• 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!

Take the blind challenge! 300B SET vs. Straight Wire with Gain

Choose ALL of the statements that apply.

  • I prefer #1 (over 3)

    Votes: 20 45.5%
  • I prefer #2 (over 5)

    Votes: 7 15.9%
  • I prefer #3 (over 1)

    Votes: 9 20.5%
  • I prefer #4 (over 6)

    Votes: 22 50.0%
  • I prefer #5 (over 2)

    Votes: 18 40.9%
  • I prefer #6 (over 4)

    Votes: 13 29.5%
  • I hear no difference between 1 and 3

    Votes: 14 31.8%
  • I hear no difference between 2 and 5

    Votes: 18 40.9%
  • I hear no difference between 4 and 6

    Votes: 9 20.5%

  • Total voters
    44
  • Poll closed .
Yes...any objectivist who said "tube sound destroys music" would be wrong. But, I can't recall that being siad by any objectivists very often at all. What I do recall hearing (and saying myself) is "hey if you like that tube sound more power to you but I don't want to use my amp to add a layer of something to the equation even if in some way it might seem to make things sound better." The issue generally that arises with tube amps is that tube amp advocates declare that the sounds coming from their tube amps are "warmer, more musical, more dynamic, and so on and so on." The implication being that the tube amp is providing a better, more complete musical experience than can be achieved via an amp that does notihing other than amplify. As you say yourself, the saturation is built into the recording already. The objectivist position is simply that (as much as is possible) that's what we want to hear - the recording, and only the recording.

But again, what this test shows (to me at least) is that "tube sound" is a pretty iffy proposition anyway.

I think this aligns fairly closely with what I've said on tubes, from my experience (and looking at discussions between people more technically competent than I am).

In the "objectivist" side you can get views that span from "tube sound is a myth" to "tubes distort the sound, why would you ever want to to that?"

And on the subjective side you certainly get what come across as exaggerations in the sonic differences between tube and solid state amps.

My position has always been that *insofar as a tube amp might make an audible difference* it is in the big picture a very subtle one. Audiophiles tend to care about subtle differences, so a subtle difference can be important to an audiophile. But just as the subjectivist may exaggerate the sonic differences, so can the objectivists who decry adding a "distortion generator" at the amp stage. Any character added by a tube amp is generally, from my experience, utterly swamped by the greater sonic characteristics of a recording - there's no gross change of a recording where you are "no longer hearing the artists intent" or whatever.
 
I think this aligns fairly closely with what I've said on tubes, from my experience (and looking at discussions between people more technically competent than I am).

In the "objectivist" side you can get views that span from "tube sound is a myth" to "tubes distort the sound, why would you ever want to to that?"

And on the subjective side you certainly get what come across as exaggerations in the sonic differences between tube and solid state amps.

My position has always been that *insofar as a tube amp might make an audible difference* it is in the big picture a very subtle one. Audiophiles tend to care about subtle differences, so a subtle difference can be important to an audiophile. But just as the subjectivist may exaggerate the sonic differences, so can the objectivists who decry adding a "distortion generator" at the amp stage. Any character added by a tube amp is generally, from my experience, utterly swamped by the greater sonic characteristics of a recording - there's no gross change of a recording where you are "no longer hearing the artists intent" or whatever.

While I have really no prior experience, I take the comments by @computer-audiophile and the detailed measurements and explanation by the OP in the review thread to indicate that I'm getting a taste of the 300B SET sound in these comparison samples, but not the full effect.

Peeking into the spoilered stuff (don't do that if you haven't listened yet) I can see one by @dasdoing that matches my impressions rather closely, among others that describe/prefer perceived differences (and the precise opposite preference set to mine from @Blumlein 88, which also makes sense).

I had a somewhat different impression trying two different sets of reproduction equipment and described same. For those that listened and heard no difference—or anyone really who hasn't done so—does anyone care to note the gear they used for their test?
 
Last edited:
Thanks @GXAlan for the work you've done on the examples.
I think it is very enlightening for many of us that we can still listen to music even with a SINAD of 20.

The recorded examples differ only very slightly, but differences are audible. In my opinion, however, the music material provided is only partially suitable for really working out the pattern of the B300.

SINAD subsumes noise and distortion. I can't hear the strength of the noise with the supplied examples, since none of them contain any quiet, critical passages.

In fact, the distortion of a triode can only be heard at very high levels. Since the B300 produces regularly graded harmonics, this is difficult to detect. On the other hand, noise is audible with the right choice of music, and in my experience, noise is what costs resolution, detail an clarity. Well graded harmonics are less likely to do this.

I think everyone here has a collection of favorite tracks that allows us to focus on certain aspects. Anyway, I always use these tracks when I compare power amps.

My recent candidates have been an Usher Reference R6, a Passlabs XA30.5 and a Krell Evolution 2250e which I currently run.

My first test is always to put my ear to one of my horns to hear the hiss. Unfortunately, I cannot do this test with the provided examples.

So far, in contrast to the Usher and the Pass, the Krell has convinced me a lot. Noise on the horns is not audible and it sounds much cleaner than the two others. The Pass delivers something I call single-ended sound but too much hiss, and the Usher not bad but a bit too snappy. Longer ago I also had some IcePower amps, but they sounded somehow wrong, so I never used them for a longer period. I ordered a NCx500 from Rouge Audio that will arrive next week let's see how it compares to the Krell.

In summary, I think that for a fair comparison of power amps (and yes, the ones I've had so far were acoustically much distinct from each other) I must hear a lot more than three tracks. This means joining the challenge makes no sense to me.

Despite of that the comparison nicely demonstrates that you could listen to music even with a small SINAD. In my opinion everyone has to find his own way here because it's a hobby! My way is clean electronics, because the speakers produce enough harmonics and with that I at least know where the "sound" is coming from.

When I was developing the Audiovolver, I also dealt with the topic of distortion. I have added one of my papers about the triode subject, how it works and how to artificially create harmonics. Also a listening test and some pointers to literature are provided.

Have fun reading.

DrCWO

Edit:
I also like to share this link. Especially the listening test on page two says much about how difficult it is to hear the regular distortion pattern of a triode (Pattern 1).
Here the first harmonic was 30dB down so SINAD also would be about that.

This paper for those of you who are willing to read scientific papers:
 

Attachments

  • Public Whitepaper on Tube Simulation.pdf
    983 KB · Views: 66
Last edited:
I think everyone here has a collection of favorite tracks that allows us to focus on certain aspects.

I really fully agree with you. Everyone here needs to have his/her own consistent "reference playlist" consists of at least ten of your-preferable music tracks for every step of your audio journey/exploration.

Just for all of your interest and reference, just as an example case, you can find summary of such a "reference playlist" in posts here and here on the thread of my long-journey audio project.
 
Last edited:
My position has always been that *insofar as a tube amp might make an audible difference* it is in the big picture a very subtle one. Audiophiles tend to care about subtle differences, so a subtle difference can be important to an audiophile. But just as the subjectivist may exaggerate the sonic differences, so can the objectivists who decry adding a "distortion generator" at the amp stage. Any character added by a tube amp is generally, from my experience, utterly swamped by the greater sonic characteristics of a recording - there's no gross change of a recording where you are "no longer hearing the artists intent" or whatever.
.
sutle diferences become bigger over time. after you get used to the "effect" you will miss it once you take it away
 
Gave it a shot since I appreciate the effort…tougher than I thought though I think (lol) I can hear differences in the highs on two of the comparisons. The other sounds the same to my tin ears.
 
Technically, random number generators are based upon the seed, and the seed often is based upon epoch time which increments by seconds. If I ran it three times in a row, I wouldn't really know if the seed failed to increment up or if the repeated random number generated the same value as before.
That's not how random number generators work. You only seed it once, then repeatedly ask for the next random number in the sequence (which will eventually repeat, after many millions of number have been requested); you don't pass it a seed each time you want a random number.
 
Only me that took an ABX-test so far? Or?
 
That's not how random number generators work. You only seed it once, then repeatedly ask for the next random number in the sequence (which will eventually repeat, after many millions of number have been requested); you don't pass it a seed each time you want a random number.

Right. But implementation matters. I just didn’t want to run into a scenario like this:


What happens is that some servers use server side caching so it doesn’t actually execute the random function even though you asked it to. It thinks it’s just different users asking for the same thing, so a cached result is used.

Then, if I asked it to randomize a boolean for 300B or D90, and I got the same answer twice, I would not know if it was really random or if it was just showing the same cached results twice or thrice.

In a way, it doesn’t really matter since the blind listening should be OK, but if I didn’t randomize things, then maybe 1,3,5 end up being the same because I would have loaded the music files in matched pairs or 1,2,3 would be the same since I would have recorded the music from the same hardware in that way.

Generating the 6 number sequence at all once avoids the server side caching issue which I would have no control over since I didn’t implement my own sequence generator.

Just going for the maximum possible entropy when asking people to compare two products with a nearly 100 dB difference in SINAD :)
 
Only me that took an ABX-test so far? Or?

Good question. I just did the blind AB comparison, I listened to the OP's original samples.

For my purposes/curiosity that was sufficient to address: 1) can ! hear a difference 2) what were the subjective sonic characteristics of any differences, and 3) did I prefer one or the other in each pair.

An ABX comparison would of course add statistical rigour. For two of the pairs I heard an audible tell (which differentiated, but didn't necessarily reveal which amp) so ABX would likely be redundant. For the other pair, ABX would be informative.

The samples were very short so I simply played them pair-wise via QuickTime Player and adjusted volume freely until I had no further subjective impressions to note. I tried two different reproduction chains (driving two different headphones) which gave somewhat different experiences.
 
Test Setup:
I connected a Topping D90, set it at 0 dB volume, to a budget 300B single-ended integrated tube amplifier "based on the 91B" and connected it to a pair of Infinity Infinitesimal IV speakers (rated at 90 dB/1W/1m) and set the volume to a comfortable listening level for music. This ends up being in the <1W range for me. I then switched out the speakers for a E1DA Cosmos ADC with a resistive load and recorded real music at 32-bit 44.1 kHz. The out-of-the-box Psvane tubes were used for the recordings. @John Atkinson 's comment on a different 300B amplifier applies to this 300B amplifier also: " Its measured performance is what I would expect from an amplifier with a single-ended output stage that uses a single 300B tube."

Then, for comparison, I took the Topping D90, set it at 0 dB volume, and recorded it directly to the E1DA Cosmos ADC (via RCA adapters) using the DIP switches to get volume matching that was closer than 0.5 dB. The recording was then further digitally volume matched according to peak dBFS as calculated by @pkane's DeltaWave. for my initial 30 second recordings. This is your theoretical straight wire with gain.

My issue with the entire test is you replaced the loudspeaker with a resistor. Why didn't you connect the E1DA in parallel with the already attached loudspeaker? If you want to know the difference between what you actually 'hear' with the 300B amp, the interaction with the load cannot be excluded.
 
My issue with the entire test is you replaced the loudspeaker with a resistor. Why didn't you connect the E1DA in parallel with the already attached loudspeaker? If you want to know the difference between what you actually 'hear' with the 300B amp, the interaction with the load cannot be excluded.

Agree 100%. I didn't run it through a loudspeaker as I had not yet discovered the load-dependent treble gain at the time I made these recordings. My initial thought was to run the recordings through a resistor, and then run the 300B through a few speakers to capture the variable frequency response as a first-order estimate of how it might sound with speakers. The assumption was that one could just assign a PEQ on top of the resistor to imagine how it might sound with a particular speaker.

Of course, I now know that not to be true -- so doing some blind testing with a known electrical speaker load is a good idea. I'm not sure how many ASR members would actually participate given the anemic response to this test.

The main reason for the different volume testing was more to make sure I wasn't overloading the pre-amp somehow. That's when I stumbled upon the findings that I found, which I had not seen or read anywhere before as a potential explanation for "sweet treble." I won't claim I'm the first to figure this out -- but certainly, maybe the first to have an opportunity to publicize it in the modern age on a website like this.
 
The samples were very short

Agree. I made my musical choices and recordings with the assumption that 30 seconds was appropriate based upon a cursory review of US copyright law and fair use. Then I did a double-check before posting and realized that it is 30 second excerpt or 10% of the total track, whichever is less. So my musical choices were then truncated further since I happened to pick short songs :(

They happen to be some of my reference tracks in part because they're short and I can listen to the whole piece!

Lots of lessons learned/opportunities for improvement -- but in my opinion, already some worthwhile discoveries.
 
So I expressed a preference each time - but I am far from certain I was actually hearing a difference.
 
Except that most of us objectivists don't worship at the SINAD alter like the subjectivists want to believe. We know we likely can't hear the difference between 90 and 110db SINAD. We just figure all things being equal we might as well get the 110db SINAD amp if we have the option.
I'm pretty sure I can't hear the difference between 60dB (or maybe even 40dB as long as it is distortion dominated) and 110dB.

This test sort of re-enforces that.
 
Thank You for creating the poll.

The differences were certainly smaller than I expected, but of course, one can always use the perennial favorite excuse: "The sytem used for listening was not revealing enough" :)
 
It's the challenge -- I suppose if I didn't say what was being tested, no one would be excited to try.

Given this audience, I would expect most people to expect the 300B SET to sound worse, especially when looking at the measurements of the 91E which is similar.

@computer-audiophile, your vote still encourages people to try a 300B for themselves if there are a lot of votes. If the recording takes the magic away, but in the absence of magic, it is as good as a straight wire with gain, it probably will make listeners curious to hear the real thing! The recording into a resistor takes away some of the speaker interaction which can also be preferred.

It’s really hard to tell the difference, which means that a very low SINAD and very high SINAD is not nearly as different as we think, if the low SINAD measurement is from a 300B.

This amp sounds great to me, but it will be interesting to see how a very transparency focused crowd perceived these recordings.
That , why not record with the speaker in circuit , you get speaker interaction and microphonics in the amp and was not speaker interaction the special sauce in your measurement :)

In this test you test for preference I would like the option “I hear no differences” and “I hear differences” but I don’t really prefer any over the other . And 3 choices the original should be there to for completeness A,B,X style.
Even if I suspect that recording the D90 is practically the same as the original to us listeners

Preference in this case is a bit untethered in both ends as in I probably won’t know how many of these songs should sound as I have no reference .
 
The challenge:
How does a budget 300B SET compare to a "straight wire with gain"?

You tell me. It's time to put your ears to the challenge.

I listened for preference only and used OLLO S5X for my comparison using @danadam uploads.

Alas, I agree with @restorer-john and @computer-audiophile and did not hear the 300B magic through speakers which appears so massively different than the resistor (given the FR responses) so that a retry with speakers (ONLY the used speakers) will yield a different result and that same tube amp would have, very likely' sounded different yet again on my or other speakers.

It would also depend on the response of speakers as well because what sounded 'right' on one speaker might sound too bright or dark on another skewing preference.

Personally I could care less if it were tube or SS or a mix. The results with speakers might have been different and as I don't listen to the actual recorded speaker in that room that too will not be conclusive either.

Should be repeated with A1, A2, A3 and even that won't tell anyone here how it sounded in your room to you. (resistor vs SS (or source) vs speaker load)
Even more interesting could be: original, SS+resistor, SS+speaker, tube+resistor, tube+speaker. That will tell you something about how variable the 'results' can be with SS and that particular tube amp (on those particular loads).

I thought 3 sounded 'rougher' than 1. The problem is to find whether it was due to less treble or better treble quality so preference for A1
There was a very distinct difference in B. B2 was brighter and had some sibilance which is in the recording IMO B1 was too dull and like brighter sound so preference for B2
C1 sounded 'smoother' (less gritty). The problem is to find whether it was due to less treble or better treble quality but preference for C1
 
Even more interesting could be: original, SS+resistor, SS+speaker, tube+resistor, tube+speaker. That will tell you something about how variable the 'results' can be with SS and that particular tube amp (on those particular loads).

That’s a great point. As I mention a few posts earlier, we stil get the benefit of “this a what 22 dB 5W 1 kHz SINAD sounds like” but even though we recognize the load dependency, the level dependent effects weren’t what I predicted.

What I think I need to do is
1) find longer source material so I can legally share 30 seconds instead of less
2) run through a speaker to see if I can capture that source dependent boost.

This of course adds another complication of not being sure what volume to set it at.
 
Set the volume to the listening level you would listen to 'actively'.
 
Back
Top Bottom