• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Listening Comparison Test: DSP Phono vs SUT

OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
It might be worth left SUT/right direct using sweeps and also the reverse.

However, you would have been better not to select mono in the Devialet. What I'd hoped to find is if there are channel FR differences. When you switched to mono you made that impossible. So repeating the SUT on both channel sweeps, and Direct on both channel sweeps, but keeping them in stereo would tell us the answer I think. It will let us know if there are differences in the channels anyway. That also likely will let us see if there are FR differences in the two signal paths SUT vs Direct. Finally with music, at upper frequencies the noise may be intruding on the measurements enough to be a false reading if the sweeps look just fine. That could apply to the below 50 hz frequencies if there isn't much music there and we are simply seeing noise mostly there as well. Most mastering mono'd frequencies below 50 hz on LP.

So, for the next rip, just to clarify:

1. Sweep or music?

2. If music, stereo or mono?

3. If mono music, via mono LP or mono button?

4. Would you prefer the dual 1 channel through the SUT, 1 direct? Or something more normal?

P.S. I went it to the Devialet and adjusted the L/R balance for the cartridge, which is a pain in the butt because you can't do it on the fly. You have to eject the SD cards, go to a computer, stick them in, go the website configuration, create a new config, download it, go back to Devialet, reboot it. Total PITA. They need to add a web server to it, given it has a network connection. Anyway....it's not perfectly balanced, but it's closer than it was (dropped left, added to right).
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,759
Likes
37,612
So, for the next rip, just to clarify:

1. Sweep or music?

2. If music, stereo or mono?

3. If mono music, via mono LP or mono button?

4. Would you prefer the dual 1 channel through the SUT, 1 direct? Or something more normal?

P.S. I went it to the Devialet and adjusted the L/R balance for the cartridge, which is a pain in the butt because you can't do it on the fly. You have to eject the SD cards, go to a computer, stick them in, go the website configuration, create a new config, download it, go back to Devialet, reboot it. Total PITA. They need to add a web server to it, given it has a network connection. Anyway....it's not perfectly balanced, but it's closer than it was (dropped left, added to right).
1. I'm greedy I suppose, but one sweep and one of music.

2. Mono music is more useful for comparing what is happening.

3. Don't engage the mono button. So use mono disc.

4. I think the music with one channel via SUT and one via direct is useful. On the sweep, one with both thru SUT and another with both thru Direct would be the most informative. And actually one sweep with one channel SUT and one Direct. So that would be four recordings.

A mono music through one SUT and one Direct channel. You don't need more than 3 minutes of music.

A sweep with both channels SUT and another sweep with both channels Direct.

Followed by a sweep with one channel SUT and one Direct.

Maybe @RayDunzl has some input on it too.
 
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
1. I'm greedy I suppose, but one sweep and one of music.

2. Mono music is more useful for comparing what is happening.

3. Don't engage the mono button. So use mono disc.

4. I think the music with one channel via SUT and one via direct is useful. On the sweep, one with both thru SUT and another with both thru Direct would be the most informative. And actually one sweep with one channel SUT and one Direct. So that would be four recordings.

A mono music through one SUT and one Direct channel. You don't need more than 3 minutes of music.

A sweep with both channels SUT and another sweep with both channels Direct.

Followed by a sweep with one channel SUT and one Direct.

Maybe @RayDunzl has some input on it too.

Caveat: I don't think I have a mono sweep disk.

Plenty of mono music.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,759
Likes
37,612
Caveat: I don't think I have a mono sweep disk.

Plenty of mono music.
Okay, I'm assuming the sweep is a mono signal even on a stereo test disk, but either way the sweep is useful.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,250
Likes
17,191
Location
Riverview FL
So, for the next rip, just to clarify:


For me, and REW, two tone sweeps.

No "adjustments" at the Devialet. Just normal stereo. Don't try to match levels.


First sweep recording:

1564810944210.png


Second sweep recording, move the SUT to the left channel.



1564811329371.png


Easy.


Caveat: I don't think I have a mono sweep disk.

The "stereo" sweep tone should be "mono" - in that it has the same signal cut into both channels.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,759
Likes
37,612
For me, and REW, two tone sweeps.

No "adjustments" at the Devialet. Just normal stereo. Don't try to match levels.


First sweep recording:

View attachment 30568

Second sweep recording, move the SUT to the left channel.



View attachment 30569

Easy.




The "stereo" sweep tone should be "mono" - in that it has the same signal cut into both channels.
Why don't you do it as Ray described and let us see what that shows. It will be useful, and perhaps we'll not gain anything doing every possible permutation.
 
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
For me, and REW, two tone sweeps.

No "adjustments" at the Devialet. Just normal stereo. Don't try to match levels.


First sweep recording:

View attachment 30568

Second sweep recording, move the SUT to the left channel.



View attachment 30569

Easy.




The "stereo" sweep tone should be "mono" - in that it has the same signal cut into both channels.
Why don't you do it as Ray described and let us see what that shows. It will be useful, and perhaps we'll not gain anything doing every possible permutation.

I love Ray's drawings.....not just here, but generally. Very helpful. Our artist in residence!

Okay, here are the new rips.

WARNING:

1. There are 1 khz lead-in tones at the beginning of the tracks.
2. I didn't undo the cartridge balance adjustment attempts I made last night. So the channels would normally be better (but not perfectly) balanced than before...but of course, the gain differences between SUT vs direct will pretty much swamp all that.
3. As a reminder, the side that is bypassing the SUT will be be loaded completely wrong, which should probably lead to a giant spike in the treble.
4. I can't hear the 15khz-20khz end of the track, so have to go by visual inspection of the cart in the groove, the time stamp, and the meters to make sure I didn't cut it off prematurely before hitting the run-out groove. I think I got it right, but maybe I didn't.

Right SUT, Left Direct:

https://drive.google.com/open?id=1wX5G-qHXs1GZRoOtrO3D_eZlsN0S9hUK

Left SUT, Right Direct:

https://drive.google.com/open?id=1Y6q872DFGLrhPNECQV-3hgvUeo1x43MO
 
Last edited:

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,250
Likes
17,191
Location
Riverview FL
I love Ray's drawings.....not just here, but generally. Very helpful. Our artist in residence!

Okay, here are the new rips.

WARNING:

1. There are 1 khz lead-in tones at the beginning of the tracks.
2. I didn't undo the cartridge balance adjustment attempts I made last night. So the channels would normally be better (but not perfectly) balanced than before...but of course, the gain differences between SUT vs direct will pretty much swamp all that.
3. As a reminder, the side that is bypassing the SUT will be be loaded completely wrong, which should probably lead to a giant spike in the treble.
4. I can't hear the 15khz-20khz end of the track, so have to go by visual inspection of the cart in the groove, the time stamp, and the meters to make sure I didn't cut it off prematurely before hitting the run-out groove. I think I got it right, but maybe I didn't.

Right SUT, Left Direct:

https://drive.google.com/open?id=1wX5G-qHXs1GZRoOtrO3D_eZlsN0S9hUK

Left SUT, Right Direct:

https://drive.google.com/open?id=1Y6q872DFGLrhPNECQV-3hgvUeo1x43MO


---

The files have to be "shortened" - REW will only anallyz 87.4 seconds of data.
Audacity "Change Tempo" tool used to shorten the files, with best conversion and no dither.

SUT is around 23dB "louder"

There's a little hump in the right channel above 6kHz

Left Channel orange and red
Right Channel blue and green

Raw Data, 1/48 octave smoothing:

1564867211352.png


1/6 octave smoothing

1564867687255.png


Levels "normalized" at 1kHz relative to Left SUT (red trace)

Left SUT - add 0dB
Right direct - add 22.4dB
Right SUT - add 23.1dB
Left Direct - subtract 0.7dB

Looks like a .7dB imbalance in channels from whatever cause, either cartridge or Devialet

1564868267633.png


Normalized at 1kHz and 1/6th octave smoothing:

1564869258795.png


Zoom - 20-20kHz 1/48 smoothing

1564868585644.png


Zoom - 2kHz to 20kHz 1/48

1564868658076.png


Don't know what you guys are hearing as a difference, at least in terms of frequency response.
 
Last edited:
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
First off, thanks again for your work. I really appreciate it


SUT is around 23dB "louder"

Hypothetically, it should be 24 dB of gain, so with some vagaries in cart output, resistance losses, etc, that sounds on target.

There's a little hump in the right channel above 6kHz

I would have expected the way-out-of-spec 47k loading for direct to have the treble spike show up on the direct channels. But it seems to be MIA.

Hmmm.....


Left Direct - subtract 0.7dB

Looks like a .7dB imbalance in channels from whatever cause, either cartridge or Devialet

I'm going to blame the cart, because it's consistent with the other data and the corrective measures I tried to take in the cartridge config (cut left by 20%, boost right by 20%). It looks like it's still off a hair, but a lot better than the 5 dB gap from before.

I'll probably just leave it as-is, because that's probably tighter than any of the tolerances for any of my headphones or speakers, not to mention room interactions.

Don't know what you guys are hearing as a difference, at least in terms of frequency response.

Me, either.

So here are some hypotheses:

1. Following the manufacturers directions, I broke-in the SUT over a few over nights, trying to accumulate 48 hours of time, by pumping in regular line level signals into the *output*. Maybe that saturated or magnetized the coils in some way for a while, and now it's faded.

2. Maybe it wasn't done being broken in (if that's even really a thing).

3. I dropped the SUT a few times. Maybe that goofed it up temporarily in some way by mashing things together.

4. The sweeps were run with speakers off. Music was recorded with speakers on. Maybe the in room vibrations affected things. I could almost hypothesize that for the bass sounding less tight or bloomier....but I can't really see how that could affect the treble.

5. Distortion components that don't show up by just looking at FR.


Next steps:

Maybe I record a new track of music, normally (i.e. both channels to SUT vs both channels direct), leave the file label names mysterious, and see if people can tell the difference in an ABX test?
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,759
Likes
37,612
I didn't find anything to add to what Ray has already shown. I did find a kind of "hey I wonder why it does that" oddity.

In the left regardless of whether SUT or Direct, you get sidebands 250 hz above and below the main tone in the left channel. They are about 40 db below the signal. In the right channel you get a single lower sideband 300 hz below the main tone. Looked at it with 3 different softwares, with FFT and spectrograms, and different size FFT and it is there in all of those. Doesn't explain any sound difference.

Here is the left.
1564890445146.png


Now the right. Looks this way in both files and stays in each channel.
1564890591751.png

Sorry should have included the scaling of the spectrogram colors. Here it is and it is the same for both of the above pics.
1564890752017.png
 
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
I didn't find anything to add to what Ray has already shown. I did find a kind of "hey I wonder why it does that" oddity.

In the left regardless of whether SUT or Direct, you get sidebands 250 hz above and below the main tone in the left channel. They are about 40 db below the signal. In the right channel you get a single lower sideband 300 hz below the main tone. Looked at it with 3 different softwares, with FFT and spectrograms, and different size FFT and it is there in all of those. Doesn't explain any sound difference.

Here is the left.
View attachment 30644

Now the right. Looks this way in both files and stays in each channel.
View attachment 30645
Sorry should have included the scaling of the spectrogram colors. Here it is and it is the same for both of the above pics.
View attachment 30646


Wow...that's weird.

At first I was thinking some kind of tonearm resonance....but only in one channel would be bizarre...

So what should we do next?
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,759
Likes
37,612
Wow...that's weird.

At first I was thinking some kind of tonearm resonance....but only in one channel would be bizarre...

So what should we do next?
I like your idea of recording some music with direct and SUT, but don't tell us which is which. You'll need to level match it closely. I'd probably suggest adjusting to the same RMS level for the entire track.
 
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
I like your idea of recording some music with direct and SUT, but don't tell us which is which. You'll need to level match it closely. I'd probably suggest adjusting to the same RMS level for the entire track.

Can Audacity do that?
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,454
Likes
15,806
Location
Oxfordshire
4. The sweeps were run with speakers off. Music was recorded with speakers on. Maybe the in room vibrations affected things. I could almost hypothesize that for the bass sounding less tight or bloomier....but I can't really see how that could affect the treble.
There is always a difference between sound on and sound off. It is not only low frequencies which are added, structure borne and airborne sound excites resonances anywhere. If you have a detachable headshell arm the hf could be headshell resonances being excited, or anywhere else coupled to the cartridge body as well as anything like record modes being excited.
Fixed headshell arms tend to be better than detachable. The SME type with a single draw-pin are susceptible and sensitive to how tight they are and the nature of the washer.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,759
Likes
37,612
Can Audacity do that?
Yes, just select the track in its entirety. Under analyze click Contrast. It will give you the RMS level. Plus it has another box where you can select a different track if you have more than one open in Audacity. And it gives you RMS level of that and the difference in db between them. I'd suggest this will work better than normalize because pops or ticks on an LP can be high enough to corrupt normalizing. I also suggest going in and manually reducing any large ticks or pops like especially the one you get when you sit the need down in the groove.

1564913087414.png
 
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
There is always a difference between sound on and sound off. It is not only low frequencies which are added, structure borne and airborne sound excites resonances anywhere. If you have a detachable headshell arm the hf could be headshell resonances being excited, or anywhere else coupled to the cartridge body as well as anything like record modes being excited.
Fixed headshell arms tend to be better than detachable. The SME type with a single draw-pin are susceptible and sensitive to how tight they are and the nature of the washer.

Yes, my Jelco SA-750D uses detachable headshells in the SME style. I'm using an Audio Technica headshell, which allows for azimuth and overhang adjustment, too, so the possibilities for even more wiggling is there, too.

I 100% agree that speakers on vs not makes a difference.

The question I have: does it make *more of a difference* with the SUT?

Give up on record players ? ;)

And miss out on all this analog mystery hunting and tweakery fun?

My digital front end has nothing for me to do.

It's like a vintage stick shift vs a self-driving car.
 
Last edited:
OP
watchnerd

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,414
Location
Seattle Area, USA
Yes, just select the track in its entirety. Under analyze click Contrast. It will give you the RMS level. Plus it has another box where you can select a different track if you have more than one open in Audacity. And it gives you RMS level of that and the difference in db between them. I'd suggest this will work better than normalize because pops or ticks on an LP can be high enough to corrupt normalizing. I also suggest going in and manually reducing any large ticks or pops like especially the one you get when you sit the need down in the groove.

View attachment 30679

Thanks. I'm thinking of making 4 files.

I could then make a poll amongst the choices.
 
Top Bottom