Recording Vinyl using Griffin Powerwave - clipping?

[thearne3]

Hi,

I'm new here and can't find the answer I need, so I apologize if this has been covered!

I am using a Linn LP12 turntable and Linn Kairn pre-amp into a griffin Powerwave USB (using Tape Out and RCA cables) and an iMac Core 2 Duo. I'm getting good recordings, but the input (with Gain at lowest) still registers some yellow and occasionally red, indicating clipping - I presume.

Is there any way to limit/lower the input signal? Other alternatives?

Also: any alternatives to the griffin that would improve overall quality?

Thanks for any input!

Tom

[CDJonah_alt]

Take something you have recorded -- select the whole thing up and use
the analysis-waveform stats and see if it is clipping. If you are
lucky, the color was just warning you that you were getting close. If
it shows clipping, you can use a menu item to find the clipping points
-- might just be a crack.

You don't want to limit the input but possibly lower it. If it is
really clipping, you need an attenuator but lets hope you don't

Chuck

thearne3 wrote:

Hi,

I'm new here and can't find the answer I need, so I apologize if this has been covered!

I am using a Linn LP12 turntable and Linn Kairn pre-amp into a griffin Powerwave USB (using Tape Out and RCA cables) and an iMac Core 2 Duo. I'm getting good recordings, but the input (with Gain at lowest) still registers some yellow and occasionally red, indicating clipping - I presume.

Is there any way to limit/lower the input signal? Other alternatives?

Also: any alternatives to the griffin that would improve overall quality?

Thanks for any input!

Tom




_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

[thearne3]

I checked. Definitely not cracks - but also not that prevalent: about 350 instances/channel in an album of 35 minutes. From what I can see the max duration of a clip is 0.4 milliseconds (if I'm reading correctly at max zoom: measuring the flat part of the peak, with each marker is a millisecond - right?).

So...if the only option is an attenuator - where do I look for that? Would an alternative to the griffin give me a possibly lower input signal?

Thanks for your help, Chuck!

[Sonic Purity]

So...if the only option is an attenuator - where do I look for that? Would an alternative to the griffin give me a possibly lower input signal?

Well... looks like you have a very high-end preamp feeding IMO a very low-end A/D device. From http://www.griffintechnology.com/articl ... -powerwave we see:

Maximum Input Signal: RCA in (Line Level Mode): 0.0dBu

or about -2.25 dBV. I could not find specs for the Linn Kairn, yet from my couple-decades-ago audio component repair career, i clearly recall that the higher-end gear tended to run a bit "hotter" in terms of line levels, so odds are good that your Linn is putting out peaks that exceed this (unimpressive) maximum level of the Griffin device.

5 or so years ago, unable to get good results out of a Griffin iMic, a friend of mine got an M-Audio Transit USB interface. For something like $79 it was a good bit more than the Griffin, yet in no way a bank-breaker, and was/is a vastly superior device. There was also an Edirol model at the same price point that was/is probably as good. Now there are a whole fleet of options. Get yourself to your favorite pro audio retailer (i like Sweetwater.com) and shop around to see what best meets your needs.

Yes, you could go with an attenuator, yet i think you'll be better served overall with an A/D device more in keeping with the grade of your turntable and preamp, and specifically one designed with a maximum input level closer to +6 dBV (about 2Vrms) or ideally even higher.

Just looked at http://www.m-audio.com/products/en_us/Transit.html on the Specifications tab, and found:

maximum input voltage: +6dBV

Given your reported results with your current setup, i expect this will be fully sufficient. Might be worth verifying the maximum line output of the Linn preamp, though.

[thearne3]

Thanks, Sonic.

I may be missing something. The Griffin device is clearly putting a signal to the computer that is too strong. You have (I think) correctly identified one problem: the Griffin can't handle the input from the Kairn. Is it possible these are two separate issues?

Right now the clipping is literally showing as the wave peaks 'hitting' the outside boundaries of the channel - all with Gain at minimum. If I Normalize, those flat peaks 'shrink' away from the channel maximum, but remain clipped.

Could it be that the M-Audio, which can handle the Kairn output, might still overdrive the inputs to Maestro? There does not seem to be a Gain switch/knob on the M-Audio device.

What do you think?

Tom

[CDJonah_alt]

That looks like a real nice box -- I would be tempted if I could
justify it. With 24 bits instead of 16 and what appears to be another
25-30 db lower noise level, you should find it much easier to get a good
recording without obsessing about getting the signal to fill the entire
A/D range (not that seems to be your problem).

You suggest testing to see if the compliance would be sufficient.
Personally, I don't know how to do it without some (relatively) nice lab
equipment or some trickery (see below).

Not knowing the equipment he/you has/have, one possibility is to put the
pre-amp into a power-amp receiver and then take the signal out the
headphone jack. Those often have an attenuator so one could attenuate
the signal to keep it on scale and then use wavestats to measure the
peak value and some other reference value. One can then work out what
the peak should have been if it hadn't been clipped.

If I recall all the terms, one should be able to use almost a factor of
3 greater peak signal with the M-audio unit.

Chuck



Sonic Purity wrote:

thearne3 wrote:

So...if the only option is an attenuator - where do I look for that? Would an alternative to the griffin give me a possibly lower input signal?

Well... looks like you have a very high-end preamp feeding IMO a very low-end A/D device. From http://www.griffintechnology.com/articl ... -powerwave we see:

Maximum Input Signal: RCA in (Line Level Mode): 0.0dBu

or about -2.25 dBV. I could not find specs for the Linn Kairn, yet from my couple-decades-ago audio component repair career, i clearly recall that the higher-end gear tended to run a bit "hotter" in terms of line levels, so odds are good that your Linn is putting out peaks that exceed this (unimpressive) maximum level of the Griffin device.

5 or so years ago, unable to get good results out of a Griffin iMic, a friend of mine got an M-Audio Transit USB interface. For something like $79 it was a good bit more than the Griffin, yet in no way a bank-breaker, and was/is a vastly superior device. There was also an Edirol model at the same price point that was/is probably as good. Now there are a whole fleet of options. Get yourself to your favorite pro audio retailer (i like Sweetwater.com) and shop around to see what best meets your needs.

Yes, you could go with an attenuator, yet i think you'll be better served overall with an A/D device more in keeping with the grade of your turntable and preamp, and specifically one designed with a maximum input level closer to +6 dBV (about 2Vrms) or ideally even higher.

Just looked at http://www.m-audio.com/products/en_us/Transit.html on the Specifications tab, and found:

maximum input voltage: +6dBV

Given your reported results with your current setup, i expect this will be fully sufficient. Might be worth verifying the maximum line output of the Linn preamp, though.

------------------------
))Sonic((




_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

[CDJonah_alt]

No, the box is not putting too big a signal to the computer. Let's
review what happens. An analog signal comes into the Griffin. It is
converted to a digital signal. Say that the maximum voltage that the
A/D converter can convert is 1 volt (which it isn't), any signal greater
than 1 volt will give exactly the same output from the converter. So 1
volt, 1.1 volt or 2 volts will give 32767 (or -32768). This will then
make a constant value and then the clipping. The USB input is digital
and can only be one of the values given above.

Will the M-audio box have the same problem -- possibly but personally, I
doubt it. Check the previous e-mail for more details.

Chuck

thearne3 wrote:

Thanks, Sonic.

I may be missing something. The Griffin device is clearly putting a signal to the computer that is too strong. You have (I think) correctly identified one problem: the Griffin can't handle the input from the Kairn. Is it possible these are two separate issues?

Right now the clipping is literally showing as the wave peaks 'hitting' the outside boundaries of the channel - all with Gain at minimum. If I Normalize, those flat peaks 'shrink' away from the channel maximum, but remain clipped.

Could it be that the M-Audio, which can handle the Kairn output, might still overdrive the inputs to Maestro? There does not seem to be a Gain switch/knob on the M-Audio device.

What do you think?

Tom




_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list

[thearne3]

OK. I think I'll try the Edirol device, rather than the M-Audio - both are about $80but the Edirol has Input gain control. Otherwise seems to be comparable.

Thanks Sonic and Chuck!

[Sonic Purity]

Thanks, Sonic.

I may be missing something. The Griffin device is clearly putting a signal to the computer that is too strong. You have (I think) correctly identified one problem: the Griffin can't handle the input from the Kairn. Is it possible these are two separate issues?

Right now the clipping is literally showing as the wave peaks 'hitting' the outside boundaries of the channel - all with Gain at minimum. If I Normalize, those flat peaks 'shrink' away from the channel maximum, but remain clipped.

Could it be that the M-Audio, which can handle the Kairn output, might still overdrive the inputs to Maestro? There does not seem to be a Gain switch/knob on the M-Audio device.

What do you think?

Tom

Just wanted to add to what Chuck wrote... he's already covered this well. As he pointed out, the A/D hardware is in charge of changing the analog signal value to a numeric representation within a fixed range, for a given bit level. As far as i know, it would only be possible for a level in the digital realm to be "too high" if the software following the A/D hardware is not in agreement with the numeric range determined by format standards and the particular bit value. I have not personally seen this.

Software such as Amadeus can only control the levels to the degree the OS audio infrastructure and the A/D hardware allow it to be controlled (as far as i know... corrections welcomed). The real, true level adjustment is happening in the analog stages of the A/D converter, via gain adjustment to those stages. The abilities and options here vary quite a bit with different hardware. I have read in the past that there is or was some hardware whose levels Amadeus could not control, due to how the hardware was made: levels had to be controlled externally in analog hardware, as i vaguely recall.

There is no gain switch/knob in hardware on the M-Audio Transit. The Transit comes with its own software which does include level controls that i assume are adjusting the Transit analog stages via USB signaling. The software for OS 9 is quite bad, yet the software for OS X (at least through Tiger) is quite good (per using my friend’s and subsequent reports from him). Some (many?) competing products may not come with their own software, instead using what is part of OS X. There should be a way to set levels on any decent (pro-grade) device, and many less decent devices. Whether that is only the level control in OS X and/or Amadeus Pro, or in some other software, i could not say, and would likely be case-specific.

You suggest testing to see if the compliance would be sufficient.
Personally, I don't know how to do it without some (relatively) nice lab
equipment or some trickery (see below).

I may not have expressed myself well. If you are referring to "Might be worth verifying the maximum line output of the Linn preamp, though", i was thinking of putting more effort into looking up the published specifications than i did. For example, it may be in the owner's manual, which Tom may have.

Not knowing the equipment he/you has/have, one possibility is to put the
pre-amp into a power-amp receiver and then take the signal out the
headphone jack. Those often have an attenuator so one could attenuate
the signal to keep it on scale and then use wavestats to measure the
peak value and some other reference value. One can then work out what
the peak should have been if it hadn't been clipped.

I have long used a variant of this, especially on Vintage Mac (OS 9 and earlier) systems, where level control and display was more bizarrely uncorrelated (with the Sound Manager) than it is in OS X with Core Audio: there was far less relation between actual clipping, software knob setting, and meter display... at least on my systems. I tend to use a standard audio CD player with a variable output (line or headphone) as the signal source, playing a 0 dBFS (Full Scale) mid-frequency tone off any of many test discs that have such a tone. Making sure that the digital input levels on the Mac software are plenty low enough, i run up the level on the CD player until i hear (or see, looking at an actual waveform) clipping, then back off just below the clip point. I can then see what the digital level controls and displays do, and find my options for setting them.

Here's an example which i think goes a long way to prove Chuck's point, and may hopefully clarify things. As noted above, sometimes calibration needs to be performed with certain software and hardware. One such case is the freeware program Coaster for the Vintage Mac OS, which is a recorder only: not an editor, and does not even play back! Just records, with some of the best meters i have seen.

Coaster has a level control, with a nominal 1.00 reference gain mark. Yet i quickly found that when using this level, even though the Coaster meters never showed clipping, the recording was clipping. With this software, the level controls affect both the recording level and the meter display. The problem was that the A/D hardware was clipping due to too high an input level, and Coaster could not "reach in" to adjust that level (this was the built-in A/D on a Power Mac 8600). Using the procedure above, it was easy to adjust the input level to just below the clip point, then adjust the Coaster levels so its meters display just below clipping (and then run the CD player volume up and down a little to check clipping indicator tracking. I then noted the level mark in Coaster: that became my fixed reference point: i needed to reduce levels with an external analog attenuator. It was safe to raise the gain above that reference mark (meters would accurately represent clipping), yet never to go below that mark with the Coaster level controls. (I actually wound up making a whole level calibration chart... another story.)

Someday (probably not today, nor even soon), i'll get my act together and again bring up, here, the topic of the level meters in Amadeus. There are great alternative meters, yet i'd still like to see something better built into Amadeus Pro. I've found the meters substandard since Amadeus II, which is why i still use Coaster and its vastly superior meters for the A/D conversion (then AMII or AP for the editing and so on). Heck, i haven't even delivered the CD recording interface discussion i promised Martin i would bring here, back in February... coming as soon as i can, i promise!

[CDJonah_alt]

It wouldn't surprise me if such a useful spec does not exist for the
following reason. It isn't hard to get a lot of voltage range on a low
power pre-amp and the designers may well have designed to to be linear
for a large range of input signals, as the sensitivity of the cartridge
can vary quite a bit. So the maximum linear output may be much larger
than what he would get with his particular cartridge.

Chuck

CDJonah_alt wrote:

You suggest testing to see if the compliance would be sufficient.
Personally, I don't know how to do it without some (relatively) nice lab
equipment or some trickery (see below).

I may not have expressed myself well. If you are referring to "Might be worth verifying the maximum line output of the Linn preamp, though", i was thinking of putting more effort into looking up the published specifications than i did. For example, it may be in the owner's manual, which Tom may have.

--
Charles D. Jonah CDJonah@anl.gov
630-252-3471
Chemistry Division
Argonne National Laboratory
Argonne, IL 60439



_______________________________________________
Amadeus forum mailing list
Unsubscribe / change settings at http://two.pairlist.net/mailman/listinfo/forum_list