This is a review and detailed measurements of the audio performance of the JBL SDP-55 AVP home theater processor. It was kindly purchased by a member and drop shipped to me for testing first. It arrived about two months ago. The SDP-55 costs US $6000.
Note: our company Madrona Digital is a Harman (and hence JBL) dealer. Our business though is almost entirely custom installation in commercial and high-end real estate. So we don't do much business with them for typical products I review here (and the company is totally separate from this website). That said, we can source Harman gear and the member indeed purchased this through us. So assign as much bias as you like to this review.
The SDP-55 has a very elegant, modern and serious look:
The menus and graphics are super responsive which is great. The large rotary volume control is nice but a bit too light for my taste. Then again it allows you to rotate it quick and watch the display respond quickly to it.
The back panel shows what a modern AV processor should look like:
16 channels are balanced output are nicely provided and no legacy video inputs and such. Harman acquired Arcam a few years ago and this unit shares a lot with that platform with the exception of Dante digital audio which I did not have the ability to test.
The unit comes with a lit remote control. To my surprise, after putting a pair of Duracell batteries in it, it would not work. I checked the batteries and they were fresh. Still, I tried another pair and it still did not work. On a hunch, I put in the no-name batteries that came with it and it worked with those! The reason? The Duracell batteries' positive tip is shorter than Chinese no-name batteries. With some battery holders, e.g. cheap Chinese LED flashlights, the molding keeps the tip just far away enough to not make a contact with the positive terminal. Well, the same is true here sadly. So best to stuck on no-name batteries or use a different remote. This is the first AV product I have seen with this problem by the way!
The unit comes with a letter saying it was opened and firmware upgraded. So I proceeded to do my testing that way. Once finished, with owner's permission and warning that last time I tried to update an Arcam AVP it hung, I proceeded to update the firmware here from JBL Synthesis website. I put the update on a flash thumb drive and put it in the back. Followed the procedure and update started. It said updated the "Net" and then some USB error popped up and it rebooted to operating form. Knowing that this was not the full update, I followed the instructions once again to update using the network (yes, including setting HDMI and IP to "ON").
As with Arcam, it took a good 2 hours (!) to do its thing, only to hang in the same spot: "LCD writing...." And writing it did for another hour or more.
On Arcam I power cycled the unit and it recovered by restoring the old version. On the SDP-55 it simply said "Erasing LCD...." and that is all it would. Multiple power cycles, disconnecting power cables, etc. did nothing. Search online produced no hits either. On a hunch again, I put the thumb drive back in there and this time, it proceeded to update the LCD reliably!!! It rebooted and all was well with it. A quick dashboard run showed audio performance was the same.
I don't think there is a worse thing than an expensive product bricking itself and that is almost where I was. Looking at long, log list of bugs they addressed, this product from software and testing point of view seems like quite a mess.
There were other issues that had me going mad at first. Every time I would start a new test with my Analyzer, the unit would refuse to produce audio. I was going crazy until I realized if I changed inputs it would produce other again. I had to wait until it fully locked onto the unused input, switch back to HDMI to have it produce audio.
AVP DAC Audio Measurements
When I tested the ARCAM AVP, I went back and forth with the company a few times and in the process narrowed some of the performance issues there. I took advantage of that knowledge to test the AVP. Let's start with Coax S/PDIF input, bypassing any issues with HDMI:
I opted for 3.8 volt output as the next notch up on volume (79) gave me 4.2 volts and reduced performance a bit.
Most of you don't use Coax but HDMI so let's plug in the HDMI cable BUT NOT USE IT AS AN INPUT:
Notice how our SINAD sunk 4 dB. T hat is due to noise level shooting up some 15 db! Again, let me remind you that this is an identical test with Coax as input. All I have done is hook up my PC HDMI output to the SDP-55 UHD input but not selecting. Mere connection of my HDMI output to the SDP-55 causes noise to bleed into the output of the DAC.
Arcam's response to this was that I had created a "ground loop" with my HTPC. Never mind that there is no mains frequency peak that would indicate a ground loop. The noise is broadband and encompasses a range up to a few hundred hertz. They said that they test using a portable HDMI signal generator that is operated on batteries. Naturally that device is generating a cleaner source than my PC. I explained to them that consumers don't use portable HDMI generators as their source and that they need to design their equipment so that it is immune to vagaries of HDMI. They did not respond. And here we are with the JBL version doing exactly the same thing.
Anyway, to make forward progress let's test with HDMI as the input and adjusting the input to give us 4 volts:
Naturally the noise is there still and reduces the performance of the unit to below a number of mass market AVRs:
FYI grounding the SDP-55 to my normally floating Audio Precision balanced inputs (as it should be), reduced the noise a bit and raised SINAD a couple of dB. If you have noise issues in your system, you may want to do the same (I used the pre-out RCA shield for grounding, other places did not work).
There is some good news here in the form of very nice levels of output without clipping:
While you do lose performance above 3.4 volts, it is not too bad up to some 8 volts which should let you drive very low gain amplifiers.
Signal to noise ratio naturally is not that great:
IMD distortion test versus level shows typical design issue with ESS implementation which we have been free of in desktop DACs in the last year:
I had to run the Multitone test using Coax. AP software doesn't have good output buffering and it underruns sometimes as it did here with HDMI causing errors. So this is with Coax input which was very similar to HDMI while it did work:
Very strange shape with high noise floor to boot.
DAC filter performance is poor:
I like to see at least 90 to 100 dB attenuation, not 60 dB. This allows fair bit of out of band signal to be produced by the unit which screws up the performance of the THD+N versus frequency:
We know a lot of that is due to filtering because when I change the sample rate to 192 kHz (green), most of the problem goes away (since out of band signals are outside of the measurement bandwidth of 90 kHz). But even then performance is too low relative to even cheap desktop DACs (dashed blue).
Linearity is rather poor:
Note how Coax did much worse here. The reason is that the device seems to mute its output with signal below -90 dBFS. It would flash the sample rate on the display. I reported this to Arcam as well with no response.
Finally, jitter and noise is very bad on Coax input:
The high level noise floor of the unit should have masked a lot of sins but the problem is bad enough where that did not occur (with HDMI and certainly not with Coax).
Conclusions
Some people think I relish in seeing poor measurements but I have to tell you it is the opposite: it is depressing to see a device fail so much both operationally and from performance of point of view. I usually don't share my work with my wife but in this case I had to express my frustrations to her last night as the thing bricked itself. Ruined my evening and morning for that matter until I figured out how to get it working again.
From performance point of view, there is no question the device lacks proper design verification, more than a year after it was announced. And surely has very low target performance where it blends in with mass market products at much lower price.
There is some good news in there with high output level before clipping -- something we struggle to see in just about any AV processor receiver.
I debated a lot what rating to give the SDP-55: beheaded Panther or "I don't know" one. From functionality and reliability, the beheaded one is definitely deserved and hence its inclusion. On general performance, you have a balanced processor with general SINAD performance that is "not broken." So I gave it up one step up by including the I don't know panther as well.
Personally I would not buy this unit, nor recommend it. But the picking is slim in high-end processors with 16 channels of balanced output so I let you decide if you want to buy it.
-----------
As always, questions, comments, recommendations, etc. are welcome.
Per above I am depressed after doing this review. I am going to drawn out my sorry by going and picking tomatoes in the hot sun from our garden. To the extent you think some money may cheer me up as well, here is how you can donate to the site : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
Note: our company Madrona Digital is a Harman (and hence JBL) dealer. Our business though is almost entirely custom installation in commercial and high-end real estate. So we don't do much business with them for typical products I review here (and the company is totally separate from this website). That said, we can source Harman gear and the member indeed purchased this through us. So assign as much bias as you like to this review.
The SDP-55 has a very elegant, modern and serious look:
The menus and graphics are super responsive which is great. The large rotary volume control is nice but a bit too light for my taste. Then again it allows you to rotate it quick and watch the display respond quickly to it.
The back panel shows what a modern AV processor should look like:
16 channels are balanced output are nicely provided and no legacy video inputs and such. Harman acquired Arcam a few years ago and this unit shares a lot with that platform with the exception of Dante digital audio which I did not have the ability to test.
The unit comes with a lit remote control. To my surprise, after putting a pair of Duracell batteries in it, it would not work. I checked the batteries and they were fresh. Still, I tried another pair and it still did not work. On a hunch, I put in the no-name batteries that came with it and it worked with those! The reason? The Duracell batteries' positive tip is shorter than Chinese no-name batteries. With some battery holders, e.g. cheap Chinese LED flashlights, the molding keeps the tip just far away enough to not make a contact with the positive terminal. Well, the same is true here sadly. So best to stuck on no-name batteries or use a different remote. This is the first AV product I have seen with this problem by the way!
The unit comes with a letter saying it was opened and firmware upgraded. So I proceeded to do my testing that way. Once finished, with owner's permission and warning that last time I tried to update an Arcam AVP it hung, I proceeded to update the firmware here from JBL Synthesis website. I put the update on a flash thumb drive and put it in the back. Followed the procedure and update started. It said updated the "Net" and then some USB error popped up and it rebooted to operating form. Knowing that this was not the full update, I followed the instructions once again to update using the network (yes, including setting HDMI and IP to "ON").
As with Arcam, it took a good 2 hours (!) to do its thing, only to hang in the same spot: "LCD writing...." And writing it did for another hour or more.
On Arcam I power cycled the unit and it recovered by restoring the old version. On the SDP-55 it simply said "Erasing LCD...." and that is all it would. Multiple power cycles, disconnecting power cables, etc. did nothing. Search online produced no hits either. On a hunch again, I put the thumb drive back in there and this time, it proceeded to update the LCD reliably!!! It rebooted and all was well with it. A quick dashboard run showed audio performance was the same.I don't think there is a worse thing than an expensive product bricking itself and that is almost where I was. Looking at long, log list of bugs they addressed, this product from software and testing point of view seems like quite a mess.
There were other issues that had me going mad at first. Every time I would start a new test with my Analyzer, the unit would refuse to produce audio. I was going crazy until I realized if I changed inputs it would produce other again. I had to wait until it fully locked onto the unused input, switch back to HDMI to have it produce audio.
AVP DAC Audio Measurements
When I tested the ARCAM AVP, I went back and forth with the company a few times and in the process narrowed some of the performance issues there. I took advantage of that knowledge to test the AVP. Let's start with Coax S/PDIF input, bypassing any issues with HDMI:
I opted for 3.8 volt output as the next notch up on volume (79) gave me 4.2 volts and reduced performance a bit.
Most of you don't use Coax but HDMI so let's plug in the HDMI cable BUT NOT USE IT AS AN INPUT:
Notice how our SINAD sunk 4 dB. T hat is due to noise level shooting up some 15 db! Again, let me remind you that this is an identical test with Coax as input. All I have done is hook up my PC HDMI output to the SDP-55 UHD input but not selecting. Mere connection of my HDMI output to the SDP-55 causes noise to bleed into the output of the DAC.
Arcam's response to this was that I had created a "ground loop" with my HTPC. Never mind that there is no mains frequency peak that would indicate a ground loop. The noise is broadband and encompasses a range up to a few hundred hertz. They said that they test using a portable HDMI signal generator that is operated on batteries. Naturally that device is generating a cleaner source than my PC. I explained to them that consumers don't use portable HDMI generators as their source and that they need to design their equipment so that it is immune to vagaries of HDMI. They did not respond. And here we are with the JBL version doing exactly the same thing.
Anyway, to make forward progress let's test with HDMI as the input and adjusting the input to give us 4 volts:
Naturally the noise is there still and reduces the performance of the unit to below a number of mass market AVRs:
FYI grounding the SDP-55 to my normally floating Audio Precision balanced inputs (as it should be), reduced the noise a bit and raised SINAD a couple of dB. If you have noise issues in your system, you may want to do the same (I used the pre-out RCA shield for grounding, other places did not work).
There is some good news here in the form of very nice levels of output without clipping:
While you do lose performance above 3.4 volts, it is not too bad up to some 8 volts which should let you drive very low gain amplifiers.
Signal to noise ratio naturally is not that great:
IMD distortion test versus level shows typical design issue with ESS implementation which we have been free of in desktop DACs in the last year:
I had to run the Multitone test using Coax. AP software doesn't have good output buffering and it underruns sometimes as it did here with HDMI causing errors. So this is with Coax input which was very similar to HDMI while it did work:
Very strange shape with high noise floor to boot.
DAC filter performance is poor:
I like to see at least 90 to 100 dB attenuation, not 60 dB. This allows fair bit of out of band signal to be produced by the unit which screws up the performance of the THD+N versus frequency:
We know a lot of that is due to filtering because when I change the sample rate to 192 kHz (green), most of the problem goes away (since out of band signals are outside of the measurement bandwidth of 90 kHz). But even then performance is too low relative to even cheap desktop DACs (dashed blue).
Linearity is rather poor:
Note how Coax did much worse here. The reason is that the device seems to mute its output with signal below -90 dBFS. It would flash the sample rate on the display. I reported this to Arcam as well with no response.
Finally, jitter and noise is very bad on Coax input:
The high level noise floor of the unit should have masked a lot of sins but the problem is bad enough where that did not occur (with HDMI and certainly not with Coax).
Conclusions
Some people think I relish in seeing poor measurements but I have to tell you it is the opposite: it is depressing to see a device fail so much both operationally and from performance of point of view. I usually don't share my work with my wife but in this case I had to express my frustrations to her last night as the thing bricked itself. Ruined my evening and morning for that matter until I figured out how to get it working again.
From performance point of view, there is no question the device lacks proper design verification, more than a year after it was announced. And surely has very low target performance where it blends in with mass market products at much lower price.
There is some good news in there with high output level before clipping -- something we struggle to see in just about any AV processor receiver.
I debated a lot what rating to give the SDP-55: beheaded Panther or "I don't know" one. From functionality and reliability, the beheaded one is definitely deserved and hence its inclusion. On general performance, you have a balanced processor with general SINAD performance that is "not broken." So I gave it up one step up by including the I don't know panther as well.
Personally I would not buy this unit, nor recommend it. But the picking is slim in high-end processors with 16 channels of balanced output so I let you decide if you want to buy it.
-----------
As always, questions, comments, recommendations, etc. are welcome.
Per above I am depressed after doing this review. I am going to drawn out my sorry by going and picking tomatoes in the hot sun from our garden. To the extent you think some money may cheer me up as well, here is how you can donate to the site : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
Last edited:
