Tape Hiss removal
Michael Dobony
dolby B. The software I originally used is Total Recorder, but I also have
Audacity. The tape is a Magnetics. Any hints on removing or at least
reducing the tape hiss, any addins for Audacity that could help? The tape
level is a little low, record level was at max and nowhere near clipping on
record. Translation: S-N level was a little low on the tape.
Peter Larsen
Do nothing if you can't do it right.
Kind regards
Peter Larsen
Steve King
news:199l7iiw4v6x9.1...@40tude.net...
Is it your assessment that the original recording was made using Dolby B?
If it wasn't dolby will exacerbate the tape hiss. Beyond that tape hiss is
pretty evenly distributed from the mid-range up. It is difficult to reduce
with EQ without doing more damage to the recording than is justified. Like
Peter said, sometimes doing nothing is the best choice. It is what it is.
Steve King
Michael Dobony
Dolby B sounded best. I found out how to use Audacity to remove the hiss.
It sounds better, but hardly production. The instrumental is clear, but
many of the vocals are "hollow drum" sounding and there is a lot of
clipping. It was evidently recorded with too high a gain on channel
inputs, but recorded on the tape at low levels. I couldn't get my recording
to clip at all. The tape meter never went above -3, nowhere near 0. I doubt
it was recorded in a studio. The mismatch between the vocals and
instrumental is too drastic in quality and sometimes in levels. Guitar was
probably on a pickup, but everything else was poorly miced.
John Williamson
You *might* get a marginally better result by using whatever noise
reduction plugin you're using to do the reduction in stages, say
removing a quarter of the noise with each pass. There are also peak
restoration plugins for the clipping, but their success rate is
variable. Possibly an expander might help pull some of the higher level
signals out of the murk, but will add a whole new set of problems.
Intelligibility can be improved, but that's probably about it.
Sorry not to be more hopeful.
--
Tciao for Now!
John.
Ian Bell
> "Michael Dobony" <sur...@stopassaultnow.net> wrote in message
> news:199l7iiw4v6x9.1...@40tude.net...
> |I was given a tape to convert to CD. My player is a Pioneer CT-W205R set to
> | dolby B. The software I originally used is Total Recorder, but I also have
> | Audacity. The tape is a Magnetics. Any hints on removing or at least
> | reducing the tape hiss, any addins for Audacity that could help? The tape
> | level is a little low, record level was at max and nowhere near clipping
> on
> | record. Translation: S-N level was a little low on the tape.
>
> Is it your assessment that the original recording was made using Dolby B?
> If it wasn't dolby will exacerbate the tape hiss.
No it won't. Dolby replay always reduces hiss not matter how the tape
was recorded.
Cheers
Ian
Peter Larsen
> Dolby B sounded best.
Playback with dolby b enabled and then boosting the treble can also be one
of the better ways to get rid of noise with a non-dolby tape.
> I found out how to use Audacity to remove the
> hiss. It sounds better, but hardly production. The instrumental is
> clear, but many of the vocals are "hollow drum" sounding
Noise is uncertainty, and if you want to get rid of it you have to exchange
that uncertainty for some other untertainty, in this case a temporal one,
ie. fake and not very well sounding reverb.
> and there is
> a lot of clipping. It was evidently recorded with too high a gain on
> channel inputs, but recorded on the tape at low levels. I couldn't
> get my recording to clip at all.
It's not "simple", problem tapes rarely are, but "advanced". Magix has a
modestly priced, wizard based, over the counter audio restoration package,
it may be more suited for the task and with no time for a learning curve and
no budget to pay somebody who has been on one that may be your best bet.
> The tape meter never went above -3,
> nowhere near 0. I doubt it was recorded in a studio. The mismatch
> between the vocals and instrumental is too drastic in quality and
> sometimes in levels. Guitar was probably on a pickup, but everything
> else was poorly miced.
So basicly this is useless and will never sound good, make a clean transfer
and suggest that it is used as basis for re-recording or ask what the
outsourcing budget is and tell them that it's gonna cost to try to make it
sound good and that there is no guarantee.
Kind regards
Peter Larsen
Scott Dorsey
1. Make sure the azimuth is correct. If the top end is muffled, you will
lose a LOT of S/N.
2. Make sure the Dolby levels are correct. If you are losing the top end
at times because the Dolby decoder is mistracking, you will lose a lot of
perceived S/N.
3. Take it to someone with a good multiband hiss reduction system. CEDAR
has a good one. Waves Restoration Bundle has one too.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Michael Dobony
> Michael Dobony wrote:
>
>> Dolby B sounded best.
>
> Playback with dolby b enabled and then boosting the treble can also be one
> of the better ways to get rid of noise with a non-dolby tape.
>
>> I found out how to use Audacity to remove the
>> hiss. It sounds better, but hardly production. The instrumental is
>> clear, but many of the vocals are "hollow drum" sounding
>
> Noise is uncertainty, and if you want to get rid of it you have to exchange
> that uncertainty for some other untertainty, in this case a temporal one,
> ie. fake and not very well sounding reverb.
>
Running it through Audacity's noise reduction one time made the vocals a
little clearer. EQ'ing the first few tracks helped out a lot. The rest are
not that bad, or rather, EQ'ing will not help them much. the first 5 or 6
needed more hi-mids and 2 need some lows (sub 100 hz).
>> and there is
>> a lot of clipping. It was evidently recorded with too high a gain on
>> channel inputs, but recorded on the tape at low levels. I couldn't
>> get my recording to clip at all.
>
> It's not "simple", problem tapes rarely are, but "advanced". Magix has a
> modestly priced, wizard based, over the counter audio restoration package,
> it may be more suited for the task and with no time for a learning curve and
> no budget to pay somebody who has been on one that may be your best bet.
>
>> The tape meter never went above -3,
>> nowhere near 0. I doubt it was recorded in a studio. The mismatch
>> between the vocals and instrumental is too drastic in quality and
>> sometimes in levels. Guitar was probably on a pickup, but everything
>> else was poorly miced.
>
> So basicly this is useless and will never sound good, make a clean transfer
> and suggest that it is used as basis for re-recording or ask what the
> outsourcing budget is and tell them that it's gonna cost to try to make it
> sound good and that there is no guarantee.
>
> Kind regards
>
> Peter Larsen
Budget, what I have on hand or free, for a tape that was a gift to someone
who does not have a tape player. This is for personal use, but the gift
giver wants me to make CD's for him to hand out. These are original songs
and very well done, artistically speaking. It reminds me of an old, 50's
"hot wax." Considering it is kind of southern Gospel style, the recording
quality almost compliments the songs, but it is a bit over the edge.
The group is retired, from what I hear, so there will be no re-recording.
It is what it is and all I can do is minimal cleanup. I would love to at
least mellow out some of the clipping, especially one of the songs.
Michael Dobony
> In article <199l7iiw4v6x9.1...@40tude.net>,
> Michael Dobony <sur...@stopassaultnow.net> wrote:
>>I was given a tape to convert to CD. My player is a Pioneer CT-W205R set to
>>dolby B. The software I originally used is Total Recorder, but I also have
>>Audacity. The tape is a Magnetics. Any hints on removing or at least
>>reducing the tape hiss, any addins for Audacity that could help? The tape
>>level is a little low, record level was at max and nowhere near clipping on
>>record. Translation: S-N level was a little low on the tape.
>
> 1. Make sure the azimuth is correct. If the top end is muffled, you will
> lose a LOT of S/N.
Varies with the track. Some are decent, some need work, mostly in the 1-4k
range and some in the under 100 need either boost or cut.
>
> 2. Make sure the Dolby levels are correct. If you are losing the top end
> at times because the Dolby decoder is mistracking, you will lose a lot of
> perceived S/N.
>
Again, varies with song.
> 3. Take it to someone with a good multiband hiss reduction system. CEDAR
> has a good one. Waves Restoration Bundle has one too.
> --scott
For the purpose of the transfer, Audacity did an adequate job of noise
reduction. The aforementioned peak restoration plug ins would have been
nice, but I was unable to find any for Audacity. Clipping really only hurts
on one song and would be helpful in about 3 or 4 of the other 14 tracks.
The rest have no noticeable clipping and Audacity doesn't find any,
possibly because the levels are good on the transfer. If I locate them
later I can hit the problem tracks.
William Sommerwerck
> at times because the Dolby decoder is mistracking, you will lose a lot
of
> perceived S/N.
Only if the mistracking is with the Dolby calibration level set too low.
Laurence Payne
Has anyone mentioned head alignment on the cassette player yet? A
quick Goggle reveals that the Pioneer mentioned is a dual-deck machine
with auto-reverse. I could suggest that, almost by definition, the
heads are not going to be aligned correctly, even with themselves!
Maybe there's a lot more to be got off the tape than he suspects?
Licensing issues aside, I find it inconceivable that a modern computer
can't emulate 30-year old Dolby technology. Someone must have done
it?
John Williamson
They have, there's a versatile plugin for Winamp (And the Linux
equivalent) that emulates Dolby A, B, and C, with all sorts of control
over levels and such. Winamp can then write a .wav file direct to HD.
DSP tape restorer live, I think it's called.
Scott Dorsey
>On 28 Nov 2009 18:12:16 -0500, Scott Dorsey wrote:
>
>> In article <199l7iiw4v6x9.1...@40tude.net>,
>> Michael Dobony <sur...@stopassaultnow.net> wrote:
>>>I was given a tape to convert to CD. My player is a Pioneer CT-W205R set to
>>>dolby B. The software I originally used is Total Recorder, but I also have
>>>Audacity. The tape is a Magnetics. Any hints on removing or at least
>>>reducing the tape hiss, any addins for Audacity that could help? The tape
>>>level is a little low, record level was at max and nowhere near clipping on
>>>record. Translation: S-N level was a little low on the tape.
>>
>> 1. Make sure the azimuth is correct. If the top end is muffled, you will
>> lose a LOT of S/N.
>
>Varies with the track. Some are decent, some need work, mostly in the 1-4k
>range and some in the under 100 need either boost or cut.
How do you know? Have you adjusted the azimuth on each track to peak
the top end?
If you lose it in playback, and you have to boost it afterward, what you
get is more noise. Set the machine for proper playback.
>> 2. Make sure the Dolby levels are correct. If you are losing the top end
>> at times because the Dolby decoder is mistracking, you will lose a lot of
>> perceived S/N.
>
>Again, varies with song.
It shouldn't, if they were all recorded at the same time on the same
machine. If it's different from track to track, you're going to have to
readjust the Dolby levels on each track to try and keep the pumping down.
When the noise floor pumps, it makes it that much more annoying and that
much harder to remove with broadband NR systems.
>> 3. Take it to someone with a good multiband hiss reduction system. CEDAR
>> has a good one. Waves Restoration Bundle has one too.
>
>For the purpose of the transfer, Audacity did an adequate job of noise
>reduction. The aforementioned peak restoration plug ins would have been
>nice, but I was unable to find any for Audacity. Clipping really only hurts
>on one song and would be helpful in about 3 or 4 of the other 14 tracks.
>The rest have no noticeable clipping and Audacity doesn't find any,
>possibly because the levels are good on the transfer. If I locate them
>later I can hit the problem tracks.
I found the Audacity NR to be a real problem as far as annoying artifacts
go.... but again your goal is to get as clean a signal _into_ the DAW as
possible and from that point on the differences are somewhat reduced.
Michael Dobony
Head alignment would not explain why the guitar is extremely clear, but all
the other instruments and vocals are off and only in a few songs.
Michael Dobony
> Michael Dobony <sur...@stopassaultnow.net> wrote:
>>On 28 Nov 2009 18:12:16 -0500, Scott Dorsey wrote:
>>
>>> In article <199l7iiw4v6x9.1...@40tude.net>,
>>> Michael Dobony <sur...@stopassaultnow.net> wrote:
>>>>I was given a tape to convert to CD. My player is a Pioneer CT-W205R set to
>>>>dolby B. The software I originally used is Total Recorder, but I also have
>>>>Audacity. The tape is a Magnetics. Any hints on removing or at least
>>>>reducing the tape hiss, any addins for Audacity that could help? The tape
>>>>level is a little low, record level was at max and nowhere near clipping on
>>>>record. Translation: S-N level was a little low on the tape.
>>>
>>> 1. Make sure the azimuth is correct. If the top end is muffled, you will
>>> lose a LOT of S/N.
>>
>>Varies with the track. Some are decent, some need work, mostly in the 1-4k
>>range and some in the under 100 need either boost or cut.
>
> How do you know? Have you adjusted the azimuth on each track to peak
> the top end?
Define "top end." Adjusting for every track? Get real. Also if the azimuth
was off the guitar would be off along with everything else. The guitar
sounds great. The rest of the instruments and the vocals are off.
>
> If you lose it in playback, and you have to boost it afterward, what you
> get is more noise. Set the machine for proper playback.
Are you certain it is lost in playback as opposed to lost in original
recording or re-recording? Boosting the hi-mids did not add noise.
>
>>> 2. Make sure the Dolby levels are correct. If you are losing the top end
>>> at times because the Dolby decoder is mistracking, you will lose a lot of
>>> perceived S/N.
>>
>>Again, varies with song.
>
> It shouldn't, if they were all recorded at the same time on the same
> machine. If it's different from track to track, you're going to have to
> readjust the Dolby levels on each track to try and keep the pumping down.
I doubt they were recorded at the same time.
>
> When the noise floor pumps, it makes it that much more annoying and that
> much harder to remove with broadband NR systems.
Again, the Audacity noise removal did an adequate job. With or without the
tape hiss, a few of the songs needed to be EQ'ed.
>
>>> 3. Take it to someone with a good multiband hiss reduction system. CEDAR
>>> has a good one. Waves Restoration Bundle has one too.
>>
>>For the purpose of the transfer, Audacity did an adequate job of noise
>>reduction. The aforementioned peak restoration plug ins would have been
>>nice, but I was unable to find any for Audacity. Clipping really only hurts
>>on one song and would be helpful in about 3 or 4 of the other 14 tracks.
>>The rest have no noticeable clipping and Audacity doesn't find any,
>>possibly because the levels are good on the transfer. If I locate them
>>later I can hit the problem tracks.
>
> I found the Audacity NR to be a real problem as far as annoying artifacts
> go.... but again your goal is to get as clean a signal _into_ the DAW as
> possible and from that point on the differences are somewhat reduced.
> --scott
Audacity is what I have and I do not have a budget to buy anything else for
this single project. It has been ages since I have converted tape to CD and
they were always store-bought tapes. This is definitely NOT store bought. I
am free to give these away, but not sell any copies, per grandson of
artists. He wants to give them away, but nobody has tape decks anymore.
Laurence Payne
<sur...@stopassaultnow.net> wrote:
Might be worth trying anyway. I've rarely experienced a transfer from
cassette where head alignment didn't improve things. If the
alignment's WAY off, you could even be getting problems with
cancellation between the two stereo tracks. IS the recording stereo?
What does just one channel sound like?
Peter Larsen
>>> Licensing issues aside, I find it inconceivable that a modern
>>> computer can't emulate 30-year old Dolby technology. Someone must
>>> have done it?
no, the link someone kindly supplied seems to be a to a pro-logic emulator.
>> Head alignment would not explain why the guitar is extremely clear,
>> but all the other instruments and vocals are off and only in a few
>> songs.
> Might be worth trying anyway. I've rarely experienced a transfer from
> cassette where head alignment didn't improve things. If the
> alignment's WAY off, you could even be getting problems with
> cancellation between the two stereo tracks. IS the recording stereo?
> What does just one channel sound like?
There is a simple test here: try summing the tracks, make a fft analysis,
cancellation if any should be obvious. Remember the context before chasing
grails and don't mess wit a taperecorder if you do not have the capacity to
alight it to a standardized aligmnent tape again.
There may not be a problem that matters, some of the time just be happy that
the cassette didn't disintegrate and that the tape didn't arrive broken.
ALWAYS DISCLAIM, cross fingers, whatever, casettes are a quagmire worse than
Omaha.
Kind regards
Peter Larsen
Scott Dorsey
>On 29 Nov 2009 12:59:58 -0500, Scott Dorsey wrote:
>
>> Michael Dobony <sur...@stopassaultnow.net> wrote:
>>>On 28 Nov 2009 18:12:16 -0500, Scott Dorsey wrote:
>>>
>>>> In article <199l7iiw4v6x9.1...@40tude.net>,
>>>> Michael Dobony <sur...@stopassaultnow.net> wrote:
>>>>>I was given a tape to convert to CD. My player is a Pioneer CT-W205R set to
>>>>>dolby B. The software I originally used is Total Recorder, but I also have
>>>>>Audacity. The tape is a Magnetics. Any hints on removing or at least
>>>>>reducing the tape hiss, any addins for Audacity that could help? The tape
>>>>>level is a little low, record level was at max and nowhere near clipping on
>>>>>record. Translation: S-N level was a little low on the tape.
>>>>
>>>> 1. Make sure the azimuth is correct. If the top end is muffled, you will
>>>> lose a LOT of S/N.
>>>
>>>Varies with the track. Some are decent, some need work, mostly in the 1-4k
>>>range and some in the under 100 need either boost or cut.
>>
>> How do you know? Have you adjusted the azimuth on each track to peak
>> the top end?
>
>Define "top end." Adjusting for every track? Get real. Also if the azimuth
>was off the guitar would be off along with everything else. The guitar
>sounds great. The rest of the instruments and the vocals are off.
Top end = high frequencies. The azimuth is off, because with every cassette
the azimuth is _always_ off.
Yes, it might be a bad mix too, but that's no reason not to adjust the azimuth
correctly.
>> If you lose it in playback, and you have to boost it afterward, what you
>> get is more noise. Set the machine for proper playback.
>
>Are you certain it is lost in playback as opposed to lost in original
>recording or re-recording? Boosting the hi-mids did not add noise.
It is lost because the azimuth of the playback machine does not match the
azimuth of the recording machine. The record azimuth is wrong, because it
is always wrong on every cassette machine, and it doesn't stay put for very
long anyway. That's why it's important to match azimuth for playback.
>>>> 2. Make sure the Dolby levels are correct. If you are losing the top end
>>>> at times because the Dolby decoder is mistracking, you will lose a lot of
>>>> perceived S/N.
>>>
>>>Again, varies with song.
>>
>> It shouldn't, if they were all recorded at the same time on the same
>> machine. If it's different from track to track, you're going to have to
>> readjust the Dolby levels on each track to try and keep the pumping down.
>
>I doubt they were recorded at the same time.
If they were recorded on different machines or if the internal Dolby levels
on the machine drifted, you may have to adjust the Dolby points on each
track by ear for the least audible pumping. This is a pain in the neck.
Again, it's an issue that is caused by recordings made on cassette decks
that were not properly aligned, because people never aligned them right
for the tape they were using. There are no line-up tones on a cassette
like there are for normal 1/4" tapes.
Arny Krueger
> Head alignment would not explain why the guitar is extremely clear, but
> all
> the other instruments and vocals are off and only in a few songs.
Depends what the various instruments are. Some instruments don't have the
kind of strong high frequency response that is very critical of azimuth
alignment. Acoustic guitars recordings can be very tolerant of HF loss,
again depending on the instrument and other details of the recording.
For example I was watching some of the documentaries of the Beetles and
contemporaneous groups on TV, and noticed that the electric guitars could be
heard but some of the videos also showed someone playing a tambourine that
you really had to listen for and when you picked it out it was more like a
rattle. The guitars sounded pretty horrible by modern standards, but every
note could be heard. I mention the tambourine because as a rule its a
pretty raucous instrument with fundamentals going way down into the
midrange.
Michael Dobony
So you think adjusting the azimuth on my machine, with possibly 3rd
generation recordings, will correct the azimuth drift from previous
generation off-azimuth recordings?
Laurence Payne
<sur...@stopassaultnow.net> wrote:
>So you think adjusting the azimuth on my machine, with possibly 3rd
>generation recordings, will correct the azimuth drift from previous
>generation off-azimuth recordings?
What's "azimuth drift?" You can match alignment with the machine that
made THAT recording. Presumably you'd agree that was a good idea?
Scott Dorsey
>On Tue, 1 Dec 2009 11:49:01 -0600, Michael Dobony
><sur...@stopassaultnow.net> wrote:
>
>>So you think adjusting the azimuth on my machine, with possibly 3rd
>>generation recordings, will correct the azimuth drift from previous
>>generation off-azimuth recordings?
Wait a moment? We are talking about recordings that are not first generation?
That changes everything completely. It is ABSOLUTELY worth whatever it takes
to get the first generation recordings if at all possible. You did not mention
that you were dealing with multiple generations of cassette nastiness.
>What's "azimuth drift?" You can match alignment with the machine that
>made THAT recording. Presumably you'd agree that was a good idea?
Problem with cassettes is that the azimuth isn't stable. You'll find it
shifts from the beginning of the tape to the end... you will often have to
ride it. That's why a machine with auto azimuth like the Dragon or a machine
with a vernier azimuth control like one of Saul Mineroff's retrofits is so
valuable.