Jump to content


Photo

Question On Recording And Output Levels


  • Please log in to reply
19 replies to this topic

#1 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 27 July 2013 - 06:48 AM

I have had a HD500 for about 2-3 years. I am questioning my levels though out the signal path. Generally, I set the mixer to center on both A and B with the level at 1.5 or 2.0. The amp level at 60 to 65. I use red and blue comps a lot, level set at 60 to 70. And the master out level on the top of the HD500 at 12:00 (middle). I am connected to a D box with a 1/4 " cable into the left mono 1/4" out put on The HD500. The D box then is connected to the PA system.

I got concerned with the possibility of some digital distortion. So I lowered levels in the preset, mix level 0.0 red and blue levels 50 to 55, amp level 37 to 43, and increased the master out to 1:00 or 2:00. I down loaded a graphical analyzer to check the peaks (clipping) between the two different level setups. But this analyzer is monitoring the digital signal through the USB port.

Question: If I set levels that don't clip in the graphical analyzer. Would that insure that there is no digital distortion at the 1/4" output to the D box (then to the PA system)?


  • 0

#2 mazuwa

mazuwa

    Iknowathingortwo

  • Members
  • PipPip
  • 412 posts

Posted 27 July 2013 - 07:59 AM

Interesting question. I always thought it would. But it is important to know that the usb signal runs through the Asio driver. So I set the Asio level to maximum as well before I decide to compensate my preset or master level. Furthermore even if the hd500 level is without unwanted clipping at the outputs, there is to question if this 0db level is perfect for your input. For the guitar is never the loudest device I always keep a distance of at least -6db to prevent unexpected clipping. Now I can adjust the input level. That is my way.
  • 0

#3 phil_m

phil_m

    Uber Guru

  • Line 6 Expert
  • 5484 posts
  • LocationMinneapolis, MN

Posted 27 July 2013 - 08:02 AM

It seems to me that you're talking about two or three different things here. It is possible to induce digital clipping in some of the effects blocks within the HD500's signal chain. For example, it seems like putting an EQ block after the amp block or mixer block with too high of a level can induce clipping. But this clipping won't necessarily be reflected as a clipped signal at any of the outputs, including the USB output. If, for example, you had the HD500's master volume low, the signal being recorded could still be far from clipping in the recording, but it could still have internal clipping. The graphical analyzer you have monitoring the USB signal will tell you what the levels of the overall signal are like, but it won't tell you anything about internal clipping. Since there's no meter for the internal signal path, you just have to listen for it. Personally, I've never had it be an issue, but it is a possibility.

 

As far as clipping on the other outputs, you monitor those as you would any other analog or digital signal going into a DAW. You have to look at your meters and make sure you're not going into the red.


  • 0
Time is a train
Makes the future the past
Leaves you standing in the station
Your face pressed up against the glass

 


#4 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 27 July 2013 - 02:49 PM

The ASIO level is set to max. And the higher level preset hit the - 1db and -3db (in the red) mostly in mid frequency. The lower level presets run near the -6db. I know that can be adjusted in the ASIO. And would do so for recording. I would agree with you mazuwa. I do the same, with good resaults.
As Phil_m menstioned , I am taking about to different things. Because the master output adjust the level through the 1/4" out. And does not affect the graphical analyzer. But, if an effect block (EQ) causes clipping. It would change the over all signal. How can it clip and not do anything in the over all signal.
My opinion would be this could be a good gauge to go by or set point. I noticed the low level preset are a little cleaner, less scratchy, harsh drive then the higher level preset. I would think the difference would be a lower signal to noise ratio. The lower levels internally gives more head room on the master output level. It is not a fix all. Some thing that would help.
Do you think I am on the right track? I would really like to find a way other than what I hear. I am get up there in age and don't trust my hearing. I have done well so far, I am getting some very good tone from the HD500.
  • 0

#5 saTa

saTa

    Just Startin'

  • Members
  • Pip
  • 39 posts

Posted 28 July 2013 - 01:07 PM

What phil_m is saying is: If clipping occurs within your POD, it won't necessarily be reflected by your DAW. It is important to distinguish clipping from signals hitting 0dbfs.

 

Yes, hitting the 0dbfs mark will clip your signal but if you turn it down by 2dB afterwards, the next stage won't recognise that because the signal has been clipped and then turned down. Your EQ would then show up peaks at -2dB and everything would look fine but still sound digitally distorted.

 

As a general rule, clipping is an issue that should have a priority much higher than SNR with digital devices such as the HD500. You could easily track all your guitars with peaks at -20dB and not have to worry about noise... tracking hot, that is to say, as loudly as possible without going into the red if a left-over from analogue days that doesn't really apply to nowadays practice anymore... but that's another subject altogether. But even then, it had more to do with the saturation-behaviour of analogue recording equipment rather than SNR.


  • 0

#6 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 28 July 2013 - 07:06 PM

Thanks sata, mazuwa, and phil_m, that is why my graphical analyzer is not a good idea. I will just keep doing what I have been doing. Keeping the levels to the lower side at the beginning.  Thank for the info. 


  • 0

#7 saTa

saTa

    Just Startin'

  • Members
  • Pip
  • 39 posts

Posted 28 July 2013 - 10:49 PM

You could also bypass all the fx blocks and then add them back in from left to right while monitoring the output... put simply, the concept of "gain staging" suggests that each gain stage in a signal chain (comps, eqs, pres etc.) should make the signal louder - never drastically turn it down. So if after your stomp box the signal is much louder than after the next effect, tweak the stomp box output. If you stick to this, SNR will be optimal and digital clipping avoided.


  • 0

#8 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 29 July 2013 - 04:27 AM

I was thinking the same thing. And how much of a pain that would be. A rule of mine is to set the first effect block at 40 to 50. then the next the same or a little more. It gets tricky when trying to set each preset to the same level, so I'm not driving the sound man crazy wihen changing from preset to preset.
All of my presets have the foot peddle assigned to the drive of the amp or the distortion effect. Up lowers the drive ( gain), down raises the drive. With clean to drive and everything between. But I have to assign the amp level to composite for the different levels. Foot peddle up drive down volum up. Foot peddle down drive up volum down. At the same time keeping it from clipping.
  • 0

#9 MartinDorr

MartinDorr

    Iknowathingortwo

  • Members
  • PipPip
  • 228 posts
  • LocationArizona, U.S.A.

Posted 29 July 2013 - 08:08 AM

Once you consider that digital models doing the tone shaping it makes no sense to think of them as gain staged elements increasing output to a louder and louder level. All those models have an ideal signal level range to run in. Signals in the digital domain do not need to to be made louder as a whole, they are transformed in their frequency content and per-frequency volume by the models. Independent from an aggregate signal level all models have limits in respect to what they do when the output of their normal function exceeds what can represented by a number (e.g., when your Amp setting would duplicate the volume and you have a spike of input level at 60% the model simply cannot put out more than 100%, i.e., it must use some internal mechanisms to limit such a signal to less than 100%.

From what I measured there seem to be a -12dB magic limit in most of the Line 6 models (I have evidence for some EQs and all Amps). Once you go higher they start to soft clip (compress) and if you push them towards 0dBFS they distort more and more till they get into hard clipping mode(their internal processing can no longer limit the output to something less than 0dBFS and they simply cut off at 0dBFS. As far as I can tell the only exception is the mixer and possibly the volume pedal. They seem to do no soft clipping and simply cut when the output exceeds 0dBFS.

With that in mind one would probably want to take your guitar input signal as soon as possible to a healthy high level but not exceeding the -12dB at peaks unless you actually want some amount of clipping. All subsequent model stages should be focused on shaping the tone, but not to change the level of the signal unless changing the signal level is essential to achieving some intended effect (e.g., get some soft compression w/o an effect model as a side effect of going above -12dB output).

Good luck. Your millage may vary ;-),
Martin

PS: I gathered my data using USB and DAW for measurements.
  • 0

#10 pfsmith0

pfsmith0

    Iknowathingortwo

  • Members
  • PipPip
  • 118 posts

Posted 29 July 2013 - 10:16 AM

I've done quite a bit of measurements on the HD500 and summarized the results in http://line6.com/sup...uency#entry5206 and http://line6.com/sup...s/?hl=frequency. I've found that digital clipping occurs long before any of the analog output stages clip. Also, except for the the Mic input (which has it's own trim control), digital clipping occurs before any of the analog input stages clip.


  • 0

#11 MartinDorr

MartinDorr

    Iknowathingortwo

  • Members
  • PipPip
  • 228 posts
  • LocationArizona, U.S.A.

Posted 29 July 2013 - 03:35 PM

Not sure whether your feedback is in addition to my comments or stating a diverging opinion. By the way, your measurements are extremely helpful understanding what's going on. I just wished I had them before I started looking closer at the preamps and some EQs.

My comments where in regard to what the models appear to do in the digital domain. You can easily verify this by doing a few measurements around the -12dbFS input level for say the Studio EQ. If you dial in a +6dB gain on a -18dB input signal you'll get about -12dB output. I say about because the Studio EQ is pretty bad at matching the dialed in gain and it may be up to almost 2dB off depending on what gain you dialed in. But I think this does not matter because you measure some gain and that can be used as your baseline. Now if you increase the input level by say 6dB you will not get an increase of 6dB output but less (and compression as well as some distortion). If you go higher it gets worse. If go lower from the initial -18dB input and the +6dB gain the resulting output gain increase stays the same, i.e., the Studio EQ gets into non-linear model behavior (just like the real thing) when you go above the -12dB intput or output level. Not sure what really happens if you come in high and then lower the gain, but I'd gues the damage is already done by whatever is in front of the Studio EQ.

I measured all preamps and a 2 EQ models. They all do this as far as I can tell. Just above the -12dB not much happens but once you push beyond -6dB the effect becomes pretty obvious. Thus my advice is stay below the -12dB on all internal signal paths unless you want that soft clipping / compression effect. If you need more output use the mixer as very last stage to raise your level higher than -12dBFS.

Martin
  • 0

#12 bjnette

bjnette

    Iknowathingortwo

  • Members
  • PipPip
  • 450 posts

Posted 29 July 2013 - 04:20 PM

Interesting and good to know.

 

If you have the time someone could use the FX send to feed a peak meter and analysis tools, then  by inserting the FX loop send block after each FX. 

So long as you keep the FX loop output the same  any FX could be observed for peak limiting using something like TT Dynamic Range Meter. http://www.pleasurizemusic.com/ This will reveal how much peak limiting is occurring. But all sorts of tools could be placed from the Loop send.

 

By completing the FX loop with a return on side B panned hard and comparing to the main outputs on the HD the comparison will have more value.

 

The is still plenty to find out. The DAC digital to analogue converters will have a say in what goes out.

 

Audio Diff maker could also reveal a thing or two http://www.libinst.c...o DiffMaker.htm


  • 0

#13 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 31 July 2013 - 02:31 AM

MartinDorr, could you possibly give a general level setting. You suggest staying in the -12db rang. I think it would simplify setting up preset. I like others don't have the resources to monitor the db level. Something like "x" level for the different FX models. Posibly make a table of some sort. Or EQs level setting is 40 to 50 for best tone quality.
  • 0

#14 MartinDorr

MartinDorr

    Iknowathingortwo

  • Members
  • PipPip
  • 228 posts
  • LocationArizona, U.S.A.

Posted 31 July 2013 - 06:56 AM

I am afraid there is no general gain level that would make sense to know for a number of reasons:
For one it is unknown what level range of your guitar input signal is.
Second, any effect or amp model you select and dial parameters in will have a different input to ouput level transfer function at a neutral gain setting.
In some cases even the default model parameters do not produce a 0 gain transfer function, e.g., Mid Focus EQ.
For amps the gain transfer function depends on your guitar input signal level, drive, and channel volume even if there would be a magical 'neutral' tone stack setting (it is not known how the actual amps tone controls are modeled with the models tone controls, some controls are added some are left off).
Even if you use a JTV and come in using the digital connection, different guitar models produce different output levels with same volume/tone settings.

I think the best you can do (if you want to go about this 'scientifically') is to try and get a feeling of the guitar input level you produce while selecting your favorite pickup configuration and volume/tone controls. If you know that you come in low or high (relative to some choosen peak, e.g., -12dB) you can by default use a preferred first effect model (like a Mid Focus EQ) to raise or lower your signal to feed subsequent models while adjust frequency content (which is often something you'd want anyway) to get specific tones.

A simple experiment w/o measuring would be to setup up a tone with no amp and no effects except a Studio EQ (to keep the sound uncolored) at 0 gain and the Mixer at 0dB, left or right channel centered, other channel muted and input 2 set to an unused source (if you use Variac select Aux, if you use guitar use Variax) to get the cleanest and least touched signal through the HD500. If you have a DAW or other measurement equipment see what level you get out and adjust the the Studio EQ's gain so that your hardest hit or strumming does not exceed your target dB level (I think slightly less than -12dB for peaks would be a good target). You should hear no distortion whatsoever. If you can't measure you need to trust your ears, add gain and listen for distortion. If your hear a hint of it I would back off at least 3 if not 6 dB. Obviously this whole process only matters if you want to maximise output and work with the healthiest undistorted guitar input signal you can produce within the HD.

Once you have that 'optimal' signal I think you should do nothing in any subsequent model to raise the level. if the output gets louder you probably entered a models soft clipping, compression, distortion range. But remember there is nothing wrong doing that if you want this effect.

Sorry to be of no more help. Good luck, and let us know how this works for you,
Martin
  • 0

#15 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 31 July 2013 - 07:37 AM

That sounds like a good idea. I have a preamp in my strat that might be out of the norm. No volume, adjustable gain and two tone controls. I also have a Les Paul traditional. Would be interesting to see the difference. I will post what I find.
  • 0

#16 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 31 July 2013 - 06:01 PM

MartinDorr, I tried out your test. I was not surprised with the outcome, kind of reassuring what I thought. I used a graphical analyzing software, a preset with no effects or amp (default preset). I put a studio EQ with the gain a 0db, mixer at 0db (both A&B) and pan set to center. I tried my Strat that has a preamp, gain set to 1. The 364Hz seemed to be a focus point. Other frequencies like 185,272.5, 545.5Hz peaked but not as much as the 364Hz. That peaked at -21db (guitar gain 1) and -15db (guitar gain 10). And that was the bridge pickup. The Les Paul peaked at -19db at 364Hz and that also was the bridge pickup. I would think that the input signal from just about any guitar would not clip unless there was a effect peddle before the HD500.
A lot of my presets hit the -3db to 0db, that is with the same graphical analyzer. Not good for recording, that can be adjusted. Just seems to me that if the presets are on average in the area of -3db. It would affect the tone quality. Yes, I can adjust the master out so it does not clip in the PA system. But the internal clipping has all ready affected the tone quality.
  • 0

#17 MartinDorr

MartinDorr

    Iknowathingortwo

  • Members
  • PipPip
  • 228 posts
  • LocationArizona, U.S.A.

Posted 01 August 2013 - 08:21 AM

Thanks for feedback.

 

Couple questions to put this in perspective with my own measurements and possibly other member's situation if they do this experiment too.

  1. It looks like you did not mute one of the A / B channels in the mixer. If you do, you should see 6dB less output level. So dependent on what you do with mixer in your presets (and where the mixer is in your effects chain) you may not compress the tone as much as you may think when the output hit -3 to 0 dB. Can you confirm that you muted or not muted one of the channels in the mixer for the experiment and what your mixer does and where it is in the presets?
  2. You did not mention the input setting. If you did not follow my description (not to use Same for Input 2) let me/us know. No using Same should lower your guitar input signal by 6dB in the experiment.
  3. I assume from your description that you did not use USB and one of the analog output options to measure. I have no experience with that and can't tell how the DA conversion and line level conversion impacts volume levels. In other words, it is not clear at all whether you exceeded the internal digital threshold (e.g., my assumed -12dBFS) even though the analog signal you measured was peaking somewhere between -3 and 0dB. There is also the potential for measurement errors due to uncalibrated equipment and level miss-match problems. I am not saying you ran into these things, but it would help if you can be specific of what analyzer equipment you connected to what outputs, whether the equipment was calibrated and anything else you may have done to asure reliable results.

Thanks again for digging in. It's always good to hear what others run into actualling repeating experiments with slightly different variables. I am a JTV user and will try to dig up what I found exactly when doing this and measuring via USB. Too bad I don't have anything reliable to measure what comes out on my analog outputs to compare with the USB results.

 

Martin


  • 0

#18 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 01 August 2013 - 04:34 PM

OK, I see some confusion about my last post. Most of my presets (ones I use every day) are hitting the -3db. The test tone as we will call it, never exceeded -20db or so. To answer your questions:

 

I did not mute either channel, I have input 1 set to guitar and input 2 to variax making that a moog point.

 

Did not use same for input 2, I always set input 2 to variax.

 

I used the USB output for the test. The software I used is open source, so I don’t think there is any calibration. If I where submitting this somewhere as an official test I would be more concerned.

 

I have taken some screen shots of the test you mentioned. I also took some screen shots of a preset I use from time to time. It is called PK75 EFFECT 3. I tested that preset also, with screen shots of the test. This is one of the presets I said is hitting the -3db range.

So, looking at the two different tests, you can see that my Les Paul does not have too high of an input. Because your test tone shows that. Also, that there is clipping in my preset PK75 EFFECT 3. Because my test shows that. And that this clipping causes adverse effect on the tone quality. Regardless of the master output level.

So, as I asked the question in my first post. Using a DAW or other metering software, like the graphical analyzer. It will show internal clipping that affects the tone quality. And when using that preset live, will have a poor tone quality.

Attached Files


  • 0

#19 MartinDorr

MartinDorr

    Iknowathingortwo

  • Members
  • PipPip
  • 228 posts
  • LocationArizona, U.S.A.

Posted 02 August 2013 - 07:53 AM

Thanks for the update and nice picture posts. Now everything is much clearer.<br /><br />Because you used USB I am not concerned about calibration. You may want to check into any SW driver default dB adjustments though to make sure you captured the 'real deal' and not some adjusted numbers (my OS-X drivers and DAW played that trick on me before I discovered weeks later). Compared to my JTV measurements your guitar input level experiment measurements seem low, but I have never measured using the analog guitar input on the HD and everything may be quite normal.<br /><br />I am just not so sure about the clipping you think to see. The screen shot shows levels in orange and red if they exceed -20 and -10 dBFS, respectively. I don't think this means that they are clipping in the digital domain. They are just exceeding some meter display thresholds. I don't know anything about your SW, but it may simply be a meter setup that causes the coloring and not a sign of real clipping. Whithout knowing more (e.g., cut-off level plateaus when looking at highly zoomed in waveforms) I would not have assumed that you ran into real clipping because you did not hit 0dBFS. On the other hand there are 2 peaks captured that exceed my assumed -12 dB threshold and I would thus assume you may have run into slight soft compression for those frequency bands (all assuming my threshold theory is true for all models).<br /><br />What might be interresting to check is what peak levels you get when you turn effects off, one by one from right to left if you are concerned about adverse tone impact due to compression. You could have somewhere in between a transfer that is higher level than the final output signal. If true, this model to model transition would have likely more impact on your tone than the final level you get. Not sure how much more time you want to sink in to this (I got kind of tired measuring rather than playing when i did this ;-), but if you do you might get some insight into how to keep the signals in a healthy range all the way through your amp and effects chain.<br /><br />I would not want to spend my life constantly checking sound changes via measuring, but it's nice to have a backup method to check if parameter changes just don't seem to sound right.<br /><br />Again, thanks for posting your results. Was very interresting (at least for me and apparently you ;-).<br />Martin
  • 0

#20 jcspirited

jcspirited

    Just Startin'

  • Members
  • Pip
  • 68 posts

Posted 02 August 2013 - 11:29 AM

I think I will try your suggestion, turning off effects from right to left. And I didn't what to spend the amount of time that I had already. So, thank you for your time and info. It will not go to wast
  • 0




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users