Can you hear a difference between 2 solid state preamps?

[PMA]

New music samples

Also found the recording to be a curious choice for a 'technical' comparison, but I must admit I enjoy the music very much.

Yes, I understand you completely, but I had some good reasons to pick that 1953 Sinatra recording
====================

Anyway, for those who might be interested, there are new samples

http://pmacura.cz/claire.zip

Now, it is a contemporary recording issued by a highly regarded Glasgow based record label. To make it not so easy, there are 3 sample recordings now. The same preamps were used in the test as for Sinatra files. Please comment and bring your sound impressions of the new samples.

 

[PMA]

I don't think Samuel would mind me showing this excerpt. The text made it clear that there was some randomizing of the gain ranging going on, as you can see each cycle can be different. It is possible (I am speculating) that using a differential circuit the glitches can be made to cancel in some way. I have no idea without looking at the internal signals. Op-amp rolling is not the way to approach these real engineering problems.

Thanks for bringing this, Scott.

 

[PMA]

Some info on the new samples:

foobar2000 1.4.8 / Dynamic Range Meter 1.1.1
log date: 2019-11-08 09:12:34
--------------------------------------------------------------------------------
Statistics for: ?-claire1
Number of samples: 7009187
Duration: 1:13 
--------------------------------------------------------------------------------
                 Left              Right
Peak Value:     -0.16 dB   
Avg RMS:       -12.54 dB   
DR channel:     10.82 dB   
--------------------------------------------------------------------------------
Official DR Value: DR11

Samplerate:        96000 Hz
Channels:          1
Bits per sample:   24
Bitrate:           1462 kbps
Codec:             FLAC
================================================================================

foobar2000 1.4.8 / Dynamic Range Meter 1.1.1
log date: 2019-11-08 09:12:38
--------------------------------------------------------------------------------
Statistics for: ?-claire2
Number of samples: 7009187
Duration: 1:13 
--------------------------------------------------------------------------------
                 Left              Right
Peak Value:     -0.16 dB   
Avg RMS:       -12.55 dB   
DR channel:     10.81 dB   
--------------------------------------------------------------------------------
Official DR Value: DR11

Samplerate:        96000 Hz
Channels:          1
Bits per sample:   24
Bitrate:           1462 kbps
Codec:             FLAC
================================================================================

foobar2000 1.4.8 / Dynamic Range Meter 1.1.1
log date: 2019-11-08 09:12:48
--------------------------------------------------------------------------------
Statistics for: ?-claire3
Number of samples: 7009188
Duration: 1:13 
--------------------------------------------------------------------------------
                 Left              Right
Peak Value:     -0.16 dB   
Avg RMS:       -12.56 dB   
DR channel:     10.79 dB   
--------------------------------------------------------------------------------
Official DR Value: DR11

Samplerate:        96000 Hz
Channels:          1
Bits per sample:   24
Bitrate:           1461 kbps
Codec:             FLAC
================================================================================

None of the 2 samples are identical.

 

[billshurv]

It saddens me that even Linn are releasing stuff with DR less than I consider acceptable. I will be interested to listen to these if I get a quiet couple of minutes (big if).

 

[PMA]

Pavel, can you post detail pics of construction of both line stages please ?.

Dan.

Sorry Dan, a) I do not want to misuse your talent to expertise internal wiring and construction of the audio amplifiers, b) maybe I would like to keep something for myself

On the other hand, I would greatly appreciate if you could give a listen to the new music samples and told your impressions, it might be also possible to tell which of the samples belongs to #1 and #2 preamplifiers? The preamps used are the same as in the worldstring Sinatra test.

 

[PMA]

It saddens me that even Linn are releasing stuff with DR less than I consider acceptable. I will be interested to listen to these if I get a quiet couple of minutes (big if).

Yeah, even they do, but still quite listenable. Thank you Bill and your time will be appreciated.

 

[jan.didden]

Do you think that the AP does not use opamps? Really, the suspicion of AD797 hump was funny and yes, I got irritated a bit. There are still not many parts, if any, to beat AD797 parameters in audio band. Hats off, Scott.

My AP2722 inside looks like someone emptied a bag of 797's in it ;-)
AP uses a 47pF neutralisation cap in all places I have checked.

Jan

 

[Tournesol]

the worldstring Sinatra test.

Yes "I’ve Got The World On A String", of Sinatra.
I have both of them, sorry for the stupid confusion (search made in my computer by the song title) . I apologize.
I was thinking at Harry Connick Jr. at the moment I was writing, because, in the same spirit, it is so much better produced on a technical point of view:
Just listen to this: one of the best big band Jazz recording I know:
YouTube

May-be, with a good recording (This record "France I wish you love" is a master piece), some could change their preferences with a revealing system that enlight quality and defaults of the source.

 

[PMA]

AP uses a 47pF neutralisation cap in all places I have checked.

Jan

Understood, and discussed with Scott few times, AFAIK.

 

[Max Headroom]

Sorry Dan, a) I do not want to misuse your talent to expertise internal wiring and construction of the audio amplifiers, b) maybe I would like to keep something for myself

Don't be shy.

On the other hand, I would greatly appreciate if you could give a listen to the new music samples and told your impressions, it might be also possible to tell which of the samples belongs to #1 and #2 preamplifiers? The preamps used are the same as in the worldstring Sinatra test.

Ok, in the next day or two.
I had another quick listen today to the WS files and I might revise my initial findings.
I am currently healing from fractured L3 vertebra (wearing corset, I may have to get a rigid back brace) since three weeks now so I'm in a lot of pain.....and without meds, don't like them.
So, I'm saying my ears are different lately to what they usually are and the pain is a distraction and modifier, stay tuned and I let you know what I find.


Dan.

 

[PMA]

These new files
http://pmacura.cz/claire.zip

not only that they have very different technical quality compared to the vintage 1953 Sinatra WS recordings, but I also changed the DAC for the one that has also balanced output. THD vs. amplitude record of the complete DAC - ADC loop with preamps #1 and #2 inserted is attached, same level settings as used for Claire sample recordings. Now we can see both lower noise and distortion with #1 preamp. Frequency response and rise time remains unchanged.

P.S.: I am proud D) how well matched are the levels in claire1-3 samples, taking into account that all setting is made only by an analog pot without digital level equalization.

 

[scottjoplin]

I'm guessing that Dan didn't listen to the samples through headphone. I think I have mentioned to him to change his amplifier.

At some point I will listen through my stereo, see if it has a smoothing effect.

 

[scottjoplin]

Clearly different in initial impressions.

I wish I had ABX to see if there was a consistently recognizable difference.

Also found the recording to be a curious choice for a 'technical' comparison, but I must admit I enjoy the music very much.

Do you have a preference?

 

[Jakob2]

@PMA,

if you would use the newer Foobar version, then in the trial logs the erroneous conclusion "probability that you were guessing" would be exchanged for the correct one "probability to get your result by random guessing" .

Of course you could do the edit each time yourself, but it is more tiresome and might be forgotten sometimes...

 

[Soldermizer]

Has anyone here (or indeed, anywhere) ever tried the null test as suggested by Winer ("The Audio Expert" book) and others? This would seem a much easier test to do compared to anything involving human listening. I don't have the equipment or knowledge to do the test. I do, however, know about the unreliability of human perception and the result that 99% of equipment differences heard are complete and utter bull---t

 

[edbarx]

The problem with this question, is the inherent assumption, that playback of these audio files, does not add anything of its own. To answer this question objectively, playback, amplification and electronic current to sound conversion must be of exceptionally high quality. Moreover, there is also the issue of the human listener. Can the latter, be assumed to have an absolutely balanced and accurate sound perception?

 

[PMA]

@PMA,

if you would use the newer Foobar version, then in the trial logs the erroneous conclusion "probability that you were guessing" would be exchanged for the correct one "probability to get your result by random guessing" .

Of course you could do the edit each time yourself, but it is more tiresome and might be forgotten sometimes...

Thank you, I will check it. I was re-installing the ABX plugin recently, so it is an interesting news.

 

[PMA]

Has anyone here (or indeed, anywhere) ever tried the null test as suggested by Winer ("The Audio Expert" book) and others? This would seem a much easier test to do compared to anything involving human listening. I don't have the equipment or knowledge to do the test. I do, however, know about the unreliability of human perception and the result that 99% of equipment differences heard are complete and utter bull---t

The problem with this question, is the inherent assumption, that playback of these audio files, does not add anything of its own. To answer this question objectively, playback, amplification and electronic current to sound conversion must be of exceptionally high quality. Moreover, there is also the issue of the human listener. Can the latter, be assumed to have an absolutely balanced and accurate sound perception?

Yes, the null test was tried. And also I made several tests against original data, without loops or anything inserted, they can be searched in "Everything Else" forum.

In this current test, I am rather interested in listeners subjective opinion on sound differences of the samples (and I am not going to criticise it) rather than insisting on ABX results, as I do in other tests. Of course any ABX protocol is appreciated, but not necessarily requested now.

 

[mountainman bob]

I applaud your interest in the actual sound/perception........if people can be taught how to listen then maybe some schooling is in order.

I suppose live sound engineering or recording engineering would hit on that mark?

Edit.....I wonder if local tech or community colleges offer introductory courses in such things.....or maybe even online?

 

[PMA]

Of course you could do the edit each time yourself, but it is more tiresome and might be forgotten sometimes...

It is impossible, because when you change the txt protocol, the signature check tool

foobar2000 ABX Log Signature Tool

will announce that the signature is invalid. It is a good tool to reduce cheating.