Do Tubes Make that much of a Difference

[Jeffw5555]

Back in the 70's, I did extensive tube rolling with my 1971 Marshall JPM50. The only tube that seemed to respond was the phase inverter tube, but it really didn't result in a change that I would say was better or not. Back then, there weren't any Chinese or Russian tubes, so my experiments were with the established western brands like RCE, GE, Mullard, Sylvania. My conclusion at the time was that tube rolling was a pointless exercise. I currently have 3 Mesa amps, Mark IV Rev B, Mark V 90, and a JP-2C. I did some rolling on the Mark V when I first got it, and my conclusion is that tube rolling is even more of a pointless exercise with Mesas than it is with Marshalls.

Speakers are a whole different story. If you want to change the sound of your amp, speakers can give you a major shift in your sound. When I bought my used Mark V, it was a combo. I have owned combos in the past, and I hate the sound of open-back speakers. I bought a headshell and converted it to a head and sold the combo cabinet. It all depends on the genre of music you like of course, but I play hard rock and metal, and IMHO you cannot get a decent sound with those genres of music with open-back. Open back is fine for cleaner music, like Indie, jazz, country. I currently have a Marshall 4X12, a Mesa 4X12, Mesa 2X12, and a Mesa 1X12. All sound different, quite a bit different., from each other. The two that sound closest to each other are the 4X12 and 2X12 Mesas because they both have VIntage 30's in them.

 

[Jamme1961]

if you have a bunch of tubes laying around- roll away- I just wouldn’t buy tubs to roll not worth it IMHO

 

[DECEMBER]

I play hard rock and metal, and IMHO you cannot get a decent sound with those genres of music with open-back.

I play a lot of heavy, chuggy high gain stuff with a Mark V 90W, and I find my cab to sound too boxy with the back closed, with V30's or MC-90's, so I keep the back off.

 

[Jamme1961]

I play a lot of heavy, chuggy high gain stuff with a Mark V 90W, and I find my cab to sound too boxy with the back closed, with V30's or MC-90's, so I keep the back off.

Best sealed 2x12 I’ve used is an old Marshall 1936 model- it’s a little oversized for a 2x12 so it sounds nice not compressed or boxy

 

[DECEMBER]

I said I would post a comparison I did, so here it is:
Mesa/Boogie 6L6GC power tube comparison. Mark V on 90W, ch 3 Mk IV mode. EMG 81X, drop-C. EQ in the loop. Recorded direct from the slave out into a mix of 5 IRs (per channel, double-tracked stereo) I made of my speakers (MC-90 + V30 left, Jensen Tornado Stealth 80 + V30 right).
The only thing that changes is the power tubes. A rimshot indicates a tube change. Order is:
1: 4 STR-440
2: Inner 2 STR-441, Outer 2 STR-445
3: Inner 2 STR-445, Outer 2 STR-441
4: Inner 2 STR-441, Outer 2 STR-440
5: Inner 2 STR-440, Outer 2 STR-441
6: Inner 2 STR-445, Outer 2 STR-440
7: Inner 2 STR-440, Outer 2 STR-445

I think the biggest difference is between the first two.

 

[Jay Omega]

I agree. Example from personal experience: I changed a JJ ECC83S in V1 for a Mesa Branded 12AY7. That's both different brand and different type of tube. Should be a clear difference, right? If I set the gain on the amp to get the same total output and distortion, nobody can tell the difference in a blindfold test.

If tubes had a sound, the manufacturers would have frequency response curves, just like speakers do. That's is not and has never been the case. Tubes, like capacitors and resistors, don't have a sound of their own, the overall circuit does.

They certainly do publish curves and have for quite some time. Here is an example from a current production JJ:

Although these are not frequency response curves, which I can't say I've seen for a discrete electronic device, they depict the inherent nonlinearity which indicates how the tube will distort when you're above small-signal range and this certainly influences the sound.

That being said, I agree that the audible difference between versions of a particular tube type within the same circuit is overstated as much of the performance is influenced by the associated circuitry. If feedback is involved and you're operating linearly there is even less influence. You will hear the most difference between tubes when signal levels are relatively large and feedback is minimized.

I've certainly been able to perceive a difference when swapping a 12AX7 with a 12AY7 as mu is quite different, some places in an amplifier will be more noticeable than others.

 

[DrKev]

"Although these are not frequency response curves, which I can't say I've seen for a discrete electronic device"

That's what I said. If there was any meaningful fundamental difference in frequency response between types or brands of tubes, it would have been published. That has never been the case.