Tape Tests

percy · 4th February 2008, 09:14 PM

Recently I was doing some audio performance tests on a "tape" - a ¼" analog magnetic tape. Was just curious to see what kind of response, distortion, noise, etc.. a tape really is home to. Have not really 'seen' it yet.

So I pulled my trusty RT-707 out of the rack and decided to do these tests in a loopback mode using RMAA on my PC. The recording was in monitor mode i.e. whatever was being recorded was getting played back right away. And that played back signal was being analzed by RMAA.

I have some questions about a couple of results. First lets look at the Noise/SNR. Just for reference I included the LineIn->LineOut loopback test of just the electronics (minus the tape) which is in green. The tape response is in white.

A couple of things that I'd like to figure out are -

What could be causing the sudden loss of snr below about 600hz ?

As you can see this is accompanied with a 10hz harmomic all across from 10hz to about 600hz, and then it suddenly disappears. What could this be ?

Could these two be related ? Although I think I probably know what that 10hz harmonic is. Its the motor (or the speed servo control) because I can hear a very faintly audible tick in the headphones/speakers and the audible occurense seems to match the visual occurence on a realtime scope function.

Another thing I am trying to confirm is -
While doing these test I realized that RMAA is using a Logarithmic sweep/tone signal to check frequency response. Would it be correct to say that a Logarithmic Sweep sine signal - the one in which amplitude level drops with increase in frequency - is unsuitable for measuring the frequency response of a tape ? (because of the tape's variable sensitivity to frequency ?)

Heads had been cleaned and properly demagnetized before start of these tests. The tape however was an old but lightly used Maxell UD35-90, visibly in good condition nevertheless.

TiA!

oshifis · 5th February 2008, 11:40 AM

percy,

The low frequency "noisiness" is probably related to FFT windowing. Have you tried using pink noise instead of sweep tone? That should give more realistic view of frequency response.

percy · 5th February 2008, 12:51 PM

this is a noise level test - there is no signal.
And it is not fft windowing. It is a very clear 10hz harmonic. Otherwise why isnt windowing a problem in the "reference" (green) chart ?

oshifis · 6th February 2008, 11:19 AM

Hm, sorry then. Could you upload the graph of an 1 kHz test tone recorded at 0 dBm?

percy · 6th February 2008, 03:43 PM

Why 0dbm ? Wouldn't that be too much for that tape ? I'd imagine -10db to -20db would be where I'd like the signal to be ?
I can post the thd measurement later today, but how would that help ? I guess I am trying to understand what exactly are you looking for from the thd measurement ?

oshifis · 6th February 2008, 04:50 PM

You will see the distortion and noise characteristics at 0 dB. 0 dB gives 3% distortion if well calibrated. The S/N can also be evaluated from that graph. On the other hand, frequency response can be measured at -10 dB (reel-to-reel) or at -20 dB (cassette).

percy · 9th February 2008, 04:02 PM

ok here's the thd image. I took three measurements from -10 to -20db. The only difference at higher levels like close to 0db is that the skirt widens.
What does the skirt indicate though ? tape saturation ?

percy · 9th February 2008, 04:04 PM

I'd also like to post the waveform of the spectrum.
This is a small portion of the recorded waveform - not zoomed in all the way - as I wanted to show you the "notches" at the top and bottom of the waveform.

What does that indicate ? An issue with the tape, or the heads, or bias ??

Higher the frequency of the signal the more rougher they get. As you can see in the next one which is a 16Khz signal.

percy · 9th February 2008, 04:13 PM

16khz -

oshifis · 11th February 2008, 10:00 AM

Percy,

The "skirt" is phase modulation of the test signal, caused by wow and flutter. It does not widen as the level is increased, just shifted upwards along with the test tone. If that is a Pio RT707, the dual-capstan explains the low level of wow and flutter. The notches in the waveform indicate worn tape, inadequate tape tension or both (dirty head excluded). Another interesting test would be to connect the L/R channel outputs to the X/Y inputs of an oscilloscope. You should get a stable 45° straight line. If there is some phase difference between the channels or the playback head is misaligned re the record head, the line becomes elliptical. It is a tough test at 16 kHz... Also the playback level at 1 kHz and at 16 kHz should be about the same, if recorded at -10 dB.

4th February 2008, 09:14 PM	#1
percy diyAudio Member Join Date: Apr 2004 Location: MN	Tape Tests Recently I was doing some audio performance tests on a "tape" - a ¼" analog magnetic tape. Was just curious to see what kind of response, distortion, noise, etc.. a tape really is home to. Have not really 'seen' it yet. So I pulled my trusty RT-707 out of the rack and decided to do these tests in a loopback mode using RMAA on my PC. The recording was in monitor mode i.e. whatever was being recorded was getting played back right away. And that played back signal was being analzed by RMAA. I have some questions about a couple of results. First lets look at the Noise/SNR. Just for reference I included the LineIn->LineOut loopback test of just the electronics (minus the tape) which is in green. The tape response is in white. A couple of things that I'd like to figure out are - What could be causing the sudden loss of snr below about 600hz ? As you can see this is accompanied with a 10hz harmomic all across from 10hz to about 600hz, and then it suddenly disappears. What could this be ? Could these two be related ? Although I think I probably know what that 10hz harmonic is. Its the motor (or the speed servo control) because I can hear a very faintly audible tick in the headphones/speakers and the audible occurense seems to match the visual occurence on a realtime scope function. Another thing I am trying to confirm is - While doing these test I realized that RMAA is using a Logarithmic sweep/tone signal to check frequency response. Would it be correct to say that a Logarithmic Sweep sine signal - the one in which amplitude level drops with increase in frequency - is unsuitable for measuring the frequency response of a tape ? (because of the tape's variable sensitivity to frequency ?) Heads had been cleaned and properly demagnetized before start of these tests. The tape however was an old but lightly used Maxell UD35-90, visibly in good condition nevertheless. TiA!

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
New 7"driver tests	ucla88	Multi-Way	4	2nd February 2007 02:01 PM
X-BOSOZ first tests	wuffwaff	Pass Labs	116	16th January 2006 04:00 PM
Some tweeter tests	ucla88	Multi-Way	18	19th April 2004 04:49 PM
First tests with my 5ch GC	Mika	Chip Amps	39	24th November 2003 01:05 PM
Listening tests	Rob M	Everything Else	37	4th February 2003 04:37 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

5th February 2008, 11:40 AM	#2
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	percy, The low frequency "noisiness" is probably related to FFT windowing. Have you tried using pink noise instead of sweep tone? That should give more realistic view of frequency response.

5th February 2008, 12:51 PM	#3
percy diyAudio Member Join Date: Apr 2004 Location: MN	this is a noise level test - there is no signal. And it is not fft windowing. It is a very clear 10hz harmonic. Otherwise why isnt windowing a problem in the "reference" (green) chart ?

6th February 2008, 11:19 AM	#4
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	Hm, sorry then. Could you upload the graph of an 1 kHz test tone recorded at 0 dBm?

6th February 2008, 03:43 PM	#5
percy diyAudio Member Join Date: Apr 2004 Location: MN	Why 0dbm ? Wouldn't that be too much for that tape ? I'd imagine -10db to -20db would be where I'd like the signal to be ? I can post the thd measurement later today, but how would that help ? I guess I am trying to understand what exactly are you looking for from the thd measurement ?

6th February 2008, 04:50 PM	#6
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	You will see the distortion and noise characteristics at 0 dB. 0 dB gives 3% distortion if well calibrated. The S/N can also be evaluated from that graph. On the other hand, frequency response can be measured at -10 dB (reel-to-reel) or at -20 dB (cassette).

11th February 2008, 10:00 AM	#10
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	Percy, The "skirt" is phase modulation of the test signal, caused by wow and flutter. It does not widen as the level is increased, just shifted upwards along with the test tone. If that is a Pio RT707, the dual-capstan explains the low level of wow and flutter. The notches in the waveform indicate worn tape, inadequate tape tension or both (dirty head excluded). Another interesting test would be to connect the L/R channel outputs to the X/Y inputs of an oscilloscope. You should get a stable 45° straight line. If there is some phase difference between the channels or the playback head is misaligned re the record head, the line becomes elliptical. It is a tough test at 16 kHz... Also the playback level at 1 kHz and at 16 kHz should be about the same, if recorded at -10 dB.