I can't agree more. Jasinski's, Z's, and Hokey's tunes (and I'm sure others) usually sound great regardless of bit rate/encoding/whatever. I think worrying about bit rate and encoding can be a distraction from improving ones mixes.Taron wrote:Oh, I might be risking to lose sympathies here, but I really don't think the highest quality sound reproduction is a necessity for a good tune.
One Synth Challenge #83: MPowerSynth by MeldaProduction (Z.Prime Wins!)
- KVRist
- 396 posts since 7 Dec, 2006 from Richmond, VA, USA
-
- KVRAF
- 2396 posts since 7 Mar, 2014
For sure, I agree. It's just sometimes we get very very disappointed when we listen back to an uploaded song which just sounds dreadful compared to the original we have been mixing and hearing.
I did 6 versions before the one up now is at least satisfactory to my ears, fixing little bits here and there - the fixes were mainly in the bass vs high frequency and transients area. It's worth noting that most of my songs transcoded really well.
The last thing I do is obsess over any one patch. I too like to layer up sounds so they work together - very much "into" soundscape and quite dense sounds, when the composing takes me there.
But, watch out for the usual suspects this time round - I bet their tracks will be awesome compositions and sound design to blow us all away .
dB
The last thing I do is obsess over any one patch. I too like to layer up sounds so they work together - very much "into" soundscape and quite dense sounds, when the composing takes me there.
But, watch out for the usual suspects this time round - I bet their tracks will be awesome compositions and sound design to blow us all away .
dB
-
- KVRian
- 928 posts since 13 Jan, 2013 from United States
I agree with DB. After spending many hours fine tuning sound design and mix, it would be nice if those details came across properly in the Sound Cloud version of the song. Of course a good song works across any kind of reproduction, and for things that are going to be released to the world it makes sense to check your mix on small and big speakers and headphones, but as a component of this competition is sound design, it would be appropriate if the base level reproduction of the songs accurately reflected our efforts.doctorbob wrote:For sure, I agree. It's just sometimes we get very very disappointed when we listen back to an uploaded song which just sounds dreadful compared to the original we have been mixing and hearing.I did 6 versions before the one up now is at least satisfactory to my ears, fixing little bits here and there - the fixes were mainly in the bass vs high frequency and transients area. It's worth noting that most of my songs transcoded really well.
![]()
The last thing I do is obsess over any one patch. I too like to layer up sounds so they work together - very much "into" soundscape and quite dense sounds, when the composing takes me there.![]()
But, watch out for the usual suspects this time round - I bet their tracks will be awesome compositions and sound design to blow us all away .![]()
dB
- KVRist
- 379 posts since 29 Jul, 2013 from Toronto
Folks, just to keep the discussion thoroughly sidetracked on soundcloud encoding, could the issue be that SC is using Joint Stereo? Maybe encoding to 128kbps VBR Joint Stereo would exhibit the same artifacts? Would be nice to know what the exact soundcloud encoding is, for cases like this, to avoid 7angram's situation. Alas, I most definitely don't have time this month to do the science myself. (Admittedly, it'd be even nicer if they used a better encoding; srsly it's 2016 ppl)
-
- KVRist
- 427 posts since 24 Feb, 2015 from Stockholm, Sweden
Yes, Joint Stereo would explain the stereo issues. I'm guessing CBR though since that's implied (to me anyway) when you say 128kbs MP3.
I would be nice of Soundcloud actually told their user's exactly which algo they used. That way you could actually test the same settings yourself and experiment when mastering.
General advice I've read (but rarely apply...) is to band limit ~30-~17000Hz before transcoding. Sure, you'll lose some clarity, but the end result might be better than the scrambled mess the hihats sometimes end up sounding...
I would be nice of Soundcloud actually told their user's exactly which algo they used. That way you could actually test the same settings yourself and experiment when mastering.
General advice I've read (but rarely apply...) is to band limit ~30-~17000Hz before transcoding. Sure, you'll lose some clarity, but the end result might be better than the scrambled mess the hihats sometimes end up sounding...
-
Touch The Universe Touch The Universe https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=190615
- KVRAF
- 5988 posts since 2 Oct, 2008
Had to switch to fl studio, are there tips to make it play nicely there. I recall something but can't seem to find it after skimming a few pages.. Getting some timing issues and way too much cpu.
100 High Quality Soundsets: Omnisphere 2, Dune 3, Tone 2 Synths, Pigments, Uhe Synths, Halion, Spire, and others.
TTU Youtube
TTU Youtube
-
- KVRist
- 112 posts since 22 Apr, 2015
Works fine for me in FL 12 but eats a bit of RAM. Maybe you read about going to: SETTINGS=>Settings and deactivating GPU acceleration.Tjgoa wrote:Had to switch to fl studio, are there tips to make it play nicely there. I recall something but can't seem to find it after skimming a few pages.. Getting some timing issues and way too much cpu.
- KVRist
- 396 posts since 7 Dec, 2006 from Richmond, VA, USA
Entry:
https://soundcloud.com/hennessey/bh9090-reedwounds
34 instances in Live 9.5
DAW Effects: EQ8, Compressor, Auto Pan, Auto Filter, Saturator, Simple Delay, Filter Delay
3rd Party: Voxengo OldSkoolVerb, Tal Reverb 4, Camel Crusher, Molot, Limiter6
(BJ - Sorry for the multiple Soundcloud submits.
Hope that doesn't foul you up)
https://soundcloud.com/hennessey/bh9090-reedwounds
34 instances in Live 9.5
DAW Effects: EQ8, Compressor, Auto Pan, Auto Filter, Saturator, Simple Delay, Filter Delay
3rd Party: Voxengo OldSkoolVerb, Tal Reverb 4, Camel Crusher, Molot, Limiter6
(BJ - Sorry for the multiple Soundcloud submits.
- KVRist
- 68 posts since 30 Dec, 2015 from Sydney, Australia
Yes, a good idea to do this from the get go. On my laptop(quad core i7 Win7) with Reaper4 , I couldn't get more than 3 instances before I got an error message. After disabling the GPU acceleration, I never had any problems.Meakaale wrote:Works fine for me in FL 12 but eats a bit of RAM. Maybe you read about going to: SETTINGS=>Settings and deactivating GPU acceleration.Tjgoa wrote:Had to switch to fl studio, are there tips to make it play nicely there. I recall something but can't seem to find it after skimming a few pages.. Getting some timing issues and way too much cpu.
If you are on a Windows laptop with a dedicated GPU, then switching your DAW to use you dedicated GPU instead of the integrated GPU would probably also work. Not sure if it is using CUDA or OpenCL for acceleration though?
The nice thing is that you only have to disable it on your first instance, and subsequent instances of the plugin are automatically disabled.
-
- KVRAF
- 2396 posts since 7 Mar, 2014
Aha, I smell a great entry here. The electronic pads are great, good deep sub bass, percussion is very acceptable.Tjgoa wrote:Had to switch to fl studio ...
Cheers,
dB
-
- KVRAF
- 2623 posts since 20 Oct, 2014
https://soundcloud.com/hanzmeyzer/a-dar ... powersynth
My entry for One Synth Challenge #83. Most presets created from scratch. Tried to let it sound analog-like
It's not mastered or finalized.
Genre: Dreamwave, 80s
BPM: 88
DR: 12
DAW: Renoise 3.1, mostly purely MPowerSynth (with its fx section) + some DAW internal fx: 8x Compressor, 4x multi tap, 3x EQ5, 2x digital filter
My entry for One Synth Challenge #83. Most presets created from scratch. Tried to let it sound analog-like
Genre: Dreamwave, 80s
BPM: 88
DR: 12
DAW: Renoise 3.1, mostly purely MPowerSynth (with its fx section) + some DAW internal fx: 8x Compressor, 4x multi tap, 3x EQ5, 2x digital filter
Last edited by Hanz Meyzer on Sat Jan 23, 2016 7:39 pm, edited 2 times in total.
- KVRAF
- 2237 posts since 29 Sep, 2011
Yes, I think a couple things are in play here:bh9090 wrote:I can't agree more. Jasinski's, Z's, and Hokey's tunes (and I'm sure others) usually sound great regardless of bit rate/encoding/whatever. I think worrying about bit rate and encoding can be a distraction from improving ones mixes.Taron wrote:Oh, I might be risking to lose sympathies here, but I really don't think the highest quality sound reproduction is a necessity for a good tune.
- Listening to your mix after transcoding and/or on other systems generally reveals deficiencies in the mix
- Does the bass sound funky? It may not be leveled properly across the track or there may be multiple bass frequencies overlapping - make sure to HPF ALL the other tracks that shouldn't be overlapping the bass!! Probably at 120-140Hz+. Is the bass stereo? Make it mono < 100Hz! Is there very low frequency bleed < 30-40Hz? HPF it out if you need to!
- Have you properly compared it to a 'pro' track? The overall EQ audibly and through a spectrum analyzer?
- SoundCloud encoding seems to be optimized for louder volumes
- Is it a listenable level? Does it sound about as loud as other tracks on SoundCloud after you compare apples to apples with the uploaded SoundCloud version?
But most importantly, is the mix good? Have you spent the time to automate levels of the parts so they're not really loud in some places but quieter in others - so they don't 'get lost' at times when they shouldn't because of other competing parts? If you kill all the plugins on the master channel and listen to your mix raw, is it going over 0 a lot? for certain parts but not others?
I've started to mix with nothing on the master (no limiter, nothing). This forces you to get a good sounding track *in the mix*. After that, simply boosting a bit with the limiter takes it to just about where it needs to be with no other fiddling. Some of the best advice I've read is very simple and seems obvious now: learn your level meters. This seems silly, perhaps, but it's invaluable and will lead to a much faster mixing process. What I mean by 'learn your level meters' is: learn where the different sounds should fall. Once you get your tracks falling at the "right" level, the track basically 'mixes itself' as they say. High frequency sounds are easier to hear so low frequency sounds need to be louder. For example, a kick maybe would be -10db, a snare -11db, sub bass -12db, other basses -16db, hats/other perc -30/-20db, pads/leads/other stuff -20/-16db... as a very rough ballpark. Then the main melody would be boosted up just a bit to be louder. This way you'd not be going much over 0 anywhere, ideally, and the limiter can just have the signal boosted into it to get your levels up.
Anyway, hope that helps.
Last edited by z.prime on Sun Jan 24, 2016 2:39 pm, edited 2 times in total.
- KVRAF
- 2237 posts since 29 Sep, 2011
Side note: I've not started a track this month. I might tomorrow and see how far I can get. Chances of me finishing an OSC track in a week are slim, though it's happened before. 
- KVRAF
- 2237 posts since 29 Sep, 2011
Last thing... I highly recommend the 'Xilentify' plugin. It does wonders for your mix:
-
- KVRist
- 427 posts since 24 Feb, 2015 from Stockholm, Sweden
Very subtle effect but certainly makes everything shine a bit more.
