Monday, December 13, 2010
The XONO has replaced the GSP combo, and while the comparison has been made slightly closer (i.e., improved bass frequencies), differentiation between systems is still possible. We found that the mid-high frequencies were more clearly represented (increased volume perhaps) with the expensive setup as compared to the less expensive setup. Our next step is to record the less expensive turntable and cartridge through the Steelhead to determine the overall effect of the turntable and cartridge.
Last week David and I met to test out additional audio with the most recently modified setups (i.e., Ortofon Kontrapunkt A on less expensive and Kontrapunkt B on more expensive). In total we tested 4 participants. Using both the Michael Jackson (MJ) clip and a newly recorded Pink Floyd (PF) clip, we were still able to determine a difference between the 2 systems. Interestingly, the preference switched between the clips for all participants. For the MJ clip, those participants that could tell the difference chose the more expensive setup, but for the PF clip, those that could tell a difference preferred the less expensive setup.
As a result of these findings, we have chosen to attempt to replace the GSP combo preamps (Jazz Club and Elevator EXP) with the Pass Labs XONO. We will record the same audio and perform a comparison between the 2 systems as we have done previously. If these systems are also found to be distinct, the next step will be to simply record both outputs of the turntables through the same preamp, to determine the effect of the turntable/cartridge combination.
As a result of these findings, we have chosen to attempt to replace the GSP combo preamps (Jazz Club and Elevator EXP) with the Pass Labs XONO. We will record the same audio and perform a comparison between the 2 systems as we have done previously. If these systems are also found to be distinct, the next step will be to simply record both outputs of the turntables through the same preamp, to determine the effect of the turntable/cartridge combination.
Tuesday, November 16, 2010
11/16/2010
David and I have met and tried out the patch with newly recorded testing material. The test music consisted of excerpts from Michael Jackson's Off the Wall, Dave Grusin's Discovered Again, Prokofiev's Romeo and Juliet, and The Weather Report's Heavy Weather.
The results were that we found difficulty in discerning the difference between both the Prokofiev and Grusin excerpts, while it was considerably easier to identify which of the test trials were comparing similar systems, i.e., Michael Jackson recording from the more expensive system to an second recording of Michael Jackson from the same system. It is our opinion that this discrepancy in results is founded in the repeatable phrases in the excerpts chosen: The strong reoccuring rhythms in the Michael Jackson and Weather Report create a grid within which we were able to repeatably hear and analyze the difference between the two recordings. Moreover, it was easy to use the switch button within a single playback and hear a similar if not the same phrase repeat immediately. This is not to say that the Grusin or the Prokofiev were absent of rhythm! Rather, the rhythmic repetition was either less apparent or on longer time scales.
Now what to do? Its clear that the two systems need to be made closer for this comparison. While they are certainly close, it is our opinion that we must change the components, then retest using these same audio examples. The options available to us for adjusting the systems are as follows:
We already have a duplicate of Dark Side of the Moon. Our plan for this week is to adjust the system setup and retest using the albums of difficulty, then once we are satisfied, also include Dark Side of the Moon.
The results were that we found difficulty in discerning the difference between both the Prokofiev and Grusin excerpts, while it was considerably easier to identify which of the test trials were comparing similar systems, i.e., Michael Jackson recording from the more expensive system to an second recording of Michael Jackson from the same system. It is our opinion that this discrepancy in results is founded in the repeatable phrases in the excerpts chosen: The strong reoccuring rhythms in the Michael Jackson and Weather Report create a grid within which we were able to repeatably hear and analyze the difference between the two recordings. Moreover, it was easy to use the switch button within a single playback and hear a similar if not the same phrase repeat immediately. This is not to say that the Grusin or the Prokofiev were absent of rhythm! Rather, the rhythmic repetition was either less apparent or on longer time scales.
Now what to do? Its clear that the two systems need to be made closer for this comparison. While they are certainly close, it is our opinion that we must change the components, then retest using these same audio examples. The options available to us for adjusting the systems are as follows:
- cartridges: we have already adjusted the cartridge, and in doing so have added a step-up amplifier (because we were using an MM cartridge originally, and now are using a MC cartridge). This change significantly improved the difficulty of the comparison. Perhaps an additional adjustment to another MC cartridge more akin to the Ortofon in use on the expensive system might be an appropriate next step, especially considering that we are using the DL-103, which is on the less expensive side of our cartridge collection (see cartridges here).
- pre-amplifer: before changing cartridges from MM to MC, we tested the comparison of the expensive turntable and cartridge through the GSP combo (Jazz Club and Elevator EXP) and Manley Steelhead using the variable output (set to a similar perceptual volume loudness to the GSP combo). This comparison produced a very difficult comparison.
We already have a duplicate of Dark Side of the Moon. Our plan for this week is to adjust the system setup and retest using the albums of difficulty, then once we are satisfied, also include Dark Side of the Moon.
Monday, November 8, 2010
11/8/2010
Our testing software is now just about ready. As a first step to assure that the patch will in fact behave correctly, we will be conducting the experiment ourselves, using audio recorded from the 2 systems (but not the audio from the actual experiment). If there is a noticeable difference between the 2 systems, we will begin to modify the setup of the less expensive system such that it is more similar to the more expensive one.
Here is a screenshot of the present patch:
The patch opens first for the experimenter to enter the participants ID. Upon entering the ID and hitting OK, the patch in the background (purple) opens in full screen. Once the participant selects OK, the audio begins for the first trial. During the initial playback of both A and B, all buttons and dialogs are hidden from view, leaving only the purple box with the A and B boxes, trial number and help button visible. Once these 2 files have played, the additional functionality appears and the user may use the patch as required (e.g., replay A or B, switch preserving the timing between the 2, and pause). Button presses and selections are recorded and stored with the time from the trial inception. Each block consists of a single musical passage, recorded twice on each system. This creates 4 versions of the passage, which will be compared in 8 trials per block, presented in a double-blind order. The trials are as follows (order is *not* preserved):
Here is a screenshot of the present patch:
The patch opens first for the experimenter to enter the participants ID. Upon entering the ID and hitting OK, the patch in the background (purple) opens in full screen. Once the participant selects OK, the audio begins for the first trial. During the initial playback of both A and B, all buttons and dialogs are hidden from view, leaving only the purple box with the A and B boxes, trial number and help button visible. Once these 2 files have played, the additional functionality appears and the user may use the patch as required (e.g., replay A or B, switch preserving the timing between the 2, and pause). Button presses and selections are recorded and stored with the time from the trial inception. Each block consists of a single musical passage, recorded twice on each system. This creates 4 versions of the passage, which will be compared in 8 trials per block, presented in a double-blind order. The trials are as follows (order is *not* preserved):
- A1, A2
- A2, A1
- B1, B2
- B2, B1
- A1, B1
- B1, A1
- A2, B2
- B2, A2
Tuesday, October 19, 2010
10/19/2010
I have met with David to discuss the consent form, questionnaire and instructions form. We have made the first round of changes to these forms and sent them to Catherine Guastavino for review. Upon return to us, we will send the forms to Ich.
We are now waiting for the MAX patch to be configured for our test. We should receive this within the week, and we can begin the testing process.
We are now waiting for the MAX patch to be configured for our test. We should receive this within the week, and we can begin the testing process.
Monday, October 11, 2010
week of: 10/05/2010
Potential test audio for the project has been provided to project members for review. The four albums are: Pink Floyd: Dark Side of the Moon, Holst: The Planets, The Alman Brothers Band: At Filmore East, and Santana: Abraxas. We will now select from the 4 albums excerpts that will be appropriate for the study. Previously we discussed finding musical examples that express a clear picture of the dynamic range of the recording, space, as well as rich and deep bass, and shimmering in high percussion.
The next stage of the process is to perform a pilot double-blind test to determine if the audio from the 2 tests is similar enough to begin actual testing. As well as taking the test ourselves, we will ask my labmates in the DDMAL to take this test in the next week. At present we are adjusting the MAX patch used in prior studies to work with our current study. Once done, we would like to use a listening room in the MPCL. I will contact Ichiro about this tomorrow.
NOTE: looking back I cannot find anywhere that I have mentioned that we have decided against a comparison of MM vs. MC cartridges. We have replaced the Ortofon OM20 (MM cartridge) with a Denon DL-103 (MC cartridge). We have also added an Elevator EXP as a step-up amplifier. This decision was made in an attempt to find a closer match between the outputs of both systems. After connecting the expensive turntable (w/ expensive cartridge) to the moderately priced GSP Jazz Club and GSP Elevator EXP. The result was that the output of this hybrid system sounded much more similar to the expensive system than the fully moderate system. We then assumed that the only remaining variables were the turntables themselves and the cartridge/step-up amplifier.
Potential test audio for the project has been provided to project members for review. The four albums are: Pink Floyd: Dark Side of the Moon, Holst: The Planets, The Alman Brothers Band: At Filmore East, and Santana: Abraxas. We will now select from the 4 albums excerpts that will be appropriate for the study. Previously we discussed finding musical examples that express a clear picture of the dynamic range of the recording, space, as well as rich and deep bass, and shimmering in high percussion.
The next stage of the process is to perform a pilot double-blind test to determine if the audio from the 2 tests is similar enough to begin actual testing. As well as taking the test ourselves, we will ask my labmates in the DDMAL to take this test in the next week. At present we are adjusting the MAX patch used in prior studies to work with our current study. Once done, we would like to use a listening room in the MPCL. I will contact Ichiro about this tomorrow.
NOTE: looking back I cannot find anywhere that I have mentioned that we have decided against a comparison of MM vs. MC cartridges. We have replaced the Ortofon OM20 (MM cartridge) with a Denon DL-103 (MC cartridge). We have also added an Elevator EXP as a step-up amplifier. This decision was made in an attempt to find a closer match between the outputs of both systems. After connecting the expensive turntable (w/ expensive cartridge) to the moderately priced GSP Jazz Club and GSP Elevator EXP. The result was that the output of this hybrid system sounded much more similar to the expensive system than the fully moderate system. We then assumed that the only remaining variables were the turntables themselves and the cartridge/step-up amplifier.
week of: 9/29/2010
David and I met to record a few records, and adjust the parameters of each system. An immediate observation is that the fixed output of the Steelhead (the preamp from the more expensive system) is much louder than the output of the Jazz Club and Elevator EXP (less expensive preamp+step-up amplifier). As the number of possible parameters to tweak on the Jazz Club and Elevator EXP combination was limited to resistance and capacitance settings (which are going to be held fixed due to the cartridge manual specifications), we have outlined 2 potential adjustments:
Repaired links on MAQ site in recording section here, and changed categories from Classical, Jazz, Rock, Pop to Classical, Jazz, and Contemporary.
David and I met to record a few records, and adjust the parameters of each system. An immediate observation is that the fixed output of the Steelhead (the preamp from the more expensive system) is much louder than the output of the Jazz Club and Elevator EXP (less expensive preamp+step-up amplifier). As the number of possible parameters to tweak on the Jazz Club and Elevator EXP combination was limited to resistance and capacitance settings (which are going to be held fixed due to the cartridge manual specifications), we have outlined 2 potential adjustments:
- use the fixed out of the Jazz Club and Elevator EXP combination as the reference level for the volume setting using the variable out of the Steelhead.
- attach the volume knob to the output of the Jazz Club so that each of the systems have variable output. Following this step, perform a group of experts tests to determine the most agreed-upon volume level for the listening tests.
Repaired links on MAQ site in recording section here, and changed categories from Classical, Jazz, Rock, Pop to Classical, Jazz, and Contemporary.
week of: 9/19/2010
Collected recordings used for developing comparison between 2 systems. Created recordings page on the MAQ site here, and edited the equipment page to have a link to the recordings page here.
Created account for David Weigl, and shared blog information with him. Discussed the AP2700 and various preamplifier tests run previously on this blog.
David and I will meet in the next week to do some vinyl recordings, and adjust the preamp volume levels for the comparison.
Collected recordings used for developing comparison between 2 systems. Created recordings page on the MAQ site here, and edited the equipment page to have a link to the recordings page here.
Created account for David Weigl, and shared blog information with him. Discussed the AP2700 and various preamplifier tests run previously on this blog.
David and I will meet in the next week to do some vinyl recordings, and adjust the preamp volume levels for the comparison.
Meeting: 9/14/2010
This meeting was held to discuss the next stages of our current perceptual test, which is to determine if expert listeners are able to differentiate between 2 vinyl record playback systems of differing price ranges. Each system includes a cartridge, turntable, preamplifier, and interconnect cables. The meeting was attended by Ichiro Fujinaga, Catherine Guastavino, David Weigl, and myself (Jason Hockman). The following are notes take from the meeting:
We will perform a study that will test participants' discrimination through preference testing.
This test will be based on a preference of system 1 or system 2, however subjects will also be tested using a comparison of different recordings of the same system, due to the variability inherent in the analog recording process. More specifically, we feel it necessary to include this "A1-A2" testing because playback of vinyl records can result in clicks and pops which may be found in different places within the digitally recorded timeseries, as well as possibly audible record degradation. Subjects will be informed in the preamble of the experiment that this will be part of the test, so they will not feel deceived by the test setup.
In a previous listening test (see the MP3 study here) started w/ 40 mins - 1 hour of audio material to listen to. Subjects were allowed to listen as many times as they wished, and could move between listening examples. We will adopt a similar structure for our experiment.
So how will we choose the files? is there a particular type of sound we are looking for? how long will the files be? The previous study used 5-12 sec. clips. Our study will instead utilize longer clips, as audiophiles will most likely require longer clips for analysis, but this makes the number of different clips heard overwhelming.
The final decision was to identify from our library 8 audio excerpts, of approximately 15 seconds each.
The participant should be able to switch between the 2 systems with a button. Upon pressing the button, there should be a fade-out of the first track followed by a 200-ms silence buffer and fade-in.
The interface should consist of this A/B switching button, a start (from beginning) button, as well as play, pause. Not discussed in the meeting, but assumed by myself is the need of "stop" and "next sample" buttons.
We will need background and response questionnaires, and receipt forms. Catherine has these forms from the previous MP3 experiment, and she will send these to David.
Timeline:
To do:
This meeting was held to discuss the next stages of our current perceptual test, which is to determine if expert listeners are able to differentiate between 2 vinyl record playback systems of differing price ranges. Each system includes a cartridge, turntable, preamplifier, and interconnect cables. The meeting was attended by Ichiro Fujinaga, Catherine Guastavino, David Weigl, and myself (Jason Hockman). The following are notes take from the meeting:
We will perform a study that will test participants' discrimination through preference testing.
This test will be based on a preference of system 1 or system 2, however subjects will also be tested using a comparison of different recordings of the same system, due to the variability inherent in the analog recording process. More specifically, we feel it necessary to include this "A1-A2" testing because playback of vinyl records can result in clicks and pops which may be found in different places within the digitally recorded timeseries, as well as possibly audible record degradation. Subjects will be informed in the preamble of the experiment that this will be part of the test, so they will not feel deceived by the test setup.
In a previous listening test (see the MP3 study here) started w/ 40 mins - 1 hour of audio material to listen to. Subjects were allowed to listen as many times as they wished, and could move between listening examples. We will adopt a similar structure for our experiment.
So how will we choose the files? is there a particular type of sound we are looking for? how long will the files be? The previous study used 5-12 sec. clips. Our study will instead utilize longer clips, as audiophiles will most likely require longer clips for analysis, but this makes the number of different clips heard overwhelming.
The final decision was to identify from our library 8 audio excerpts, of approximately 15 seconds each.
The participant should be able to switch between the 2 systems with a button. Upon pressing the button, there should be a fade-out of the first track followed by a 200-ms silence buffer and fade-in.
The interface should consist of this A/B switching button, a start (from beginning) button, as well as play, pause. Not discussed in the meeting, but assumed by myself is the need of "stop" and "next sample" buttons.
We will need background and response questionnaires, and receipt forms. Catherine has these forms from the previous MP3 experiment, and she will send these to David.
Timeline:
- recordings and pilot will be done in October
- study will take place in November
To do:
- book headphones
- once headphones are booked, verify the difficulty in differentiating between 2 systems.
- contact Harold about our timeline and requirements for the critical listening room for November
- set David up with a MAQ page account
- make copies of music for Catherine, Ich, and David for selecting musical clips
Monday, August 30, 2010
Both systems have been fully set up, and I have recorded test audio from various records (pop, piano solo, quartet).
The below analysis of the recorded audio was made after consulting our resident expert in perceptual testing. A key point that was emphasized was that we should first set up the 2 systems, and listen to the audio. If we ourselves could tell the difference, then the test is too easy: we should increase the difficulty of the task (i.e., replace components in the signal chain with "better" devices).
My initial impression was that the difference between the 2 systems was quite obvious in 2 regards: overall loudness (advantage: expensive), and brightness (advantage: cheaper). The first of these can be somewhat reduced using the Steelhead front panel controls, reducing the output gain from 55 -> 50 dB, and increasing the load from 50 -> 100 ohms.
Once this was done, samples were re-recorded, and the second of these 2 issues was still a problem. While I have used the term "brightness", it is as much of a HF emphasis that can be heard in the less expensive setup, as it is a mid-range emphasis that is heard in the more expensive setup. Note that I am not making a judgment of the quality of either representation, but rather just a comparison between the 2 systems, which is quite obvious from even the first listening. If we plan to present these 2 versions to listeners, the difference may be too obvious, and perhaps we should be looking for a more apt comparison.
That said, I decided to try to identify the contribution of each component in both systems to the overall sound. First, I recorded samples of the less expensive system using the cables that were originally connecting the more expensive setup. No difference. Next I referred to the RIAA error charts. The Steelhead does have a greater mid-range boost than the GSP preamp. Perhaps this was the cause? So I connected the output of the HR-X to the GSP. The result still sounded like the original output of the more expensive setup. So the only remaining factors are the turntable or cartridge (most likely) itself.
To truly test these, we will need to either swap cartridges on the turntables, or accept that the cartridge comparison that we are making is too discernible.
Audio examples for the above conditions are available upon request.
The below analysis of the recorded audio was made after consulting our resident expert in perceptual testing. A key point that was emphasized was that we should first set up the 2 systems, and listen to the audio. If we ourselves could tell the difference, then the test is too easy: we should increase the difficulty of the task (i.e., replace components in the signal chain with "better" devices).
My initial impression was that the difference between the 2 systems was quite obvious in 2 regards: overall loudness (advantage: expensive), and brightness (advantage: cheaper). The first of these can be somewhat reduced using the Steelhead front panel controls, reducing the output gain from 55 -> 50 dB, and increasing the load from 50 -> 100 ohms.
Once this was done, samples were re-recorded, and the second of these 2 issues was still a problem. While I have used the term "brightness", it is as much of a HF emphasis that can be heard in the less expensive setup, as it is a mid-range emphasis that is heard in the more expensive setup. Note that I am not making a judgment of the quality of either representation, but rather just a comparison between the 2 systems, which is quite obvious from even the first listening. If we plan to present these 2 versions to listeners, the difference may be too obvious, and perhaps we should be looking for a more apt comparison.
That said, I decided to try to identify the contribution of each component in both systems to the overall sound. First, I recorded samples of the less expensive system using the cables that were originally connecting the more expensive setup. No difference. Next I referred to the RIAA error charts. The Steelhead does have a greater mid-range boost than the GSP preamp. Perhaps this was the cause? So I connected the output of the HR-X to the GSP. The result still sounded like the original output of the more expensive setup. So the only remaining factors are the turntable or cartridge (most likely) itself.
To truly test these, we will need to either swap cartridges on the turntables, or accept that the cartridge comparison that we are making is too discernible.
Audio examples for the above conditions are available upon request.
Tuesday, August 3, 2010
We are presently preparing a listening test of 2 full systems--expensive and moderate. These systems are comprised of the following pieces of gear:
According to several reviews of the Kontrapunkt a, optimal sound is achieved after a number of hours of listening (many reviews state 10+ hourse are adequate, and some accounts list >100 hours!), so we are currently working towards this goal (we have also made a recording with the fresh cartridge for a potential later comparison).
More news soon...
- expensive: VPI HR-X turntable, Manley Steelhead Preamp, and Ortofon Kontrapunkt a
- moderate: VPI Aries 2, GSP Jazz Club, and Ortofon OM20
According to several reviews of the Kontrapunkt a, optimal sound is achieved after a number of hours of listening (many reviews state 10+ hourse are adequate, and some accounts list >100 hours!), so we are currently working towards this goal (we have also made a recording with the fresh cartridge for a potential later comparison).
More news soon...
J, Atkinson RIAA error
J. Atkinson of Stereophile has graciously provided us with his code for a RIAA measurement from 10Hz-100kHz. Our current RIAA measure only goes from 20Hz-20kHz because our RIAA deemphasis curve is limited to these values. Thank you Mr. Atkinson! We will test this once back from Utrecht/Athens.
Monday, June 14, 2010
RIAA error MM
The graphs below depict the log-volt distance from an RIAA curve normalized to the output voltage of the MM output of the preamps under test at 1 kHz, provided a 1 mV input signal.
1) Pass Labs XONO
2) Manley Steelhead
3) Bryston BP26
4) GSP Jazzclub
1) Pass Labs XONO
2) Manley Steelhead
3) Bryston BP26
4) GSP Jazzclub
Tuesday, June 8, 2010
RIAA error
next up in the curves is RIAA error. Without much introduction here are the curves (explanation to follow)...
These curves explain the output of each preamplifier in comparison to a reference RIAA curve. The reference RIAA curve was created using an industry standard RIAA 45-point curve provided by Audio Precision. The RIAA curve is then made to reference the 1 kHz output voltage of the first device (Pass Labs XONO, in blue). The reference is then held constant against the other devices, to demonstrate the level output difference as well as the curve disparity.
Much of the steep increases in both the Manley Steelhead and GSP Jazz Club + Elevator can be seen as the result of settling noise from variability in the AP. While this method of RIAA error does not seem to be exactly the same as the one Mr. Atkinson uses in Stereophile, it seems to be a reasonable assessment of the distance from a standard RIAA curve. However, as the measure seems better taken from comparing the freq response to a curve which is normalized to the standard output for a given device under test, the next graphs for MM will be plotted individually. The graph in this post however serves to demonstrate not only the RIAA error, but also the level differences between devices.
These curves explain the output of each preamplifier in comparison to a reference RIAA curve. The reference RIAA curve was created using an industry standard RIAA 45-point curve provided by Audio Precision. The RIAA curve is then made to reference the 1 kHz output voltage of the first device (Pass Labs XONO, in blue). The reference is then held constant against the other devices, to demonstrate the level output difference as well as the curve disparity.
Much of the steep increases in both the Manley Steelhead and GSP Jazz Club + Elevator can be seen as the result of settling noise from variability in the AP. While this method of RIAA error does not seem to be exactly the same as the one Mr. Atkinson uses in Stereophile, it seems to be a reasonable assessment of the distance from a standard RIAA curve. However, as the measure seems better taken from comparing the freq response to a curve which is normalized to the standard output for a given device under test, the next graphs for MM will be plotted individually. The graph in this post however serves to demonstrate not only the RIAA error, but also the level differences between devices.
Thursday, June 3, 2010
delta functions!
Learned how do perform delta functions with the AP. The RIAA error curve can now be displayed with a single waveform showing a preamp's de-emphasis deviation from a standard RIAA curve!
Curves to come! :)
Curves to come! :)
Tuesday, June 1, 2010
Monday, May 10, 2010
site update
Over the weekend I added the cartridges, Bryston BP26, and Cedar Duo Declickle to the equipment page on the project website.
http://coltrane.music.mcgill.ca/MAQ/equipment
http://coltrane.music.mcgill.ca/MAQ/equipment
Monday, May 3, 2010
Frequency response (*w/o additional RIAA curve*) for all 4 preamps. After reviewing the graph from the THD+N v. amplitude (along with the average output of the cartridges which will be used in future stages of this project), a comparison could be made by choosing an input voltage that was within the optimal input range of all the preamps. I therefore chose .5 mVrms. The output level references were chosen as relative measures for each device to produce approximately .5 Vrms from the input.
Below is the frequency response for all four preamps:
Below is the frequency response for all four preamps:
- Blue = Pass Labs XONO
- Red = Manley Steelhead
- Yellow = Bryston BP26
- Green = GSP Audio Jazz Club w/ Elevator (various boost and r/c settings)
THD+N v. amplitude for all 4 preamps. Same color scheme and setup as the previous post (that demonstrated THD+N v. frequency). The tested range was 10uV - 10 mV, sufficent to test the range of possible input values for moving coil cartridges. The XONO and Steelhead had fairly similar traces, while the Bryston showed slightly greater noise and distortion than the XONO and Steelhead until approximately >750 uVrms, at which point the noise and distortion continued to decrease until the test neared its highest test value of 10 mVrms. The GSP setup (Jazzclub and Elevator [used for stepping up the amplification for MC cartridges]) followed a curve that was somewhat in between the others, however with significantly higher noise and distortion until approximately 1.5 mV.
Several trials were run to ensure that the curves were in fact being followed exactly as presented here. The Bryston volume knob setting was as in the previous post (i.e., adjusted until ~ .5 V output was produced from input voltage of .5 mV). The Elevator parameters were chosen to produce the lowest distortion/noise (r=high/c=high). Setting the Elevator to r=medium/c=medium and r=low/c=low resulted in only slightly higher distortion/noise within the entire range, following the same curvature.
As in the previous post:
Several trials were run to ensure that the curves were in fact being followed exactly as presented here. The Bryston volume knob setting was as in the previous post (i.e., adjusted until ~ .5 V output was produced from input voltage of .5 mV). The Elevator parameters were chosen to produce the lowest distortion/noise (r=high/c=high). Setting the Elevator to r=medium/c=medium and r=low/c=low resulted in only slightly higher distortion/noise within the entire range, following the same curvature.
As in the previous post:
- Blue = Pass Labs XONO
- Red = Manley Steelhead
- Yellow = Bryston BP26
- Green = GSP Audio Jazz Club w/ Elevator
Thursday, April 29, 2010
A new approach. One measurement, all preamps.
THD v. Frequency for all four preamps w/ MC selected (input: .5 mVrms)... (in dB). This is somewhat arbitrary, as the Bryston doesn't have a fixed output voltage. For an appropriate level of comparison, I adjusted the volume knob to reach ~ .5 V output (the approximate mean of the others).
THD v. Frequency for all four preamps w/ MC selected (input: .5 mVrms)... (in dB). This is somewhat arbitrary, as the Bryston doesn't have a fixed output voltage. For an appropriate level of comparison, I adjusted the volume knob to reach ~ .5 V output (the approximate mean of the others).
- Blue = Pass Labs XONO
- Red = Manley Steelhead
- Yellow = Bryston BP26
- Green = GSP Audio Jazz Club w/ Elevator
Monday, April 26, 2010
So the next round of testing was performed...
I began with the XONO, with my main aim being to correct the THD+N v. amplitude graphs from the previous tests. This time, the measurement was taken with standardized values for the input voltages (200 uV - 1mV for MC, and 600-5mV for MM) and capacitive loading (100 ohms for MC, and 47 kohms for MM).
This THD+N (MC) looks a lot more like the XONO factory pic. I think my error in the previous one was that the reference voltage wasn't corrected...the range here was 10 uV - 10mV.
This is THD+N vs. frequency for the MC input.
Frequency response for the XONO MC input.
and the phase response.
Finally crosstalk.
The Moving Magnet response was pretty much similar. Below are screenshots of these graphs.
This is the correct THD+N for the MM input. The range is 600 uV - 250 mV. Looks like the factory pic.
The THD+N vs. frequency for the MC input looks just like the MC input.
as does the frequency response...
and phase response.
The cross-talk is a bit different though. The values are about 10 dB lower in the MM.
IMD and idle channel noise for both the MC and MM inputs can be seen in the previous posts with screenshots. Next up tomorrow is the Steelhead.
I began with the XONO, with my main aim being to correct the THD+N v. amplitude graphs from the previous tests. This time, the measurement was taken with standardized values for the input voltages (200 uV - 1mV for MC, and 600-5mV for MM) and capacitive loading (100 ohms for MC, and 47 kohms for MM).
This THD+N (MC) looks a lot more like the XONO factory pic. I think my error in the previous one was that the reference voltage wasn't corrected...the range here was 10 uV - 10mV.
This is THD+N vs. frequency for the MC input.
Frequency response for the XONO MC input.
and the phase response.
Finally crosstalk.The Moving Magnet response was pretty much similar. Below are screenshots of these graphs.
This is the correct THD+N for the MM input. The range is 600 uV - 250 mV. Looks like the factory pic.
The THD+N vs. frequency for the MC input looks just like the MC input.
as does the frequency response...
and phase response.
The cross-talk is a bit different though. The values are about 10 dB lower in the MM.
IMD and idle channel noise for both the MC and MM inputs can be seen in the previous posts with screenshots. Next up tomorrow is the Steelhead.
Saturday, April 24, 2010
new cartridges (unless otherwise noted, output @5.0 cm/s):
(1) Benz Ace H (MC): (2.5 mV @ 3.54 cm/s)
(2) Clearaudio Aurum Beta (MC): 3.5 mV
(3) Denon DL-103 (MC): 0.3 mV
(5) Dynavector 17D2MK2 (MC): 0.26 mV
(7) Dynavector DV-10X5 (MC): 2.5 mV
(6) Dynavector DV-20X (MC): .3 mV
(4) Dynavector XV-1s (MC): 0.3 mV
(11) Grado DJ-100 (2-4 mil stylus) ("Grado hybrid"/MC): 5.0 mV
(9) Grado Prestige 78C ("Grado hybrid"/MC): 5.0 mV
(8) Grado Prestige Gold ("Grado hybrid"/MC): 5.0 mV
(10) Grado Prestige 78E ("Grado hybrid"/MC): 5.0 mV
(12) Ortofon Super OM 30 (MM): 4.0 mV
(13) Ortofon Kontrapunkt Bach (x2) (MC): 0.45 mV
(14) Ortofon Super OM 20 (MC): 4mV
(15) Shelter Model 7000 (MC): (0.55 mV-0.65 mV)
(6) Shelter Model E501 II (MC): 0.4 mV
(4) Sumiko Blue Point Special EVO III (MC): 2.25 mV
(1) Benz Ace H (MC): (2.5 mV @ 3.54 cm/s)
(2) Clearaudio Aurum Beta (MC): 3.5 mV
(3) Denon DL-103 (MC): 0.3 mV
(5) Dynavector 17D2MK2 (MC): 0.26 mV
(7) Dynavector DV-10X5 (MC): 2.5 mV
(6) Dynavector DV-20X (MC): .3 mV
(4) Dynavector XV-1s (MC): 0.3 mV
(11) Grado DJ-100 (2-4 mil stylus) ("Grado hybrid"/MC): 5.0 mV
(9) Grado Prestige 78C ("Grado hybrid"/MC): 5.0 mV
(8) Grado Prestige Gold ("Grado hybrid"/MC): 5.0 mV
(10) Grado Prestige 78E ("Grado hybrid"/MC): 5.0 mV
(12) Ortofon Super OM 30 (MM): 4.0 mV
(13) Ortofon Kontrapunkt Bach (x2) (MC): 0.45 mV
(14) Ortofon Super OM 20 (MC): 4mV
(15) Shelter Model 7000 (MC): (0.55 mV-0.65 mV)
(6) Shelter Model E501 II (MC): 0.4 mV
(4) Sumiko Blue Point Special EVO III (MC): 2.25 mV
Tuesday, April 20, 2010
Things I have learned about THD+N measurements:
The above plot of MC THD+N demonstrates this 10x effect (XONO factory traces are not shown as they are not public images).
Here is the MM THD+N vs. Amplitude. Again, we are seeing a 10-fold increase in distortion+noise. A logarithmic scale would draw out the subtlety of the effect given lower signal.
I intend to retake these measurements when I use the AP again.
- use a logarithmic scale
- let the preamplifier warm up again between measurements
The above plot of MC THD+N demonstrates this 10x effect (XONO factory traces are not shown as they are not public images).
Here is the MM THD+N vs. Amplitude. Again, we are seeing a 10-fold increase in distortion+noise. A logarithmic scale would draw out the subtlety of the effect given lower signal.
I intend to retake these measurements when I use the AP again.
Monday, April 19, 2010
Wednesday, April 14, 2010
Below are some examples of the measurements taken from the experiment explained in the previous post.
Frequency Response
This first figure (above) shows the stereo frequency response of the Steelhead for its MC input (and 400 ohms), with input values of .1, .4 and .6 mV. There is an interesting upwards curve starting at ~4kHz.
In comparison, here (above) is the frequency response of the XONO (@ 475 ohms). The curve exhibited by the Steelhead (although small) is not present here.
2. Phase Response:
Above is the phase response for the Steelhead. Phase response is most basically the ratio of the phase output to the phase input.
For comparison, here is the phase response of the XONO.
3. Idle Channel Noise:
Above is Idle Channel Noise for the Steelhead.
Above is idle channel noise for the XONO. The Steelhead has almost a 10-20 dB lower noise floor than the XONO for much of the auditory spectrum. Also interesting is that this how much louder this is than the idle channel noise for the MM cartridge (see below).
4. Intermodulation Distortion (w/ 2 frequencies: 19 kHz and 20 kHz [both at 0.6mV])
Above is the IMD for the Steelhead.
Notice the peak at 1KHz that is above the otherwise similar noise floor to the idle channel noise above. There are also some interesting things going on at 8 and 16 kHz. Not sure what that is, as it appears in the other measurements of the XONO (such as the idle channel noise for the MM cartridge [see below...]). The 1 kHz tone is caused by non-linearities in the device that cause the input signal to create beat frequencies (largest of which is the often the difference of the 2 signals).
5. Crosstalk vs. Frequency
Here is crosstalk vs. frequency. It is performed by measuring the signal in the an idle channel when sending signal through the opposite channel. Seems strange that we're seeing a few of those bumps, but those were present in both the Steelhead and XONO as is seen below.
The crosstalk for each of these is pretty good. Each line in the above graph is a channel trace of the amount of signal in the empty channel from the one that has signal. The low values of the red curve, and perhaps also the higher values of the blue curve at LOW frequencies could be the result of timeouts that occured during testing. As a sidenote, a timeout occurs with the AP as it records several values, and uses a settling algorithm to provide a final value for the measurement being taken. If the results are too varied, it has difficulty producing the settled value and instead provides an average of the last six values. The low frequency values were marked with the timeout symbol "T" when conducting this test.
Also of note, crosstalk is frequency-dependent: there is a +6dB increase in crosstalk for frequency octave.
Below are selected results for the moving magnet level inputs. Both units were set to 100 pF. The XONO was set to 47 ohms and the Steelhead was set to 50 ohms.
1. Frequency Response
That's a big jump! The above frequency response for the Steelhead demonstrates the output for swept frequencies at input voltages of 5, 7, and 10 mVrms.
The gain increase of the XONO was not as significant. However, the XONO also has a 10dB boost setting which was not tested. Also of interest is the intervals between the 3 input signals tested. Seems like quite a difference between 7 and 10 mVrms for the XONO, but less so for the Steelhead. Its the other interval (5-7 mVrms) that seems to be significant for the Steelhead.
2. Phase Response
The Steelhead MM phase response.
The XONO MM phase response is very similar to its MC response.
3. Idle Channel Noise
The above figure is the Steelhead idle channel noise. Strange peak at 16 kHz.
Notice how the noise floor of the MM setting is lower than the above MC setting. Also notice the peaks at 8 and 16 kHz. Strange.
4. Intermodulation Distortion (IMD)
This is the IMD for the Steelhead. The 1 kHz frequency peak is pretty substantial. The input for the IMD test was again 19 and 20 kHz test tones.
No difference beat frequency (1kHz) for the XONO. Instead take a look at that regime of beat frequencies surrounding the two main peaks. This is because the two frequencies interact not just as the difference, but frequencies f1 and f2 produce new frequencies at (and not limited to) f3 = f1 + f2, f4 = f1 - 2f1, and f5 = f1 + 2f2.
5. Crosstalk vs. Frequency
This is the crosstalk for the Steelhead MM. Looks pretty good at -80 dB, 20-20kHz.
Crosstalk for the XONO seemed better than the Steelhead. Again, as with the MC XONO settings above, this test was plauged by AP2700 timeouts at low frequencies; possibly affecting the difference in channels.
more to come: THD+N vs. Frequency and THD+N vs. Amplitude.
Frequency Response
This first figure (above) shows the stereo frequency response of the Steelhead for its MC input (and 400 ohms), with input values of .1, .4 and .6 mV. There is an interesting upwards curve starting at ~4kHz.
In comparison, here (above) is the frequency response of the XONO (@ 475 ohms). The curve exhibited by the Steelhead (although small) is not present here.
2. Phase Response:
Above is the phase response for the Steelhead. Phase response is most basically the ratio of the phase output to the phase input.
For comparison, here is the phase response of the XONO.
3. Idle Channel Noise:
Above is Idle Channel Noise for the Steelhead.
Above is idle channel noise for the XONO. The Steelhead has almost a 10-20 dB lower noise floor than the XONO for much of the auditory spectrum. Also interesting is that this how much louder this is than the idle channel noise for the MM cartridge (see below).
4. Intermodulation Distortion (w/ 2 frequencies: 19 kHz and 20 kHz [both at 0.6mV])
Above is the IMD for the Steelhead.
Notice the peak at 1KHz that is above the otherwise similar noise floor to the idle channel noise above. There are also some interesting things going on at 8 and 16 kHz. Not sure what that is, as it appears in the other measurements of the XONO (such as the idle channel noise for the MM cartridge [see below...]). The 1 kHz tone is caused by non-linearities in the device that cause the input signal to create beat frequencies (largest of which is the often the difference of the 2 signals).5. Crosstalk vs. Frequency
Here is crosstalk vs. frequency. It is performed by measuring the signal in the an idle channel when sending signal through the opposite channel. Seems strange that we're seeing a few of those bumps, but those were present in both the Steelhead and XONO as is seen below.
The crosstalk for each of these is pretty good. Each line in the above graph is a channel trace of the amount of signal in the empty channel from the one that has signal. The low values of the red curve, and perhaps also the higher values of the blue curve at LOW frequencies could be the result of timeouts that occured during testing. As a sidenote, a timeout occurs with the AP as it records several values, and uses a settling algorithm to provide a final value for the measurement being taken. If the results are too varied, it has difficulty producing the settled value and instead provides an average of the last six values. The low frequency values were marked with the timeout symbol "T" when conducting this test.
Also of note, crosstalk is frequency-dependent: there is a +6dB increase in crosstalk for frequency octave.
Below are selected results for the moving magnet level inputs. Both units were set to 100 pF. The XONO was set to 47 ohms and the Steelhead was set to 50 ohms.
1. Frequency Response
That's a big jump! The above frequency response for the Steelhead demonstrates the output for swept frequencies at input voltages of 5, 7, and 10 mVrms.
The gain increase of the XONO was not as significant. However, the XONO also has a 10dB boost setting which was not tested. Also of interest is the intervals between the 3 input signals tested. Seems like quite a difference between 7 and 10 mVrms for the XONO, but less so for the Steelhead. Its the other interval (5-7 mVrms) that seems to be significant for the Steelhead.2. Phase Response
The Steelhead MM phase response.
The XONO MM phase response is very similar to its MC response.3. Idle Channel Noise
The above figure is the Steelhead idle channel noise. Strange peak at 16 kHz.
Notice how the noise floor of the MM setting is lower than the above MC setting. Also notice the peaks at 8 and 16 kHz. Strange.
4. Intermodulation Distortion (IMD)
This is the IMD for the Steelhead. The 1 kHz frequency peak is pretty substantial. The input for the IMD test was again 19 and 20 kHz test tones.
No difference beat frequency (1kHz) for the XONO. Instead take a look at that regime of beat frequencies surrounding the two main peaks. This is because the two frequencies interact not just as the difference, but frequencies f1 and f2 produce new frequencies at (and not limited to) f3 = f1 + f2, f4 = f1 - 2f1, and f5 = f1 + 2f2.5. Crosstalk vs. Frequency
This is the crosstalk for the Steelhead MM. Looks pretty good at -80 dB, 20-20kHz.
Crosstalk for the XONO seemed better than the Steelhead. Again, as with the MC XONO settings above, this test was plauged by AP2700 timeouts at low frequencies; possibly affecting the difference in channels.more to come: THD+N vs. Frequency and THD+N vs. Amplitude.
So last week (~April 4-10), we conducted the first of several comparisons of audio equipment. This particular test involved 2 preamps: The Manley Steelhead and the Pass Labs Xono. Both are high-quality and expensive amplifiers that have received praise in several reviews.
The point of the experiments is to use features that characterize the devices, towards a better understanding of their (potential) differences in later subjective evaluations.
To measure these preamplifiers, we are using an AP2700. The measurements being used are:
For MC we tested 100 uVrms, 400 uVrms, and 600 uVrms
For MM we tested 5 mVrms, 7 mVrms, 10 mVrms.
These values were determined after consulting the Handbook for Sound Engineers. In future test it has been suggested that we instead use an average output voltage of the cartridges purchased for this project.
Below is a picture of our test setup:
The AP generator was connected to the audio RCA inputs (either MM or MC) on the device under test (DUT), and the output of the DUT was then connected to the audio input of the AP 2700. For cable connections, we chose to use the Kimber Kable, KCAG cables, as in both the cable tests from a couple months ago, and additional informal testing at CIRMMT, we found that these performed (tested using THD+N) as well or better than than others.
Both the XONO and Steelhead have several settings. The Steelhead settings are easily controlled with knobs, while the the XONO settings are controlled by changing jumpers and switches inside the unit itself. For each test then, the switches had to be changed in accordance to the settings we selected for the test.
From top to bottom, the gold pairs on the right of the unit are the 2 possible input terminals (MM and MC) for each channel, followed by the output terminals. The switches to change settings are located directly left of the gold terminals.
The different tests for the Steelhead were as follows:
Gain setting: 55
A = Moving Coil:
A1: loading: 25 Ohms
A2: loading: 50 Ohms
A3: loading: 100 Ohms
A4:loading: 200 Ohms
A5: loading: 400 Ohms
B = Moving Magnet:
B1: loading: 25 Ohms
B2: loading: 50 Ohms
B3: loading: 100 Ohms
B4: loading: 200 Ohms
B5: loading: 47 Ohms
and for the Xono:
XONO test configuration:
high gain (10 dB boost): off
A = Moving Coil:
A1: loading: 10 Ohms
A2: loading: 47 Ohms
A3: loading: 100 Ohms
A4: loading: 475 Ohms
A5: loading: 1000 Ohms
B = Moving Magnet:
B1: loading: 100 PF
B2: loading: 220 PF
B3: loading: 330 PF
The point of the experiments is to use features that characterize the devices, towards a better understanding of their (potential) differences in later subjective evaluations.
To measure these preamplifiers, we are using an AP2700. The measurements being used are:
- frequency response
- phase response
- idle channel noise
- intermodulation (IMD)
- THD+N vs. amplitude
- THD+N vs. frequency
- crosstalk
For MC we tested 100 uVrms, 400 uVrms, and 600 uVrms
For MM we tested 5 mVrms, 7 mVrms, 10 mVrms.
These values were determined after consulting the Handbook for Sound Engineers. In future test it has been suggested that we instead use an average output voltage of the cartridges purchased for this project.
Below is a picture of our test setup:
The AP generator was connected to the audio RCA inputs (either MM or MC) on the device under test (DUT), and the output of the DUT was then connected to the audio input of the AP 2700. For cable connections, we chose to use the Kimber Kable, KCAG cables, as in both the cable tests from a couple months ago, and additional informal testing at CIRMMT, we found that these performed (tested using THD+N) as well or better than than others.
Both the XONO and Steelhead have several settings. The Steelhead settings are easily controlled with knobs, while the the XONO settings are controlled by changing jumpers and switches inside the unit itself. For each test then, the switches had to be changed in accordance to the settings we selected for the test.
From top to bottom, the gold pairs on the right of the unit are the 2 possible input terminals (MM and MC) for each channel, followed by the output terminals. The switches to change settings are located directly left of the gold terminals.
The different tests for the Steelhead were as follows:
Gain setting: 55
A = Moving Coil:
A1: loading: 25 Ohms
A2: loading: 50 Ohms
A3: loading: 100 Ohms
A4:loading: 200 Ohms
A5: loading: 400 Ohms
B = Moving Magnet:
B1: loading: 25 Ohms
B2: loading: 50 Ohms
B3: loading: 100 Ohms
B4: loading: 200 Ohms
B5: loading: 47 Ohms
and for the Xono:
XONO test configuration:
high gain (10 dB boost): off
A = Moving Coil:
A1: loading: 10 Ohms
A2: loading: 47 Ohms
A3: loading: 100 Ohms
A4: loading: 475 Ohms
A5: loading: 1000 Ohms
B = Moving Magnet:
B1: loading: 100 PF
B2: loading: 220 PF
B3: loading: 330 PF
New site for MAQ blog
After wrangling with html every time I wanted to report on the project, I have decided to join the 21st century and just use a tried and true method. So here it is, the new blog.
On this site, you will find the day-by-day trials and tribulations I face in attempting to measure high-quality audio gear for a research project for which I am a research assistant.
For previous information on the work done thus far, please see the old site, that exists here.
On this site, you will find the day-by-day trials and tribulations I face in attempting to measure high-quality audio gear for a research project for which I am a research assistant.
For previous information on the work done thus far, please see the old site, that exists here.
Subscribe to:
Posts (Atom)