[Orchestral] The Painting (mp3)

[tvdude]

Hello everyone.  Here is a rough, unmastered mix of an orchestral piece.  What are your thoughts?  Here is the link:
http://213.129.73.14/members/my_tracks/audio/DT_ACFA9F9053826.mp3

The piano in the first 2 minutes is muffled on purpose for effect.  I am aware that some of my string samples are not of the highest quality, but I hope you can see the song for what it is, and not for the quality of the samples.  It also needs some equalization as once again I have had to mix in headphones and not on monitors.   I look forward to your feedback.

[Sam_Zen]

First I have to say something about mixing with headphones. I hope you once will get the opportunity to mix your production with real monitors. They're worth it.
Imo headphones give a total different impression of a piece. Sound is coming straight from the side L and R, instead of coming from the front with some distance between the speakers. It's just another experience.
So if I would use headphones for a mix, I would call it a piece which has to be listened to with headphones.

Perhaps there's a relation here with the fact that I have a problem with this song. The dynamics, could be a term.
Nothing wrong with the composition of the notes and the shape of the song. And sample quality is not at stake here.
But I'm listening to a band where every instrument has set its amplifiervolume on solo-strength all the time.
Like all are compressed and normalized with the same percentages.
So the more instruments are added in a certain phrase, the louder it gets. With big steps in volume. The risk of the max.
In this way the moments of 'impression' will depend on some math of addition, so not much control for the composer.
I know it's still rough, unmastered. A useful point to layer a volume envelope over some complete intrument tracks.
Until now, there was amplification of each instrument. Now maybe attenuation, of some of them, during some phrases ?

[bvanoudtshoorn]

I like this track. It's almost minimalist in its construction, which isn't a bad thing. The timpani sound as though they're playing an octave too high, though: if you can, I'd drop them down, and when you're mastering the track, pump up their volume! They're the loudest instruments in the orchestra, after all... =P And a tamtam would be nice as well. (I like big percussion sections. =D)

I like the piano construction at the beginning, especially the "pause" at around 1.30 when it goes back to just the crotchets for a moment before bringing in the strings etc. It makes their entrance more significant.

I agree with Sam_Zen that the quality of the sample's isn't what's important: you can always get higher quality samples, but you can't just "upgrade" your track. =P From a musical perspective, I think this is a good track. All it needs is to be mastered.

[KrazyKatz]

Simply Beautiful.. Timpani is at the correct octave. Get those strings fixed up and by hell get those cymbal crashes fixed up.
My own personal note would be to harmonize the ending with French horns.

Looking forward to the final version

[tvdude]

Hello everyone!  Thanks for all the responses.  

To SamZen:  Although I mix with headphones, I do try to do a final mix of all of my productions with a set of speakers.  I should clarify what I mean by that, they are Paradigm (Canadian made) stereo speakers, not near field monitors, and are not flat response.  They are typical stereo speakers, so the high end and low end are a bit "sweetened".  Although most of the production is done with headphones, the final mix is done with the speakers, situated at eight feet from my mixing console.  According to the specifications I downloaded for orchestral arrangement, the instruments should be panned somewhat accurately.  There is very little compression added to each instrument, if at all, and all instruments have been volume adjusted, so none of them should be at the maximum volume level.  However, compression is added at the final stage, and maybe I have this set wrong, as I am still learning mastering.  Most of my music is still at a much lower volume than most of what I hear on other sites, and am still figuring that out, as well, my music also seems to be a bit more muffled.  Maybe it is the quality of samples I use, I am not sure at this point.

To bvanoudtshoorn:
I agree with KrazyKatz; the timpani is at the correct octave, although I will boost them in the final mix, they were too loud when I first wrote the piece, now too soft, so I will fix that.

To KrazyKatz: I am in desperate need of an orchestral hand held symbal sample.  Do you know of any free samples on the internet that are of good quality?  Any help would be appreciateed since all of my symbols are of sub standard quality.  

Thanks everyone.

Best Regards,
tvdude

[Sam_Zen]

I don't see the need for compression of a final stage in the first place, because I like to produce the original as a WAV file.
Ok, if I decide to place it on the net, I'll need compression to make the files smaller, but that's a post-stage.
I avoid compression during the composition and production process. And if it's necessary to make an instrument sample (or a single track in a mix) louder, using compression as a tool is often chosen too quickly.
In most cases using Peak Limiting is enough to be able to normalize to a higher level.

2 tvdude
I dunno exactly what you mean by a orchestral hand held symbol , but maybe you can find one here

[KrazyKatz]

Hey TVdude,

Orchestral samples eh... Well you came to the right guy. You'll obviously need free stuff:

Theres the site that Cubaxd is always reccomending
http://www.philharmonia.co.uk/thesoundexchange/sound_samples/sample_libraries/phrases/

And for my reccomendation, google "The G-Town church sampling project". Its a project of sounds recorded in a church . You'll want the Piattish Cymbals. Bloody Brilliant Samples!
They may be hard to find, so if you cant just PM me and I'll send em to you.

You seem to be on your way with knowing about the orchestral way of mixing. One of the key aspects is the reverb. Give each instrument its own reverb, and dont compress the reverb. Don't be afraid of having a softer than usual piece. I'd suggest you listen to mixes by the great sound engineer Shawn Murphy. Particularly the "Peter Pan" soundtrack composed by James Newton Howard. Its surprisingly soft, but unbelievably clear, crisp and defined. You'd be surprised at how little little compression it uses.

The key to mastering is for the song to sound comfortable, at a comfortable volume. If its a gentle, delicate piece, it should have a soft character to it. If theres one thing I dont want to happen, its for orchestral music to get involved in the loudness wars.

Cheers!

[bvanoudtshoorn]

Of course, you can't go past the Vienna Symphonic Library for great orchestral sounds: http://vsl.co.at/... I just wish I had 9200 euros to spend on samples: http://vsl.co.at/en/211/442/484/349/216.htm. =D

I'm looking forward to the final mix.

[Sam_Zen]

I like to emphasize two quotes of KrazyKatz :
Give each instrument its own reverb (with the remark that the reverb line also can have it's own position in the panorama)
Don't get involved in the loudness wars

[tvdude]

Heya everyone.  Thanks again for all the great feedback.  This piece may take me a month or two to master, but you guys are right, it does sound like it is in competition for loudness.  I will try to make the instruments speak with a bit more emotion.  Wish me luck, as I really like this piece and hope not to mess it up.  Krazykatz, I may pm you if I can't find the sounds I need.  Sam Zen, as usual you are right on the money.  bvanoudtshoorn thanks for the tips regarding sites in addition to KrazKatz comments.  Hopefully I will be posting soon.

[tvdude]

I avoid compression during the composition and production process. And if it's necessary to make an instrument sample (or a single track in a mix) louder, using compression as a tool is often chosen too quickly.
In most cases using Peak Limiting is enough to be able to normalize to a higher level

SamZen

Why is peak limiting better on a single track or channel in a composition than compression on a single track or channel when composing to make an instrument or sample louder? I use a compressor (individually) on a number of channels when composing to either boost the volume or make the instrument a little more defined in the mix. Then, I compress the entire piece, with a peak limiter as the final piece in the chain to even out the peaks of the mastered file which is a Wav file.  Before I post it on my website I convert it to mp3.  Sam please give me your thoughts, and anyone else who would like to comment please do so.  I look forward to any comments on this, as it seems to fall into the realm of "mastering wars".  From everything I've read on the internet and audio magazines, there seems to be few rules regarding it, yet a lot of opinions based on personal taste. This is very interesting because in the United States and Canada, loudness was paramount, yet that view is just now changing within the last few years.  From what I have read, albums recorded in Europe were (and are) far less compressed than the same album released in North America.   Why would north american audio engineers do this?  Has anyone else heard of this?  Just for the record, I agree with what everyone has written about my piece by the way; it is too loud over all, there is not enough variance in the volume of certain instruments, and boy, those cymbals suck!....

So once again, if you have experience, knowledge, or an opinion on this subject, please comment.

[Saga Musix]

It is really a question of taste and what kind of track you're going to compress. For example, a strong compressor that's based on the rythm of the bass drum is often used in modern electronica to let it sound more "tight". However, I wouldn't recommend a strong compressor on classical pieces. The idea of applying a compressor on the whole song is okay, i do that aswell, to make the mix more balanced.

[Sam_Zen]

This is quite difficult to comment, because it's not only a matter of taste, it's also the shifting interpretations of when calling something a compressor.
So I will stay on the technical issues.
The main difference between peak limiting and compression is the treatment of the low volume passages.
Peak limiting doesn't do anything with this, it only scans for volume peaks above a certain value.
While compression not only dampens volume peaks, but also makes soft parts louder.
With both modulations, it is possible to increase the total volume of the sample/track/song, by e.g. normalization, once more.
But, with compression, there will be substantial loss of dynamics, because the soft/loud differences have become smaller.

I like to add, that I never use these methods as a standard procedure for soundfiles, but only if it appears to be necessary.
Peak limiting can be needed, if a few sharp peaks in a track prevent further normalization, while the rest is still too soft.
If it's only one peak, I even don't use a PL, but zoom in with my wav-editor, select the peak, and amplify it a few dB's down.
Because it's a very short part, a few msecs, the difference is not audible, things can be made louder with no loss of dynamics.
Sometimes I needed PL, because some song on an album was too soft compared to the other songs, wrong balance.
A nice simple tool for this is, yes, PeakLimiter. You can find a screenshot here.

Interesting thing about the use of compression in production on both side of the Atlantic.
I borrowed an audio single of Eminem from my granddaughter to make a copy, and when I loaded the 2 tracks in my wav-editor, I just saw one big green rectangle from start to end.
Massive compression, to make it as loud as possible, but of course any volume nuance was vanished.

About compression formats:
I never use such fileformats for the samples during the making of a module, or the tracks in a multitrack mixer, only WAV.
For web-publications I definitely switched from MP3 to OGG a while ago. For different reasons, also legally, but especially the dynamics and the stereo-panorama are quite better preserved in that lossy environment. But maybe I've stated this before.

Last but not least:
No matter if it's a single sample for a tracker, or a complete track in the mixer, the file should be just as loud as possible.
At that stage I don't care how loud it should sound in the end. It's just material and it should be as solid as possible.
No problem if I have to lower the volume setting of it in the working process. If the source is too soft, so one has to set the amplification above 0dB, then the base-noise of the file is amplified too, so the quality gets worse.

Sorry for the long post. I didn't have time to make it shorter..