MUSINGS: Thoughts on audio device "modding" (eg. a tube modded Oppo BDP-105). [New Oppo UDP-205 firmware out with MQA USB-B...]

Couldn't resist the similarity :-). I'm sure the sonic output of both devices have been affected by the modifications! Notice that this Oppo is also sitting on a "Base" device that IMO is very much of the Dubiousness Class A variety...

Have a seat and let me tell you a tale, my friends. A number of years ago (summer 2013), around the time I started this blog, I had the opportunity to check out an interesting device that a friend had just received... It was a "tube modded" Oppo BDP-105 Blu-Ray player.

One afternoon, I went over to this friend's house to have a listen to this brand new device. He obviously spent quite a bit of money on it. The BDP-105 at that time was new and had an MSRP of US$1200. He is clearly an enthusiast so he "drop shipped" an unopened brand new player to the mod-shop and it came back to him with tube sockets for the analogue stage along with a fancy new external power supply. As you would expect, the modding itself being laborious, hand-assembled, and using premium parts was rather costly - about twice the price of the Oppo.

That afternoon, I had the opportunity to hook up my old test kit (with the Creative E-MU 0404 ADC) to the device to grab a few measurements. Since another friend already had a stock BDP-105 for me to listen to and measure, it allowed me to compare the results...

The reason that it has taken me so many years to put up this post is because I know that the modding cottage industry is small and I figured I'd patiently keep an eye open for any measurements or reviews that would allow me to corroborate some of my thoughts and findings. I figured that over time maybe there will be clarity on the value of doing this as people used and perhaps measured such things. Alas, after all these years, while reviews of these modded devices come out once awhile, I have never seen a magazine like Stereophile actually measure something like this or (gasp!) objectively compare the results side-by-side with a stock unit. I believe part of Stereophile not reviewing something like this is because there is no dealership network for these audiophile "exotics". Likewise, I have yet to see objective results posted on message forums though many have verbally expressed opinions.

When I heard the device playing through my friend's system, I was certainly not impressed. Pushing up the volume, I thought the noise level was a bit too high with audible hum through his Pass Labs amp and preamp. I remembered the stock BDP-105 was dead silent at my other friend's place.

Playing music, it sounded okay with acoustic vocals, live recordings, and classic jazz from the 50-70's that my friend was mainly fond of. Albums like Kind of Blue, or Saxophone Colossus, or Way Out West or Waltz for Debby seemed okay if perhaps a tad on the "warm" side. Since I had not heard Verity Audio speakers before, I could not say how much of the warmth could be due to the speakers themselves. But once you throw in some Rolling Stones, AC/DC, electronica, or pop albums, it just didn't seem to be able to deliver the impact or dynamics I expected - a little too "polite" for rock, a bit too "lean". My friend is an LP fan and his vinyl playback sounded nicer than this fancy modded BDP-105 on the whole. Something was certainly different with the output here and I would not say it was an improvement over the stock sound!

So, did objective measurements give a hint as to what was different? Yes, of course; IMO any audible difference of this magnitude will show something... Check out these RightMark results (RCA stereo output). Remember, I'm measuring with my old E-MU ADC so don't compare these directly with my recent UDP-205 results although correlations are there:

Wow. Those modded results are ugly for an Oppo. To the left are the 16/44 results and on the right are the 24/96 results comparing the stock from the "tube mod" Oppo. I repeated the 24/96 test 3 times and received essentially identical results. Numerically the frequency response did not appear to be flat. Furthermore, the modded device could not muster better than 12-bits of resolution. Simply nasty distortion results with THD and IMD+N between about 0.3-0.6% using RightMark's methodology (1kHz tone for THD, 60Hz & 7kHz for IMD) - those kinds of distortion numbers may be okay for vinyl playback, but simply unacceptable for any decent DAC especially if one wants to reproduce high resolution!

So how ugly does all this look?

16/44.1:

24/96:

Yeah, ugly. Frequency response no longer flat with a rather bass-shy -9dB dip at 30Hz - I guess if you put your speakers too close to walls or in a bass-heavy room this could even things out. Not unusual to see 20dB swings in room responses in the low frequencies. I'm quite sure there was no EQ or bass management active in the Oppo settings to roll-off the bass like that. There's also a poor noise floor with a big amount of 60Hz hum, obviously elevated stereo crosstalk and IMD+N sweeps. Notice that the "tube" Oppo could not keep up with a 24/96 test signal. Sure, the frequency response extended beyond 22.05kHz, but obviously the noise floor was not improved. This modded device is clearly incapable of benefiting from 24-bit resolution.

For completeness, since I grabbed the data, here are some 24/192 graphs also:

A bit more ultrasonic roll-off with the tube machine (interesting but not significant). Notice in the IMD graph the slightly higher even order 60Hz harmonic sequence, and a reasonable decay in high-order (audibly offensive) harmonics.

I told my friend about these findings and in fact sent him some graphs... But having had our debates about the importance of objective analysis and knowing that he was of a strongly "subjectivist" philosophical mindset at the time, I'm not sure if he took them seriously. The next time I met him more than a year later, he told me he had sold the modded Oppo and moved on to another DAC. His impression was that the mod did not live up to his expectations and this reaffirmed to him the superiority of vinyl playback. We never spoke about this device thereafter (he had upgraded his speakers as well to IMO much nicer full-range floorstanders).

Those results have "haunted" (if one can even use that word for something like this) me ever since. Was there something especially bad about the RCA output (my friend mainly used the XLR out so maybe it wasn't as terrible)? Did his system pick up a nasty ground loop or was this exacerbated by the measurement hook-up? Was there some kind of fault in the mod installation or component failure (by the way, the anomalies were about the same in both channels)? I don't know what tubes were used - could they have been bad? IMO, unlikely that any "tube rolling" would improve the results much. As you can imagine with this player sitting on top of Synergistic's incomprehensible "Tranquility Base", his system at the time was decked out with the tweaks du jour including a host of expensive cables and power products - did any of those things affect the sound / measurements? Alas, I did not have time to better evaluate the device and given he had since sold it, there was no way to go and test it out again.

From what I have seen, the modding company doesn't provide any measurements on the website to compare and understand if what I found was (hopefully) evidence of my error, or faulty equipment rather than "normal" or by design.

If this is what's going on after the tube mod, especially with the elevated noise level, one must simply question what is the point of spending money on a linear power supply when certainly a noise-free switching power supply already allows much lower noise performance. Even more obvious - would not a $5 power cable and inexpensive interconnects do just fine for a device with this level of performance?! (This is of course assuming one believes expensive cables do anything of value - I don't.) What's the point of spending any more money when the technical ability of the device is so diminished?

Ultimately whether one supports expensive mods, I think it's obvious that objective results will be affected - not likely in a technically good way given the excellent results of the BDP-105 already. While I see a number of testimonies online that mods like this one resulted in "better" sound based on subjective evaluation, I certainly beg to differ on that opinion based on what I heard and measured that day. As I alluded to above, I cannot be clear if what I found was simply a faulty device showing IMO an unacceptable frequency response and amount of hum on the RCA output; an issue I have never otherwise found with the devices I have tested over the years. In fact, the only other audio device I found with such an early roll-off in the bass frequencies was the Gigabyte GA-Z170X-Gaming 7 computer motherboard (not exactly a hi-fi machine!). Remember, despite that poor noise floor compared to other digital devices, LP's put up a rather poor performance as well (usually around 60-80dB of dynamic range) which makes this modded BDP-105 about equivalent.

The bottom line I think is that if one wants to spend any money on product modifications, much less significant dollars that clearly will void the original manufacturer's warranty and irreversibly alter the stock device's sound...

1. Make sure to ask the mod designer before hand to show you what some basic measurements look like. What kind of frequency response is he (likely not a she) aiming for? How is the noise floor after the mod? Can it still benefit from a hi-res signal? Can the company give you some comparisons between stock and modded performance specs and graphs? Surely this should be part of the "informed consent" process should a customer request more information.

2. Ask how the modder performs quality control. This is certainly not a large volume product so I assume there can be comprehensive quality control done to make sure things are working right! Furthermore, being hand-assembled, I assume there must be some significant variability between devices; maybe even slight differences in the parts used over the months. Does the company then put each completed machine through some kind of test? Are there measurements with quality control testing or is it purely from listening? Does the modder retain records of the tests and measurements in the event you think there's something wrong and need to make sure an issue as common as damage during shipping did not occur? Remember, for many reputable companies selling expensive products, they will test each individual device off the assembly line. For example, you can ask Sennheiser for the frequency response certificate for your specific HD800 headphone.

3. I would encourage reviewers to do objective testing and show us some results, especially for "mods" and devices where there are no distribution networks or showrooms for the consumer to listen for themselves. Honestly folks, if all we see are reviews and glossy product pictures from typically complementary magazines for expensive products or online sites selling advertising space, I think customers should remain cautious. I see many reviews out there of "tube modding" based on various kinds of devices over the years... Isn't it a little suspicious when none of these wordy reviews bother to even show a single graph or demonstrate distortions added? Are audiophile reviews nothing more than a little bit of entertainment as writers hope to capture the confidence of their readership, requiring them to believe that the writer is unbiased, is immune to placebo effects, has "golden ears" with precise acuity and fantastic auditory memory to make comparisons over time? Ultimately, does one always need to spend money to "try it out" since there is no attempt made to engage the reader intellectually to demonstrate the effect; what a device or mod truly is and does? IMO, unless reviewers "up the game", obviously, rational audiophiles with even a modicum of insight should be asking the question "Is this snake oil?"

LOL. I see there's even some competition among the modders. Check out this mod outfit criticizing their competitor's tube mod. They claim that "the original mod was pretty weak and did not sound that good" and "the signal wiring in this machine for the tube circuitry was double stranded and sounded terrible". Good grief... Who to believe!? Such is the nature of testimony. Thankfully in the world of well-engineered products, we can consider the facts through measurements and confirm through listening as well.

Check out this web page for even more fun with BDP-105 mods; this time utilizing "quantum stickers" among other things. Yeah. Seriously "fun".

If someone is going to take hundreds (even thousands) off your hands to supposedly provide a "high end" experience, it would at least be nice to check if the device will still be "hi-fi" once it's bought and paid for, right?!

So ends our story. Good luck friends... BTW, if you have access to a modded machine, I would love to see what your measurements look like!

--------------------------

As I mentioned a couple weeks back, there were issues with Oppo's beta MQA DAC firmware for the UDP-205 and the muting of the XLR output. Nice to see that this has been fixed with their official June 29th build "UDP20X-60-0625". Remember to update the USB DAC to "USB-0118" as well for MQA through the rear USB-B port. At this point I haven't done any tests but everything seems to work well. It's decoding 16/44.1 MQA-CD FLAC rips, as well as 24/44.1 and 24/48 typical MQA streams as through Tidal. Tidal users, have fun streaming MQA to the UDP-205...

I am curious about the "Dolby Vision" mode in HDR Setting that apparently converts video to Dolby Vision (from HDR10?) when the system allows. Kudos to Oppo for continuing to update the device and bring out new features (please consider the suggestions discussed last week :-).

I see that Oppo is now starting to contact folks over E-mail who have signed up to order the last UDP-205 batch(es). Still a damn shame that Oppo Digital has decided to shut down...

Happy Canada Day and Independence Day.

Have a wonderful summer!