Too much tape hiss when recording (R2R)

Maros

Maros

New member
Hi everyone,

this really bothers me. I have 8-track Tascam ATR60 calibrated to 250nW/m. When I playback all 8 tracks and I am using a NEW RTM SM911 blank tape the tape hiss is quite normal. Nothing strange. When I hit record for all 8 track tape hiss is really significant even all inputs on mix console are muted. When I change calibration to +4dB above 250nW/m situation is really better, of course. Also I understand that unused track should be muted but I am thinking about the worst scenario. Tascam manual says that this machine should be aligned to 250nW/m.

I have NR but I do not want to use it for now. My question is, if ATR-60-8 "must" have NR enabled and it was common practise? Because I dont know imagine record something on that machine insted of metal genres :D

Maybe some bias problems? I set it around -3dB lower after peak. Or different tape? Heads are demagnetized.

Today maybe I record some samples:
1. Hiss of blank tape (all tracks, 250nW/m)
2. Hiss of recorded tape (all tracks, muted inputs, 250nW/m)
3. Like 1 but +4dB according to 250nW/m.
4. Like 2 ------- // ------------ // -------// --

All samples will be aligned to 0VU when mixdown so it will be like master record. I am quite new in analog recording and I also have Studer A80 multitrack but I am waiting for calibration tape so any test will be after calibration :)

Thank you!
 
What do you mean "even if all inputs on the mix console are muted?" You mean with all input channels muted you suddenly hear noise? How?

The other thing is that I think you are saying when you replay virgin tape that's never been recorded on you are his free, but replaying a tape that's been recorded has loud hiss? This suggests you are actually recording noise?

You'll have to explain a bit more this is very confusing?
 
rob aylestone said:
The other thing is that I think you are saying when you replay virgin tape that's never been recorded on you are his free, but replaying a tape that's been recorded has loud hiss? This suggests you are actually recording noise?
Click to expand...

Yes, thats what I am talking about.

With muted channels I mean that noise its NOT from mix console, or microphones. Also if TASCAM is in INPUT mode there is no hiss transfered to outputs, so its definitely from recording process. When I record silence there is BIG step in hiss.
 
Maros said:
Yes, thats what I am talking about.

With muted channels I mean that noise its NOT from mix console, or microphones. Also if TASCAM is in INPUT mode there is no hiss transfered to outputs, so its definitely from recording process. When I record silence there is BIG step in hiss.
Click to expand...

Why do you want to record silence?
 
You should apply a normal level audio signal to the track(s)...record that, and compare the signal level to the tape hiss level.

Way back, when I got my first "serious" console...just to amuse myself when checking it out for the first time...I pushed all the channel faders up without any audio signals, and I noticed that as the faders were all raised the hiss noise was pretty obvious.
I thought something might be wrong with the console...called the company and explained...and the tech said that I was creating a situation that doesn't exist with normal use...because when applying audio signals, the level of the audio would be much greater than the background hiss noise with the faders up.

That's what they mean by signal-to-noise ratio.
 
There’s also always more noise when connections are unterminated. Mic amps in particular, when “running wild”, are noisier than when you have a mic plugged into them. Don’t know if that is at all at play here.

You *did* calibrate the machine and set the bias, correct? What kind of record frequency response are you getting when you record and reproduce your tone ladder?
 
I recorded something yesterday but I leave it in studio so today I will upload it for you today. It is recorded like this:

5sec - 1kHz / +4dBu signal to all tracks from generator and all tracks REPRO mode. Calibration 250nW/m.
5sec - 1kHz / +4dBu signal to all tracks from generator and all tracks REPRO mode. Calibration +4dB over 250nW/m.
5sec - Generator is switched off but still recording silence and all tracks REPRO mode. Calibration 250nW/m.
5sec - Generator is switched off but still recording silence and all tracks REPRO mode. Calibration +4dB over 250nW/m.
5sec - Virgin, blank tape, all tracks REPRO mode. Calibration 250nW/m.
5sec - Virgin, blank tape, all tracks REPRO mode. Calibration +4dB over 250nW/m.

Mixdown is set to -0.5dB on my soundcard so it is like mastered recording for really correct comparision of S/N ratio :)


You wil clearly see BIG difference in hiss figure. And why I am talking about this? I just want to know (from guys with skills) if the hiss is OK or there is problem with ATR-60-8. I really like +4dB over 250nW/m. I also recorded drums, guitar and bass and after mixdown hiss is realy low and I cannot notice it (rocknroll song). So maybe its OK and I care too much because I am new it this teritory :)

Stay tuned.
 
I suspect you are applying digital thinking to an analogue problem. Hiss was always considered a problem when it intruded on the signal, but I don't remember ever trying to quantify hiss like this, because analgue hiss was never a ruler flat line on a screen anyway - we used to get quite annoyed about using the wrong response curves for the tape, just because some of them seemed to help, or hinder. Find a piece of music where the peak level is close to maximum, but it has some very quiet sections - CD mastered classical music. Ravel's Bolero is a great one to choose. This piece frequently caught out hobbyist recordists because on tape you could not record the dynamic range as written. Record this and see how it sounds. Your test conditions don't appear to provide much useful data. If you record a tone and then cut the tone generator you can measure the available dynamic range. If in your view it's not enough, or too much, then you do the recovery with fingers on faders or in automation in or outside a computer - but we're talking about ratios aren't we - so all removing the signal does is leave a somewhat meaningless noise figure. Realistically, the only purpose seems to be a subjective one. would it not be more revealing to record music with a peak level near maximum, turn up the monitors and then kill the music and see what kind of control room volume the hiss is, compared to the peak level.

I suspect most of my test tapes I used in the 70s were probably far from any magnetism level detailed printed on them for first use. I don't remember it even being a problem? Tweaking bias current, experimenting with levels and eq curves was part of the technique we all developed quite individually. I had a Ferrograph test set obtained from the BBC in a clear out. Never really found it that helpful as it was by that time uncalibrated, and I could not afford to have it re-done. It was great for tracking gradual fall-off, and of course great for re-aligning when you'd repaired something but I always viewed it as subjective. All we had were ears back then.

When we hear your samples, we'll listen and probably say it's fine. If we say it's dreadful, then you have a proper fault, or a serious flaw in your gain staging. I suspect we'll listen and say it sounds like tape, and is fine.
 
rob aylestone said:
Your test conditions don't appear to provide much useful data. If you record a tone and then cut the tone generator you can measure the available dynamic range.
Click to expand...

Why? I think that its really useful. It is really S/N ratio because there is Signal with nominal level (nominal +4dBu input signal = 0VU on the meters = +4dBu on the outputs). After that I set gain on mixing console to 0VU for all 8 tracks. After that I equally set all 8 faders to get 0VU on the master section of the console and after that I set -0.5dB signal on my soundcard. So all levels are set properly and I could "measure" by ear hiss (Noise) from recording. And this is my S/N ratio.

Maybe I know what are you trying to suggest me - Nevermind the hiss if whole song sounds OK. I was surprised because when I listen some analog recording I never notice hiss too much even in quiet places. But when I recorded someting I really got hiss to much from multitrack. Ok, maybe they have NR enabled but I do not want to enable it. And also I do not want to reduce like in digital world. I really know that hiss is part of the tape. I just want to hear "Yeah man, thats OK and without NR it never be better" or "SM911 tape hiss more then SM900 for the same flux setting" or "its too much and my tascam, otari or similar 1/2" track doesnt hiss like yours"

:)
 
Remember, that setting the calibration for a single track gives the lowest noise level. But when you enable all 8 tracks, the noise level is additive, and you will get a significantly higher noise floor. Thats one reason that DBX and Dolby were almost universally used with multitrack recorders. It was the only way to avoid the increases in tape hiss when you arm 8, 16 or 24 tracks at once.
 
I’m still wondering about your record and playback frequency response. 1K tone doesn’t tell you anything about that. Maybe you need to trim your HF eq or your bias level is too low. You did a full calibration?
 
Of course, I did. There is no problem with it. Equalisation is IEC1. Now I listen it again and hiss seems to be normal for tape :)

Anyway, this is my "hiss" track:

View attachment hiss.mp3

5sec - 1kHz / +4dBu signal to all tracks from generator and all tracks REPRO mode. Calibration 250nW/m.
5sec - 1kHz / +4dBu signal to all tracks from generator and all tracks REPRO mode. Calibration +4dB over 250nW/m.
5sec - Generator is switched off but still recording silence and all tracks REPRO mode. Calibration 250nW/m.
5sec - Generator is switched off but still recording silence and all tracks REPRO mode. Calibration +4dB over 250nW/m.
5sec - Virgin, blank tape, all tracks REPRO mode. Calibration 250nW/m.
5sec - Virgin, blank tape, all tracks REPRO mode. Calibration +4dB over 250nW/m.

(Sorry for sine quality and "ticking" noise, it is from the generator from the mix console.)
 
I thought that inadequate bias would cause the tape hiss to be too much. I remember years ago when a friend of mine and i were replacing the record/play head in a recorder, the service information we had told us to increase the bias if we had too much hiss or the recording was too "bright." That information also said that we would lose high-frequency response if we overbias the tape. There is apparently a "sweet spot" where the signal from the bias oscillator is strong enough to put the tape in the portion of its curve to give a linear response to the signal being recorded. ,
 
hautbois16 said:
I thought that inadequate bias would cause the tape hiss to be too much. I remember years ago when a friend of mine and i were replacing the record/play head in a recorder, the service information we had told us to increase the bias if we had too much hiss or the recording was too "bright." That information also said that we would lose high-frequency response if we overbias the tape. There is apparently a "sweet spot" where the signal from the bias oscillator is strong enough to put the tape in the portion of its curve to give a linear response to the signal being recorded. ,
Click to expand...

This is all correct, which is why I was asking the OP if the bias had been set for the tape that is being used. But the extension of that is asking about the playback and record frequency response, what the response curve looks like when tone ladders are played off a test tape, and then what those also look like reproducing from tones recorded to the tape being used. There will be more hiss without n/r on a narrow-format machine, but I was trying to ask some guiding questions to determine if the frequency response is off and could be tweaked for better performance, but the OP seems hesitant to reveal that information.
 
That level of machine should not give you that much noise but that was exactly why the dBx noise reduction was made for.
Noise from many tracks will be additive and so as you use 8 tracks then the noise level will increase.
If you do not want to use the noise reduction then you are saying you want this hiss.
It is also possible that the play EQ is set improperly as this will increase tape hiss noise. Did you do the job right with a MRL tape or do things like
many amateurs and start in the middle of the calibration process? The signal to noise ratio can be measured per each track and determined if outside the norm. At +3sB or 250nWb/M flux density the S/N ratio should be around 50 unweighted or better.

If you use dBx the S/N ratio as well as headroom is enhanced greatly and I have used it myself to do demo tape and a properly working system will not be a problem. I am talking the newer dBx that that would have had with it if not the DX-4D.

You might be able to set the machine depending on speed to a higher flux density of maybe 320nwb/M or 370nWb/M as I have seen switches for this.
 
sweetbeats said:
... I was trying to ask some guiding questions to determine if the frequency response is off and could be tweaked for better performance, but the OP seems hesitant to reveal that information.
Click to expand...

I did calibration with MRL tape and I also set -3dB bias, because of SM911 datasheet. Now, I downloaded a manual, where there are "a new pages" for me and in my previous manual they were missing. So now I know that there is REC EQ trim pot and today I will check overal frequency response. So thank you for going more deep into the calibration. I really set levels, IEC equalisation and bias and then my manual ended :( Maybe this could cause my hiss problem if there is some bump on HF.

skywaveTDR said:
Did you do the job right with a MRL tape or do things like
many amateurs and start in the middle of the calibration process? The signal to noise ratio can be measured per each track and determined if outside the norm. At +3sB or 250nWb/M flux density the S/N ratio should be around 50 unweighted or better.
Click to expand...

I have MRL calibration tape. Honestly, I did not measure S/N ratio but I can check it. Because for me it was not important what meter says but what my ears heard and it seemed like too much hiss. Thats why I posted this thread. But I got your point. If its in the user manual, any debate around it is useless, because I hit the limit :)

skywaveTDR said:
You might be able to set the machine depending on speed to a higher flux density of maybe 320nwb/M or 370nWb/M as I have seen switches for this.
Click to expand...

In my post above with hiss.mp3 track you can see that I re-calibrate to +4dB over 250nW/s, so ~396nW/m and the result is more satisfying, of course :)

Today I will try to set bias differently but in the time of calibration I get really best results in case of noise (by ears). I downloaded a new manual with overall frequency information and REC EQ trim pot which I had no idea because of missing pages. You know, if this situation is really new for me I will make mistake easily, because I really do not know what I have to be aware. Thats why I am asking help or share my results/ideas with older, experienced guys like you. Its very important for me, because its hard to find relevant information about "this era of tape".

Thanks
 
Last edited:
This seems to be one of the techno rules topics where partial physics gets applied with no appreciation of real world results. I think most of us experimented with increasing and decreasing bias current when presented with tape that was new to us, or often just Unknown. I failed to get an engineering job at the BBC because my understanding of the purpose of bias current was misunderstood, demonstrated at the interview when I discovered the interviewer was standing in and was head of engineering! Finding the elusive linear portion of the S curve was the aim, no matter if this meant adjustments not on the list. Bias current was not to be considered a means to reduce hiss, but a method to produce linearality, which had hiss reduction as a common but not absolute by product. Music content is key. A classical recording might be advantaged by a subtle adjustment one way, but a speech recording needed something different. All the noise reduction systems needed precise alignment or they simply function as eq. So many people recorded with Dolby B on and then replayed with it off because they liked the result with a bit of HF tone control adjustment, and I had a reel to reel that disabled, by licence, Dolby B at 15IPS. I tried disabling it, but it wrecked compatibility with other machines, so I put it back.

In this topic there seems no mention at all of replay sound quality with these adjustments. Tones let you measure at specific frequencies, for adjustment, but you also need to listen, and I see little mention of this happening, just measuring?
 
Are you talking about tape hiss that only occurs while recording, in sync? Or while playing back what you recorded previously?
 
lonewhitefly said:
Are you talking about tape hiss that only occurs while recording, in sync? Or while playing back what you recorded previously?
Click to expand...

Both and it seems to be equal and higher then hiss from the blank tape. I have to try REC EQ setting, but I did not have a time today. When I was doing calibration of the playback/bias setting I had connected oscilloscope and AC milivoltmeter. Unfortunately I did not check overall frequency response so there could be a problem with HF (I did not have complete user manual I really did not realize it. Now, its clear for me and I dont know why I was not thinking about that, it is so obvious :) )

Did you listen my sample?
 
You must log in or register to reply here.
Back
Top