When recording acoustic instruments, the prerogative of the technician must be to capture a realistic impulse, a true to life transient response and delicate details, on the medium of his choice, which is mostly tape.
Transients need electronics with a wide frequency band and a minimal phase shift.
Detailing the signal to the full asks for a high degree in resolution.

GL 211 developed by Bell Laboratories for Western Electric Co. and RCA is a mono triode.

The current digital formats of CD and DAT are limited in the extension of the frequency band and are limited in resolution. These restrictions do shift the phase. Hence they have difficulty in establishing a true to life transient.
The nature of the analogue recording technique can provide transient responses to a greater extend.

Nevertheless many technicians and music lovers report that they hear a loss in natural harmonics and hear hampered transients when old tape recordings that were made on machines with valve equipment are played back on modern, transistorized tape recorders.
This phenomenon can be heard by everybody when listening to Lp's which were remastered from tapes originally made before the transistor came into use. This also counts for the recent mastering on vinyl of those famous vintage recordings of DECCA, Deutsche Grammophon, RCA and the like by the specialized companies like Speakers Corner and Testament. As a matter of fact these technicians should use the old tape recorders if they are still available and should restore them if necessary.

Cymbals
It is clear that the 'sound' of valve equipment differs greatly from that of transistor circuits. 
But not too many people do talk about the fact that this is more true for certain instruments than for others. 
Each instrument has its own colors and specific character. They depend on the nature and the amount of harmonics which are arranged in a special blend of dynamics.

For example: let's take two cymbals which are made of different sorts of brass. The differences depend on the relationship of copper and tin in these alloys. The difference in warmth of sound these cymbals produce is clearly discernible.

Well, let's now take two different tweeters. One has a 1" dome of plastic and the other is of the aluminum coated type. Now you can hear a difference too, not only in frequency response but also in 'sound' that only partly depends on the tweeter's inability to produce square waves.

Another example. 
Let us consider three loudspeaker units for midrange reproduction, one with a cone made of bextrene, another with a cone of coated paper and a third is molded of polypropylene. For our test we assume that their frequency responses and the technical parameters are identical. Yet all three loudspeaker units do sound differently. Vibrations are measurable. The material that is used however has its own 'color' and this color cannot easily be measured but can clearly be heard. 
The same is true for the many musical instruments, some do sound warmer than others. As long as we make recordings with microphones that also have their own character, we will be unable to discern the real differences in color and warmth of these instruments as we listen to the recording. The issue gets more complex because the monitor speakers have their own specific 'sound' also. 

Loudspeaker characteristics
When we make a recording of a harpsichord and a flute (traverso) and reproduce the recording via a loudspeaker system that has a metallic characteristic, then both instruments will sound metallic. The result is that we will not be able to hear the difference between a cold and warm character to the full, while during a live performance however, we are fully capable to interpret the harpsichord as warm and the traverso as cold. While we can perfectly perceive the difference in sound between the two instruments,  we cannot hear the different colors of sound and the warmth to the same degree. Although there are loudspeaker systems that have extended, flat frequency responses and the equal dynamic capabilities for nearly all frequencies of the audio band, listening to the reproduction of music via extremely well conceived Hi-Fi systems still remains an illusion. Music lovers as well as manufacturers of high end products do not readily accept this truth. 

Photography
The same discrepancy occurs in photography. 
If you look at a black and white photograph and compare it to a color photograph which depicts the same object, you will agree that the color picture looks more like reality. Hence the heading of my article.
Well, I have made a study of 'characters of sound', listening via various hi-fi systems and the results were peculiar. It is of importance to remember that the speaker-amplifier-combination always plays a dominant role. That is to say that certain combinations work better than others. Nevertheless, I want to list some findings with both transistor and tube amplifiers. I mention instruments and groups of  instruments that I listened to and noted my preferences as follows. 

 

 

Preferences
From these listings you can conclude that most of today's instruments have their place with transistor reproduced sound. The sound of transistors is generally experienced as cold and the sound reproduced via tube circuits as warm. If you listen to piano via tubes the sound could be easily too warm. Hence my preference for transistors in this case. Also it is interesting to know that old tube amplifiers reproduce the brass section and harpsichord for example beautifully, and sometimes with a timbre that reminds me of real to life performance; which indicates that the ability of rendering second harmonics is of great importance. 

Reference
You must understand that these findings do not establish a worldwide reference since they are based on personal, practical experiences that only I can subscribe to. In reality we have to guess the true nature. 
I do not think that these experiences can be explained technically. The perception as well as the wish to experience the reproduced sound as a reality make listening such a personal affair. 
At the end it all remains to a great extend a matter of illusion. 

Paul Hattink, 1990.

Paul Hattink is the composer of a/o. 'The Rose Garden Suite'. Text translated and edited by Rudolf A. Bruil.