A question for users who are also musicians

[jaydeshan]

Greetings,

This may seem an odd question for some.

I've spent years producing music. Serious amounts of times spent on not only the master mix, but individual channels (i.e. snare track.) When I engineer/produce a song, I spend a great deal of time on "getting control" of those individual tracks. Such as with the snare track example, adding compression and limiting to keep the peaks from cause issues in the master mix.

My question: Are there those who spend time on RG&B (or Narrow bands,) before combining them? I don't mean the work with reference imagaes (darks/flats/bias.) I mean once those images are processed and combined. Are there "tricks" that apply more to Ha than OIII? I hope you get my point.

Thanks for your time.

Jay

[RickS]

I occasionally do a little adjustment to one of the R/G/B colour channels, e.g. say I'm getting red halos around stars because the FWHM is worse in red then I might apply MorphologicalTransformation just to the R channel.  That's pretty uncommon though.  Usually I combine the linear R, G and B masters straight after integration and operate on the combined colour image from then onwards.

With narrowband I do a bit more.  I remove the stars from copies of the Ha, Oiii and Sii masters.  Then, if there's a big difference in SNR between the different masters I'll hit them each with different amounts of noise reduction.  After that I usually LinearFit the Oiii and Sii to the Ha before I do a 1:1:1 colour combine.  For the luminance I often combine the original Ha (with stars) with the starless Oiii and Sii.  That gives a luminance with nice, tight stars that still represents structures from the Oiii and Sii.

Cheers,
Rick.

[jaydeshan]

This exactly what I'm looking for, and I hoped my speculation would bear fruit. It did. Thanks so much.

[Juan Conejero]

Hi Jay,

adding compression and limiting to keep the peaks from cause issues in the master mix

By applying compression to a sound track, you are solving a dynamic range problem. Essentially, this is the same task we perform on images with tools like HDRMultiscaleTransform, or with CurvesTransformation to define nonlinear functions. Morphological transformations, wavelets, linear filtering, operations in the frequency domain, etc, also find their equivalents in the world of digital sound processing. This is signal processing in its widest sense.

The main problem is always the same: respecting the documentary value of the data you are processing. If you are processing a music track, especially if you are working with your own music in modern musical genres, you have freedom to do many transformations creatively. However, if you are working on tracks pertaining to a documentary work—for example, a track of bird sounds, to be used as footage for a documentary film, or a recording of classical music—, you cannot transform the data as you want. The same happens, exactly, with images. The same difference exists between processing an artistic visual work and processing a photograph of the sky. Astrophotography is documentary photography. Unfortunately, this is not understood or accepted by a large fraction of the people involved in this discipline.

By the way, I also make music and work with DAW hardware and applications (mainly Focusrite interfaces and Logic Pro X on an iMac), time permitting (which does not happen lately; too much work on PI). If Vicent sees this post I'm sure he'll chime in (he's a professional musician and astrophotographer).

[Niall Saunders]

Hi Jay,

As Juan says - you have to consider not just what you are processing, or how you are processing, but - perhaps most importantly - why you are processing.

My first thoughts, on reading your original post about performing some post-processing steps on ondividual channels, was that you would have to be very 'sympathetic' to how you processed the individual channels.

I am sure that you could see parallels in your own audio-processing experience where a single 'instrument', when processed in isolation from the remainder of the 'orchestra' will always be in discord thereafter - either standing out because you didn't really have enough data to allow you to apply the processes that you performed, or because the data was 'too good' to match the remainder of the data.

This is why I prefer to process channels 'together' (and the tools within PixInsight are very powerful in this respect). You get to see, more or less in real time, how the slected process actions have affected the overall image.

But, just keep in mind that there are no rules 

[jaydeshan]

This is signal processing in its widest sense.

Yes, I'm starting to see that. There are parallels with my images (especially early on,) with my early music mixes after my first purchase of a stereo bus compressor. At the time, I thought they sounded great, but after years at the craft, I can't listen to them. The compressor (or my rookie use of,) caused the mix to "pump." Horrible. But I did just connect it and turn it on. My ears weren't trained yet to hear the difference. My friends loved them (still do,) but ugh. The equivalent might be the urge to over-saturate (etc.) your image, but in reality it looks like a 4 year old playing with lipstick.

True, the better the signal, the less likely you are to mess with it. Recording (insert your favorite seriously talented artist here,) straight into a high end microphone and signal chain needs very little. A crappy singer recorded through a cheap mic isn't fixable, and the closer one might get making this poor singer sound better, the less natural it sounds.

To get better, you spend a lot of time learning how to not let people see your impact on the mix. I want people enjoying the song, not noticing how well the guitar "fits" into the mix. So you learn about the knobs for everything, why things are applied (exactly Niall!) Gradually, mixes you used to kick out in an evening are now taking days. What do you do? You fine tune your workflows. Optimizing every step. You document hardware settings, anything to reduce the time, yet still produce a great sound. The same is happening with my "photon mixes."  heh

My goal is to present images of our beautiful universe that appear untouched and "natural." This thread is music to my ears. 

Jay

[John_Gill]

Very interesting guys!!!  A different, but the same way of looking at art and music.  I have noticed that I now look at photographs/images very differently than what I was doing a couple of months back.

Look up
John