Topic: Stock FR2LE vs. Busman T-Mod FR2LE true side by side comparison

[Church-Audio]

Here we go.

Rastasean recently sent me his FR2LE because he suspected something may be wrong with his recorder. Turns out everything seems to be working fine. While I had it in my possession, I took the opportunity to do a side by side comparison with my Busman T-Mod 2LE.

I set up my Avantone CK-1s with the cardioid caps in front of my home stereo. I chose two songs to record and recorded both songs on each recorder, thus getting a true side by side comparison. Here are the details of my comparison.

Playback: CD>Sony PS3>Pioneer receiver> Technics speaker cabinets (10" sub and 3" tweeter)
Source: Avantone CK-1 Cardioid>Canare Starquad> FR2LEs @ 24bit 44k

Granted that playing a CD through a stereo is a limitation in this comparison as it is not a true live recording scenario (in which we would be using these recorders). I listened to the recordings using my desktop computer > my KRK RP6 studio monitors. I have formulated my opinions about the two recordings, but I will refrain from sharing until others have had an opportunity to give it a listen.

Songs recorded:
Breathe - Pink Floyd Live Pulse album
Cold Hard Facts - Del McCoury Band

I felt these two songs were very well recorded on CD and thought it would give a good comparison of rock vs. acoustic bluegrass.

The file names are ambiguous so as to not reveal which recorder was used. Give it a listen and PM me for the answer. I would like to hear everyone's thoughts.

Files located here:

http://www.box.net/shared/avzfeld2id

Thanks to rastasean for sending me is 2LE to run this test and also for supplying the web space to post these files

-Christos

Great job with setting the test up imo this is how you conduct a test of two devices objectively.

Chris

[chris319]

Got a link regarding what the busman T-mod is all about?  Google provided nothing.  Thanks. 

http://www.busmanaudio.com/mods.html

Fostex FR2-LE
Vintage and Transparency $175.00 USD

What exactly does the $175 buy you?

[jacobmyers]

Oh, fun! I'll have to download those when I'm on my "real" computer. Thanks for posting them. I love this stuff!

[aleatoric]

Got a link regarding what the busman T-mod is all about?  Google provided nothing.  Thanks. 

http://www.busmanaudio.com/mods.html

Fostex FR2-LE
Vintage and Transparency $175.00 USD

What exactly does the $175 buy you?

I am wondering too.  I emailed Chris asking this.  When he gets back to me I will post.  I am also wondering how the Busman mod stacks up against the Oade mod's that are out there.  Maybe that topic is worthy of another thread though. 

[page]

What exactly does the $175 buy you?

I am wondering too.  I emailed Chris asking this.  When he gets back to me I will post.  I am also wondering how the Busman mod stacks up against the Oade mod's that are out there.  Maybe that topic is worthy of another thread though. 

It's my understanding he does the same thing Oade does and rebuild/rewire everything up to the A/D stage. Its the same physical pre-amp, but he replaces any opamps along the way and uses the van den hul wiring in place of the stock path.

Will: I'll see if I can re-create it. I synced stuff up so I'd at least come close. If your looking, I noticed it seemed to be in the upper echelons where I noticed minor differences. Could be any number of reasons, hence my remark that it doesn't confirm.

What I've seen, between this test, Will's, and mine, are that the more gain and sensitivity you are utilizing, the more you will notice a difference. It's not to say you will audibly notice a difference (as low level noise is still low level noise), but spectrally it seems to become more apparent of a difference. ymmv.

[Church-Audio]

What exactly does the $175 buy you?

I am wondering too.  I emailed Chris asking this.  When he gets back to me I will post.  I am also wondering how the Busman mod stacks up against the Oade mod's that are out there.  Maybe that topic is worthy of another thread though. 

It's my understanding he does the same thing Oade does and rebuild/rewire everything up to the A/D stage. Its the same physical pre-amp, but he replaces any opamps along the way and uses the van den hul wiring in place of the stock path.

Will: I'll see if I can re-create it. I synced stuff up so I'd at least come close. If your looking, I noticed it seemed to be in the upper echelons where I noticed minor differences. Could be any number of reasons, hence my remark that it doesn't confirm.

What I've seen, between this test, Will's, and mine, are that the more gain and sensitivity you are utilizing, the more you will notice a difference. It's not to say you will audibly notice a difference (as low level noise is still low level noise), but spectrally it seems to become more apparent of a difference. ymmv.

I have never been one to subscribe to the whole mod thing.. For one reason its pretty hard to rework a circuit board... The traces are permanent and replacing opamps... Never really ever heard a difference in my own gear. Maybe a bit less noise 2-3 db maybe little less distortion but you really are at the mercy of the existing circuit design you cant go making large changes to a circuit that has already been built unless you want to get into adding another circuit board altogether IMO. There are some cases where I have repaired soundboards and replaced the old TLO72 opamps with Ne5532 or OPA2227 and changed a few coupling caps to better values and maybe replaced a few carbon resistors here and there.. But in the end the difference was pretty good. But we are talking a console that was made 30 years ago not something that is pretty modern..

I would be very interested in finding out what some of these companies actually change to perform the mods they do.. But we will never know because that is what they are charging for the sum total of parts is maybe $30 its the labor and the knowledge that your paying for in a mod. I hope that some of these guys would do an objective test here at least in the before mod and after mod on the websites they own so others could hear the differences and decide if they are worth the $$$ being charged for the mod.

I applaud the guy that started this thread because I feel this is one of the only ways to objectively evaluate a mod and possibly be able to hear if there are in fact any differences... One other note.. I think it would be very valuable if the guys doing this mod could produce SNR graphs and FFT graphs and distortion graphs of the before and after mod so that besides listing we could physically see the differences of the before and after I wonder why nobody does this?
Chris

[Church-Audio]

Chris has a point about noise and distortion testing; that should certainly be measured and stated, and if there is an improvement, it's an obvious selling point.

That said, some other comments about use of FFT to analyze the differences are less likely to find differences, for the simple reason that you can't see the noise on an FFT while there is a high-level signal present (you may or may not be able to hear it), and you can't see things like distortion or transient response differences on a length of program material on an FFT.  Really all you can see there is frequency response, and if the frequency response is different between the two files I'd suspect something is amiss with the test or the original or modded box.  It really shouldn't be an issue for a preamp to test flat 20Hz - 20kHz, quality of that response notwithstanding.

Anyway, there are lots of tests you can do that can be analyzed and will strongly suggest whether differences should be audible, but the manufacturers ought to do that for you.

FFT might be useful for example if we were introducing a better anti aliasing filter or if you wanted to see wide spectrum noise floor of course only useful if we have averaged it out Anyway as a frequency response test you might still be able to see effects of your circuit changes in the sonic finger print of the DUT. Providing you used a very high sample rate and a large average and even larger window. But to the average person this might not mean much.. I also like looking at the full spectrum for distortion measurements certainly a good place to use FFT. To see the even and odd harmonic distortion content. When we look at a wave form under super high resolution we can see differences will these differences translate into better audio??? Who knows but one thing for sure it would be nice to see graphs.

[tenesejedd]

Here is a portion of a PM that Chris Busman sent me about a year ago when I was going to get my recorder modded by him. Here is describes the two mods available and what they entail.

Well I have 2 mods. Vintage- a warmth mod. Helps to give more bass and low-mid frequency response while keeping the high end clean, sounds great with bright mics.
Transparent - is clean, fast and detailed throughout the frequency band. This sounds great with all types of mics and is my favorite of the two.

Both mods change out all op amps to the A/D. All input capacitors are changed for very high quality caps and the phantom DC blocking caps are also changed for much higher quality caps. All of this makes for a much quieter more musical sound that will make your recordings the best they can be.


this may shed a little light on the discussion at hand.

[illconditioned]

Chris has a point about noise and distortion testing; that should certainly be measured and stated, and if there is an improvement, it's an obvious selling point.

That said, some other comments about use of FFT to analyze the differences are less likely to find differences, for the simple reason that you can't see the noise on an FFT while there is a high-level signal present (you may or may not be able to hear it), and you can't see things like distortion or transient response differences on a length of program material on an FFT.  Really all you can see there is frequency response, and if the frequency response is different between the two files I'd suspect something is amiss with the test or the original or modded box.  It really shouldn't be an issue for a preamp to test flat 20Hz - 20kHz, quality of that response notwithstanding.

Anyway, there are lots of tests you can do that can be analyzed and will strongly suggest whether differences should be audible, but the manufacturers ought to do that for you.

FFT might be useful for example if we were introducing a better anti aliasing filter or if you wanted to see wide spectrum noise floor of course only useful if we have averaged it out Anyway as a frequency response test you might still be able to see effects of your circuit changes in the sonic finger print of the DUT. Providing you used a very high sample rate and a large average and even larger window. But to the average person this might not mean much.. I also like looking at the full spectrum for distortion measurements certainly a good place to use FFT. To see the even and odd harmonic distortion content. When we look at a wave form under super high resolution we can see differences will these differences translate into better audio??? Who knows but one thing for sure it would be nice to see graphs.

I've found obvious distortion by putting in a 1kHz tone, recording it, and seeing the distortion.  For example, I saw obvious distortion on an Edirol UA5 when I drove it anywhere above -6dB.  The distortion appeared as harmonics (not sure even or odd, but obviously there).  This was not the result of a mod (it happened on both stock and modified units), but *I believe* some defect of the design.  I never did find the cause.

This is not a "full spectrum" test, but it is surprising what defects you can discover, even though they are difficult to hear.

So, in conclusion, some sanity checks are always useful.

  Richard

[Will_S]

Chris has a point about noise and distortion testing; that should certainly be measured and stated, and if there is an improvement, it's an obvious selling point.

That said, some other comments about use of FFT to analyze the differences are less likely to find differences, for the simple reason that you can't see the noise on an FFT while there is a high-level signal present (you may or may not be able to hear it), and you can't see things like distortion or transient response differences on a length of program material on an FFT.  Really all you can see there is frequency response, and if the frequency response is different between the two files I'd suspect something is amiss with the test or the original or modded box.  It really shouldn't be an issue for a preamp to test flat 20Hz - 20kHz, quality of that response notwithstanding.

Anyway, there are lots of tests you can do that can be analyzed and will strongly suggest whether differences should be audible, but the manufacturers ought to do that for you.

FFT might be useful for example if we were introducing a better anti aliasing filter or if you wanted to see wide spectrum noise floor of course only useful if we have averaged it out Anyway as a frequency response test you might still be able to see effects of your circuit changes in the sonic finger print of the DUT. Providing you used a very high sample rate and a large average and even larger window. But to the average person this might not mean much.. I also like looking at the full spectrum for distortion measurements certainly a good place to use FFT. To see the even and odd harmonic distortion content. When we look at a wave form under super high resolution we can see differences will these differences translate into better audio??? Who knows but one thing for sure it would be nice to see graphs.

I've found obvious distortion by putting in a 1kHz tone, recording it, and seeing the distortion.  For example, I saw obvious distortion on an Edirol UA5 when I drove it anywhere above -6dB.  The distortion appeared as harmonics (not sure even or odd, but obviously there).  This was not the result of a mod (it happened on both stock and modified units), but *I believe* some defect of the design.  I never did find the cause.

I think this relates to another important point.  Some mods, which address an immediately obvious shortcoming (even if you couldn't fix this particular fault), do make sense.  For example, the stock Marantz PMD660 has great ergonomics but brickwalls easily and is very noisy, so a mod there makes a lot of sense.  Likewise it's relatively easy, and definitely useful, to add standalone digital out capability to the UA5.

I just don't hear anything obviously wrong with a stock FR2LE that argues for spending 175+ and voiding the warrantee.  Maybe there would be a worthwhile effect on the noise floor for nature recording, I need to re-listen to the samples focusing exclusively on the noise floor.  From the quoted description, the vintage mod sounds like an equalizer you can't adjust or turn off.  It might make for a pleasing sound to some folks, but I'd rather have more control.

[Will_S]

Will: I'll see if I can re-create it. I synced stuff up so I'd at least come close. If your looking, I noticed it seemed to be in the upper echelons where I noticed minor differences. Could be any number of reasons, hence my remark that it doesn't confirm.

What I've seen, between this test, Will's, and mine, are that the more gain and sensitivity you are utilizing, the more you will notice a difference. It's not to say you will audibly notice a difference (as low level noise is still low level noise), but spectrally it seems to become more apparent of a difference. ymmv.

Aaah, I was looking at the integrated result over the whole file.  Kind of in line with mshilarious' comment, wouldn't expect a big difference there unless one unit was seriously defective.  I'll repeat it looking at a part that's as much just background noise as possible.

[tenesejedd]

Will: I'll see if I can re-create it. I synced stuff up so I'd at least come close. If your looking, I noticed it seemed to be in the upper echelons where I noticed minor differences. Could be any number of reasons, hence my remark that it doesn't confirm.

What I've seen, between this test, Will's, and mine, are that the more gain and sensitivity you are utilizing, the more you will notice a difference. It's not to say you will audibly notice a difference (as low level noise is still low level noise), but spectrally it seems to become more apparent of a difference. ymmv.

Aaah, I was looking at the integrated result over the whole file.  Kind of in line with mshilarious' comment, wouldn't expect a big difference there unless one unit was seriously defective.  I'll repeat it looking at a part that's as much just background noise as possible.

If I recall correctly, i deleted several seconds of silence at the beginning of each recording of "Breathe." If that is the case and if you are interested, I can post those short clips for you folks to analyze.

Let me know

[digifish_music]

Noise is easy!  Set the two units at the same settings, and terminate the inputs with a low-value resistor (150 ohms is standard for a mic preamp, but anything in the neighborhood will do).  Record the noise on both units and compare.  Trying various different gain levels may or may not be different.

It helps if you calibrate the gain on both units to make sure the noise comparison isn't thrown off by any difference there.  Feeding a test tone or other simple signal into both of them should suffice.

Guy posted a good analysis of a noise comparison for some other unit here . . . that thread is nearby, I just have to remember where . . .

Tickets are gone I can say what I think, without instantly dropping to -5000

I think it's about time these modding companies provided technical measurements to back up their claims, otherwise it is just snake-oil. If you can hear something then it will be measurable. However not all measured differences are audible, so the bar has not been raised all that high. At least show us a measurable difference that correlates with the claimed quality improved. Noise is the easiest one as noted above. Some of the other properties should show up as slew-rate or ringing differences or something. IMO, if it can't be measured, it's not real.

This whole area reminds me of this...

http://mixonline.com/recording/mixing/audio_emperors_new_sampling/

digifish

[illconditioned]

It is not scientific, but I have found a great way to test mics is to just record ambient stuff (like family members talking, clanking dishes, running water, etc).  Even better if you can have music playing (in the background) at the same time.

Then listen on good headphones (like Sony 7506). Does is sound "real"?  Do the voices sound nice.  Do the ambient noises sound correct.  Does the music playing sound pleasant to listen to?

Some mics capture ambient sounds well, others sound muffled or less lively.  Some mics don't sound that lively, but capture music well.  A very good mic works on both types of sources.

Some examples:
- DPA4060: drop dead incredible realism, but don't sound nice on music, sounds too "harsh".  *My opinion*.
- Sennheiser KE4: sounds great on music, but not as realistic as above, somehow "colored".
- Countryman B3 & Nevaton MKE400: very realistic, and sound great on music.  Just sounds "natural".

This might be what people are getting at when they compare "warm" vs "transparent" mods of gear.

  Richard

[Will_S]

  Listening to just the "silence" on both tracks, I can't hear any differences at normal playback level, but when I really crank them I can hear (and see on the spectral analysis) a definite difference - although at least for Pink Floyd a small part seems to be differences in the actual room noise.  Also at least for the Del song, the noise floor sounds different across the samples, but I wouldn't say one is obviously much louder than the other, just tonally different.  More on that in a bit, hopefully more folks will post about what they can hear first before I post what may well be a spoiler.

I wonder to what extent a lower noise floor can contribute to a perception of increased transparency on louder musical passages...obviously at some point noise can obscure detail but how quiet is quiet enough for taping amplified music in a noisy room?