Does tube-rolling actually make a difference in sound?

I cant offer any science why, even though I know there is some, but there is a definite change in sound when rolling tubes.

 

You may want to start by reading this wiki. http://en.wikipedia.org/wiki/Tube_sound 

Edited by KingStyles - 6/19/11 at 8:26pm

IMO the "SS Sound" is more the sound of 40db or more global feedback than the sound of transistors. Transistors sound a lot like tubes when you take the feedback loops out. Not exactly the same, but closer than the rules first brigade would care to admit.

 

Regarding tube rolling:

Look for data on different distortion levels & distributions between various brands & styles (model years, different production techniques, etc) of the "same" tube. It is easy to see this having an effect. In some circuits the DC operating point of the tube or transistor (which could actually vary from sample to sample {applies to transistors moreso than tubes!}) could also have substantial effects on the operation of the circuit. 

Edited by nikongod - 6/19/11 at 9:01pm

well plain and simple rolling tubes in a tube amp absolutely makes a difference in sound.. its not really something that can be argued against.. the differences are quite obvious.

 

amps sound different.. thats just the way it is.. 2 solid state amps can sound different than each other, 2 tube amps can sound different than each other, so therefore...

ive had this experience myself.. and again, like roling tubes, the differences are way beyond deniability in some cases. 

 

i think the real argument is that "at some level" there are SS amps that can be made to sound like tube, and visa versa..  i dont know this for sure but i've heard it said.. 

Yes, different tubes offer different sound. Different brands and different production runs can be electrically different.

However, I still don't think tuberolling is worth the expense or trouble. Tubes are moving targets, they change value over their lifespan and I don't think it's worth the trouble. If a tube works and the amp sounds good, I'm happy. I'm not going to buy more tubes and experiment. I leave a tube in until it dies.

I'm glad you put quotes there because it's actually the lack of a particular (added) sound. The feedback brings op-amps closer to the ideal. Nonlinearities (distortion) will be reduced a lot and the output impedance also will be very small (ideal = 0).

Of course, you can add resistors in series with the load to get a higher output impedance and reduce the damping factor, similar to what you get with tube amps which usually have a couple of ohms Zout. This can cause huge differences depending on the headphones' impedance curve.

I can very well imagine how different tubes with different electrical properties (that also change over time) cause changes at the output, such as distortion. This is also the case for SS designs, i.e. replacing a "weak" (regarding output current) op-amp with a more beefy one could improve a number of measurement results.

The question is, whether these measureable differences are actually audible to our hearing. If your headphones have a THD of 0.01% and you have to do a blind test between two amps with 0.001% and 0.000001% THD I'm pretty sure that you'll have a hard time trying to find audible differences.
Edited by xnor - 6/20/11 at 4:52am

I discovered from this test how difficult hearing harmonic distortion is:

http://www.klippel.de/listeningtest/lt/

 

And that is with a minimum THD amplitude of -45db for the hardest bit of the test, which is a huge amount of THD by the standards of decent solid state equipment.

 

Anyone who claims they could tell the difference between 0.001 and 0.0001 THD is either lying or God. 

Edited by Willakan - 6/20/11 at 8:54am

Exactly. And this poses another problem I've seen a couple of times here and in other forums as well: some people look at measurements of several pieces of equipment, for example with the 2nd harmonic ranging from -110 dB for device A to -80 dB for device B. The latter seems largely inferior compared to the former, yet in a listening test the difference could very well be completely inaudible. I've seen people concluding that therefore, measurements do not correlate with what they hear, the "truth", "reality" (actually with their expectations). The situation gets worse if product B is a praised/hyped, "audiophile-grade" device which again people expect to sound a lot better..

Sorry if that was a bit off-topic.. I do get carried away sometimes.

Perceived differences after tube rolling could very well also be the product of expectation bias.
Edited by xnor - 6/20/11 at 10:22am

 

No way. You might say that about cables or something along those lines, there is a difference with tubes. it is very easy to hear differences. No golden ears needed, no special training needed. When 1 tube has tight strong bass and you put another one in that has weak, bloated bass that dosent extend, I doubt preconceived expectation were causing this. Do you have any experience with tube rolling?

 

+1   The differences often truly are obvious sometmes.  Same with changing amps sometimes.. i say sometimes b/c it depends on which you are comparing. some make literally huge differences, while some may end up sounding very similar.

 

 

 

 

Well, maybe one with tube-rolling are trying what type of distortions are pleasing to the subjects..

http://www.aes-uk.org/past-meeting-reports/harmonic-phase-the-missing-factor-in-distortion-measurement/

 

DBT is an instrument and one uses the brain differently (more critical -> CNV) than one usually do when hearing a musical piece..

http://www.ncbi.nlm.nih.gov/pubmed/20153786

We could settle the question of whether different tubes sound different in a given amplifier by recording the output of the amplifier with a headphone plugged in, and then playing it back at a later date within seconds of each other!

 

Just saying. 

 

Considering the availability of good quality digital recording gear its neither difficult or expensive. 

Yes. Even a simple RMAA measurement could show differences in frequency response that would for example explain the differences in bass as KingStyles mentioned.

I haven't seen any kind of such measurements but would be very interested. Unfortunately, I don't have a single tube amp.

@KingStyles: I'm not saying there never is an audible change, just that, as usual on head-fi, things get exaggerated and in some cases the changes might be smaller than expected.

I agree that changes are usually exaggerated on headfi. Tube changes is usually what I equate to a big change, even though they can be small. Depends on the tubes being rolled. The only thing that makes a bigger change is the headphone. Amps and dacs can be anywhere from almost no change to a lot of change. Usb/spfid converters are no change to a medium change, and cables usually small. That was just to show where I would hold tubes as far as there potential to change the sound. I have never used the expression night and day difference, because all changes are small in the grand scheme of a system. Its just what degree of a small change there is.

It probably depends on how the amp is designed, with my current soon to be replaced LC3 amp there is a huge difference between tubes. It's beyond subtle or subjective, the difference tube rolling on this amp is as clear as swapping headphones.