Mismatched power tubes sound?

We are routinely told that power tube matching is critical even with amps that allow for individual biasing. I am unsure what the sound of mismatched tubes is, any suggestions?
hello i thought i give my opinion,and hopefully no gives me hardtime.but with individual bias you can compensate for mismatched tubes.i dont pay extra for matched tubes and my amps sound great.i do try to test them well on my test bench them put in my amps and bias again.works for me .i use cary monoblock amps.most match sets are for fixed bias amps.if you can adjust bias you set the current each tube draws,therefore it balanecs the tubes to a certain extent.
I lack the words to describe it, but I have experienced the sound of mismatched tubes - one set drawing about 50% more current than the other. Despite biasing, the sound was "odd", less natural, perhaps "phasey" describes it.
Matched Output Tubes, are Mandatory for good results in serious Tube Power Amplifiers. Individual Bias adjustment for each tube is also mandatory to allow the cathode current of each tube to be the same value. This prevents magnetisation of the output transformer. Magnetisation of the output transformer lowers your available Low Frequency Power Output, as the transformer will start to saturate earlier thai if it was not magnetised. The actual Bias Voltage applied to each tube should also be nearly the same. If this differs significantly between tubes, you have a problem with one or more of the tubes, (probably worn out). Despite tube matching by the supplier, tubes still differ somewhat in gain, so some form of AC Drive balance control is of immense benefit. Be aware that tubes wear out, and change their characteristics drastically as they wear. I would discard, or re-match tubes showing as little as 20% difference, 50% are throwaways.
Is the the core magnetization a dynamic situation only, or is this a static condition resulting in permanent degradation? (I would suspect it's hopefully only a temporary condition)? And if it is a permanemt problem, then can you degauss the core via playing a "burn-in record" through the system? This phenomenon never occurred to me before, but certainly it is a valid concern!
Matching of tubes is important, or not, more often depending on the design and type of amp, and the tubes you choose. A perfect example of shooting the matching theory in the head just happened to me. I have finally broken in my Wolcott Presence amps, and last week, I installed all the NOS Mullard EL 34 output tubes I could find (Circa 1960's). The sound is so much better with three quarters of the tubes as Mullards and the remaining as stock Sovtecs, that I can hardly believe my ears! This is not to say that 100% Mullards will not be better, it simply points to the fact that there is no absolute rule, especially if the quality differences are vast. As far as tube bias, I had to switch some tubes from socket to socket to get them all to bias, and where this is far from a ideal situation, it produced the best sound I have ever heard from these amps, and perhaps as good as I have ever heard in any amp.
Bob. Not to worry about permanent effects of output transformer magnetization. Quality output transformars are made from fairly soft iron, and does not permanently magnetize, and if they did, driving the amplifier hard would soon demagnetize them anyway. The magnetization is caused by the unbalanced plate current of the output valves making the transformer act as an electromagnet. Gain unbalance between tubes, or valves as we call them here in Australia is another real concern and can to a degree be compensated for by adjustable AC drive levels. This was a trick used in tube type Cinema Amplifiers, and in the Modulator Stages in Tubed AM Broadcast Transmitters. We have used the same trick in a high quality tube KT 88 or 6550 Commercial Studio Monitor Amplifier. If you are interested, we can send you some details on this amplifier which we call the "Enigma", 1 because we do not know why it sounds so good, 2 we do not know why we went and designed it.