When changing gain, I would expect to see the traces move up (or down).
any reason, why there is no (visual) feedback?
One more change I'll like to see:
In addition to changing color on a trace, why not open for change in linestyle and weight.
regards
Flemming
Gain adjusment, visual feedback
Flemming,
The gain does not change the vertical position of the spectrum nor should it. The reason we chose this approach is two-fold.
Firstly, when performing tonal matching of a track with a reference you don't want the trace moving up/down as you change the gain given that you overlayed the track spectrum on the reference. It just makes the whole process more difficult.
Secondly, the vertical positioning of the spectrum is rather irrelevent since HarBal ignores any pure gain applied to the filter and chooses the overall gain on the basis of maintaining constant perceived loudness. If this sounds confusing just try this little experiment. Open a track without a filter. Now using the parametric EQ tool lift the spectrum up 10dB or so using a very low Q (flat) response such that the filtered version just looks like an amplified version of the original. Now play it back and toggle the EQ in and out. What differences do you hear? Nothing is the response I'm expecting from you! Har-Bal has chosen the overall gain to maintain constant perceived loudness so your filter change was laregely ignored. In that sense the vertical position of the spectrum is largely irrelevent and it also allows you to place your spectrum above a reference without altering the loudness, which is a definite plus.
Regards,
Paavo.
The gain does not change the vertical position of the spectrum nor should it. The reason we chose this approach is two-fold.
Firstly, when performing tonal matching of a track with a reference you don't want the trace moving up/down as you change the gain given that you overlayed the track spectrum on the reference. It just makes the whole process more difficult.
Secondly, the vertical positioning of the spectrum is rather irrelevent since HarBal ignores any pure gain applied to the filter and chooses the overall gain on the basis of maintaining constant perceived loudness. If this sounds confusing just try this little experiment. Open a track without a filter. Now using the parametric EQ tool lift the spectrum up 10dB or so using a very low Q (flat) response such that the filtered version just looks like an amplified version of the original. Now play it back and toggle the EQ in and out. What differences do you hear? Nothing is the response I'm expecting from you! Har-Bal has chosen the overall gain to maintain constant perceived loudness so your filter change was laregely ignored. In that sense the vertical position of the spectrum is largely irrelevent and it also allows you to place your spectrum above a reference without altering the loudness, which is a definite plus.
Regards,
Paavo.
When talking about filters, I do agree with you; what I am refeering to, is the gain-control (slider).
If you have 2 versions of the same track, only with a gain difference, loading one as reference and the other as source, these will show up as parallel traces; now do a loudness match, the gain will change, the tracks will sound the same, so I expect to see identical trace for track and reference.
regards
Flemming
If you have 2 versions of the same track, only with a gain difference, loading one as reference and the other as source, these will show up as parallel traces; now do a loudness match, the gain will change, the tracks will sound the same, so I expect to see identical trace for track and reference.
regards
Flemming
Flemming,fbc wrote:When talking about filters, I do agree with you; what I am refeering to, is the gain-control (slider).
If you have 2 versions of the same track, only with a gain difference, loading one as reference and the other as source, these will show up as parallel traces; now do a loudness match, the gain will change, the tracks will sound the same, so I expect to see identical trace for track and reference.
regards
Flemming
The problem here is that you can't easily have both behaviours at once. It's an eother or situation. Now if I were to support what you suggest then it would not be possible to apply a low Q filter change to a track to bring the spectrum up on-top of the reference.
Or perhaps you could a a partial fudge where the traces go up and down with the gain but then in the case of reference matching you've now coupled the gain with the spectrum and then every gain change you make will move the track spectrum off the reference spectrum. This, to my mind would be particularly annoying.
I can see your point of view but I'm not convinced that it is the way the product should behave. It is a matter of perspective. The way I envisage it is that the spectrum display corresponds to the signal before the gain block. Thus gain changes aren't reflected in the spectrum. This is much the same as a spectrum analyzer on an integrated amp. The volume control doesn't affect the display.
I'd be interested to hear if anyone else has any thoughts on this.
Regards,
Paavo.
-
mflorio
impossible
I could be wrong but...
Wouldn't raising the Gain (especially into saturation) essentially change the eq response traces ? You wouldn't get the same curve simply transposed up the y-axis, it would actually be a mis-representation of whats going on sonically. To verify this, raise the gain 6db or more, save eq'd file, then open up eq'd file back into HarBal for analyzing with the original non-eq version as a reference. I suspect the curves won't be the same shape...
Gain should be post-eq, as it already is.
Mike
Wouldn't raising the Gain (especially into saturation) essentially change the eq response traces ? You wouldn't get the same curve simply transposed up the y-axis, it would actually be a mis-representation of whats going on sonically. To verify this, raise the gain 6db or more, save eq'd file, then open up eq'd file back into HarBal for analyzing with the original non-eq version as a reference. I suspect the curves won't be the same shape...
Gain should be post-eq, as it already is.
Mike