OPPO BDP-95 Review (Blu-ray/DVD/SACD/CD Player)

[NTTY]

Magnetar players have been measured by third parties (more or less detailed measurements published in English or German):

• UDP800 (Secrets of Hometheater)

Ugly high level of THD!

• UDP900 (Secretsp of Hometheater)

Not much better, and worst than most CD players, and again dominated by distortion. Much more on XLR, BTW. How do they manage to do that? The DACs are not so bad, so it’s from the output stage.

 

[aagstn]

Ugly high level of THD!

Not much better, and worst than most CD players, and again dominated by distortion. Much more on XLR, BTW. How do they manage to do that? The DACs are not so bad, so it’s from the output stage.

My brother-in-law had the misfortune of spending a couple of grand on one of these for his home theater. It was his lesson in price tag doesn't guarantee performance. Movies would freeze at the layer change and it was apparent pretty quickly it wasn't any better than a Sony player that was thousands cheaper. I believe it is currently in his garage in the box all these years later after he got it back from them for "repair" and did all the same things he sent it in for. He learned to stay away from expensive boutique hi-fi stores.

 

[PristineSound]

Magnetar players have been measured by third parties (more or less detailed measurements published in English or German):

• UDP800 (Secrets of Hometheater)
• UDP800 (Audio and AudioVision)
• UDP900 (Secrets of Hometheater)
• UDP900 (Stereo)
• UDP900 (Audio and AudioVision)

Thanks @Scytales, didn't know about these test. Well, I guess Magnetar can't stand in to the Oppo void. I am so glad I was still able to get my hands on the Oppo 105D, not just one, but two and in perfectly mint condition.

 

[PristineSound]

My brother-in-law had the misfortune of spending a couple of grand on one of these for his home theater. It was his lesson in price tag doesn't guarantee performance. Movies would freeze at the layer change and it was apparent pretty quickly it wasn't any better than a Sony player that was thousands cheaper. I believe it is currently in his garage in the box all these years later after he got it back from them for "repair" and did all the same things he sent it in for. He learned to stay away from expensive boutique hi-fi stores.

He was probably desperate for an Oppo replacement!

 

[rnrmf]

Thanks @Scytales, didn't know about these test. Well, I guess Magnetar can't stand in to the Oppo void. I am so glad I was still able to get my hands on the Oppo 105D, not just one, but two and in perfectly mint condition.

Hopefully the recently announced MK II Magnetar players address the video issues that some experienced as well as the distortion on the balanced outputs.

 

[NHL99]

A second hand OPPO at bargain price, otherwise it's still possible to find new OPPO devices in online stores.

 

[pma]

Maybe I was unclear but that’s what I meant yes, in the previous message you quoted from me.

Oh no, you were very clear. Nice test, thank you. I may try similar test with another player. Right now I have only a test disc with -80, -90, -100, -110, -120dB shaped dither CD. The problem is not the file creation, but the physical disc, I have to buy them somewhere to burn them.

 

[pma]

Maybe I was unclear but that’s what I meant yes, in the previous message you quoted from me.
All my CD test files are tripled without dither, with rectangle dither and shape dither. Those without dither stop at -90dFBS while I go down to -110dBFS with rectangle and shape dither. I described them all here.
So I could have had a look at them but it was equally fast to record and play 24bits IMD at different levels from the USB key.
But since I have them (16bits files with shape dither), rapid copy/paste:

Once again thanks for the plots. I was curious of test results of such test when a SOTA DAC would be used with the CD player as a S/PDIF coax signal from the CD player. So I connected my Marantz SA7001 to Topping DX5 and made a set of measurements.

First, 1kHz 0dBFS dithered:

There is no distortion visible.

Then, a set of noise shaped dithered 16-bit 1kHz signals form CD with -80, -90, -100, -110, -120dB:



Again, there are no distortion component visible. To me, rather than trying to buy an unobtainium player, just take the oone you have at home and put a good DAC behind it and you are done .

CCIF IMD 19+20kHz

 

[Scytales]

OPPO BDP-95 Measurements (SACD)
[...]

View attachment 477133

The SACD wins, audiophiles are saved. But note we get "only" 1.5bit more than what can be achieved with the Audio CD and shape dither. So no, not 4 bits more as advertised by Sony (if I recall). We don't go from 16bits to 20bits, non.

I just want to add some comments to this.

To my understanding, the resolution in the audio pass-band of a sigma-delta modulated 1 bit stream (what DSD is) is not an absolute thing, but it depends on the order of the modulator and the response of the filter in the feedback loop of the modulator as well.

As an illustration, here is a small quote of digital analysis of the response of some old sigma-delta modulator designs for SA-CD production out of a very interesting series of two articles published in an Italian magazine with the help of an university professor who wrote a special software tool:

Obviously, different sigma-delta modulators produce different performances and have different behaviors. The following table was published in the Italian magazine to illustrate the equivalent resolution (in bits) of the many tested modulators in a 400 Hz - 20 kHz bandwidth:

Sony used to use its own Feed Forward type sigma-delta modulators (at least two different types) at the end of the 1990s and in the early 2000s, but there were not available to the Italians, unfortunately. But Prof. Malcolm Hawksford has published some studies using this type of modulator, notably this graph of his own about one of this Sony modulator design (from AES Convention Paper 6303):

It's seems to corroborate your memory about Sony's claim, although I have not found any material about it.

But time has passed, and it would be interesting to know the state of the art in Sigma Delta delta modulation. One of the fascinating thing about this technique is that it is able to evolve, yet the output of a 1 bit modulator is always a signal that toggles between only two levels!

 

[NTTY]

Once again thanks for the plots. I was curious of test results of such test when a SOTA DAC would be used with the CD player as a S/PDIF coax signal from the CD player. So I connected my Marantz SA7001 to Topping DX5 and made a set of measurements.

First, 1kHz 0dBFS dithered:
View attachment 477820
There is no distortion visible.

Then, a set of noise shaped dithered 16-bit 1kHz signals form CD with -80, -90, -100, -110, -120dB:

View attachment 477825 View attachment 477824 View attachment 477823 View attachment 477822 View attachment 477821

Again, there are no distortion component visible. To me, rather than trying to buy an unobtainium player, just take the oone you have at home and put a good DAC behind it and you are done .

CCIF IMD 19+20kHz
View attachment 477827

Yeah I did the same several times with a more modest SMSL PS200, and suggested (I think) that the TASCAM CD200 would make good use of it as a robust transport, for instance.
That’s also the reason why I always test the digital output of CD Players I review to ensure it’s not adding unwanted artifacts.

I ordered the Topping D50 III and I should receive it in couple of days. I wanted something to test up to where my measurement capabilities stop, as this one should get to the bottom of my Cosmos combo with 24 bits data.

 

[pma]

I ordered the Topping D50 III and I should receive it in couple of days. I wanted something to test up to where my measurement capabilities stop, as this one should get to the bottom of my Cosmos combo with 24 bits data.

Congrats! It seems that in case of CD players we are able to get "perfect" analog data from the 16-bit files.

1. test signal on CD


2. analog output from SA7001+Topping DX5

 

[NTTY]

View attachment 477830

Obviously, different sigma-delta modulators produce different performances and have different behaviors.

Interesting to see that the profile I got from the Denon Test SACD is the one shown in the middle above. It can be seen in the previous page, but I’ll update it with 20kHz upper limit for better viewing. EDIT:

And yes modulators vary depending on their order and oversampling applied for a given number of bits (in this case 1).

 

[NTTY]

Congrats! It seems that in case of CD players we are able to get "perfect" analog data from the 16-bit files.

1. test signal on CD
View attachment 477866

2. analog output from SA7001+Topping DX5
View attachment 477865

Indeed, and you get the feeling to be measuring the digital output, not the analog one

 

[Kiko1974]

If I recall correctly, the 93, 103, and 203 models were focussed on the video side. It was the 95, 105 and 205 that had superior analogue side for hifi use.

I rarely use my 95 but maybe I need to try it again soon. That said, I have had the same Tascam CD-200 as raved about here, so I’m very fortunate to have some options!

I got an OPPO BDP-93 as I couldn't afford the 95 at the time.
I was told by OPPO at the time that if I wasn't using the analogue outputs much but its HDMI outputs most of the time, as It was my case, the 93 was the right choice for me. But I connected the stereo analogue outputs anyway and 'though its audio section wasn't as good as the 95, the 93 sounded fine to me.
You state that the 95 can play DSD audio files. My 93 couldn't, but that could have been a mistaken from my end. I'll elaborate: when I extracted the DSD audio tracks from an SACD ISO with ISO2DSD, I always chose Philips' dff, not Sony's dsf. Maybe the 93 did play DSD files as dsf, but not dff. I have no way to check this out as I sold my 93 6 years ago, to my regret.

 

[Kiko1974]

So did you!

Now that I am thinking about that, let me take the opportunity to make an observation about the comparison between the test signals at 1 kHz @ -16 dBFS on CD and -16 dB SA-CD on SA-CD.

Why I always insist to write '-16 dB SA-CD' instead of '-16 dBFS' is not only a case of pedantry, but because there is a subtlety concerning levels in the SA-CD format and one must be aware of it.

Annex D, point D.2 of the 'Super Audio CD System Description, Part 2, Audio Specification', Version 2.0 (so the called 'Scarlet Book'), defines the 0 dB SA-CD as the peak amplitude of a sine wave at 50% modulation index of the DSD signal. That means that the 0 dB level on an SA-CD is not the absolute maximum digital level, contrary to what happen in the PCM world, where there can be no level above 0 dBFS because, by definition, 0 dBFS is the biggest (or smallest) usable number in a given PCM signal.

The important consequence of that fact is that even when analogue levels at the output of a DAC or a digital player is the same (ie perfectly aligned) when said devices reproduce a PCM signal at 0 dBFS or a DSD signal at 0 dB SA-CD, one must always keep in mind that DSD can go higher that.

By how much is a matter of standardization. The above mentioned Annex D, point D3.1 and 3.2, of the Scarlet Book acknowledges short term peak levels of +3.1 dB SA-CD over 28 consecutive DSD samples and long term peak levels calculated after a peak hold with slow down having a 258 ms time constant followed by a low-pass filter having a 64 ms time constant of +1.51 dB SA-CD. But even the standard set in the Scarlet Book is not at the maximum theoretical DSD level of 100% modulation index which, obviously, is 6 dB higher than 0 dB SA-CD.

This complex state of affair is the reason why when conversion (or more precisely decimation) of a DSD stream to PCM is done, the latter must usually accommodates 6 dB of headroom. That means that 0 dB SA-CD is set to equal -6 dBFS. Without that alignment, it is possible to clip the signal in the PCM domain when the DSD peak level signal exceeds 0 dB SA-CD. It is possible to dispense with this attenuation only if the DSD signal never goes above 0 dB SA-CD.

The best illustration of that necessity I know of is this graph I borrow from the datasheet of the NPC SM5819 DSD to PCM converter chip (comments in red are mine):

Of course, it is possible to optimize the alignment between a known DSD signal and the decimated PCM signal by incorporating less than 6 dB of headroom if it's feasible, but that oviously complicates things a bit.

Why I think all that I just wrote about is relevant when comparing the performance of a digital players in PCM and DSD mode is because the reference level to which the performance metrics is measured is never fully comparable in the two modes. A qualification about levels must always be kept in mind.

For instance, I have noticed on many datasheets of dual purpose DAC chips that there usually is a somewhat 3 dB discrepancy of performance metrics in dynamic range or SNR between the PCM mode and the DSD mode to the detriment of the second. But if we look at the conditions a little more carefully, we often realize (though not always) that this difference is just the consequence of differences of the reference digital levels between the two modes of operation.

Moreover, one must always be sure that there is sufficient analogue headroom after conversion of DSD signals to take care of possible peak signal levels way above 0 dB SA-CD in order not to clip any analogue stage. Especially since some SA-CDs have escaped the checks at the pressing plants to rule out some over-modulated SA-CD masters. The most famous example of over-modulated SA-CD is the 1999 issue of the Michael Jackson's Thiller album (and probably the reissues of it from the same master). I can add to this one another case I personally encountered : the multichannel area (I didn't check the stereo area) of the 2006 reissue of the opera Turandot in the RCA Living Stereo collection.

At least, it is my understanding of things, but I can be mistaken or overly cautious!

I own, but It hasn't been used in over 3 years, a Sony UBP X-8OO, and I currently use an UBP X-8OO M2. None of them play the 1999 Thriller SACD. At first I thought It was my copy of It, so I borrowed a friend's copy. It played the same, It skipped at exactely the same points on the same tracks (I remember some skipping on Beat It). I Guess these skipping were because this 1999 SACD didn't fully complied to the SACD standard despite being released by Sony Music.

 

[raif71]

I got an OPPO BDP-93 as I couldn't afford the 95 at the time.
I was told by OPPO at the time that if I wasn't using the analogue outputs much but its HDMI outputs most of the time, as It was my case, the 93 was the right choice for me. But I connected the stereo analogue outputs anyway and 'though its audio section wasn't as good as the 95, the 93 sounded fine to me.
You state that the 95 can play DSD audio files. My 93 couldn't, but that could have been a mistaken from my end. I'll elaborate: when I extracted the DSD audio tracks from an SACD ISO with ISO2DSD, I always chose Philips' dff, not Sony's dsf. Maybe the 93 did play DSD files as dsf, but not dff. I have no way to check this out as I sold my 93 6 years ago, to my regret.

Not sure about 93 but 103d for sure can play the dsd (sony dsf format) but hmmm could it be that my 103d is using modded fw

 

[Scytales]

I own, but It hasn't been used in over 3 years, a Sony UBP X-8OO, and I currently use an UBP X-8OO M2. None of them play the 1999 Thriller SACD. At first I thought It was my copy of It, so I borrowed a friend's copy. It played the same, It skipped at exactely the same points on the same tracks (I remember some skipping on Beat It). I Guess these skipping were because this 1999 SACD didn't fully complied to the SACD standard despite being released by Sony Music.

Most probably true.

Other than this specific Sony player, some DAC chips have a level processing stage which checks the compliance of DSD signals to the Scarlet Book level specifications and when it finds over-modulation, it attenuates and/or mute the output of the DAC. The AKM AK4493 and AK4499, and Cirrus Logic CS4398 and CS43198 come to mind. This feature can be disable in the Cirrus Logic chips.

It took me some time and effort to finally find an SA-CD player which is able to play my beloved Turandot RCA Living Stereo SA-CD release without muting its output and without distorting when playing through the over-modulated passages.

 

[eddantes]

Well... proof I guess. These were always great players.

Hmmm... I picked up a second hand Sherwood Newcastle SD-860 DVD & SACD for my CD/SACD needs for the price of a case of beer some years ago. Hardly ever use it because of streaming. I wonder if it's any good?

 

[NTTY]

I can add to this one another case I personally encountered : the multichannel area (I didn't check the stereo area) of the 2006 reissue of the opera Turandot in the RCA Living Stereo collection.

Ordered, thanks!
Let me know who to use for testing and I can include in my future reviews of SACD players. Some of them output only stereo channels, though.

 

[NTTY]

Well... proof I guess. These were always great players.

Hmmm... I picked up a second hand Sherwood Newcastle SD-860 DVD & SACD for my CD/SACD needs for the price of a case of beer some years ago. Hardly ever use it because of streaming. I wonder if it's any good?

Need to put an analyzer on the back of it to know more. We can’t assume anything with these multi-standard players