Ribbon pics, tube preamps, and resistance

The mic seems considerably brighter when you remove the load resistor, which is interesting. Can't say I'd have a defined preference, as on a more sibilant voice it might be problematic. Lovely low shelf filter there i will say, once applied you can quickly hear the vocal sound how it might need to in a mix.

By habit I don't watch videos, so guessing you refer to a Jensen input, but yes, 90+% of vintage tube preamps use unloaded secondaries, the capacitive input value of the tube grid being the predominate load. A few of the early Altec preamps have matching load resistors, that's about it. Many modern transformers are designed with a secondary load resistor in mind, or a Zobel, and don't give design spec response without a load. Most older transformers don't require a secondary load resistor, though there were a few companies that issued a similar design spec transformer for both loaded and unloaded use. Modern design expectation versus any vintage is fairly apples to oranges in comparison, so applying ancient thinking to modern parts frequently leads to a 'fail'. Many of those modern parts expect a bridging relationship source Z, and don't give good response without it. Many of those old parts are happy with a matching condition, or even a somewhat slightly heavier load. All of that plays into exactly what a load will do to either one. Reflected Z from an ancient unloaded input transformer can look extremely high, it can also look extremely low. It is always like a speaker impedance curve, with wide variations with respect to frequency. A modern part may be much more linear, and that load being part of the recipe for flat response. Removing it may also do what RCA frequently did in preamp design, which was to add treble boost networks to compensate for ribbon high frequency rolloff, delivering a combined flat response.  Removing the load will also change level, so some sort of null check is required for truthful listening.  

The secondary resistor removed and second position low shelf just opened up the microphone beautifully! Still had a ribbon's warmth but it was so nicely cleaned up. Positively useful!!

Yes, and there are a few EQ switches I'd like to see standard in some preamps too...
1. the 'ribbon' switch which shelves 250hz on down at a 6db slope.
2. the 'drum' switch which puts a Q.50 -4.5db cut at 350hz
3. a 'presence' switch around the 3-6k range
4. an 'air' switch that does...well you know..

Any other things? Maybe we can persuade jsteiger and others. API should have done #2 YEARS ago.

Raising the bar... I love it!

Here is a link to 24/96k files of the comparison. I did not match levels after the fact because that seems to be part of the difference.

www.dropbox.com/sh/0ul0y84ygd6zxf2/AAAH-Ig3U4I7ShII3tqc48HQa?dl=0

Awesome video demo!! Thanks for posting. (I watched the video on my desktop computer speakers) This is certainly extremely interesting to me. Like all things in audio, its a compromise. When I heard the first take and judged the dynamics, I felt the R-84 sounded wonderful on your voice. Especially when in the higher register. Had an "old" quality to it. Perhaps a little dull, or dry/warm. Etc. However you say it, it felt flattering to your voice, but also not super detailed.

During the Second example, the quieter parts had more "life" and air. Which I really liked over the first example. However, the louder parts where not as "ear cozy", so the sacrifice here is that you will get more detail, but a bit harder to handle transient.


Thanks again for sharing! Super Cool. The Loading off the plate is nice. If the circuit could use some type of Non-Linear Compression after the input amplifier, I might get the "match made in heaven" feelz!

Digging a bit deeper tonight...

I don't have an example to show, but without the resistor there were all sorts of ringing anomalies throughout the sweep. It seems clear, as Doug pointed out, that the resistor is helpful in stabilizing the input transformer.

So, here are the frequency responses of the current prototype. The first photo shows the reference signal compared to the right and left channels of the preamp. The second photo shows the low shelf RC networks in action.

I plan to adjust the medium shelf a bit lower. (Or maybe I won't. Need to play around with various instruments.) The graph can be visually deceiving, as the maximum cut is only 5.5 dB at 50Hz.

Thoughts?