Earn $100 by testing your reel to reel tape recorder for me :)

  • Thread starter Thread starter mshilarious
  • Start date Start date
miroslav said:
I'm curious...why do you feel that way?
Sure, 1/2" 4 track has some serious width per track, but it's not just about that, and of course, with only 4 tracks, if you want to do more involved productions, you have to bounce, and bounce...which then undermines the "track width" factor completely...IMHO.
I've recorded on 4-track 1/4 back when the term "home/project studio" didn't even exist...and I had a HUGE amount of fun doing it, and man, I did all kinds of bouncing maneuvers between the 4-track and my 1/4" 2-track...but at this point, unless i was going to track to something like a 1/2" 4 or 8 track and then dump to DAW so I could add more tracks...the whole bouncing thing just doesn't work for me, though it can be OK if you are very careful and do a lot of pre-production so that you minimize the bounces and plan your sub-mixes accordingly.
Anyway...I was just wondering why you think a 4-track trumps a 16 or 24....?
Click to expand...

I like bouncing. In my endless quest to make records that sound like the '60s, I have realized that multiple tape generations contributed a good amount to the sound. I also discovered that in order to get the 'right' sound, you have to go all the way.

It also helps me keep my mixing under control. The idea that 'mixing' is separate from 'recording' is somewhat a modern concept. Back in the '60s, the mixing process (sometimes called things like 'final dubdown', etc.) was really just part of the recording process. Parts were added during mixdown on the fly to gain an extra track.

My recordings are actually quite layered, but I'm able to get 'em all into 8 or 4 tracks. Ultimately, it's all ending up mono or 2 track anyway; you have to squeeze everything in at some point ... I'd rather not wait until the very end.

There are also practical concerns. 16 & 24 track mixers are large and I don't have a lot of space. When I decided on getting a large-format deck, I decided for sure I wanted a Scully 280 because I wanted a '60s deck w/ mike preamps ... and, yes I had a 'magical feeling' about it as well. I would have preferred 8-track, but those units are just too large ... about 400 lbs, and the transport is just really wide & heavy. I also was coming from 1/2" 8-track, so I had a good amount of 1/2" tape on hand, etc.

oh, and I think 1/2" 4-track, bounces and all, still sounds a lot better than 1/2" 8-track with no bounces. You get extra hiss and noise, but it retains the thick, punchy quality lacking on the smaller-format decks.
 
mshilarious said:
Sorry, the technical quality of the test is too low for me to use. The 1kHz sine in sample #1 has about 5% THD. I think this is probably because the Roland input is getting overloaded, because the playback is sample #4 has less distortion, likely because its output is lower level (it is impossible for a following device to remove distortion, so the distortion in sample #1 must be occurring primarily in the Roland). Also on sample #4 in the silence between tracks there is a curious amount of residual signal from other tests that were tens of seconds distant.

About all I can tell is the Tascam is fairly well calibrated for frequency response . . .
Click to expand...

So you are saying that the digital device is ruining the quality? :D

I guess -4dB is still to 'close' to 0 for the digital recorder... This is why I think analog is so much better.

Could you still send or post some graphs for me to see? I'm interested how it performed.
 
The "you" in that post can refer to either you or VP. Mostly VP, because he is thick.

Very plainly: tape works fine, but I don't think all tape is remotely equal. You don't think that either, but VP seems to think that any tape is better than any digital. VP continues to insist digital has all kinds of flaws that it doesn't. He admits he has never tested the HF fidelity of his digital systems (he's getting to that), but still he knows that something is wrong. Well, if there is something wrong he needs to upgrade his converters--consumer-grade stuff of the '90s did have more significant passband attenuation, but still nowhere near as bad as cassette, and probably as good as some better tape machines. Still, lower quality filters mean more aliasing. So buy something produced for at least the prosumer market in the last 10 years, or if you go older stick with the professional gear.

The Ampeg engineers? I bet they could provide me with measurements and I bet it wouldn't take them longer than a week when they already had a scope on their bench. I bet they don't say silly things about digital sampling that can be proven false with basic tests.

I am pretty sure the Ampeg engineers would agree that neither tape recording nor digital recording results in a signal with infinite resolution, because that's physically impossible. They would also understand that no audio signal will ever approach infinite bandwidth, and for an acoustic signal that is even more preposterous.

I am fairly certain they would agree that tape has an effective bandwidth and dynamic range, just like digital does. I also have a pretty good feeling they realize that digital jitter is several orders of magnitude less of a problem than flutter--I mean it's right there in the specification that they wrote. I would also guess that they know 24 bit digital recording has greater dynamic range than any tape recorder--this is also in the specifications. And less crosstalk and less distortion. They would also know that digital has better LF response. None of this is remotely controversial to a design engineer.

They would also understand that digital signal analysis gear works within its design limitations (bandwidth and dynamic range are always stated for scopes, because you pay a *lot* more for more of it). They would quickly understand the validity, again within the constraints of signal bandwidth and dynamic range, of the converter vs. cassette tests I did on page 2 of the other thread. Why did I test cassette? Because people on that thread were claiming it was better than digital. I am fairly certain the Ampeg engineers would not agree that cassette is a higher resolution, higher fidelity medium than CD.

They might argue that the best quality tape recorders rival and even sometimes exceed the resolution of CD, and I might even agree with them. That can be possible with NR, but that's not a fair fight because CD can use NR too (and there is a little-used AES standard for it). I don't think they can or would argue that tape can exceed the resolution of a good, modern 24/96 converter, although I am willing to see that disproven by miro's test.

What else is there to argue about? I've said probably a dozen times that professional quality tape recordings sound fine. Ty Ford did a test where engineers could only distinguish the two by paying close attention to head bump (if they were fastidious they could have EQ'ed that out, shouldn't have been more than a few dB). Bobbsy referenced an earlier study with the same conclusion. To me, that does not prove that tape is the technical equal of high-resolution digital, just that both systems can exceed the ability of the human ear to discriminate--which is all that really matters. Yes, the ear is the final arbiter, not because it is better than test gear but because it places a limit on how good a system needs to be (bandwidth, dynamic range, distortion). Once a system greatly exceeds that limit, further improvement is not really required.

What exactly do you want me to challenge the Ampeg engineers with? I don't care which medium people like better, that doesn't interest me in the slightest. I also take no position of the efficacy of tape emulation algorithms because they don't interest me at all. I don't own one and likely never will.

As a general rule I think DSP can be very successful, but often it isn't because implementation is compromised. Take my UAD-1 card, for example, their only compressor plug (as of 2009 or so, I stepped off their train years ago) that has an acceptable level of aliasing is their LA2A. I don't know or care if it sounds like the real thing, I just need a compressor that doesn't alias, and that is a pretty good one. So clearly they know how to write a compressor plug that doesn't alias (they are pretty smart) . . . so why don't all of their plugs not alias? Beats me . . .I actually started to write my own, I might just finish it someday . . . that said, with their UAD-2 they seem to be devoting a lot more DSP to their plugs, so presumably they perform better now, I don't have one of those so I don't know. I bet I could get somebody to test one for me though . . .

Back to Ampeg . . . please quote my exact statement from this thread that you think contains a scientific error to which they would object. Maybe it's the one that tape recorders can be fully described by analysis. That's not quite the same thing as saying they *have* been, but I feel comfortable that they are probably familiar with an immense amount of research on that topic--most of which isn't laying around at my fingertips, which is why I asked people to test their recorders. The Ampeg engineers probably already know all of those answers, and could tell me in one post rather than dodging the issue for over a week. Or maybe they don't care to, that is their prerogative, in which case I would continue to learn what I can the usual way, through research.

I don't mean that statement to a VP-level of "infinite" silliness. I don't need to try to describe it down to the quantum level, just within its practical constraints of dynamic range and bandwidth. Yes, tape may have some errors that are not periodic, but they can still be well described and understood.

How about the guy on the other thread who said (paraphrased) that tape engineers eventually gave up on accuracy and just went for euphonics? Would the engineers agree with that? I doubt it, seems to me the best machines were the last ones made (high-end, of course). That statement could be an insult to an engineer.
 
Chilljam said:
So you are saying that the digital device is ruining the quality? :D

I guess -4dB is still to 'close' to 0 for the digital recorder... This is why I think analog is so much better.

Could you still send or post some graphs for me to see? I'm interested how it performed.
Click to expand...

It's not that -4dBFS is digitally clipped, that doesn't happen (unless there is a setting for digital attenuation after the ADC input, which there might be). So it's either an overload of an analog buffer stage or that. Or maybe the line output is really dodgy.

The odd thing is the difference in performance between 100Hz and 1kHz--both digital and analog circuits tend to be very flat response, certainly across that bandwidth. The peak level is the same, the 1kHz is badly distorted but 100Hz is not. That is rather unusual, to say the least . . .
 
lonewhitefly said:
It also helps me keep my mixing under control. The idea that 'mixing' is separate from 'recording' is somewhat a modern concept. Back in the '60s, the mixing process (sometimes called things like 'final dubdown', etc.) was really just part of the recording process. Parts were added during mixdown on the fly to gain an extra track
Click to expand...

Yeah, but now I have to choose between Beach Boys in stereo and Beach Boys in mono with all of the vocals :(
 
OK, here are my carefully formulated questions for the Ampex engineers:

- flutter specs do not seem to improve much as recorder quality (using price as a proxy) improves. That seems counterintuitive; to what degree are actual unweighted improvements in flutter masked by the spec? Does the weighting really make sense? Does it favor the lower end for some odd reason?

- Is intermodulation with bias frequency a practical issue, and if so what methods of filtration are commonly employed? I note that Ampex tended to push the bias frequency pretty high maybe to avoid that. Why didn't others? Isn't that kind of the low hanging fruit? Or does that really only matter if bandwidth >20kHz is desired?

I think that's about it . . . both of these questions get at the primary flaws of digital: that of bandwidth limitation and aliasing. If tape suffers from similar concerns, and the magnitude is equal or greater than digital, then that's really about it for VP's arguments.
 
OK, looking at the 100Hz sine again that was a useable signal, so you've earned at least $20. Probably also the white noise, I am not posting that because of the weird 100Hz bleed-though, but there were no other surprises there, so let's say $40.

Here is 100Hz--this isn't the best digital performance ever, but not too bad for a computer line out. THD, mainly 2nd and 3rd, of about 0.03%. The Tascam manages a very respectable 0.4% at +3, all odd-order, mostly 3rd.

(as an aside, especially guitarists seem to think that only even-order distortion is good, but then they will say that it's the output stage of a tube amp that is good, where a push-pull amp will mostly be odd-order, and so is tape saturation, and transformer saturation, etc. I dunno, I like some 3rd and 5th, sounds reedy to me)

Flutter is reasonable, seems to be at 6.2Hz, about 0.8% unweighted (the little peaks that float around -52dB, right around the 100Hz peak). Also there is a smaller set of peaks +/-16Hz, but those are low enough that they don't affect the total measure. I read up on weighting but I'm still a bit vague on it, so I'll leave it there.

I don't have any idea what the 72Hz peak is. Maybe some sort of mechanical vibration or VLF interference. If it's constant and low-level it's very hard to hear against a signal anyway, I usually don't worry about it.

In general I would say this is the sort of thing people want out of tape, the LF saturation. I have also said I would just do that with a transformer, which doesn't do some of the other things tape may do, but this is all we are looking at for the moment:

100Hz.GIF
 
mshilarious said:
You both implied it, so stop the whining.

There is this amazing invention called "language", humans created it so we could share the knowledge of our experiences, which avoids the requirement that every human being have every experience in the world to understand the knowledge gained from that experience. You should try it sometime.

For example, because I own a Porta you both assumed that I took all tape quality to be equal. That is manifestly false, as evident in the many, many posts on the other thread where I said positive things about professional analog recordings that you somehow refuse to acknowledge. In fact, from that experience (wait for it) I *know* that larger decks must test better. I mean it's right there in the specification section in their manuals even if I had never heard a recording made with one.

In this thread I am having other people run the bench tests I would if I had the gear laying around. Unlike you, I have no desire to collect more machines than I can reasonably use in a single session, so there is no other way for me to gather that information (this is faster and cheaper too). By sharing images and text, I will gain a fuller understanding of the technical specification of this equipment.

What's even better is that based upon my experience in my primary occupation of analog circuit design, I will know how the equipment sounds based upon its test results and nothing else. I know you two think that tape is magical, but it isn't really, if it was you probably wouldn't use it because its results would be unpredictable. One day you'd get a nice warm sound and the next day a tape reel would turn into a bouquet of roses!

Whatever it does is predictable, repeatable, and thus can be fully described through analysis. I know you don't like that idea, but I have bad news: the engineers that designed these recorders believed that.
Click to expand...

Incorrect Mr Strawman!

VP
 
Humans did not invent language eh? It was probably dolphins then. Just this week I was watching an episode of Flipper! On Youtube though, so Flipper was digital, and pretty heavily compressed digital at that.

I love Flipper.
 
I wish I could participate in your project and I don't care about the $100 even. It would just be fun to see what you see coming off my Ampex MM-1000, but everybody pretty much knows its not yet operational. Maybe when it is I can contact you and do the test for funsies. The 3M M-64 halftrack would be interesting too.
 
Yeah, go ahead whenever I will analyze as many as are posted, and test file is on my server so it ain't going anywhere.
 
Well it ain't *my* server, just some space I rent. So long as that company doesn't piss off Anonymous I should be OK . . .
 
from what i know about human nature, you would get a ton of more participents if you substituted one or two of your mics for the $100.
Even if the retail value of the mics were less.
 
zorf said:
from what i know about human nature, you would get a ton of more participents if you substituted one or two of your mics for the $100.
Even if the retail value of the mics were less.
Click to expand...

This was about crass bribery and not promotion. Anyway, I've had two, miroslav will participate when he is ready.
 
I will participate if I am "Eligible". I need the money for a deposit on a "Spectrum Analyzer"

VP
 
There are plenty of freeware spectrum analyzers and I'm not buying gear for anybody. You just need to have a converter capable of 96kHz operation and the ability to record a loop through it.
 
You must log in or register to reply here.
Back
Top