Comparison of stock 12AX7 vs. JJ ECC83S Tubes

Tarekith · Post by **Tarekith** » Tue May 20, 2008 6:29 pm

I wanted to create a quick and simple test to see how much of an audible difference replacing the tubes in my new Korg EMX-1 would make. After looking into what tubes to replace the stock 12AX7 tubes with, I decided on a pair of matched triode JJ ECC83S which I bought from Eurotubes.com:

https://ssl.eurotubes.com/cart/index.ph ... gory_id=12

I wanted to see how much difference there would be in level (volume) between the tube types, as well how much the distortion varied as well. The first thing I did was create a simple 1 bar pattern with a held sine wav to measure the output of the EMX using the stock tubes. The EMX-1 main outs were then routed to the inputs of my MOTU Ultralite soundcard, and I used Ableton Live 7.07 and the plug in Inspector XL from EAS to measure the levels. Inspector XL was set to use the K-14 meter scale, which is my preferred metering scale. I recorded the RMS meter readings at 5 difference positions of the Tube Gain knob on the EMX-1, based on the picture below:

Here are the readings I obtained with the stock Korg tubes:

Knob Position 0 = -11.30dB
Knob Position 2 = -9.50dB
Knob Position Half = -0.55dB
Knob Position 8 = 3.67dB
Knob Position Full = 3.93dB

Here are the readings I obtained with the JJ ECC83S tubes after the swap:

Knob Position 0 = -17.20dB
Knob Position 2 = -15.80dB
Knob Position Half = -6.49dB
Knob Position 8 = -2.22dB
Knob Position Full = -2.00dB

As you can see, the JJ tubes are quieter by roughly 5-6dB at all knob positions. Next to the tubes in the EMX-1 are pots to adjust the output levels of each tube, but there was not enough range to successfully match the levels of the two tube types, so be aware that swapping your tubes will potentially lower the output level of your EMX-1.

My next step was to record audio examples of the two tubes in action so people could hear the differences with a more real-world example. For this test, I used an 8 bar EMX-1 pattern from my last song, "Puled" which you can read about here:

http://www.korgforums.com/forum/phpBB2/ ... b9714a869b

I recorded 8 bars of audio from the EMX-1 into Ableton Live 7.07 using the same Tube Gain settings listed above. Live was set to record 24bit/44.1kHz wav files, and these were then converted to 320kbps MP3's once I was done to save on traffic on my site. I think even the MP3's make the differences plain to hear, so likely nothing would have been gained anyway by posting wav file formats. Here's the audio examples:

Knob Position 0:

http://tarekith.com/mp3s/01_Stock_Tubes_0.mp3

http://tarekith.com/mp3s/02_ECC83S_Tubes_0.mp3

Knob Position 2:

http://tarekith.com/mp3s/03_Stock_Tubes_2.mp3

http://tarekith.com/mp3s/04_ECC83S_Tubes_2.mp3

Knob Position Half:

http://tarekith.com/mp3s/05_Stock_Tubes_Half.mp3

http://tarekith.com/mp3s/06_ECC83S_Tubes_Half.mp3

Knob Position 8:

http://tarekith.com/mp3s/07_Stock_Tubes_8.mp3

http://tarekith.com/mp3s/08_ECC83S_Tubes_8.mp3

Knob Position Full:

http://tarekith.com/mp3s/09_Stock_Tubes_Full.mp3

http://tarekith.com/mp3s/10_ECC83S_Tubes_Full.mp3

As you can hear, not only are the stock tubes louder, but they also distort much more at higher Tube Gain settings on the EMX-1. As to which tube type is better, I'll leave that for you to decide, as it really depends on how you use the tubes and the sound you personally prefer.

Hope this was helpful to people. Thanks for taking the time to listen, and feel free to let me know if you have any questions on the testing I did.

11hz · Post by **11hz** » Tue May 20, 2008 7:32 pm

Nice work. Another comparison idea could be all the files normalized, to compare tone differences of the tubes at the different tube gain levels while keeping volume constant.

the EH tubes definitely distort more than the jj tubes, which some people may be into, I'm not. At knob possition 0 for the EH tubes they sound pretty much the same, in terms of distortion, as the JJs at any possition.

another thing to compare would be the noise floor of both tubes with the tube gain turned to 0 or half and no sounds playing. I noticed that the JJ tubes (ex803s) I put in my esx significantly reduced the noise floor compared to the EH tubes. For this test you would not normalize the files because you'd want to evaluate the volume differences. For me this is the clincher I prefer giving up the distortion and getting less noise.

good post Tarekith.

Tarekith · Post by **Tarekith** » Tue May 20, 2008 8:03 pm

11hz wrote:Nice work. Another comparison idea could be all the files normalized, to compare tone differences of the tubes at the different tube gain levels while keeping volume constant.

I debated that, but I figured in the end people might want to hear the difference in volume more so than tone. Easy enough to get a close approximation by turning up your soundcard while playing back the JJ samples I guess. If I get some more free time maybe I'll try that too though.

11hz · Post by **11hz** » Tue May 20, 2008 8:35 pm

Tarekith wrote:Easy enough to get a close approximation by turning up your soundcard while playing back the JJ samples I guess.

That is what I did.

So what do you think about the different tubes?

anselmi · Post by **anselmi** » Tue May 20, 2008 9:15 pm

excellent work takerith!

I can see that if I retube the gain knob would be stucked at full position all the time!...
the sound quality is excellent and I notice a really nice compression that´s really cool

it´s true that you got far less volume...full JJs are about the same volume of the EH at half gain, but I never use the EH at any point beyond 2 because of the ugly distortion...this disto could be good for some individual parts if you´re looking for a smashed tone, but not for a whole mix

due to the umpaired tube level some users report I think the matched pair is a must

Tarekith · Post by **Tarekith** » Tue May 20, 2008 9:16 pm

I like the new ones MUCH more personally. I don't want the tubes for distortion purposes, and I felt the Korg's distorted way too low in the range of the Tube Gain knob. The JJ's are definitely more of what I was looking for in this type of feature.

With the Korg stock tubes, I was leaving my Gain know at 2 all the time. Now I'll likely leave it closer to half or a tad less than that. Just a bit of tube warmth to the signal, nothing drastic.

Tarekith · Post by **Tarekith** » Tue May 20, 2008 9:47 pm

My stock tubes were mis-balanced too, and even the JJ's were tough to get exactly balanced in terms of left and right volume. I noticed each time I power on the EMX one channel is slightly louder or quieter, and it changes each time. Certainly the trim pots make it all but unnoticable unless you're watching a detailed meter, but it's there. Just a limitation of using a tube output stage I guess.

anselmi · Post by **anselmi** » Wed May 21, 2008 3:18 am

I was looking at the link you posted and notice there´s a "high gain" model too...I wonder what´s the difference other than more output level...I mean, of course is very probable they distort the signal at high tube gain levels but maybe they did it in a more pleasant way than the stock EH tubes

11hz · Post by **11hz** » Wed May 21, 2008 4:25 am

I like mine better for the same reasons tarakith and anselmi do. Not so drastic distortion, just a little warmth to the whole mix.

I used to keep the tube gain knob at 0 or at most 2 with the EH tubes now I keep it at half and can go up to full and it still sounds good.

Ruso · Post by **Ruso** » Thu May 22, 2008 3:22 pm

it's not even so much about the distortion it makes everything sound "hifi" for the lack of a better expression and it brings out every detail out of every sound... you begin to hear like filter pops and attack envelopes. It just shows all the detail that is there.

The Puppeteer · Post by **The Puppeteer** » Tue May 27, 2008 9:40 pm

I think it would be best to tune the tubes in diagnostics mode before doing the comparison. The tuning utility adjusts the gains for the tubes, and if I ever changed tubes it would be the first thing I did. You will need to build some custom cables to do it.

reddone · Post by **reddone** » Tue May 27, 2008 10:27 pm

it would be nice .. but sadly we dont hav that option at least most of us . So ppl hav to simply do what they feel they must regardless of a 100 percent accurate test .

Im happy for now with how my machine sounds , but when it goes in to hav the cutoff pot replaced , i'll ask them to tune my tubes n see what happens .

Tarekith · Post by **Tarekith** » Tue May 27, 2008 11:07 pm

do you have more info on this diagnostic mode?

The Puppeteer · Post by **The Puppeteer** » Wed May 28, 2008 8:53 am

I do, but I'm in Europe at the moment so won't be back in the studio for a bit. I'll post a step by step when I get a chance.

TrondC · Post by **TrondC** » Wed May 28, 2008 11:18 am

hm, my stock tubes (for the ESX) start to distort at 35-40%, while in your demo's I could only hear the same distortion at 100%.. Are my tubes broken? I've always thought the tubes were a gimmick, as they almost never produce a sound I'd like to use. Didn't quite see why I would distort the entire output....

alos, I didn't hear much difference besides the volume, may be due to the crappy laptop-speakers, but I didn't really think there was a big enough difference to justify the 50-60$ for the other tubes..

but then again, this may be because of my speakers, would love to hear the difference live in person..

anyway, great demos, although I didn't notice very much difference