You need to know the OUTPUT impedance of the preamp, not the input impedance, to match it to the amp.
All of the preceding responses are correct. A theoretically ideal audio amplifier or preamplifier will have zero output impedance and infinite input impedance. That is obviously not practical for many reasons. The numbers cited above (several hundred ohms output impedance, or perhaps even one or two thousand ohms output impedance, and 50K or 100K input impedance) are typical for preamp inputs and outputs and power amp inputs, and are completely compatible.
Power amp output impedances have to be far lower (a fraction of an ohm) in order to drive speaker impedances, which are usually less than 8 ohms.
If an output impedance is not much lower than the input impedance of the device it is driving, a consequence is that a fraction of the voltage put out by the driving device will be wasted across its own output impedance, instead of appearing across the load device. The fraction lost would equal the output impedance of the driving device divided by the sum of that output impedance and the input impedance of the driven (load) device.
That assumes that stray capacitances and inductances in the circuit are negligible. That is not always the case, and if it is not the situation can be much worse. If interconnect cable capacitance or inductance, or the input capacitance of the load device, are significant, and the driving device's output impedance is too high, than the loss across that output impedance will be frequency dependent, which will cause the system's overall frequency response to be non-flat.
All of this applies only to audio frequency devices. At radio frequencies (such as antenna inputs to FM tuners, or high speed digital signals) completely different considerations apply, which usually require source, cable, and load impedances to be matched. Different considerations can also apply in certain specific situations involving audio frequencies, such as differential balanced outputs of professional microphones, which are often specified to work into load impedances that essentially match their output impedances.
It is not unusual for tube preamps to greatly exceed their reported (at 1K hz) output impedance at lower frequencies. This can be heard as a bass roll off with some match-ups, usually into SS amps with input impedance below about 20K. John Atkinson provides useful information regarding impedance matching in the "Measurements" section of Stereophile reviews, and here is what he said about your SLP 05, in Sept of 2006:
"The SLP 05's output impedance is specified as a usefully low 400 ohms. However, I got significantly higher values for the unbalanced output of 15001600 ohms in the midrange and treble, rising to 3400 ohms at 20Hz, with similar if slightly lower figures for the balanced output. The Cary preamp needs to be used with power amplifiers having input impedances of 10k ohms or greater if the bass is not to sound a little lean."
The spec you want to know is the worst case output impedance and at what frequency (this is the great service that JA provides in his measurements). As Mitch quoted, the impedance typically rises as frequency decreases and as Al states (the voltage divider definition) this will cause the bass to roll off.
Saki70, you are welcome. I am glad to help with things I have learned, as many others here have helped me along the way. To Nick778, I am a little surprised at JA's comment about using the Cary with amps having input impedance of at least 10K ohms. I consider the 1:10 ratio recommended by many to be a minimum at the lowest frequency where your speakers provide useful output. Personally I prefer a ratio of about 1:20 or better when possible. Since the 05 was measured to have output impedance of 3,400 ohms at 20hz, I would probably look for an amp having input impedance of at least 47K ohms (seems to be a common input impedance) to use with the 05, and I believe your Spectron amp at 50K ohms should work fine.