Analog Tape specs

If one can pinpoint the answer to this query, then they will have also answered the same question about photography with film vs. digital. I've done photography for 22 years and I can see the difference between Kodak, Fuji, and Agfa films, but I couldn't explain why they look the way they do or what makes them different. Something with the composition of the silver halide emulsion I assume. Just like analog tape changes the sound when it is saturated, film changes when it is saturated, or overexposed. Film looks better when it is a touch underexposed and analog tape sounds better when it is a touch saturated. Go figure.

I know, off topic but I think it's relevant.:p
 
bozmillar said:
I started this thread in a different section, the realized it probably belongs here.

I've read about a million threads on analog vs digital, but they seem to generally be religious arguments rather than factual. I do not want to start another one. What I do want are some cold hard specs on reel to reel sound quality. SNR, Dynamic range, frequency response, THD and the likes.

We all know the how this works in digital audio, but as far as analog tape goes, I've never seen a spec written. Can somebody please provide this?

Also, the same for vinyl.

(I don't want this to turn into an analog vs digital argument. I just want some real information. The words warm, brittle, sterile and so forth are useless terms. Please don't use them.)
Click to expand...

Too many parameters still not measurable or difinable to state the nature of the analogue sounds superiority to digital sound.

To justify my statement:

Give me a specific measurement and definition on:
Attack
Room Definition
Depth
Width
Textural effect
Instrument recognition
Distinction of instruments
Voice naturality

You can continue yourself.

It's not a religion.
It's just beyond sience to explain why the human ear choose analog sound as the best and most natural when equipment and media conditions are present to pursuit the potential.

"dolph"
 
niklasthedolphi said:
Too many parameters still not measurable or difinable to state the nature of the analogue sounds superiority to digital sound.

To justify my statement:

Give me a specific measurement and definition on:
Attack
Room Definition
Depth
Width
Textural effect
Instrument recognition
Distinction of instruments
Voice naturality

You can continue yourself.

It's not a religion.
It's just beyond sience to explain why the human ear choose analog sound as the best and most natural when equipment and media conditions are present to pursuit the potential.

"dolph"
Click to expand...

Especially well put. Thanks.

--
 
The simple fact is that sound is a PURELY ANALOG PHENOMENOM. Analog recorders capture the COMPLETE waveform, albeit in a non-linear fashion due to the properties of magnetic tape. Digital recording has extremely linear characteristics but fails to record the compelte waveform and in fact leaves out large amounts of the information, obviously dependent on the sampling rate.

Let's take a quick example:

Compact Disc Audio reproduces 20-20,000 Hz with a dynamic range of 96dB. Because of the sampling rate of 44.1 kHz, there is a very sharp roll off between 20Khz and 22.5 Khz, because nothing can be reproduced above that point AT ALL. Even worse, if a low-pass filter is not used when recording the original digital tracks (again assuming the 44.1 kHz sampling rate), an Aliasing frequency will be reproduced in the audible range of the digital recording. So if there is a frequency of 23.5 kHz that is present during the recording, it will be reproduced as a 1 khz aliasing frequencty on the recording. Again, filters are in place to prevent this, but my point is everything above 22,500 Hz is missing from the recording INCLUDING THE NATURAL OVERTONES. Obviously everything below 20 Hz is misssing as well.

Now let's tape an analog recorder - I'll use the spec sheet for my Tascam 42B since I have the manual here:

AT 15 IPS: 30Hz - 22kHz +/- 2dB Signal to noise ratio 70 dB


Already we see that the analog unit is not rated to go down to 20 Hz, but only 30...... EXCEPT the analog recorder DOES recording lower frequencies than than... they just are recorded outside the specification of +/- 2dB. The same is true of the higher frequencies in reverse: the analog recorder is still reproducing withing the 2 dB specification at 22kHz where the digital recorder has already been forced to roll off that frequency - to the point where it is likely inaudible. At even higher frequencies, the analog deck will continue to reproduce, albeit at a much reduced level, basically getting quieter and quieter the higher the frequency until the point where it fades out. This closely resembles the way the human ear actually works, which is why people say analog sounds 'better' or whatever word you want to use. All the natural overtones are also present on the tape in some fashion as well.


Now the above is VERY abbreviated; and there are many technical issues regarding how ATRs reproduce sound that can't be explained in a few paragraphs. If you REALLY want to learn how things work, both digital and analog, you need to do some reading or go to school and get a degree from a good recording program - like the University of Miami. You might also pick up and read a copy of The Recording Studio Handbook by John Woram; it is available used through Amazon.com.





AK
 
Last edited:
Beck said:
Not at all. Preoccupation with specs is an early phase people go through in sound engineering/recording. I started that phase when I was 17 and grew out of it before I was 20. There is comfort in thinking that one can quantify everything about sound, but in fact one cannot. Standard specifications as listed in the back of a manual are the tip of the iceberg -- a skeletal representation of the body that is really there. So there are things that the industry can't measure and things they just don't measure or disclose.
Click to expand...

ok, so maybe "specs" is the wrong word. When I started the thread, that's what I was looking for, and I've learned that that's not the only driver behind why it sounds the way it does.

If I were to try to make a plugin that emulated tape, I would need to know everything possible about it. I don't need to know thinks like punchy and warm or "zen," I need to know why it's punchy or warm.

Beck said:
Your statement that you've never seen a spec written for analog tape only tells us you aren't familiar with the analog world, which is ok.
Click to expand...

True. That is why I'm here asking questions.
 
analogkid said:
The simple fact is that sound is a PURELY ANALOG PHENOMENOM. Analog recorders capture the COMPLETE waveform, albeit in a non-linear fashion due to the properties of magnetic tape. Digital recording has extremely linear characteristics but fails to record the compelte waveform and in fact leaves out large amounts of the information, obviously dependent on the sampling rate.

Let's take a quick example:

Compact Disc Audio reproduces 20-20,000 Hz with a dynamic range of 96dB. Because of the sampling rate of 44.1 kHz, there is a very sharp roll off between 20Khz and 22.5 Khz, because nothing can be reproduced above that point AT ALL. Even worse, if a low-pass filter is not used when recording the original digital tracks (again assuming the 44.1 kHz sampling rate), an Aliasing frequency will be reproduced in the audible range of the digital recording. So if there is a frequency of 23.5 kHz that is present during the recording, it will be reproduced as a 1 khz aliasing frequencty on the recording. Again, filters are in place to prevent this, but my point is everything above 22,500 Hz is missing from the recording INCLUDING THE NATURAL OVERTONES. Obviously everything below 20 Hz is misssing as well.

Now let's tape an analog recorder - I'll use the spec sheet for my Tascam 42B since I have the manual here:

AT 15 IPS: 30Hz - 22kHz +/- 2dB Signal to noise ratio 70 dB


Already we see that the analog unit is not rated to go down to 20 Hz, but only 30...... EXCEPT the analog recorder DOES recording lower frequencies than than... they just are recorded outside the specification of +/- 2dB. The same is true of the higher frequencies in reverse: the analog recorder is still reproducing withing the 2 dB specification at 22kHz where the digital recorder has already been forced to roll off that frequency - to the point where it is likely inaudible. At even higher frequencies, the analog deck will continue to reproduce, albeit at a much reduced level, basically getting quieter and quieter the higher the frequency until the point where it fades out. This closely resembles the way the human ear actually works, which is why people say analog sounds 'better' or whatever word you want to use. All the natural overtones are also present on the tape in some fashion as well.

AK
Click to expand...

This is one of the main reasons that I think 44.1kHz sampling rate is too low. If you want to go flat up to 20kHz, your LPF will have to knock down your signal by 96dB by the time you reach 22.05kHz to completely avoid aliasing. Any filter this sharp is going to cause some very unnatural effects.

So what I get from this is that a high of 20kHz doesn't imply a sharp filter at 20kHz which causes it to sound more natural. Makes perfect sense to me.

Out of curiosity, do you know how sharply tape cuts off on the high end? I'm asking because I don't know and I think it would be interesting to know (I've found I have to carefully phrase my questions so people don't take offense).
 
I agree that 44.1 sampling rate is far too low; I have been saying that since the standard was adopted :P.

How quickly the high frequencies roll off is dependent on several factors - the tape being used, the equalization standard, the quality of the recording and playback heads so there is no single answer. Looking at the specs for my Pioneer RT-707 - which is a consumer deck from the late 1970s - it shows a frequency response of 20 - 28,000 Hz at 7 1/2 IPS (30 - 24,000 Hz +/- 3dB). I am assuming that at 24 kHz the signal would be the full 3 dB down, and obviously there was some measurable signal at 28 kHz. It would seem both the Tascam deck and my Pioneer (at least in the test lab under perfect, controlled conditions) are reproducing sound at 22.5 kHz with only a few dB drop-off, while CD Audio has by definition completely cut off.


AK
 
Does anyone know how the frequency response of vinyl compares for comparision to tape?
Click to expand...

Well again I don't have specific answers about vinyl onhand - but I do know that high-quality vinyl can reproduce frequencies in excess of 100 kHz. I do have the specs for my turntable cartrigde though, and they show that (at least this cartridge) reproduces frequencies from 5 - 35,000 Hz.
 
I notice the highs in my digital workstation barely even make an affect on a track. When I crank up 10k on an electric guitar track, 12 db it barely affects it, it either adds a thin tinniness or doesn't affect it at all. This seems weird to me.
 
Preoccupation with specs is an early phase people go through in sound engineering/recording.
Click to expand...

I know this is very true for me...I judged equipment by S/N ratio and frequency response so many times and then started to realize that they weren't telling the whole picture. Its like these home theatre in-a-box things you can get for $100 that have 1000watts of power...:rolleyes: Right. What weighting? What's the distortion rating at the rated power? You won't find it in the manual. Specs can be used as a relative comparison I suppose but they are dangerous because they often leave out defining parameters.

Bozmillar, check out this article. It compares a bunch of different response curves for a bunch of different analog open reel decks. The interesting thing is that even though the upper and lower threshholds are generally comparable, what happens in between is where the rubber meets the road...dips and peaks in the response curve especially in the low-end can be in the "right" places (i.e. generally sound "good" to most ears) or in the "wrong" places...I'm putting the quotes in there simply because this is all subjective stuff, at least at my level of understanding and knowledge. So you can have a response of 30Hz to 22kHz +/-3dB or whatever, but if it is +3 at 400Hz and -3 at 100Hz you're going to be constantly compensating for that, at least in my experience.

Like what Beck is saying, and I didn't even really notice that I'd stopped looking at the "specs" for the most part since returning to analog, I really don't pay attention to them anymore, with the exception of mechanical specs or other operational type specs (i.e. determining the operating level and impedance of different inputs and outputs to properly interface my gear...that sort of thing).

I hope the Jack Endino article imparts some useful information. It certainly has for me.

Here's another thing to consider:

Now I may be wrong about this...The bias frequency on my Tascam decks is 145kHz...that's 145kHz that gets printed to tape, and I know it is on the tape because there are traps on the amp cards to make sure this 145kHz tone doesn't make it out of the deck...iirc it would mess with stuff like noise reduction units and such. So that's pretty incredible. Now I don't sit around wishing I could jam to 145kHz tones all day long, nor am I bragging that my gear goes "up to 11"...my point is simply that there is lots of information in frequencies that are outside of the human range that our bodies still respond to (psychoacoustics) about which, as mentioned above, we certainly have limited knowledge. So that is a major factor driving the sound of analog tape.

But if I had to pick one thing that defines "tape sound", it'd probably be the way it distorts/saturates. Its not going to show up on a spec sheet, so that's a tough one in answering your question.
 
"Have you ever noticed, browsing these forums, that the analog only section is so much more supporting and polite than the other sections? I think it's because we really love this stuff, and sharing experiences and information with others is never a hassle, there is no elitism."

thats exactly it man. EXACTLY it. i wish this guy wouldn't mess up the supportive flow with his ignorance.
 
johnbilben said:
"Have you ever noticed, browsing these forums, that the analog only section is so much more supporting and polite than the other sections? I think it's because we really love this stuff, and sharing experiences and information with others is never a hassle, there is no elitism."

thats exactly it man. EXACTLY it. i wish this guy wouldn't mess up the supportive flow with his ignorance.
Click to expand...

I don't think he's ignorant at all. He may be a little stubborn in ways, but he's not ignorant.
 
whether you guys think so or not, this is helping me out quite a bit. I know that these specs aren't the whole picture, and I know probably nobody can explain the whole picture, but part of the picture is better than none of it.
 
That Endino article is indeed interesting, HOWEVER....... ( I just love saying HOWEVER or ACTUALLY :P) Since Jack did not specify what kind of tape he used, which EQ parameters were used (NAB or IEC), nor did he specify how much under/over-bias was set one each machine. He also did not take the time to calibrate the machines for EACH SPEED TESTED, so what we really have is some neat Ross Perot-type graphs without any way to really analyze what they are telling us. I am NOT trying to knock what Jack Endino has said here...... I am just saying that his article leaves out a lot of information that we could discuss and would allow us to get an even better understanding of the performance of each of these decks. I would love to have access to those machines so I could do a full, scientific testing program on them. We could spend the next 20 years analyzing the results!


AK
 
AK,

Yuck. I do not want to analyze anything for 20 years! :)

I thought there was some info on Endino's bias practices and the tape used in the preamble to the article...? I'll have to read it again...

HOWEVER, :D I think it is a good smorgasbord of response curves for bozmillar and whomever to see.

You're absolutely right about the information that is not there, but it does show what each machine is capable of or how it can perform.
 
Sorry, Endino does say he used 499 and 456, he just doesn't specify which decks used what, so I assume he is using 456 on the Tascam decks as they really can't utilize the 499 properly. Really his article does make one thing clear: Analog is a tricky business at times, and you can't ever assume you will get a certain result without testing first.

AK
 
What it tells you...

What it tells you is that the analog frequency response curve can be within spec and simultaneously really not flat at all!:eek:;)
 
Yes!

What it tells you is that the analog frequency response curve can be within spec and simultaneously really not flat at all!
Click to expand...

That is spot-on, Dave.

I don't condemn bozmillar's pursuit of specification information, but I want him to realize the caveat of it, and that is what I took away from the Endino article...a deviation spec from xHz to xHz says very little. It says something, but leaves out much, and the response curves of several different decks may present the same info on paper, but sound very different.
 
You must log in or register to reply here.
Back
Top