Hello Everyone,
This is a review and detailed measurements of the Marantz CD6007 stereo player and transport.
View attachment 386998
After the
Onkyo C-733, this is my second review on ASR.
Fact is I’m a lot into CD players, especially older ones, I like testing them to witness the evolution (over the last 4 decades) of the Digital to Analog conversion techniques.
Following the review of the
Marantz CD6006 here, I wanted to give a chance to Marantz and to a modern CD player, hoping it could serve as a reference for me (when it comes to compare old vs new CD player).
Anyways, follow me because it was kind of an adventure.
Marantz CD6007 - Presentation
This is still the latest iteration from Marantz (as of August 2024) of this player, replacing the previous CD6006. As you can see on the above picture, the look has not changed and has been the same for a long time. I like it a lot.
On the front face you can find phones output, which is cool. Even cooler, the amp gain can be adjusted (3 levels - High/Mid/Low - Default is Low), so you can make it match the impedance of your favorite cans. There is also a USB input which will accept a drive and higher resolution files (up to PCM 24bits/192kHz and DSD 5.6MHz).
The oversampling filter has two settings, Filter 1 (Default) and Filter 2. They are described as such in the user guide:
View attachment 387010
Back panel shows the essential:
View attachment 387744
We have RCA out and digital (Otpi + Coax). Note the coper (plated?) screws, giving a luxury touch.
I did not open it, but you'll find plenty of pictures on the web. You will see that this player features a discrete output stage from Marantz (HDAM). There was also an apparent issue with AKM to deliver the DAC (AK4490) after a fire at their factory, and it seems some of these players have an ESS DAC instead of the AK for that reason.
In terms of usage, first impression was rather good. But I was rapidly annoyed by the slowness of the drive, especially considering this is a CD player only. Indeed, if going forward is relatively fast (but no where near older players), going backwards was a nightmare. It just did not react to my instructions. When using my test CD which contains 44 tracks, it was faster to go 20 tracks forwards than 5 backwards. I don't like that.
Other than that, the finish is flawless.
Marantz CD6007 - Measurements (Analog out - From CD)
From now on, I will be consistent with my measurements as I described them on the
Onkyo C-733 review. So over time, this will help comparing the items I reviewed.
The Marantz CD6007 outputs roughly 2dB above the standard 2Vrsm, maybe Marantz has something to say THEY WANT YOU TO HEAR?
Let's start with the standard 1kHz sine @0dBFS (dithered) from my test CD (RCA out) with Filter 1 since it is the default choice:
View attachment 387009
I was honestly not expecting a change with Filter 2, but here you go:
View attachment 387011
It's minor, but seeing more harmonic distortion only because of slow/sharp filter change was a surprise (to me, maybe only).
I showed only one channel, but left and right are identical. THD is just below CD Audio resolution, which is good, but could have been better.
SINAD and ENOB are limited by the dithered noise of the test CD, but could be almost 2dB better, as it's the max of the test CD (when measured digitally).
Other results (not shown) are:
- Crosstalk : Non measurable at 1kHz, -132dB at 10kHz
- SNR : 97.5dB (1kHz @-60dBFS, no dither)
- IMD AES : -95dB
You probably already noticed that this is a very quiet CD player, with minimum power supply–related spuriae in its output (below -120dB at 50Hz):
View attachment 387012
This is really nice, and I'm happy to see near excellence in this area.
Bandwidth measurement was the point where this CD player started to give me some concerns. The two channels are matching at less than 0.05dB (very good). But have a look at the bandwidth between the two filters:
View attachment 387014
Yes, -5dB at 20kHz for Filter 1... And both have a negative effect starting at 10kHz or before... If the younger ears out here are to perceive a difference, or say smoothness into this player, this is why.
But let's zoom a little more:
View attachment 387020
Typical ringing of (old) filtering can be seen with Filter 1. You know, back in 1982, Marantz released its first CD player (CD-73) which had a 14bits only DAC and was relying on oversampling (and noise shaping) filter to compensate for the missing 2bits. This filter had significant ringing (I still have mine, I'll do a review here one day), and that might have driven Marantz this way?
Ringing is one thing, but analog filtering post conversion is another. And so, at this point, I started scratching my head, obviously in need of deep diving into what was going on here.
In the meantime, mutlitones test showed no issue, with both filters, I was happy again:
View attachment 387022
In both cases, CD Audio content is safe, from 20Hz to 20kHz.
The Marantz did not show any sign of Jitter:
View attachment 387023
This is an overlay of Analog (blue) and Digital outputs (red). The Marantz did not add jitter on top of what's potentially on the Test CD. Note the absence of additional rays, it's very good.
Before I go into digital output, I need to analyze what's about the oversampling filter.
When using the Filer 1, I saw such a slow filter that I wondered: what if I give it a square signal? Well, here you go (Square 1kHz @0dBFS ):
View attachment 387053
Whaaat?
That is theoretically not possible. Because of bandwidth limitation (22kHz max in CDA), this should show the
Gibbs Phenomenon. Filter 2 gives a more logical result:
View attachment 387054
We find here more of the bandwidth limitation of CD Audio, but still not what it should be. And for reference, this is what we should (and need) to see (from the Onkyo C-733):
View attachment 387057
The ringing is correct here. So, what's going on with the Marantz CD6007?
In order to better understand, I had to analyze up to 96kHz (I can't go higher than that with my Motu interface). And here are the results:
View attachment 387059
Sorry for the complex view, it contains a lot of information. Deep red and blue traces are white noise analysis (@16bits/44.1kHz) from 20Hz to 96kHz. They both show the oversampling filter response.
On top, there are triple tones test at 80Hz, 1kHz and 10khz (my choice). Light blue is Filter 2 and shows very high level aliases of the conversion at 78.2kHz (-46dBr), 87.2kHz (-58dBr), 88.1kHz (-60dBR), 88.3kHz (-60dBr) and 89.2kHz (-62dBr). Light red shows the same with Filter 1, but aliases are attenuated, as per the oversampling response filter.
I think this type of response recalls me a case of upsampling to DSD before conversion. I will double check later. EDIT: Found! I finally recalled my test of the Asus Xonar Essence One DAC, which offered an oversampling function. Since I still have it, time to grab and plug, here you go with its filter response when upsampling is activated or not (from CD Player digital output, that is 16bits/44.1kHz):
View attachment 387153
Even though we're not looking at the exact same effect here, you can see that activating the upsampling filter on the Asus generates a very early roll off, making it sound dull. Filter 1 of the Marantz is not that bad, yet -5dB at 20kHz is what we see.
In terms of 1kHz Square view, the Asus also reproduces a near perfect square, indicating again a large amount of
artifacts of the conversion going through (because of bad filtering):
View attachment 387154
And again, without upsampling, the square shows the
Gibbs Phenomenon, which it should because there's proper filtering:
View attachment 387155
Ok all of that is clear now, I think. The Marantz operates an upsampling (to DSD?) of the signal, be it 44.1kHz or 48kHz. It decreases the out of band attenuation which should have been avoided, on my perspective.
So you're left with two choices when playing CD audio or files at 44.1kHz or 48kHz (same response):
- Filter 1: very slow filter, kind of "NOS", non oversampling filtering, letting a lot of artifacts of the conversion go to your preamp, amplifier and speakers.
- Filter 2: sharper filter and reasonable attenuation (-70dB) at around 24kHz but with higher energy artifacts starting from 70kHz.
When I started my testing, I thought sharper filter was better, but in the end, I'm not sure.
What will happen with real music, you might ask. Well, here you go below with a long term analysis (1700 averages) of the track "
Fast Car" from Tracy Chapman:
View attachment 387062
Filter 2 (blue) shows high energy of aliases at near -50dB around 88kHz. Filter 1 (red) shows typical Sinc envelope conversion and
unwanted imaging. So with real music, the lack of post 22kHz filtering shows here.
Marantz CD6007 - Measurements (Analog out - From USB in)
If the above let me concerned, when using higher resolution files (24bits/96kHz) from the USB input, the results, at least on a filtering perspective, were good. Here below is an example of filter response (Filter 1 and 2) with a 24bits/96kHz file (played from a USB stick):
View attachment 387065
What was a concern with 44.1kHz files has vanished here. Sharp and Slow filters show good results. This tells me that the Marantz CD6007 is at better ease dealing with high-res files than CD Audio (too bad for a CD player).
The 1kHz @0dBFS with 24bits/96kHz files showed the below:
View attachment 387066
If the distortion did not improve much, the SNR did. That said, on a noise perspective, I'm close to the limit of my Motu interface which is not so at ease with only 2.5Vrms input. So the Marantz might do better. But THD is what it is, and that's the SINAD limitation here.
Marantz CD6007 Measurements - Optical Out
Last but not least, I measured the digital output of the player, from my test CD, for those who'd like to use it as a transport. The below view shows what's on the CD:
View attachment 387068
Note this is only 4 averages when I measured, compared to 32 averages I usually apply. It does not change the final results though. THD is basically non existent and rest is limited by the dithered noise of the test CD.
This is very good and I confirm this player can act as a very good transport, not modifying the bitstream.
Conclusion
This review might generate discussions. I'm eager to learn, as always. Maybe I got something wrong (about filter issue with 44.1kHz files/CD and 48kHz files too), but I'm sure of my measurements. Someone from Marantz or here might be able to assist and give some clues as to what is happening. Is it upsampling to DSD causing this?
Despite the overall good results, in audio band analysis, I can't ignore what's happening beyond 20kHz as it concerns me (but should it be of concerns?). I don't understand the benefit of the lack of filtering I saw, not to mention the significant roll of starting at 15kHz with Filter 1. Obviously Filter2 offers better response in band and also better attenuation out of band, but the energy that it lets go through at around 88kHz are artifact from low frequency in-band audio (where music has a lot of energy, ie: bass). So I'm puzzled as to what filter to use.
Since that is primarily a CD player, I expected better processing with 44.1kHz audio files. Higher resolution files are processed correctly up to what I can measure (96kHz).
It is a good Transport too.
————
Flo
