest of the mix.
Sometimes, for example, Roger Waters or Mike Oldfield solo guitar is so harsh-- just unbearable.
But KDS and ARC CD5 manage to smooth it down -- just in the same way as Naim DAC/XPS did.

Well-done Naim!

Unfortunately the demo piece was taken away for other hungry users and I was promised to get my own box in the next shipment.
I have added another PL and DC1 to the order (one PL will got on DAC and one on XPS). TP has transformer coupled BNC which I can use with DL1 and switch ground to Chassis.
Next time I shall come back after one week of burn-in and also elaborate more on hi-res tests. I hope it will happen next month.

It's still amusing to see how a 2$K TP designed in 2006 by an amateur audiophile from Silicon Valley keeps competing with 2010 top products from venerable hi-end manufacturers.
While giving up by a small margin in analogue section it still proves it's strength as a digital transport.

Bravo Sean!
Posted on: 22 February 2010 by AMA
quote:
Why don't you try rebuilding a FLAC to wave and compare on playback?

js, it will take some time -- do have any hints that TP transcoding is not accurate?
Posted on: 22 February 2010 by likesmusic
AMA - you don't have to rebuild anything; just go into Squeezecenter > Settings > Advanced > Filetypes. You can choose the stream format for each file format. WAV can be streamed as FLAC or PCM, FLAC can be streamed as FLAC or PCM.
Posted on: 23 February 2010 by AMA
Thanks, likemusic! I shall check this out.
Posted on: 23 February 2010 by js
That's interesting. I wonder if that would eliminate the processing bottleneck I hear with lossless. Smile I like that one, likesmusic. Smile AMA, I would just do 1 file you like as a stored wav to compare and see what the 3 itterations sound like.
Posted on: 26 February 2010 by DaveBk
My DAC's also arrived... so another Transporter -> DAC setup to consider. First impressions are very very good.

Setup is now:

Transporter (BNC) -> DC1 -> Naim DAC (Powerline) -> HiLine -> 252/SC2 (Powerline) -> 300 (Powerline) -> NACA5 -> Ovator S-600s

Most excellent music.
Posted on: 26 February 2010 by Dustysox
quote:
Originally posted by DaveBk:
My DAC's also arrived... so another Transporter -> DAC setup to consider. First impressions are very very good.

Setup is now:

Transporter (BNC) -> DC1 -> Naim DAC (Powerline) -> HiLine -> 252/SC2 (Powerline) -> 300 (Powerline) -> NACA5 -> Ovator S-600s

Most excellent music.


Wow Dave,

Don't you be selfish...tell all! Can't wait to hear what your thoughts are.

Do tell?
Posted on: 26 February 2010 by AMA
quote:
Don't you be selfish...tell all!

Let him burn in the DAC for couple of days...

Difficult to compare a new DAC with TP which is burnt-in for years.
The new one will always sound a bit harsh from the box.
I was not gobsmacked with a bare Naim DAC although it sounded more dynamic than TP straight from the box. Other than this there were no serious advantages over TP in other sonic elements.

But it was XPS which leveled it up to the CDS3 (or close). The sound became smoother and tighter and the background became blackish - similar to what it does on CDX2.

Dave -- you definitely have to try XPS.
It also needs time to break-in -- even more than DAC itself.
Posted on: 26 February 2010 by JYOW
Dave,

I would be very interested in your conclusion re the Transporter standalone Vs Transporter plus the Naim DAC. I have moved on from the Transporter to Macbook/Weiss and now Macbook/Naim, but never had a chance to hear how the Transporter sound when connected to these DACs.

The Transporter is such a good machine
Posted on: 27 February 2010 by js
Please play wav files when doing comparisons. When comparing DACs, you want the DAC to be the most significant influence on the sound.
Posted on: 27 February 2010 by AMA
I shall also do this test later on -- when I get a small break in the current workload. I shall report on findings.
Posted on: 27 February 2010 by Hook
quote:
Originally posted by js:
Please play wav files when doing comparisons. When comparing DACs, you want the DAC to be the most significant influence on the sound.


JS -

How far have you taken this WAV versus FLAC (and/or other lossless) comparison?

I think we all agree that the transcoding works, and is reversible.

But since you and others have reported audible improvements with WAV, the theory seems to be that processing load is to blame. The real-time transcoding by a streamer or a computer running playback software introduces noise, errors, etc. through increased demands on its power supply.

So my question is: have you compared say, two computers (one with faster cpu and/or more memory) running the same playback software, each playing WAV versus FLAC? If the theory holds, wouldn't we hear less of a difference with the faster/larger memory setup?

Was just reading the same discussion over in the Linn forum among Klimax DS users. There was no real consensus. Some hear an advantage with WAV, others don't. Of course, the lack of meta data support makes WAV impractical for ripping large libaraies.

Perhaps the answer lies in playback software that fully transcodes FLAC to a WAV temp file before playback rather than doing it on the fly. My understanding is that Klimax, Transporter, and perhaps all streamers transcode on the fly (hence the additional processing and power draw). Ditto for the different software players?

Anyway, just curious how much testing you've done.

Thanks.

Hook
Posted on: 27 February 2010 by DaveBk
I remain to be convinced that there's an audible difference between wav and flac, but I'm prepared to give it a go and report what I find. Flac takes far more processing power to encode than decode so is a relatively light load on modern processors so I don't buy the greater power consumption making the power supply noisier argument. Will report back.
Posted on: 27 February 2010 by js
quote:
Originally posted by ghook2020:
quote:
Originally posted by js:
Please play wav files when doing comparisons. When comparing DACs, you want the DAC to be the most significant influence on the sound.


JS -

How far have you taken this WAV versus FLAC (and/or other lossless) comparison?

I think we all agree that the transcoding works, and is reversible.

But since you and others have reported audible improvements with WAV, the theory seems to be that processing load is to blame. The real-time transcoding by a streamer or a computer running playback software introduces noise, errors, etc. through increased demands on its power supply.

So my question is: have you compared say, two computers (one with faster cpu and/or more memory) running the same playback software, each playing WAV versus FLAC? If the theory holds, wouldn't we hear less of a difference with the faster/larger memory setup?

Was just reading the same discussion over in the Linn forum among Klimax DS users. There was no real consensus. Some hear an advantage with WAV, others don't. Of course, the lack of meta data support makes WAV impractical for ripping large libaraies.

Perhaps the answer lies in playback software that fully transcodes FLAC to a WAV temp file before playback rather than doing it on the fly. My understanding is that Klimax, Transporter, and perhaps all streamers transcode on the fly (hence the additional processing and power draw). Ditto for the different software players?

Anyway, just curious how much testing you've done.

Thanks.

Hook
A variety of PCs and players. I really don't want this to be the case either as I like the idea of flac tags moving about with the tunes. I would think a good buffer arrangement would eliminate the issue but haven't found that to be the case. I get fine results from lossless in some setups when using Itunes etc. when things are a bit masked anyway but in the best setups, it's still quite telling. I'll give another listen with the next (latest) HDX we recieve. Last time I tried, I think I liked ALAC a bit more overall but wav was still the standard.